… · CPIA PUBLICATION 550 VOLUME I OCTOBER 1990 Reproduction is not authorized except by specific permission. AD-A240 864,iIID11011,11, il I il …

CPIA PUBLICATION 550VOLUME IOCTOBER 1990Reproduction is not authorized except by specific permission.

AD-A240 864

,iIID11011,11, il I il 1990 JANNAF

PROPULSION MEETING

VOLUME I

ANAHEIM MARRIOTT HOTELAnaheim, CA

3-5 October 1990

CHEMICAL PROPULSION INFORMATION AGENCYOperating under contract N00014-91-C-0001

. THE JOHNS HOPKINS UNIVERSITY *a G. W. C. WHITING SCHOOL OF ENGINEERING 9 COLUMBIA, MARYLAND .

Approved for public relensc; distribution Is unlimited,

91-11661 Best A.,I li~ill l!/il~illi!! l~l~llll lil

Form ApprovedREPORT DOCUMENTATION PAGE 0M No. 0704-0188

PuS• ic re g burd en foe this collctiO n of informatiO n i estimated to average I hour oer response, includimg the time for reviewing instruclions, searching ewisting data sources.

gatheig and maintaining tie data needed, and 4omletig and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of thiscotlecior of information Including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for information Operations and Reports, 1215 JeffersonCcvi Hlghw p'. Suite 1204. Arlington, VA 222024302., and to the Office of Management and Budget, Paperwork Reduction Project (0704.0 188), Washington, DC 20503,.

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED

October 1990 Meeting Proceedings, 26 May 89-5 Oct 904L TITLE AND SUBTITLE 5. FUNDING NUMBERS

1990 JANNAF Propulsion Meeting, Volume T C - N00014-91-C-0001PE - 65802 S

6. AUTHOR(S) PR - 1.0

Brown, Karen L. and Eggleston, Debra S. (Editors)

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION

Johns Hopkins University REPORT NUMBER

Chemical Propulsion Information Agency CPIA Publication 550,10630 Little Patuxent Parkway, Suite 202 Volume IColumbia, MD 21044-3200

9. SPONSORING /MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSORING/ MONITORINGAGENCY REPORT NUMBER

DLA/DTIC Office of Naval ResearchCameron Station 800 N. Quincy StreetAlexandria, VA 22314 Arlington, VA 22217-5000

11. SUPPLEMENTARY NOTES

Reproduction not authorized except by specif'ic permission from CPIA. DTIC-assignedsource code is 081100.

12a. DISTRIBUTION/ AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE

Approved for public release; distribution is unlimited.

13. ABSTRACT (Maximum 200 words)

This volume, the first of six volumes, is a collection of 34 unclassified/unlimiteddistribution papers which were presented at the 1990 Joint Army-Navy-NASA-Air Force(JANNAF) Propulsion Meeting, held 3-5 October 1990,,at the Anaheim Marriott Hotel inAnaheim, California. -Specific subjects discussed ihclude insensitive munitions,composite rocket motor cases; instrumentation; turboramjets; hypersonic airbreathingtest facilities; airbreathing combined cycle engines; electric thrusters,' andmonopropellant systems.

Bc:l. Available C&iy

14. SUBJECT TERMS 15. NUMBER OF PAGESElvctric propulsion Rocket engine cases 458Gun prop, ]I.ants Solid rocket propel. lants 16. PRICE CODELiqciid 1 ropel lant- rocket. engine 'rest: facili ties

17. SECURITY CLASSIFICATION 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACTOF REPORT OF THIS PAGE OF ABSTRACT

1Hrtr, I ;I¶-5 f i 1c s 1 v ( I tfhu. I "tI s I f* i ,d UL

'1,, ,'4001" ','Hii)O Standard Form 298 (Rev 2,89)i'i.1' ,t iy ANý l 'c.

PAGESARE

MISSINGIN

ORIGINALDOCUMENT

A( -. • ,

CPIA PUBUCATION 550VOLUME IOCrOBER 1990 •" ..

Repred•iom b noo uthodzed atcept by specific permission.

1990 JANNAF -__

PROPULSION MEETING

VOLUME I

ANAHEIM MARRIOTT HOTELAnaheim, CA

3-5 October 1990

CHEMICAL PROPULSION INFORMATION AGENCYOperating under contract N00014-91-C-0001

e THE JOHNS HOPKINS UNIVERSITY.

e G. W. C. WHITING SCHOOL OF ENGINEERING * COLUMBIA, MARYLAND.


Best !va'la'oe COPY

The Chemical Propulsion Information Agency (CPIA) is a DoD Information Analysis Centeroperated by The Johns Hopkins Univcrsity, G. W. C. Whiting School of Engineering, Charles and34th Streets, Baltimore, Maryland 21218, under Office of Naval Technology ContractN00014-91-C-0001. The applicable DoD Instruction is 3200.12-R-2, "Centers for Analysis of Scientificand Technical Information."

The CPIA also provides technical and administrative support to the Jouit Army-Navy-NASA-AirForce (JANNAF) Interagency Propulsion Committee in accordance with DoD Instruction 5030.24,"Interagency Propulsion Committee (Army, Navy, National Aeronautics and Space Administration,Air Force (JANNAF))."

The Government Administrative Manager for CPIA is Mr. Paul Klincfeltcr, Program Manager forInformation Analysis Centers, Defense Technical Information Center, Code DTIC-DF, CameronStation, Alexandria, Virginia 22304-6145. The Government Technical Manager (Contracting Officer'sTechnical Representative) is Mr. David S. Siegel, Office of Naval Technology, OCNR-213, Arlington,VA 22217-5000.

iiJ

I!

PREFACE

The Joint Army-Navy-NASA-Air Force (JANNAF) Propulsion Meetings (JPMs) are governedby the JANNAF Interagency Propulsion Committee and are conducted on an approximately annualbasis to promote the exchange of technical information in the fields of missile, space, and gunpropulsion. Sessions are organized to cover significant technology advances during the past year aswell as problem areas under current investigation on Government-funded programs. Sponsorship ofthe JPM rotates among the JANNAF agencies. This year's meeting was hosted by the AstronauticsLaboratory (AFSC), Edwards Air Force Base, California.

The 1990 JANNAF Propulsion Meeting was held at the Anaheim Marriott Hotel, Anaheim,California on October 3-5, 1990. Mr. Berge B. Goshgarian of the Astronautics Laboratory (AFSC)was the Meeting Chairman.

Proceedings of this meeting are published in a six-volume set by the Chemical PropulsionInformation Agency (CPIA). Volumes I and II are Unclassified and are approved for public release.These volumes may be purchased from the National Technical Information Service (NTIS), U. S.Department of Commerce, Springfield, Virginia 22161.

Volumes II1, IV, and V are Unclassified and carry "limited distribution" statements, andVolume VI is Confidential. Combined author, source, and subject indexes, as well as a list of meetingattendees, appear in all six volumes of these proceedings. These documents are available only toqualified subscribers of CPIA's products and services.

CPIA subscribers may request copies of the meeting proceedings by contacting The JohnsHopkins University, Chemical Propulsion Information Agency, 10630 Little Patuxent Parkway, Suite202, Columbia, Maryland 21044.

Karen L. BrownDebra S. EgglestonEditors

iii

CONTENTS

PREFACE .............................................................. in

1990 JPM KEYNOTE ADDRESS: *AEROSPACE COMPETITIVENESS FOR THETWENTY-FIRST CENTURY" .............................................

J. M. Swihart, National Center for Advanced Technologies, Washington, DC

NOZZLE VECTOR ANGLE DETERMINATION USING A LASERMEASUREMENT SYSTEM ............................................... 33

W. A. Mascio, Astronautics Laboratory (AFSC), Edwards Air Force Base, CA; J. F.Sely, Sverdrup Technology, Incorporated, Arnold Air Force Base, TN; and R. J. Torick,Jr., Arnold Engineering Devek~pment Center, Arnold Air Force Base, TN

EVOLUTION OF AN AGING PROGRAM -- MINUTEMAN STAGE II SOLIDROCKET MOTOR ..................... ................................. 45

P. W. Veit, L. G. Landuk, J. W. Simpson, Jr., and G. J. Svob, Aerojet PropulsionDivision, Sacramento, CA

INSENSITIVE MUNITIONS MASTERPLAN AND ITS IMPLICATIONS FOR THESDI COM MUNITY ...................................................... 55

F. R. Cook, Army Strategic Defense Command, Huntsville, AL, and W. C. Stone, StoneEngineering Company, Huntsville, AL

FLIGHTWEIGHTI DIVERT PROPULSION SYSTEM HOT-FIRE TEST ............ 61L K. West and P. C. Phillipsen, Aerojet Propulsion Division, Sacramento, CA

DEVELOPMENT OF THE ORBUSe I CARBON FIBER/EPOXY MOTOR CASEFOR TIlE STARS/STARBIRD PROGRAMS .................................. 69

H. N. Reynolds and M. Dhillon, United Technologies Corporation, Chemical SystemsDivision, San Jose, CA

JANNAF STANDARDIZATION OF TENSION, COMPRESSION, AND SHEARTEST METHODS TO DETERMINE MECHANICAL MATERIAL PROPERTIESFOR FILAMENT WOUND COMPOSITE STRUCTURES ....................... 89

T. L Vandiver, Army Missile Command, Redstone Arsenal, AL

SOLID ROCKET PROPULSION APPLICATIONS FOR ADVANCED POLYMERS ... 103J. S. B. Chew and J. Rusek, Astronautics Laboratory (AFSC), Edwards Air Force Base,CA

SOLID ROCKET BOOSTER INTEGRATION WITH THEAQM-37C MISSILE TARGET ............................................. 111

F. M. Cumbo, Naval Ordnance Station, Indian Head, MD

v

STABILITY TESTING OF FULL SCALE TACTICAL MOTORS .................. 119F. S. Blontshield, Naval Weapons Center, China Lake, CA; J. E. Crump and H. B.Mathes, Comarco, Incorporated, Ridgecrest, CA; and M. W. Beckstead, Brigham YoungUniversity, Provo, UT

NONLINEAR STABILITY TESTING AND PULSING OF FULLSCALE TACTICAL MOTORS ............................................. 135

F. S. Blomshield and C. A. Beiter, Naval Weapons Center, China Lake, CA; H. B.Mathes and J. E. Crump, Comarco, Incorporated, Ridgecrest, CA; and M_ W. Beckstead,Brigham Young University, Provo, UT

SHEAR STRESS TRANSDUCER CONCEPTS ................................ 151E. C. Francis, R. E. Thompson, and S. W. Heerema, United Technologies Corporation.Chemical Systems Division, San Jose, CA

NORMAL STRESS TRANSDUCER BEHAVIOR ............................. 161R. E. Thompson and E. C. Francis, United Technologies Corporation, Chemical SystemsDivision, San Jose, CA

PROCESS AND QUALITY IMPROVEMENTS FOR M7 PROPELLANT ........... 171R. A. Gott, Hercules, Incorporated, Radford, VA

MECHANICAL DESIGN OF SURFACE LAUNCHED TACTICAL MISSILES ....... 181J. J. Yagla, Naval Surface Warfare Center, Dahlgren, VA

COOLING SYSTEM AND INSULATION CONCEPT FOR A MACH 5TURBO-RAMJET AIRCRAFT ............................................. 191

S. C. Jones, Lockheed Engineering and Sciences Company, Hampton, VA, and D. H.Petley, NASA Langley Research Center, Hampton, VA

A SUMMARY OF EUROPEAN AND JAPANESEHYPERSONIC FACILITY ACTIVITIES......................................203

D. G. DeCoursin, FluiDyne Engineering Corporation, Minneapolis, MN

HIGH PRESSURE MACH 10 TO 20 ELECTROTHERMALHYPERSONIC WIND TUNNEL ........................................... 211

R. L, Burton and F. D. Witherspoon, GT-Devices, Incorporated, Alexandria, VA, and0. Rizkalla and W. Chinitz, General Applied Science Laboratories, Ronkonkoma, NY

AN EXPERIMENTAL EVALUATION OF COMBUSTOR LINER MATERIALSFOR SOLID FUEL RAMJET TESTING .................................... 227

J. B. Oppelt, Wright Research and Development Center, Wright-Patterson Air ForceBase, OH

SHOCK WAVE/BOUNDARY LAYER INTERACTION CONTROL IN A GENERICHYPEFSONIC INLET ................................................... 235

h F. White and R. E. Lee, The Johns Hopkins University, Applied Physics Laboratory,Laurel, MD, and B. D. Couch and P. Galambos, General Dynamics Corporation, FortWorth, TX

vi

ADVANCED COMPUTATIONAL MODELS FOR ANALYZING HIGH SPEEDPROPULSIVE FLOWFIELDS ........................................... 247

S. M. Dash, Science Applications International Corporation, Fort Washington, PA

MODEL 320-2: A COMPACT ADVANCED UAV TURBOJET ................... 285E. H. Benstein, B. Cassem, and K. Elliott, Teledyne CAE, Toledo, OH

THE CRYOJET REVISITED .............................................. 295B. Singh, W. Roberts, and S. Harper, The Marquardt Company, Van Nuys, CA

PREMIXED, TURBULENT COMBUSTION OF AXISYMMETRIC SUDDENEXPANSION FLOWS .................................................. 305

S. A. Ahmed and A. S. Nejad, Wright Research and Development Center, Wright-Patterson Air Force Base, OH

THE DEVELOPMENT OF A LIQUID FUELED GAS GENERATOR FOR DUCTEDROCKET RESEARCH ................................................... 313

P. L. Buckley and E. Corporan, Wright Research and Development Center, Wright-Patterson Air Force Base, OH

HYPERVELOCITY BY EXTENDED PROPELLANT BURN ..................... 321F. A. Vassallo, Calspan Corporation, Buffalo, NY

MOLECULAR STRUCTURE OF LOW TEMPERATURE FORM OFTRIAMINOGUANIDINIUM NITRATE (TAGN) ............................. 331

A. J. Bra'uti, Army Armament Research, Development, and Engineering Center,Picatinny Ai.enal, NJ

HEATS OF EXPLOSION, DETONATION AND REACTIONPRODUCTS: THEIR ESTIMATION AND RELATION TO THEFIRST LAW OF THERMODYNAMICS ...................................... 345

E. E. Baroody and S. T. Peters, Naval Oidnance Station, Indian Head, MD

BALL POWDER* PROPELLANT APPLICATIONS TOLARGE CALIBER AMMUNITION ......................................... 355

E. J. Kirschke and A. F. Gonzalez, Olin Corporation, St. Marks, FL

PROGRESS REPORT ON LIQUID PROPELLANTINJECTOR/COMBUSTOR TESTS .......................................... 363

R. E. Rychnovsky, R. W. Carling, S. K. Griffiths, S. R. Voscn, and R. F. Renzi, SandiaNational Laboratories, Livermore, CA

FLUID DYNAMIC/COMBUSTION INTERACTIONS ASDRIVING MECHANISM OF PRESSURE OSCILLATIONSIN A REGENERATIVE LIQUID PROPELLANT GUN ......................... 369

K. C. Schadow and E. Gutmark, Naval Weapons Center, China Lake, CA, and N. S.Nosseir, San Diego State University, San Diego, CA

vii

DIAGNOSTICS OF IGNITION/COMBUSTION IN A BULK-LOADED LP GUN ..... 379R. L. Talley, Veritay Technology, Incorporated, East Amherst, NY

THE EFFECT OF ACOUSTIC DAMPENING DEVICES IN REGENERATIVELIQUID PROPELLANT GUNS ............................................ 389

N. E. Boyer, J. DeSpirito, J. D. Knapton, and G. P. Reeves, Army Ballistic ResearchLaboratory, Aberdeen Proving Ground, MD

REGENERATIVE LIQUID PROPELLANT GUN AND SELF PROPELLEDHOWITZER INTEGRATION ............................................. 401

W. E.. Jacobsmeyer, General Electric Company, Pittsfield, MA

AUTHOR INDEX ....................................................... 409

SOURCE INDEX ....................................................... 415

SUBJECT INDEX ....................................................... 423

MEETING ATrENDEES ................................................. 451

INITIAL DISTRIBUTION ................................................ 457

viii

JUNE, 1990

AEROSPACE COMPETITIVENESS FOR THE TWENTY-FIRST CENTURY

PRESENTATION BY

MR. JOHN M. SWIHARTPRESIDENT

NATIONAL CENTER FOR ADVANCED TECHNOLOGIES1250 I STREET, NORTHWEST

WASHINGTON, D. C.

i i i - i i | i - 'i i1

TITLE SLIDE ON

GOOD MORNING. I AM GOING TO TALK TODAY ABOUT AEROSPACE

COMPETITIVENESS AND WHAT WE ARE DOING TO IMPROVE OUR POSITION

IN THE GLOBAL MARKETPLACE. I'LL START BY DESCRIBING WHO WE ARE

AND HOW WE REPRESENT AEROSPACE.

AIA IS THE TRADE ASSOCIATION OF AIRCRAFT AND SPACECRAFT

MANUFACTURERS. THE NATIONAL CENTER FOR ADVANCED IECHNOLOGIES

IS A NON-PROFIT FOUNDATION ESTABLISHED BY THE AEROSPACE

INDUSTRIES ASSOCIATION TO COORDINATE THE ACTIVITIES OF THE AIA

TECHNICAL AND OPERATIONS COUNCIL'S KEY TECHNOLOGIES EFFORT.

THIS EFFORT IS DESIGNED TO CREATE AN ATMOSPHERE WHERE WE CAN

DEVELOP A NATIONAL CONSENSUS ON A SET OF KEY TECHNOLOGIES PLANS.

IN GENERAL, MY REMARKS WILL COVER THE EARLY DAYS OF THE

KEY TECHNOLOGIES FOP THE YEAR 2000 PROGRAM, REVIEW OUR GOALS,

PRESENT OUR TECHNOLOGY DEVELOPMENT PLANS AND DISCUSS THE

SUBJECT OF TECHNOLOGY VALIDATION DEMONSTRATIONS.

_______og V(nt

for

America's

Future

Compe itimeftx~ssThe National center for Advanced Technologies

Aerospace Industries Association

2

SLIDE 2 ON

IN 1986, THE AIA TECHNICAL COUNCIL, CONSISTING OF THE

SENIOR VICE-PRESIDENTS OF THE 54 AIA COMPANIES, REVIEWED THE

SIrUATION IN SOME OF THE OTHER U.S. INDUSTRIES, SUCH AS STEEL,

AUTOMOBILES AND ELECTRONICS. OUR CONCERN WAS THAT THE

AEROSPACE INDUSTRY, ALTHOUGH IT WAS STILL MAKING A MAJOR

CONTRIBUTION TO OUR BALANCE OF TRADE, MIGHT GO THE WAY OF THESE

OTHER INDUSTRIES, IF WE FAILED TO TEND TO OUR BASIC

TECHNOLOG IES.

%W~ue W!%to(tical Prqedent-s

1970 - The Smoke Stacks

1975 - Appliances

1980 - Automotive Industry

1985 Consumer Electronics

"• Aerospace Technology,

SLIDE 2 CONT'D

AFTER REVIEWING THE STATUS OF VARIOUS AEROSPACE

TECHNOLOGIES, AND COMPARING CURRENT R&D EFFORTS WITH THOSE OF

OTHER COUNTRIES, WE DECIDED UNANIMOUSLY, THAT WE HAD TO EMBARK

ON A CAMPAIGN OF ACCELERATING THE DEVELOPMENT OF SOME OF THE

MOST CRITICAL TECHNOLOGIES. IN THIS WAY, WE COULD BU'.STER OUR

WORLD LEADERSHIP IN AEROSPACE TECHNOLOGY AND 'NJTITNUE TO

PROVIuE THE NATION WITH A VERY FOSITIVE BALANCE OF TRADE, AS WE

HEAD INTO THE 21ST CENTURY. IT IS IMPORTANT NOT ONLY TO

REMEMBER WHAT THE CONSEQUENCES WILL BE IF WE ALLOW OUR

TECHNOLOGY TO FALL BEHIND THAT OF OUR INTERNATIONAL

COMPETITORS, BUT ALSO OF rAILING TO TAKE THE ADVANTAGE OF

LEAP-FROGGING THEIR TECHN!CAL CAPABILITIES.

S.. . .i

SLIDE 3 ON

WE KNEW, AS I STATED EARLIER. THAT WE WOULD HAVE TO FORGE

A NATIONAL CONSENSUS, IF YOU WILL, FOR THE ENCOURAGEMENT OF OUR

FUTURE INTERNATIONAL COMPETITIVENESS. WE WOULD BE TAKING INTO

ACCOUNT OUR CIVIL TRADE, OUR INTERNATIONAL DEFENSE TRADE, OUR

DEFENSE POSTURE, AND OUR SPACE PROGRAM, INCLUDING ITS

INTERNATIONAL BUSINESS APPLICATIONS. TO ACCOMPLISH THIS, WE

NEED THE COMBINED SUPPORT OF THE GOVERNMENT, THE INDUSTRY AND

ACADEMIA, A TRULY FORMIDABLE TROIKA, WHEN PUR1SUING THE SAME

GOALS. OUR OBJECTIVE, THEREFORE, WAS TO SET UP A PROGRAM THAT

WOULD KEEP THE AEROSPACE INDUSTRY, AND INDEED THE NATION, IN

THE FOREFRONT OF INTERNATIONAL COMPETITION. WHILE IT IS TRUE

THAT OUR INTERNATIONAL EXPORTS ARE INDEED INCREASING, IT IS

ALSO IMPORTANT TO REMEMBER THAT THE RATE OF GROWTH AND THE SIZE

OF OUR INTERNATIONAL IMPORTS ARE ALSO INCREASING.

(is" all"" 2000

35

se

77 7f 79 00 61 $7 13 '$A4 'IS 86 '8 7 '8S '$O

,: i i i i i i i i i i i

SLIDE 4 ON

WE THEREFORE SET UP THE KEY TECHNOLOGIES STEERING

COMMITTEE, A TEAM SELECTED FROM THE TECHNICAL COUNCIL. THIS

COMMITTEE ESTABLISHED A SET OF CRITERIA, AS SHOWN HERE, TO

SELECT THOSE BASIC, GENERIC TECHNOLOGIES THAT WE BELIEVED WERE

ABSOLUTELY NECESSARY IF WE WERE TO MAINTAIN U.S. AEROSPACE

LEADERSHIP. ALSO A LONG-RANGE, TEN YEAR, PROGRAM WAS NEEDED TO

ALLOW THESE TECHNOLOGIES TO BE BROUGHT TO FRUITION.

2000)Needs MoreMultipleu e !LEmphasis

HighS.Leverage

Enabling -- Te 9- )

S\ -Higherj Risk

Long Genericm

6

SLIDE 4 CONT'D

WE PICKED TECHNOLOGIES THAT, IN OUR VIEW, NEEDED MORE

EMPHASIS. WE SELECTED SOME THAT WERE HIGH RISK, BUT OFFERED

GREAT PAYOFF. FROM THAT GROUP, WE SELECTED THOSE WITH A LOT OF

POTENTIAL FOR SYNERGISTIC BENEFITS AMONG THEMSELVES. THE

THOUGHT WAS THAT, WHEN COMBINED, THESE BENEFITS WOULD MULTIPLY

THE POTENTIAL FOR EXCEPTIONAL PERFORMANCE IN THE END PRODUCT.

FOR EXAMPLE, THE MAINTENANCE COST, THE LIFE CYCLE COST AND THE

ENDURANCE LIFE OF OUR PRODUCTS CAN BE SIGNIFICANTLY ADVANCED IF

WE CAN COMBINE THE TECHNOLOGIES OF ULTRA-RELIABLE ELECTRONICS,

SOFTWARE, AND ARTIFICIAL INTELLIGENCE WITH COMPUTATIONAL

SCIENCE. FINALLY, WE SELECTED TECHNOLOGIES THAT WERE ENABLING

SO THAT WE COULD DEVELOP NEW TECHNOLOGY PROCESSES THAT WOULD

GET US FROM TECHNOLOGY APPLICATION TO PRODUCT IN HALF THE TIME

IT TAKES TODAY.

I7

SLIDE 5 ON

THE TECHNOLOGIES SHOWN HERE WERE SELECTED AS THOSE MOST

LIKELY TO ENABLE THE UNITED STATES AEROSPACE INDUSTRY, THE

DEFENSE DEPARTMENT AND NASA TO REMAIN THE MOST TECHNOLOGICALLY

ADVANCED IN THE WORLD.

8I

SLIDE 6 ON

SLIDE 6A ON SLIDE 6D ON SLIDE 6G ON

SLIDE 6B ON SLIDE 6E ON SLIDE 6H ON

SLIDE 6C ON SLIDE 6F ON

THESE TECHNOLOGIES OFF-.xAEAT POTENTIAL TO INCREASE THE

PERFORMANCE OF BOTH CIVIL AND MILITARY AIRCRAFT, SPACECRAFT,

AND NEW SATELLITES. THIS SLIDE DEPICTS THE MULTIPLE END USES

OF KEY TECHNOLOGIES AND THE FOLLOWING SLIDES SHOW EXAMPLES OF

THE BENEFITS IN NEW SYSTEMS. THE BENEFITS OF KEY TECHNOLOGIES

DEVELOPMENT WILL BE FAR REACHING AND ENCOMPASS MANY OTHER AREAS

OF CONCERN BESIDES AEROSPACE. THEY WILL HELP PROVIDE NEW

SOLUTIONS FOR ADVANCING CAPABILITIES IN AIR TRAFFIC CONTROL,

CHEMICAL AND WASTE DISPOSAL, ALL LEVELS OF EDUCATION, THE

DESIGN OF "SMART" AUTOMOBILES, MEDICAL DIAGNOSTICS, THE

DEVELOPMENT OF ADVANCED PROSTHETIC DEVICES, AND A VARIETY OF

OTHER ENDEAVORS BENEFITING BOTH THE PUBLIC AND PRIVATE SECTORS.

2000

9

"C11,W let,/x+

- 33% Low-*I Coal TheanForeign Comollp Ilion

a Low Mainlenancf@ - Lo9 ng *lt

SLIDE 6a a k4Maw,. CAOCAM and bu.sines Systems

* 10o Thrusl. lo.Welght Ratio Engine

a Electronics and Fiber Optlcs Replace

'- Cables and Hydraulics

* turbulenCe (Mirtobuftl) DelectiOr'

With Radar

a Pilot Associate (AI). Utrllabl.

Electrolics and Advanced Senisor"Provide Increased salely

* SmarI Skins for SvpriSCiesih

* Ulra..liaorVly Neve( fail6 Autoiomous operalton Willh PilbviAssociate (Al) SLIDE 6b

am Iproved Faull Tolronce TI-jough

Asconflguralbllty

a Inlegratled F and EQ Operations

10:1 Improvemnent

• Multleinsor Fuslon

* 21% Cool Reduciol Thrlough CompulerInlegraled luInalla Icdm)

a Sem.-Independenl Opreralitio Thgl,

Oft-)rd Robotic Ele•entl

a Advanced Launch and lranapohlallon,

S Aulonomou• Ground Sila@

S IDE 6c • Mullit•,,no Plalorm,

"C Low-Power and High. enrldy Eleclronics

* DOaS-n 10 ManulactI•.,rlg ivirM) Will

Enable 2!1.Year tilbespan

"* EIpfCl Coil RllducllonFrom 3 lv lO.t

10

Fail-Safe Air Traffic Control*Workload assistance from aircontroller .issoc~tie (Al)

Complete multiscrisor coveragec

Onboard threat and turbulence SL IIDE 6dstatusSystem would handle 10 timespresent capacity

4 parallel processing will enablelarge-scale traffic ComputationsWorldwide (±t 100 It) coverageusing satellites to provide locationinformation for all aircraft

Chemical Processing.4 ~ optrCreated products

designed for recycling orbiodegradability

Toxic waste conversion

waste managementimprovement

SLIDE6e *Automated process conltrol at

Imfproved oil exploration and

recovery techniques

p Health ipovements*Automated vital sign monitoring

and Symptom diagnostics

*Automated disease detection andcontrol

*improved quality of fife for thec

*Sophisticated prostheses SIE6

~SLIDE 6

SLIDE 6 CONT*D

SLIDE 6A SLIDE 6D SLIDE 6G

SLIDE 6B SLIDE 6E SLIDE 6H

SLIDE 6C SLIDE 6F

NOW LET ME TELL YOU HOW WE I"PLEMENTED OUR IDEAS. THE

TECHNICAL COUNCIL APPOINTED ONE OF ITS MEMBERS TO BE A SPONSOR

FOR EACH OF THESE TECHNOLOGIES. THIS SENIOR PERSON THEN PICKED

AN AUTHOR, GENERALLY FROM HIS OWN COMPANY, AND CHARGED THIS

PERSON TO BUILD A TEAM, COMPOSED OF APPROXIMATELY 1/3 EACH OF

INDUSTRY, GOVERNMENT, AND ACADEMIC SPECIALISTS. THESE PEOPLE

WERE CHARGED TL 3UILD A RGAMDAP FOR EACH OF THE TECHNOLOGIES.

AS THE ROADMAt, '`RE BFiNG FINISHEb. WE COULD SEE THE NEED FOR

MUCH MORE i)I *.,lEj) TECHNOLOGY DEVELOPMENT PLANS, IF WE WERE

GOING TO HAVE A CONSENSUS FOR A TRULY NATIONAL EFFORT.

12

SLIDE 6 CONT'D



SLIDE 6C SLIDE 6F

WE MUST BE FLEXIBLE WITH THESE TECHNOLOGY DEVELOPMENTS.

IF WE SEE THAT A PARTICULAR TECHNOLOGY IS GETTING A GREAT DEAL

OF CONCENTRATED ATTENTION, AS WAS THE CASE WITH VLSI AND

SEMATECH, THEN WE WILL DROP IT FROM THE LIST; BUT IF A NEW

TECHNOLOG.' APPEARS THAT COULD BENEFIT FROM THIS CONSENSUS

EMPHASIS, WE WILL ADD THAT TECHNOLOGY TO THE LIST.

SUPERCONDUCTIVITY AND COMPUTATIONAL SCIENCE WERE ADDED LAST

YEAR.

13

SLIDE 6 CONT'D



SLIDE 6C SLIDE 6F

THROUGHOUT THE EARLY PHASES OF DEVELOPING THE KEY

TECHNOLOGIES EFFORT. EMPHASIS WAS CONTINUALLY GIVEN TO THE NEED

FOR A CONSENSUS THAT INCLUDES STRONG INPUT FROM THE ACADEMIC

COMMUNITY. THIS IS THE AREA WHEREIN OUR REALLY BASIC, GENERIC

TECHNOLOGIES ARE SPAWNED. INTUITIVELY, WE FELT A HIGH PRIORITY

MUST BE GIVEN TO EDUCATION, EVEN AS LOW AS BETWEEN KINDERGARTEN

AND THE S1XfH GRADE, IF WE ARE TO HAVE THE ADEQUATELY TRAINED

AND POSITIVELY MOTIVATED PEOPLE WE NEED.

14

SLIDE 7 ON

THE LACK OF MOTIVATION TOWARD A CAREER IN SCIENCE AND

ENGINEERING IS PARTICULARLY NOTICEABLE IN THE UPPER SCHOLASTIC

RANKS. STATISTICS SHOW THE NUMBER OF UNITED STATES SCIENCE AND

ENGINEERING PHDS HAS DRASTICALLY DECLINED SINCE 1970. NOT ONLY

HAS THE TOTAL NUMBER OF PHDS DECREASED, BUT THE TOTAL NUMBER OF

U.S. CITIZENS WHO OBTAIN THE DEGREES HAS DROPPED EVEN MORE.

WHILE IT IS TRUE THAT ABOUT HALF OF THE FOREIGN DOCTORAL

CANDIDATES STAY IN THE U.S., ALMOST NONE OF THEM CAN BE HIRED

BY THE DEFENSE INDUSTRY.

$0200

tuvi • -- a fte,,k+ ,,.,,ro

Number .••

O 00L. -

0 I 0 I ! ,[[ I I I A I I

t o 47 6 . 66 6 6 7 0 7 2i 7 4 7 6 ' T 0 # 0 " $ 1 4 0" 8 8

15

SOME DEMOGRAPHICS ON THE FUTURE WORKFORCE PREDICT THAT

MORE lHAN 50% WILL BE MINORITIES AND WOMEN. THIS GROUP ALSO

MUST BE PROVIDED WITH SOME SORT OF EDUCATIONAL INCENTIVE THAT

WILL MOTIVATE THEM TO CONSIDER CAREERS IN SCIENCE, MATH, AND

ENGINEERING, IF THE U.S. IS GOING TO REMAIN WORLD COMPETITIVE

IN THE 21ST CENTURY.

PRESIDENT BUSH IN HIS STATE OF THE UNION ADDRESS SAID IT

THIS WAY: "EDUCATION IS THE ONE INVESTMENT THAT MEANS MORE FOR

OUR FUTURE BECAUSE IT MEANS THE MOST FOR OUR CHILDREN. REAL

IMPROVEMENT IN OUR SCHOOLS IS NOT SIMPLY A MATTER OF SPENDING

MORE. IT'S A MATTER OF ASKING MORE, EXPECTING MORE OF OUR

SCHOOLS, OUR TEACHERS. OF OUR KIDS, OF OUR PARENTS AND

OURSELVES."

16

MANY OF THE GOVERNMENT AGENCIES HAVE PREPARED LISTS OF

CRITICAL OR KEY TECHNOLOGIES. I AM NOT GOING TO TRY TO COMPARE

OUR LIST WITH ALL OF THEIRS. BUT I CAN ASSURE YOU THAI THE DOD

LIST OF CRITICAL TECHNOLOGIES, PRESENTED TO SENATOR BINGAMANIS

COMMITTEE LAST MARCH, MATCHES VERY WELL WITH OURS. THE

COMMONALITY BETWEEN LISTS LEADS US TO BELIEVE OUR APPROACH WILL

PROVIDE EACH GROUP WITH A SET OF ADOPTABLE DEVELOPMENT PLANS,

AS WELL AS THE FULL BACKING OF THE TROIKA THAT WE HAVE

ASSEMBLED TO DEVELOP A NATIONAL CONSENSUS.

17

SLIDE 8 ON

WE MUST DEVELOP THIS CONSENSUS BECAUSE [HERE IS NO

OVERARCHING PLAN AND NO NATIONAl. PLAN TO ACCOMPLISH NATIONAL

GOALS SUCH AS IHOSE OF THE NAIIONAL SPACE COUNCIL OR THE

PRESIDENT'S SPACE EXPLORATION INITIATIVE AT THIS TIME.

2,00.... (),. .,

manVurn mu

SLIDE 9 ON

WE MUST TAKE A BOLD, NEW, APPROACH AND COMBINE ALL OF THE

ACTIVITIES SHOWN HERE- TOTAL QUALITY MANAGEMENT, CONCURRENT

ENGINEERING, ETC. IF WE CAN COMBINE ALL OF THESE WITH OUR

CONSENSUS APPROACH, WE FEEL CERTAIN THAT OUR INTERNATIONAL

COMPETITIVENESS CAN BE VERY SUBSTANTIALLY IMPROVED.

MyTohalW W 0 4taQ' 2000Combint-d Thrusts To Improve U.S. Competitiveness

Nation-Wide Thrust

19

SLIDE 10 ON

I'VE BEEN DESCRIBING HOW OUR TECHNOLOGY TEAMS WERE FORMED,

A BOTTOM UP VIEW, BUT WE ALSO FELT THAT WE NEEDED TO LOOK AT

POLICY QUESTIONS AT THE SAME TIME. FOR A LONG-TERM COOPERATIVE

EFFORT WE NEED BOTH A BOTTOM-UP AND A TOP-DOWN VIEW.

200(

Alb"

20

SLIDE 11 ON

TO PROVIDE THE TOP-DOWN VIEW OF THIS CONSENSUS EFFORT OF

INDUSTRY. GOVERNMENT, AND ACADEMIA, WE HAVE ESTABLISHED THE

AEROSPACE TECHNOLOGY POLICY FORUM. THIS BODY IS MADE UP OF

HIGH-LEVEL INDIVIDUALS FROM EACH OF THE THREE SECTORS.

21000)

* Oversees the Key Technology Process* Identifies Impediments, Policy Needs* Facilitates Cooperative Eftorts

21

SLIDE 12 ON

THE GOVERNMENT REPRESENTATIVES, AS YOU CAN SEE HERE, ARE

AT THE VERY TOP OF THE POLICY-MAKING DEPARTMENTS OF MANY OF THE

AGENCIES DIRECTLY CONCERNED WITH OUR INTERNATIONAL

COMPETITIVENESS. THE PRESENCE OF THE OSIP DIRECTOR ON THE

POLICY FORUM PROVIDES AN EXAMPLE OF JUST HOW INFLUENTIAL THIS

GROUP CAN BECOME. THIS GROUP MET TWICE IN 1988 AND, AFTER A

CHANGE OF ADMINISTRATION, ONCE IN 1989 AND ONCE SO FAR IN

1990. WE INTEND TO CONTINUE WITH 2 OR 3 MEETINGS PER YEAR.

THE AEROSPACE TECHNOLOGY POLICY FORUM WILL GIVE OVERSIGHT

TO THE PROGRAM, AS WELL AS IDENTIFY AND DEAL WITH ANY

IMPEDIMENTS 10 PROGRESS. ATTENTION WILL BE PAID TO POLICY

ISSUES SUCH AS IR&D, INCENTIVES FOR R&D INVESTMENT, THE

ADEQUACY OF THE NATION'S TECHNOLOGY BASE, ETC.

Government 2 0

I 1he President's Science Advisor

* Associate Administrator. A&ST, (NASA)

* DirtE:lor, DARPA, OSD

* Dirtrctor, National Science Foundation

* Depuly Director, Olfice of Energy Research, DOE

* Director, Defense Research and Engineering, OSD

SASSIId(ll Secttary of the Air rorcc (A)

* Ass•,itant Secrelary of the Navy (R, E&S)

* Assistant Secretary of the Army (R, D&A)

* Und.r Secretary of Commerce. Technology

22

SLIDE 13 ON

THE UNIVERSITY MEMBERS OF IHE FORUM ARE FROM VERY

DISTINGUISHED SCHOOLS IN DIFFERENT AREAS OF THE COUNTRY. WE

STRONGLY VALUE THEIR INPUT BOTH BECAUSE THEY REPRESENT THAT

INDIVIDUAL. INVESTIGATOR, WHO IS SO VERY IMPORTANT TO AMERICAN

TECHNOLOGICAL SUCCESS, AND BECAUSE THEY CAN GUIDE US IN OUR

QUEST FOR GREATLY INCREASING THE NUMBER OF STUDENTS ENTERING

COLLEGE WITH ENGINEERING AND SCIENCES AS THEIR GOAL IN LIFE.

THEY CAN ALSO TELL US HOW TO STIMULATE THE BEST AND MOST

WILLING OF THOSE STUDENTS TO GO ON AND GET ADVANCED DEGREES.

Universities

"* Chairman, Aeronautics and Astronautics, M.I.T."* Vice Chairman and Provost. University of Michigan

"* Dean. School of Engineering, Stanford University

"* Deputy (hancetor and Dean of Engineering. Texas A&MUniveisity

"* Execuhive Direclor (f W/ashington Technology Center,Univeriry of Washinglon at Se,'lllt

23

SLIDE 14 ON

THE INDUSTRY MEMBERS ARE THE PRESIDENT OF THE AIA, MR. DON

FUQUA; MYSELF AS PRESIDENT OF THE NATIONAL CENTER FOR ADVANCED

TECHNOLOGIES; THE EXECUTIVE DIRECTOR OF NCAT, AND THREE OTHER

DISTINGUISHED MEMBERS FROM INDUSTRY, MR. FELIX FENTER,

PRESIDENT OF THE MISSILES DIVISION. LIV CORPORATION, AND

CH;AIRMAN OF AIA'S TECHNICAL AND OPERATIONS COUNCIL; DR. J. R.

BURNETT, rE"CUTIVE VICE PRESIDENT OF TRW AND CHAIRMAN OF THE

KEY TECHNOLOGIES COMMITTEE, AND MR. NOEL LONGUEMARE, SENIOR

VICE PRESIDENT FROM WESTINbIOUSE AND VICE-CHAIRMAN OF THE KEY

TECHNOLOGIES COMMITTEE.

Industry

* President. AIA. Chairman

* President, NCAT, Vice Chairman

* ExeCutive Director. NCAT

* Chairmen. Technical and Operations Council (Pros. Missiles Div, LTV)"* Chairman. Key Technologies Committee

(Ex VP and Dep. Gen. Mgt. Space and Defense, TRW)

"* Vice Chairman, Key Technologies Committee(VP Systets Development and Technology Diva, Westinghouse)

24

SLIDE 15 ON

I'VE MENTIONED NCAT AS A NON-PROFIT FOUNDATION, BUT THIS

SLIDE SHOWS THAT WE ARE INVOLVED IN INTEGRATING AND

COORDINATING THE ENTIRE KEY TECHNOLOGIES EFFORT.

A liw . e i•Vat• .........

25

SLIDE 16 ON

IN ORDER TO REACH OUR GOALS BY THE YEAR 2000, WE HAVE

DEVELOPED ROADMAPS FOR EACH OF THE KEY TECHNOLOGIES. THESE

ROADMAPS ARE A FIRST CUT AT WHAT THESE TECHNOLOGIES PORTEND.

THESE ROADMAPS AND PLANS ARE LAUDABLE AND VERY NECESSARY,

BUT IN MY OPINION THEY WILL NOT BE SUFFICIENT. I BELIEVE WE

ALSO NEED TECHNOLOGY VALIDATION DEMONSTRATORS. THESE WOULD BE

THE BRIDGE FROM CONCEPT TO APPLICATION, A PROOF-OF-CONCEPT

INTEGRATION STEP THAT WOULD DEMONSTRATE RISK-ACCEPTABLE

READINESS FOR USE ON NEW PRODUCTS.

THE GOAL OF THE KEY TECHNOLOGIES FOR THE YEAR 2000 PROGRAM

IS TO HAVE THE INDIVIDUAL TECHNOLOGIES AVAILABLE FOR

APPLICATION BY THE YEAR 2000, BUT SOME ARE BOUND TO MATURE MORE

QUICKLY THAN OTHERS. WORKED INDIVIDUALLY, THESE TECHNOLOGIES

WILL NOT BE INTEGRATED INTO A MEANINGFUL WHOLE.

*Ow Poo^ 2000

26

SLIDE 17 ON

A TECHNOLOGY DEMONSTRATION PROGRAM, AS VIEWED BY AIA AND

NCAT, WILL INTEGRATE THOSE INDIVIDUAL TECHNOLOGIES AND GIVE US

AN OPPORTUNITY TO LOOK AT THE PROCESS REQUIREMENTS AND THE

INTERFACE DEFINITION. WE WILL BE ABLE TO SHOW THAT WE CAN

PRODUCE TECHNOLOGY THAT IS RELATIVELY RISK-FREE AND THAI WE ARE

READY TO MAKE A WORLD-COMPETITIVE PRODUCT.

THIS GIVES A QUICK OVERVIEW OF OUR GOALS AND

IMPLEMENTATION PLANS. WE SEE A SERIES OF SYMPOSIA GEARED TO

EACH TECHNOLOGY OR A GROUPING OF TECHNOLOGIES. WE WANT TO

DEVELOP NATIONAL PLANS SO THAT WE CAN MOVE FORWARD TOGETHER BY

WORKING COOPERATIVELY TOWARD COMMON CONSENSUS GOALS.

2000

T**w- '0a",

"p."n1 IT .deeaemie2

27

SLIDE 18 ON

LOOKING AT CAPABILITIES FOR NEW SPACE SYSTEMS AND THOSE

WHICH SURELY WILL BE NEEDED FOR THE SPACE EXPLORATION

INITIATIVE, YOU CAN SEE HOW OUR KEY TECHNOLOGIES ARE IMPORTANT

IN IMPROVING PERFORMANCE AND RELIABILITY AND REDUCING COST AND

WEIGHT.

THIS SLIDE SHOWS HOW OUR KEY TECHNOLOGIES PROVIDE

SYNERGISM. I USED 1HE SPACE EXPLORATION INITIATIVE TO SHOW

THIS SYNERGISM.

Al Lead to Lowor Cost for Payload Delivered. HigherRellabltitly and Shorteed D)evelopment Time

28

SLIDE 19 ON

THE 10 KEY TECHNOLOGIES THAT WE HAVE IDENTIFIED AS

CRITICAL TO OUR FUTURE ARE AT VARIOUS STAGES OF DEVELOPMENT.

SUPERCONDUCTIVITY HAS JUST REACHED THE STAGE WHERE IT IS

ADVANTAGEOUS TO GET A CONCENTRATED PROGRAM GOING. HOWEVER,

ROCKET PROPULSION TECHNOLOGY, WHICH WAS ONCE THE ALMOST

EXCLUSIVE PURVIEW OF THE U.S., NOW NEEDS A REDOUBLING OF OUR

EFFORTS IF WE ARE GOING TO KEEP A SIGNIFICANT PORTION OF

CIVILIAN SATELLITE LAUNCHES.

WE WILL LOOK AT THE CRITICAL PATHS FOR ROCKET TECHNOLOGY,

VALIDATED FULL SCALE ENGINEERING MODELS, EXPANDED DATA BASES,

IMPROVED MATERIALS, COMPUTER AIDED DESIGN. MANUFACTURING, AND

INSPECTION. THIS WILL LEAD US TO INCREASED RELIABILITY,

SAFETY, AND GREATLY REDUCED COST. WE ARE CONFIDENT THAT WE CAN

REDUCE THE COST OF PAYLOAD TO ORBIT, FOR EXAMPLE, BY AT LEAST A

FACTOR OF THREE AND SOME ESTIMATES ARE A FACTOR OF TEN

IMPROVEMENT OVER WHERE WE ARE TODAY.

ShS~YWNW% 2000Ready

forProduct

All Breathing PropUflonRockel PropAts

UlMIReliable Electronic Systems mpolilAdvantced 9h Arlificial IntelligenceOptical Informotlln Sollware Developinent

• lllc @""Compultltonal Science

supqconduCI~v,,y 29

SLIDE 20 ON

IT IS THIS KIND OF NATIONAL CONSENSUS ON EACH OF THE KEY

TECHNOLOGIES DEVELOPMENT PLANS THAT WILL KEEP US IN THE LEAD OF

THE WORLD'S AEROSPACE FIRMS AND KEEP THE UNITED STATES BALANCE

OF TRADE POSITIVE FOR THIS SECTOR OF OUR ECONOMY. WE WANT TO

HAVE A PLAN THAT GOVERNMENT, INDUSTRY AND UNIVERSITIES CAN

SUPPORT AND THEY CAN ALSO USE THE PLAN FOR BUDGET

JUSTIFICATION. WE WANT A NATIONAL CONSENSUS ON A FEW KEY

TECHNOLOGIES. WE FEEL THAT WE CAN ASSESS THE PROGRESS AT

SYMPOSIA AND ALSO WORK AT FILLING TECHNOLOGY GAPS AND REMOVING

INHIBITORS IF WE HAVE THE NATIONAL CONSENSUS PROCESS WORKING.

Kof I QT%QA~qgk4 tQt th~ VO Z%Implementation

Technology- Symposia

Government --- National I

ITechnology AssessmentI T* Plans

Industry - Development * Gaps

pion * Inhibitors

Iplan lhbtO~

Universities • Revisions &Updates

A National Consensus

30

SLIDE 21 ON

IT WAS AND IS OUR FIRM CONVICTION THAT OUR FUTURE

COMPETITIVENESS DEPENDS UPON OUR TECHNOLOGICAL LEADERSHIP.

WHAT AMERICA NEEDS IS A NATIONAL TECHNOLOGY STRATEGY AND THAT

IS EXACTLY WHAT THE KEY TECHNOLOGIES FOR THE YEAR 2000 PROGRAM

IS DEDICATED TO PROVIDING. WE AS A NATION MUST HAVE AN

INTEGRATED STRATEGY (ALMOST A PARTNERSHIP) THAT CONSIDERS ALL

NEEDS, MILITARY & CIVIL, AND ENCOURAGES A CONSENSUS OF PRIVATE,

PUBLIC, AND ACADEMIC LEADERSHIP.

Strength.

Security_

31

NOZZLE VECTOR ANGLE DETERMINATION USING A LASERMEASUREMENT SYSTEM*

W. A. Ma189ico. J. F. Seey, U. R. J. TorIck, Jr.. USAFAWWrAutlcs Laboratory Svsrh'Tshoo, TInc. mlEnierg

Edwr-A ir FrceSau3 CAAEMDevelopment CenterArnold Air F-aroe Base, TN

ABSTRACT

This PR addreseas, the mseasurvwnent of n oz le position during solld-propellarit rocket motor (SAM grounc.testing. The two prinicipsif measuremenit systems currently used to determi~ne nozzle position are a system of lonea poten.-tiometers11, ard oul. acta~redma systelams ) husetwo systems are mounte on fte motor and provide output whlc~i08a AXWOtlo offthei dlept4 ýSO1 These measurement systerneiare, however, sub*~Cted to mnotor-and nozzie-ptlullar'"Of(Ic caused by motor pressurization. Unsar potentiometer calculations attempt to account for these mfor and rx .zlemov~omrnen throughi appropriate data reduiclion schemes, actuator feedback systems do not. A lawe-based techniquawas doveeQpe with the Intent of providing a nonlritruslve measurementi system Independent of the test article, Y.' fichwould be capable of measuring the nozzle position with rer4,ect to the motor centaine regardless of motor- and no-zie-pecullair NInvils. The feasibility of the system was demovnstrated on rocket motor ground tests at both sea-level 3nd84a IM lA180tet1fciltie at ti " AstronaUtics Laboratory at Edwards Air Force Base. California and the Arnold Engineering De-vslopmrnen Center In Tennessee. respectively. The system was demu. ibxated at amrblent,. Cold, and hot test temperatures.The* nozzle material tarete was curbon/catbon composite. Excellent results have bpen obtained to date (the specifictase system hardware used for thii demornstration was designed, however, for a different application, Pnd Its capabilitieswore found kicking In tVe V RM environment). The results were evaluated against Jata acquired by the current measure-men!t systems Based upon this evaluation, It Is concluded that this technique Is feasible, and provides a level of accuracyon the orde rYo the currently recognized starxda'd measurement system (potentiometer- based) while maintaining theadvantages of a nonlntrusive, ground-based system.

NOMENCLATURE

AVrDC Arnold Englnwling Development C'n-"erAL Astionautlos LaboratoryCPU Central Processing Ui~ntO&AAI Gallium ArsenideIR InrgredILPMS Lawe Position Measurermen SystemLVDT Linear Variable Differential TransformerSAM Solid -Propellan t Rocket Motor

INTRODUCTION

An appoach to measurement of true nozzle position with respect to tho motor centerline was dernonstrat'jd overthe yeam 1965 -1969. This aipproach, unlike previous methods, was nonintrusivo and ground-base, and offered the ad-vantage of improved InW*lnsivlty to test article Interactions. This paper presents the results of the efforts at the AstronauticesLaborate-y (AL), Edwards AFB, CA, and at the Arnold Engineering Development Center (AEDC), Arnold AFB, TN. Devel-opmental work, data reduction codes, and tost applik-sone are discussed. The syttern was evaluated agairst currentatauiowd measur~emenittechniques, and was found to bo feasble and accurate, while providing the advas tages associated

lvth Its nonintrueive, ground-baised approach.

*'Te re~okoch reported herein was performed by the Arnold Engineering Development Center (AEDC),Air ForceSystome Command. Work "n analysis fur this research were done by Air Force personnel, personnel of Sverdrup Technot.ogy, Inc., AEL)C Group, operating contractor for the AErlO propulsion test facilities and personnel of tOe AstronauticsLabor siory, Edwaerds Air Force Base, C/.Aipsoveci for pubic release; Dlstrihut~in~ Is uni1mited.

33

SYSTEM DESCRIPTION

The Iae Position Measureerril Syslem VWPM) was cSTprI5se of two probes and a Cenitral Procssing Unit(CPU). A schemafti stiwwng fte measurment princploe is preene in Fig. 1. The proe were of off -teseftechnolo-gy vintage. oonsisting of bolh a galium~ arsende (GaAs) lInv didsourc opral:in;Ug In ft re n-wavelengt infrared OiR)spectrum, and a daectso The probes used over nwarxfcld by Seloom, probe type =35- 555. enxcled In l1qui cool-Inig )wc*ets, having a 5-In. meeamsemart range and a 28-in. skandoff distance from the ie surface. The GaAs tamedioes were operated In a pulsed mode, rmodulated at 16 kHz whth an average power of 10 mW. The emitted IR figh beam,directed onto the suulacto of the rmnozle. was, rlctdin a dlffinue nw we and focused on the detector at a spot toe subs*-querit detector outpu was related to fth displacement of the ,oze The IMP In wer seni -condulaor, analog IWisepoe~o-sereietive devrioss capable of disoerning" mtcerleced IR spo iWages from bedcground IR leols The9CPU pro-vided for slWia processing and data output; an anaelog cuipxi sigrui and valid/Invald measuaremni stabtusIndicator wereavailable for each probe.

SiGNALCONDITIONING TO DATA

WEORDING

PROBE (PUSYSTEM

LASER00SIODRE ~,POSITION -SENSITIVE

-_ NOMZE SURfACE

FIg1 LPUS Measuremnent Pkvhcpl

SYSTEM APPLICATION

Thao rr-aaswement wnormpt was first proven feaible by application to rocket motors at thie AL Edwards AFB, CA.Following demorestratlon of the LPMS concept at these sea-lervel conditions, fth probes were usw AEDC In a typicalaltitude rocket propulsion developmrrent test cell. The test environment and associated ccnfiguri v LPMS for eachtest are piresented In Table 1. A povtograph shiowing a typical test installation is presented in Fi_

34

DFIEiOR YAW PROBE

OUR(E Dif FUSER

PITCH PROBE 3-/OZZIFi. 2 T LPMS I-tll In an M Rocket Test Cll

Table 1. Tedt Conl ato

Test No. yp Meue- Avg. N &

m&ne MIR 3 Probe*f FrSomIj 1e Finter Board Reducon. Palume

&ID. PA we A _ O•F

1 LVDT. POT P 5 Parallel 9 no 500 125 -0- 2600

(72) LPMS Y 5 9 no 500 125 -0- 3500

2 LVDT, POT P 5 Perpes. 2 no 500 125 -0- 3000

(110) LPMS Y 5 " no 500 125 -0- 3500

3 LVOT P 10 Perpend. 3 yes 500 125 -0- 3000

(45) I.PMS Y 10 " yes 500 125 -0- 3500-~~ m -- m -

4 LVDT P 5 Perpenl. 3 yes 500 500 0 & 30 3000

(110) LPMS Y 10 yes 500 125 0 & 30 3500

5 LVDT, POT P 5 Perpend. 2 yes ND 4 = 0. 500 30 3000

(45) LPMS Y 5 " yes 500 500 30 3500

6 LVDT P 5 Perpend. 2 yes 500 500 30 3000

(110) LPMS Y 5 " yes 50 500 30 3500- -

7 LVDT P 5 Perpend. 3 yes 500 500 0&30 3000

(45) LPMS Y 5 " yes 50 500 0&30 35001) LVDT - Actuator Feeclback, POT - Unear Potentiometer, LPMS - Laser System2) P - Pitch Plane, Y - Yaw Plane3) M.R. - Probe Measurement Range4) ND - Neutral Density Filter

35

II

OPERATIONAL VERIFICATION AT ALTITUDE CONDITIONS

Prepraton for the LPMS application at simulated altitude co~nditions Incluided characterizatlon of system per-11orrnance in a laboratory environment. Concerns over probe operatin In a low-pressue environment (resulting frompreensiuelmation ofalftudeconditiors) were adckeesed. The probsewere insalled In the acaajmcharnber. anid a seriesof taeft were runi to characterize any changes in operation asa funmction of amblent pressue. Repeatability In system Cali-bration and the operabilty of tho systhm was verifie. No sigriflant effects of preesure on probe operation or data outputwere noted.

Additional h~bmation oblained from fth laboraory lleskc ., Iwrant from an installation viewpoint Indiae thatthe probe could be installedas much as40 degfrom norrmal io tetarget andwsill got adeqat sinl~. However, the dataoutput was affected as the argot sMope change. lb maintain calibration fidelity. a change In targot orlertation of < 5 degIncidence was reqired.

TEST INSTALLATION

Two Installations were used during the altitds e sequence. The firet Instalation, shown In Fig. 3a, was usedonly for the firet application. Subsequent toesing was performed In the test configuration shown In Fig. 3b. The changes InInstallation hardware and orientation were dictated by changes In the SRMs.

Active c, 1ing and gaseous nitrogen (GN2) lines were required. Cooling was provioded via a liquid-jackete boxwround each probe for protecio from the harsh SRM environment and to keep tie electronics within operational tempera-tue limits. GN2 lines were used for external cleanlineai of the probe lenses dui~nng the moWo firing.

DIFFUSER-

YAW PROBE SUPPORT STRUCTUREAFT ATTACH RING

TEST CELL FLOOR - /77/77 OUENCH PROBE

a. Init&IlFig. 3 Schematic of Test Coll Installation

36

DIFFUSERENTRANCE PLANE

AFT FIRINGROCKET MOTOR RINGL______________ YAw PRoBE

YAW.. - , YAW PROBE

SUPPORT STRUCTURE

PITCH PROBE

" ,-PITCH PROBE

GROUND TEST CELL FLOOR- SUPPORT STRUCTURE

b. TypicalFRg. 3 Schematlc of TestCell Insitallixtion

Vibration of the fmout was of concen. not so ruch from a probe operational viewpoint, but because any vibrationof fte probe wouId be reflOcted in t data set. Therefore. probe vibration Ivels were measured for te eM confguratindescribed above. Vibrations measured were well within desired levels (< 2 mils peak/peak). This Information Is Incl Inte measurement uncertainty analysis discussed ie.

TEST INSTRUMENTATION

Measurements were made of nozle displacement, probe vibration, probe temperatue, and probe valid/invalidirndicators. The probes were calibrated in a laboratory prior to Installation in the test cell. Calibration pointswere taken every1 mm, and were verifled po:ttest.

SYSTEM UPGRADES DURING TESTING

Measurement in ft SRM environment and with the accuracie required for definition of nozzle movementsproved challenging In two areas: (1) controlling the effect of the background radiation, and (2) reducing transient data scat-:er. The effects of these two deficiencies are shown in Fig. 4.

37

6 i- EXCESSIVE BACKGROUND RADIATION

,.,,_ " EFFECT 72 (LOSS OF DATA

.. ... , 0-f=z -2

"- -4 _-

-6 1r

6 DATA SCATTER T- - DATA SCATTER4 (PRE-IGNITION LEVEL) (DURING TEST)-

=LLJ

_ 4. -4 -ý-BAC KGROUND RADIATION EFFECT

-6 _10 0 10 20 30 40 50 60

e2

-2 ---- j~K---- -- I I

-6- I-LTIME, SEC

Fig. 4 Effect of Probe Umitations on LPMS OutputThese probes were not designed to operate in the SRM environment, and nodifications were required to in-

crease the level of backgruund IR radiation that the probes could withstand. The estimated maximum nozzle surfacetemperature targeted In this test sequence was on the order of 3,000 *F, with a corresponding plume temperature of 3,500'F However, the probes wore desgne4d for a maximum temperature of approximately 2,800 F. A combination of radiationshields and optical filters was used to reduce the background radiation to an acceptable level. Note the resultant Improve-ment seen from test 4 to test 7 (Fig. 4). Test number 7, however, still Indicated an area of time from 16 to 25 sec when the datawere again Invalid. Further Improvemerts and understanding are required before a high data acquisition succews rate canbe routinely achieved on all rocket motor tests

The LPMS transient data scatter was on the order of 100 mils peak/peak (5-In. probes). Desired levels were < 10mils peak/peak. The transient data scatter was reduced during the test sequence to 20 mils peak/peak (steady-state reso-lution was adequate at 2 mils). Control over transient data scatter was achieved by varying gain settings and digitalaveraging board rates itRhin the CPU. The scatter In the test data typically increased from the levels quoted above as afunction of time throughout the motor bum (Fig. 4). It Is unclear at this time whether the scatter was. in fact, measuremensystem-related (potentially caused by Increased detector saturation from background radiation levels), or nozzle vibration.

,38 I I II I I I II

DATA REDUCTION TECHNIQUE

A data redu~ction code was developed to translate the linear displacenment data output from the LPMS system Intonozzle vector angle. The basic problem was that of describing a 3 -di mensional surface using Information that was derivedfrom changing locaions on that surface. See Fi"~ 5.

NONVECTORED POSITION /

VECTORED POSITI N j

EWYAW TARGET

NEW PITCH TARGET

Fig. 5 Dafta Roduction Toctvdqu

The approach used was to mathematically describe the nozzle suirface In the region of interest, and to mathemet-Icaly escibeth reatinshp btwen he wo erbeam displacement output as the nozzle was vectored. Using

information obtained pretest. an accounting was made for forward and aft trarsW~ajn of fth nozzle surface by relating thismovemrrnr to motor chamber pressure (both nozzle moviement and dtvus stand movements were accounted for); addition-ally, the pivot point was allowed to move relative to the ncizzle surface as a function of vector angle.

RESULTS AND DISCUSSION

The teat series and corresponding probe hardware oorfigitioneoas conducted at AEDC are tabulated In Table 1.The system was demnonstrated at nominal (72 -F), cold (45 *F), and hot (110 *F) environmental test temperatures. Thenozze muatrial targeted was carborVcarbton copoefle. It was of Interest to compare our results with those from the current-ly accepted measurement systems.

COMPARISON OF RESULTS TO CURRENT MEASUREMENT SYSTEMS

Measurementi systems used for determination of nozzle position consist primarily of linear potentiometers andactuator feedback Unear Variable Diffe rential Transformer (LVDT) systems. Figure 6a presents a schematic of these sys-tems Installed on a rocket motor. These systemnsare mounted on the motor and provide outputs which are a functionaof their

39

dlpapeerfnes, bu te sib tedIo bt moth. and nozzle moas (Fig. 6b). The potentionvier-bd systems ard toac~ount for t•ee movemera; acMutor ftee•k systems do rn. Fig.?7 pvw ty1ical outptA data from each measuremeno system. Errof of up to 1.5 deg r noem d In the acumtor tfedec* LVDT syseom output as presented In Fig. 8. Becaise

of the hrtocomingsod the LVOT system, we will focus our discussion on a comparison tothe potentlometer-beed system.

C.ASE INITIAL

kOY(MENT I k PRESSURIZED•.\ FIEESEAL /

(ONDITION / .),ACTUATOR NULL \ IHITIAL

FEBC L i / ' of FSET \NOZZLEI / ,\POSITIOlN

-(8) LINEARPOTENTIOMETERS FINALNOT7L2 1

POSITION

a. Installed on SAM b. Motor Pressurizatoon Effect

Fig. 6 Typical Measurement Systems

LVDT

4TIME OF MAXIMUM.• 2 . ... PRESSURE EFFECT-

0

-4POTENTIOMETER

tz -2 -

-4 -

LPMS4-

TIME, SEC

Fig. 7 Measurement Systems Output

40

2

LpJ

~S~bI4.~DIFFERENCE

-10 0 10 20 30 40 50 60 10 80TIME, SEC

F.8 CorMarison of LVDT Output to Recognized Standard

POTENTIOMETER-BASED MEASUREMENT COMPARISONS

The most interesting comparison Is with the currently accepted standard for accurate, absoute nozzle positionmeasurerment: the system of placing linear potentlometersi in a circumferential mannrv about the nozzle (Fig. 6a). This sys-tem, as with the LPMS, Is intended to measure nozzle position regardless of Cae pressurization or other effects. Acomparison of potentiomet~er arid LPM.S output Is shoown In Fig. 9. Two aea In particular stand out: 11) the Initial choice ofwhere to define fte nozzle zero position, and (2) motor presvurizatIon effects.

M 1 INITIAL NOZZLE POSITION DIFFERENCESMTO____ ___ -. BURNOUT-

MOOR PRESSURIZATION EFFE(T-10 0 10 20 30 40 50 60 10 80

TIME, SEC

Fi9. 9 LPMS Comparlson to Potentlometer-Based System

The first area Identified above, that of choosing Initial zero cositlon, results esserilially In a bias between fte twosets of data. The potentiometer- based system chooses zero based upon measurements first made at the manufacturer,and retselling the actuator stroke prior to test to obtainfth Identical outpu. The choice of zero vector for the LPMS was madeby maxksuring nozzle position In the test Cell with an Inclinometer, referenced to motor hardware. Dlifferences will result as afunction of (a) the ability to make these measurements and (b) the accuracy of the different mefthdologies. This issue Isgeneally resolved by simply chooing a common zero, motor specifications on the *zero" tolerance having been ad-dressed via manufacturing tolerance allowables.

The second area identified above Involves the variances seen In calculated nozzle position resulting from motorpressurization. Figure 10 represents the differences between the LPMIVS and potentiometer- based systems, but using acommon Initial nozzle position. Tho differences seen are on the order of 0.2 - 0,3 degs. Figure 1 la shows the output from

41

the two rneesmmernt systemse asa function of motor clanter prow"ia. Theme data were obtained with the nozzle cokm-mended to fte null position (vecWo angle - 0). Note the chosen nozzle otstposition of 1.5 dog for the unpresasuftedcondition. This uriprma'utzed position Is chose such that the average pruessurized position Is nooiniaily 0Odeg. The curvespresented were clearly ea',ed 11onmotor chamber pronw".

The diffurences seen between twos two sefs of dita wre aftrlbuted to limitations in both measurement systems. Asnoted previously In the LPMAS calculation prcogam descripton. the aft movemenitof thenozzlecaUsed by mKoto preefwlra-tion is accouted for In the LPMS progrm by a linear relationship with motor chamber pressurization. This relatioinship Isprovided by the marufacturer, and is based upon motor benchtests. ifiNs reationship werenerrorby5O%. an erorof 0.1deg would be propagated throuigh the LPMS outpLA (the maximum differences seen in Fig. 1 Ia. on the order of 0.3 dft,equate to errors on the order of 1/2 in. 150% over and above the p~redicted 1/3-ti. aft movement). Pemaining differencesmay be due In ped tocother motor chamber pressure-Irnduced effects, possibly a shift In pivot point aWe characteristics ofthe nozzle aseerrbly change. Potential errosources for thee. potentiometer-bae systems soud Include movements offte poternkxmete grounld point relative to fte nozzle attach point during motor presuruization-lndcuced case and nozzle

shifts, and assumptions In fth pivot point location for those conditionis where they cannot be calculated.

-NOZZLE POSITION

LLJ~ 0

21 -10 0 10 20 30 40 50 60 10 80TIME, SEC

Fig. 10 LPMS Comparison to Potentiometer-Based System,Comnmon Initial Nozzle Position

In Fig. 11 lb the differences between these two sets of data are shown as a function of vector angle. The differen~esat null vector angle and at fth total 3-dog vector step are of roughly tie samne order of magnitude, Indicating small vari-ances because of step size. As disicussed above, the bulk of the differences between the measurement systems occur asthe motor Is pressurized. Removal of the accumulated chamber pressure- Induced differences prior to a vector step, thecauses of which were discussed above, highlights the remaining differences of nominally 0.06 dog (with one cutl~er of 0. 18dog). This is done by looking at nozzle deflectionis relative to each system's null position. These remaining differences arelikely related to residual pressauslzatbon-induiced conditions such as a shifted pivot point location, In addition to differencesin accounting for pivot point shifts due to nozzle vectoring. The one cutIle of 0.18 dlog highlights the difficulty of discerningthese extremnely small movements. it is recommended that a validation of both systems. be performed simultaneously on a'callbrated nozzle to resolve the remaining minute but significant differences.

42

310 111.0ALFMS DATA t:0 PITCH DIFFERENCEaPOTINIIOMET11RB1ASID DATA 0. 0 YAW DIFFERENCE

S2 0 DIFFlREN(I IETWFEN THEMTWO MEASUREMENT SYSTEMS 0.6

DEFLECTION ONLYT4 04 REMOVAL OF

.0NULL TOTAL VECTOR NULL DIFFERENCESCHOSEIN UNPRES5.UNHZE 102 .16

rNGURJ POSITION (I.SOEG) .1j 0 .L612B DO0 0 FIrn I LO~

0 20 40 60 g0 100 120 0 +3 -3 +3 -3NORMALIZED CHAMBER PRESSURE (OMMANDED VECTOR, DEG

a. Motor Pressure Effect b. Vector Slop Size EffectFig. 11 Parametric Evaluation of Difforencos Between LPMS and Potentiometer Systems

MEASUREMENT UNCERTAINTY

An uncertainty analysis was undertaken to quantify the LPMS measurement. The approach used included esti-mating both precision and bias ITeMe of measurement uncertainty for all potential error sources. providing an estimatewithin which the true value of the data will lie. The estimated uncertainty was [Bias + t95(Precislori)j = [0.025 + 2(0.013)]= 0.05 In. =0.11 dog (for this test configuration). A representative uncertainty estimate for"th potentiomreter system wasniot available for reporting.

PROPOSED UPGRADE

The results of the LPMS application were very encouraging. The system as tested, however, utilized only two inde-pendentd meaa&,. sments. A proposed upgraded system Is shown schematically in Fig. 12. To calculate nozzle positionwithout relying upon pretest prediction (for pivot point movements and motor pressurization effects), a total of four probesare needed If an assumption of pivot point movement alon the nozzle centerline Is appropriate; otherwise, a total of sixprobes wouid be needed to allow for p"o point movements off of the nozzle centerline.

43

LPMS PROBE IMPACTTECHNIQUE PROBES ORIENTATION

BASELINE 2 PERPENDICULAR RELIES ON PRETESTESTIMATES

NEW 4 - 6 ANY TOTALLY INDEPENDENT

MEASUREMENT

Flg. 12 Future LPMS System

The currently used LPMS hardware was not designed for the SAtM environment. Any upgrades should considerImprovements to the hardware. The probes should be designed to withstand the radiation levels associated with SRMplumes, and the system should produce data thathave transientdata scatter < 10mils peak/peak maximum.Anadditionaltbenfit of improved data scatter would be the possibility of measuring vibrational levels of the nozzle nonintusively.

SUMMARY AND CONCLUSIONS

In sumnary, we have described a system that provides a noninrusive and ground-based measurement of nozzlevectoring. Developmental work, data reduction codes, and tes applications were discussed. The system was evaluatedagainst current standard measurement techniques, and was found to be feasible and accurate. The potential of the LPMSto measure nozzle vlbration was dascussd.

Comparisons with current measurement systems indicated that (1) as expected, actuator feedback measure-ments did not provide precise nozzle position information, and (2) agreements with potentiometer- based systems werewithin 02-0.3 deg; this variance occurred primarily as a function of motor chamber pressr:e, not vector angle, and wasbelieved to be attributable to the two systems' Inability to aoount for motor pressurization and pivot point movement ef-fects.

Additioral work is required before the LPMS can achieve Its full potential. The hardware should be specificallydesigned for the SRM test environment, and a system of 4 to 6 probes should be demonstrated to Improve accounting ofmotor preasurization and pivot point movement ees.

44

EVOLUTION OF AN AGING PROGRAMMINUTEMAN STAGE II SOLID ROCKET MOTOR

P. V. Veit, L. G. Landuk, J. V. Simpson Jr., G. J. SvobAerojet Solid Propulsion Company

Sacramento, California

ABSTRACT

Aging and surveillance programs designed to verify the service life of the propellant-liner-insulation system of the Minuteman Stage II and Stage III motors have been underway atAerojet for the past 27 years. During that period the program has been continuously revised torefocus the effort and to incorporate technology advances. This paper presents a historicalreview of the program, with emphasis upon the evolutionary process which has led to the currentprogram structure. The "umbrella approach" to testing applied in the early days is discussed,as is the major impact provided by the ability to "look at the motor," made possible by thedevelopment of motor dissection and nondestructive evaluation techniques. Finally, the presentbalanced approach incorporating "the best ot both worlds" is addressed, and several importantlessons learned during the arduous process of recognizing and correcting early judgment errorsare elucidated.

INTRODUCTION

An aging and surveillance program designed to verify a 3-year service life for thepropellant-liner-insulation system of the Minutemau I Stage II motor was initiated at AerojetSolid Propulsion Company in 1961. The current program, sponsored by the Air Force Ogden AirLogistics Command (O0-ALC), is intended to validate a service life of at least 17 years for theremanufactured Minuteman II/III Stage II motor and predict age-out with sufficient letd-time toallow orderly replacement or refurbishment of the missile force.

Modified continuously over the years to respond to revisions in the required operationallife of the Stage II and Stage III motors, and to inc...:- -Avances in state-of-the-artregarding aging, the program provides a uniqt• opportunity to see lap.. _;,•,;q and changes thathave occurred over the last 27 years, and review the analyses, discoveries, errors, and good andbad luck that brought these discoveries about.

The paper will be divided into three main areas: an overview of the chronology of thtprogram with emphasis on the events that provided the impetus for change, an in-depth 4.icussionof changes in key areas, and finally a description of the program as structured today, with areview of the lessons learned.

DISCUSSION

OVERVIEW

The 27 years of aging programs for the propellant-liner-wn,'l,. a v, stea cen Le tnve-niently categorized into three general periods.

The 1960s applied "the umbrella approach," includirg v.-. . it.hantsm studies,test development in areas of mechanical properties, cF- 'ri'1 lea.-ing anl bcl-listics, and in general, an attempt to learn more a r. ;-i myt,-t 1 e-, ematerial, solid propellant.

The 1970s changed the emphasis to "look at the motor." Testi:ý, -iai Iovtoed 1cincorporate newly-developed failure criteria, and the a•': c' JzJisact In( tcprovided a new insight into the realities of aging.

The 1980s achieved a cost-effective, balanced approach that included the "best

of both worlds," using motors as well as laboratory samples.

THE EARLY DAYS

The Stage II Minuteman motor being manufactured in 1961 (Minuteman I) was cast with a44-in.-dia bipropellant grain. Although both propellants were formulated with a polyurethanebinder, differences in oxidizer blends and burn rate additives were sufficient to suggest thatdifferent aging behavior could be expected for the two propellants.

This work was performed under Contract No. F42600-86-D-0093 with the Ogden Air LogisticsCenter, Ogden, Utah.


45

To understand the program, it is important to understand the environment in the aerospaceindustry in the early 1960s. The york force at Aerojet was approximately 20,000 employees, andmany rocket motors (in addition to Minuteman) were being developed ax the time. The spirit ofthe day was "full-speed-ahead, let's get the job donel" Convenience and availability of materi-als were important factors in accomplishing any task. The one-gallon ice cream carton or thrfoil-lined "barbecue bag" therefore seemed ideally suited as containers in which to cast, cure,and age propellant.

Aging temperatures were rather arbitrarily selected as 80, 110, 150, 180, and 2200F, withthe objective of getting quick results at the high temperatures. Results at 2201F could some-times be quicker than expected; e. g., the previously-solid propellant occasionally dripped ontothe bottom of the aging oven.

Aging samples included both inner and outer-cast propellants, as well as a simulation ofthe bipropellant bond; the importance of integrity in that area was obvious. Less attention wasgiven, however, to the propellant-liner-insulation bond. The decision to age prectt bond ten-sile and peel specimens was, no doubt, based on considerations of convenience. However, in timeit became obvious that this was a bad decision, since the properties obtained from the pre-cutspecimens after a given storage history were substantially different than those for similarlyaged bulk samples. A)l test specimens were wet-cut and then dried for one week prior to test-ing, a difficult roadblock for getting data in a hurry.

Testing of aged samples was largely directed at looking at changes in mechanical proper-ties; however, some chemical tests were also included. Data analysis, based rather loosely onthe Arrhenius equation, generally took the form of looking at the time for a given property tochange by a factor of 2: a change that was arbitrarily defined as "end of useful life." Whenthe log of this value was plotted as a function of the reciprocal of absolute storage tempera-ture, the result was usually obligingly linear, but not too meaningful, since no failure crite-ria were available with which to relate the data (Figure 1). Other techniques for data analysesincluded stress-strain failure envelopes and master curves generated using WLP (Williams,Landel, Ferry) or empirically derived time-temperature shift factors. All data were hand-plotted with great enthusiaso.

The program also included firing of subscale motors which were aged at elevated tempera-tures, as well as full-scale motors aged at ambient temperatures. Laboratory samples of propel-lant were stored to be tested periodically during the life of the motor.

In August 1962, plans were solicited by the Air Force Ballistic Missile Division for a3-year aging program to support the Stage II Improved Minuteman (Wing VI) motor (52 in.-dia, 88%solids CTPB propellant). This request Inspired creative thinking from all areas. The first-cuton costs indicated a 25-man level of effort for three years. Although the program was notcompletely funded, it is interesting to rte the areas of interest at the time.

Items initially proposed for investigation included chemorheology (effects of stress,temperature, environment, antioxidant, tiller), mass spectrometry, gas chromatography, ignit-ability/ ballistics, research on aging of liner and model compounds, effects of humidity, andextensive studies of creep behavior (the latter to be conducted by Stanford Research Institute.)

When the improved Minuteman Aging Program was initiated In 1963 it was initially patternedafter the program used for the Minuteman I (Wing I-V) Stage I1 motor. A wide range of tests wasconducted on propellant samples stored at temperatures ranging from 80 to 220OF. Severe post-cure, surface hardening (oxidative), and plasticizer loss at the surface of the samples indi-cated that motor aging conditions should be simulated as closely as possible in propellant-liner-insulation samples as well as propellant samples. Therefore, when additional aging sam-ples were needed in 1965 to extend the 3-year program to ten years, the practice of aging precutspecimens and small cartons was discontinued in favor of using 9 by 9 in. foil-lined cartons, inwhich the propellant was cast to a depth of at least 5 in. onto the liner-insulation system. Inaddition, stocage temperatures for aging samples were limited to 80, 110, and 135*E, to minimizechemical reactions not representative of those that would occur in the temperature range of themotor environment.

The program was asl., xpanded to include aging samples representing other variables thought' be significant. The.-. icluded samples from both of the manufacturers of the CTPB polymer,w well as the two types of mixers used in processing the propellant.

is work was performed under Contract No. F42600-86-D-0093 with the Ogden Air Logistics'',.ntet, Ogden, Utah.

46

Three additionil progiasms wore initiated in the mid-1960s that contributed significantly tothe basic aging proLrom. TI:ese vere (I) a transportation and handling program, (2) a failurecriteria study, and (3) a stidy of chemical mechartsms underlying the chemical and physicalchanges that occurred during aging of the propellant/liner system.

The aging program for the Minuteman I 44-in.-dia motor that vas initiated under the spon-sorship of the Ali Force Ballistic Missile Division vas scheduled for transition to the cognl-sance of Ogden Air Materiel Area (Hill Air Force Base) in 1966. Lack of sufficient samples toextend the agi , p-ogram to 10 years focused .ttention on the need to dissect full-scale motorsto obtain samp'es vlth positive lead-time vith respect to the operational force.

Motor VISSOL.ion During the late 1960s, several possible methods of motor dissection verebeinX'g•h iTi"i'd "o6 currently. Leading contenders for cutting through the steel (or titanium)case wre (a) electrochomical milling and (b) grit blait. Both methods required knife-cutt'ngthrougis the internal insulation and vire cutting of the propellant grain. The grit-blast methodcame Wno general use after fires occurred during electrochemical milling at other facilities.

As a matter of historical reord, a third metLod, innovative anm effective, but not entire-ly succeosful, vat also used, but only once. Linear-shaped-charge eplosive yas successfullyused to make tvo circumferential cuts through the steel case on a Min.-to:.an I motor. Vhen athird cut (axial) yap attempted, several large pieces of propellant vere forcibly pxpelled fromthe motor, and ont of the portions ignitel. Other portions, hovever, vere unharmed and vereused over the ,cxt fey years as a source of samples to continue the aging program for theMinuteman I Stage II motor.

The first full-scale motor to be successfully dissected vas a Minuteman Stage III motor(fiberglass cage) dissected in June 1969 as part of a design-margin study. By 1973, nine StageIII motors h4d been dissected, ranging from 10 to 60 months in age. These studies gave thefirst Indication that propellant in a fu)!-scale motor might differ significantly from propel-lant aged as a laboratory sample. Analysis of stresses vithin solid propellant grains, as veil*a prediction of aging stability, had elvays been based on mechanical 'ropertias determined fromlaboratpry sample The tacit assumption yss !hat the latter vould provide a reasonable esti-mate of properties vithin a propellant grain. Results of the dissections indicated littledliffeence in bond strength betveesr the laboritory sample and the motor. The propellant, hoy-ever, wes significantly harder in the motor (modulus at least 30% higher). The bond stressesvoold therefore be significantly underestirmted by use of laboratory data (Figure 2).

In addition o differences in properties of the bulk propell,.nt, significant gradients inproperties vere measured as a i-nction of distance from the bond line (Figure 3). Anisotropicbehavio, of the propellant, noted in laboratory samples as differences in propert~es caused tyspecimen orientation, yas almo seen in the motor.

An affect that van expected, but not confirmed, vas surface hardening at the bore of themotor. The plasticizet-rich environment--in that closed area, vhich apparen ly prov dad someprotection against surface haidening, had not been simulated in laboratory samples. (This vouldto corrected later in the program). In addition, it vai found several years later that thesurface hardening characteristics of the propellant differed significantly, depending on thesource of the CTPB polymer (Figure 4).

An Improved A&ln& Projr m. In additisn to the dissection of the first full-scale motor,anothgs slin-TIlcunt event o-curred in 1q69. A oiev aging program, initiated for the Stage IIImotor, provided the opportunity to incorporate the lessons lenrned in earlier programs. By thistime th importance of Imuletin: the actual system in the motor was obvious. Great care yestaken to assure that the insulation, fiberglas3, cork, and fungicide used to prepare the casemimulant against which the propellant vould be cast vere all fully qualified materials. Thesitual mo'or case was expected to be permeable, and it was assumed that the simulant would bealso. go we were surpris,'a to see no evidence of permeability in results of the aging studies.

The truth did not become obvious until several years later when we dissected an aged StageIll motor. Measurement of the xtrength of the propellant-liner-insulation bond shoved signifi-cant Jifferences betveen areas that vere protected by broad aluminum ground-straps and thosethat vere not (Figure 5)1 clear evidence of a diffusion-related phenomenon occurring. The 'eystep that we had nvglected in our catefully simulatee aging samples yes the hydrotesting(straining) of the fiberglass, in order to produce sufficient crazitg to iake the miterialpeo eable,

Th:: york yas performed under Coritr",'t No. F4?6(i1-R6-P-0093 vith the Ogden Air I/gist!isCenter, Ogden, Iltah.

4?

The aging program for the Stage III motor included several major improvements in additionto the attempted improvements in sample preparation.

The major innovation was in a formalized logic for assessing the aging stability of thepropellant-liner-insulation system, as vell as other motor components. The program requiredidentification of all age sensitive items, determination of possible failure modes, and an eval-uation of the probability of occurrence and effects of those failure modes. After a review ofpertinent data frn-a other programs, materials to be included in the prograa were selected andtests and failure criteria were established to evaluate aging trends and service life.

Another improvement wms in establishing a uniform cutting diagram for propellant agingsamples, whereby specimens for a given test would be obtained from the same location within theaging sample at each test interval. The effect of the cutting diagram was to reduce variabilitybetween test intervals that vas due to specimen location, thereby maximizing the ability todetect changes due to aging. An example of the variation in uniaxial tensile properties withinan aging sample is shown in Figure 6.

LONG RANGE SERVICE LIFE ANALYSIS PROGRAM (LRSLA)

A significant era for aging and surveillance began In 1971 when the Air Force, realizingthat scheouled production of Minuteman motors vms soon to be completed, reviewed the adequacy ofservice life prediction capability. This review revealed that state-of-the-art techniques,based on oldest tested motors in the operational forces, were acceptable for assuring thatage-out would ,iot occur within 2 years. However, the 2-year lead time that would be adequate ifthe subsystem were still in production would be insufficient once production had terminated.Location of new vendors, qualification of material sources, and countless other problems wouldadd significantly to the time required to resume production, particularly at the high ratestypical of early Minuteman production.

A three-phase program was therefore initiated by the Air Force in 1972 to extend the ser-vice life prediction capability for the propulsion subsystem, so that the start of age-out couldbe predicted with sufficient lead time to allow orderly replacement or refurbishment of theoperational force.

During Phase 1, various areas of technclogy were investigated and an approach was selectedthat was to be validated in Phase II for application in Phase III. The approaches investigatedincluded failure mode analysis, overtesting, failure distribution, and accelerated aging. Acombination of approaches, failure mode analysis with overtesting, was selected. Dissection ofa ten-year-old Stage II motor in 1974 revealed that the material (liner) that provides the bondbetween propellant and insulation had degraded significantly in the booted area of the motor.The potential for degradation of the liner had been identified in laboratory samples in 1967,but X-ray inspection of approximately 40 motors did not reveal a problem at that time.

Two questions were immediately asked: (a) what was the condition of the missile force? and(b) could the condition cause malfunction of the motor? To answer the first question, a methodwas developed in which a sample that included propellant, liner, and insulation was excised fromthe aft end of the motor and used as a source of specimens to evaluate pertinent material prop-erties (Figure 7). Removal of the sample was not detrimental to motor performance. 1n inten-sive effort was mounted to obtain data from more than 50 motors. The results shoved clearlythat bond strength degraded significantly with time (Figure 8).

Initially there was some doubt about the significance of liner degradation in a low stressarea of the motor. The failure mode that was hypothesized was a rather subtle one, involvinghydrolytic degradation of one material, the liner, coupled with shrinkage of another, the insu-lation in the boot (Figure 9).

The failure mode was demonstrated in September 1978 in the static firing of a motor thathad been artificially aged to a condition that would be expected to occur in some field motorsafter storage periods ranging from 13 to 20 years. This catastrophic failure provided suffi-cient basis for the Air Force to reassess the service life of the Stage II motor. It was subse-quently established as 17 years, with motors being regrained as they approach that age.

TESTING EMPHASIS

In conjunction with the changes in the methodology applied to the aging and surveillanceproblem ovei the years, significant shifts in emphasis on the various testing approaches have

This work was performed under Contract No. F42600-86-D-0093 with the Ogden Air LogisticsCrater, Ogden, Utah.

48

o0 ,rred. Figure 10 shows the specxrum of tests contained by various test plans implementedover the period from 1962 to 1984.

As indicated previously, the approach in the early years was to leave no stone unturned.After a fey years it became clear that some tests were less important than others. Initiallythere was considerable concern that a grain as large as the second stage Minuteman might slumpexcessively after aging; hence the early emphasis on creep testing. Later it became obviousthat slump was not a significant problem, and the creep testing vas eliminated. Motor vibrationconcerns and associated dynamic testing followed a similar scenario.

Changes in ballistic parameters and safety characteristics were thought to be likely conse-quences of aging when the program started, so these areas received considerable emphasis in theoriginal test plans. However, it became evident early on that these properties were quiteinsensitive to aging effects in this propellant system, and therefore were less deserving ofthat level of attention.

This gradual learning process, as illustrated in Figure 10, has resulted in a testingapproach which focuses primarily upon the response and failure data required to support analysesof the critical structural failure modes and on chemical data needed to evaluate and monitor thechemical reactions associated with key degradation mechanisms. In addition, heavy emphasis isplaced upon obtaining these data from motor samples which have experienced realistic storageenvironments.

THE CURRENT PROGRAM

The integrated Minuteman aging program at Aerojet is shown schematically in Figure 11.Each type cf testing provides unique information with respect to variability and/or aging trendsof the system. Vhen combined, this information provides sai extensive data base aud a cost-effective approach to prediction of service life of the propellant-liner-insulation system.These sources of information were incorporated into the aging program in the chronological ordershown.

1. Motor Firings provide the "bottom line" demonstration that the system works (but are oflittle value iKndetecting aging trends). An Aerojet Value Eng!neering Change Proposal (VECP)was implemented which significantly reduced the number of static firings of aged Stage II andIII motors, with resulting savings to the government of many millions of dollars.

2. Laborator Spes form the basis for an economical method of measuring trends in keyproperties, as welf as the variability of propellant within and between motors during produc-tion. On the negative side, there is a bias In properties between propellant cast into labora-tory samples and propellant cast into full-scale motors. The stress/strain environment existingin a full-scale motor is not present in laboratory samples. As a test of differences amongpropellant/liner chemical lots, the analog carton can flag the need to look at a group ofrelated motors (heredity). The configuration of the aging analog currently in use simulatesboth the propellant-liner-insulation bond and the bore surface of the motor (Figure 12).

3. Dissection of Full-Scale Motors provides an excellent opportunity to look at the realthing. Stresses and strains are, of course. telieved by disection, so that aging data obtainedfrom remnants of dissected motors must be treated with caution. The storage of remnants forfuture tests must also be viewed cautiously, since sealing against moisture effects is differentthan in the motor.

4. Excised Samples, as developed in the LRSLA program, also provide an excellent, e.;'en-tially nondestruct ve, means for sampling the propellant-liner-insulation system in the full.-scale motor. This is a prime example of benefits of "looking at the motor."

Excised samples have now been tested from approximately 150 Hinuteman II and III SecondStage Motors, ranging -n age from 3 months to approximately 18 yeaLs. The current aging plancalls for testing of samples from 6 motors per year as they are returned for remanufacture.Additional samples excised by O0-ALC are also tested at Aerojet. These are from motors ofvarious ages temporarily available at O0-ALC. Samples are excised from the aft end of the motorto evaluate mechanical properties of the propellant-liner-insulation system. In addition,samb)les are excised from the surface of the forward fin slots to detect any changes in Ignit-ability of the propellant surface that may occur as a result of aging. The device used for thisevaluation (the IDM or Ignition Delay Motor) is a special piece of equipment developed at

This work was performed under Contract No. P42600-86-D-0093 with the Ogden Air LogisticsCenter, Ogden, Utah.

49

Aerojet which has been used in support of several motor programs. The device was specificallydesigned to simulate the heat flow during motor ignition.

5. Bulk Samples (samples of propellant larger than the excised sample) were removed fromthe aft bore area o 25 motors prior to remanufacture to evaluate effects of aging and motor-to-motor variability in the critical bore area. This test provided important information but amsdiscontinued when correlations with other data were established. The Aging and Surveillanceprogram is not totally fixed but is continuously examined and modified, with OO-ALC approval, todelete unneeded tests and add others as they can provide improved understanding of motor aging.

6. Nondestructive Test Methods (NDE) include not only the excised samples described previ-ously (conducted on six motors per year), but also use of the On-Surface Tester, another specialtool designed and built at Aerojet to nondestructively assess physical properties of the propel-lant surface in the bore of the motor.

A visuaL inspection is also made of all motors being returned for remanufacture. Observa-tions are made and noted in accordance with an established plan, to provide a uniform ranking ofage-related phenomena observed In motors to be remanufactured.

7. The Plu Rotor is the newest add'tion to the aging program. Through-the-case samples,Including propelant-iner-Insulation ani case, are being removed at periodic intervals fromthree full-scale production motors. Twenty-four 6 x 6 x 4 in. samples will ultimately beremoved from each motor. The motors represent both original and remanufacture motors (vintages1976, 1984, and 1986). The plugged areas are recast with propellant and sealed to maintainstructural/ environmental integrity during the scheduled storage period of 12 years. Initialtesting includes plugs from four areas of the motor to provide an estimate of within-motorvariability. Subsequent testing provides information regarding effects of aging in a realisticstress-strain environment. By coordinating the test schedules for motor remnants and plugs,valuable information can be obtained regarding variability in propellant as a function of time.The chronology of the aging program is shown in Figure 13.

CONCLUSIONS

In retrospect, a number of errors and misjudgments were made early in the program which ledto inaccurate and misleading results. However, the recognition and correction of these mistakeshas, over time, provided the impetus for refining the approach to its current level of effec-tiveness. A number of lessons have been learned in going through this rather arduous process.Those considered most valuab)e are summarized below.

Valid aging samples are the prodict of a detailed review of, and proper consider-ation for, the motor design, materials, .nd manufacturing processes.

A core of tests should be established for use throughout the program, but it'simportant to maintain sufficient flexibility to respond to unexpected developmenti.

All materials should be characterized thoroughly in the initial phase to provide areference base for aging evaluation, as well as subsequent requalification.

In assessing the significance of atparent aging trends, it's important to giveproper attention to material variaoilities.

Although in some cases carton samples have certain limitations in terms of simu-lating the motor, when prop2rly designed they can be a valuable data source forassessing variability and monitoring age-induced changes.

Motor samples should be used as much as possible to calibrate laboratory sampleresults, to verify apparent aging trends, and to discover any age-induced changesnot manifested in the laboratory samples.

An ac'.urate, well-doctimented data base should be maintained througho-it thi program.

This work was performed under Contract No. F42601 86-0-0093 with the Ogden Air LogisticsCenter, Ogden, Utah.

50

REFERENCES

Williams, Malcom L., Landel, R.F., and Ferry, J.D., J. Am. Chem. Soc., 77, 3701 (1955).

Lund, Eugene P., Minuteman Long Range Service Life Analysis Overviev, AIAA/SAE 12thPropulsion Conference, July 26-29, 1976.

Veit, P.W., Landuk, L.G., and Svob, G.J., "Experimental Evaluation of As-ProcessedPropellant Grains," Journal of Propulsion and Pover, American Institute of Aeronauticsand Astronautics, Vol'um-iT, Number 6, Novemher-December 1485, Page 494-497.

2.8 sawf1 01

Sad-O* ' T*b ~j Pve.-1. '00 " --"----.........---

ethane ropellats Ca• e Relatd to 30 PBpgherinoosthnn bra

2.400 CP rpW

Lm .. .. .. --_ /" "_ ". 4- _- _ _0J

w 2,000 0 A •

CenerOgdn, tah

S I I I I I &

he1s 0.wnC

6.".

-M..s _ 1 ___ J

Figure 1. Signadficantrdet in Properties o oyr Figure 2. GReaxaetion Propertsies Aproiat eSigetare Evidellnt s at n te bodne Relne toi0%ichet and MotyoVry tha PnLaolyme

DisseceMoosSue

ThsnorwapefreuneCotato.F2(-6D-)3vtthOgeAiLoiicCenter, ogen Utah.W~ TP) Us aR T@

a5;

An Rod a"g on" 46

Figure 5. Strength of Propellant-LnrInsulation Bond is Significantly

/ Greater Under Ground Strap

too "Na I'sa h a IL

-4-- is 0

Am"

Ill

I 0416

Figure 6. Properties May Vary Significantly iIVith Location in Propellant Sample

Ue6Aduk Lot~~ Uoduh Codosas

A.,." 1100"M

L~mL

-2 Figure 7. Propellant-Liner-Insulation SamplesCan be Removed From OperationalMinuteman Motor Vithout Adverse

4*0.meEffects on Operational Performance

An-

0 0

Figure 8. Kinetic Prediction iq Supported byoData From Field Motors ---- -- -

000

o

00Do" From. Eacised sanm1.

SltpTim*,M. th

This work vas performed under Contract No. F42600-86-D-0093 with the Ogden Air LogisticsCenter, Ogden, Utah.

52

prp~to41"iem- Figure 9. Failure in an ArtificiallyCbeh rh Asen aw " liam"andF#UWU~Aged Stage II Minuteman Motor

am vi PlowCaused by Lifting of Forward

Boot Nipple was Demonstrated

Propellant TestsTest Plan Year

* Mechanical 1962 1955 1913* 1SIP 1944'

Coolftc~sI of Tlwrrma ExpanadmOrnoDly x x x xDiltataion x x x xDyna Ic x x xO*ll;Modulua xCreep x x x

Consalan Strau x x XUmiajaw Tons" It x I x xUndaalai Tonal. W/ Prossgro and Presiraon K K x KMeantal Tonsd. x x x K

Cohesive Toam x x

* ChemicaSura~ng x x x x xExtraction x x It x

* other x

Arc Imag4/lgndll4On DelY K K xSolely

Detonation K xImpact Sensitivity x xA.uloognilhon x

Propellant-Liner-Inhultalon Bond Tests

Constant Loadl Tenade anW Shoat x x x KCoenal aw Term" nAd~d hear x K KHigh Rate Sham W/3Soionpelmped Pf**ewo x K x

Adhesive Toam xPool x

* Mechanical and Chedmica' Tests andEgcleed Semple (Atl PuopILInf/Ilnst x x

Eackaod Sammle (l"ni'on Detail x xftomments (StsagItaNW 11) 1 x 1

Gradieto in burn Rate K K

Dissection x

-Includes Componenis Also

Figure 10. Aging Test Plan:. Change as Knowledge Grows

Thils work was performed under Contract Nn. F42600-86*D-0093 with the Ogden Air LogisticsCenter, Ogden, Utah.

Figure 11. Aging and Surveillance Program Figure 12. Analog Samples Simulate Bore andEmphasizes a Unified Approach Bond Area in Weather-Sealed Motor

Ye__Ar Event SiS ificcnce

1961 Minuteman l Stage II program begn Exploratoy period; high temperature(polyurethane) aging

1963 Minuteman II Stage II program begun Emphasis changed to bulk aging(polybutadiene)

1966 Service life goal increased from 3 to 10 Service life exceeded initial expects-years t ions

1965 Special studies conducted: transporta- Aging program focused by special studiestion and handling, failure criteria.mechanisms of aging

1968 Ainutean Stage III program begun Emphasis on failure modes and failure

cri teria1969 Minuteman Stage III motor dissected Crton-motor bias identified; motor

renants added to aging progrga

1912 Planning initiated for Long Range Intensive effort to extend predictiveService Life Analysis (proga) Program capability

1974 Segraded liner found in 10-year-old Possible failure mode identifiedmotor during dissection

1975 Propellant-liner-insulation samples Problem identified in operational forceexcised from operational motors

1978 Artificially aged motor fails in static Predicted failure mode is confirmed;firing service life established as 17 years

1980 On-surface tester and Ignition Delay Use of N-m techniques strengthsMotor (IM) ndded to program prograam

1984 "Plug Motors" added to aging program Ability to monitor aging behavior in

realistic stress environment

Figure 13. Chronology for Minuteman Propellant Aging Program

This Dork vas pegformed under Contract No. F42600-86-D-0093 with the Ogden Air dgistics

1enter, Ogden, Utah.

54

INSENSITIVE MUNITIONS MASTERPLAN AND ITS IMPLICATIONSFOR THE SDI COMMUNITY*

Felicia Riggs CookUSASDC, GBI-X

Huntsville, Alabama

William C. Stone, PhD, PEStone Engineering Company

Huntsville, Alabama

INTRODUCTION

Military leaders have long known of the vulnerability of stored munitions and the desire on the part ofthe enemy to eliminate these assets by fire or explosive initiation. In past history these munitions uuallyconsisted of artillery shells, gun powder, small arms rounds, and other devices associated with ground basedwars. More recently the Introduction and use of rockets, both solid and liquid, as well as high explosivewarheads has presented an additional target for intentional initiation. The chances of and consequences ofaccidental initiation have also multiplied. Two recent incidents have served to emphasize the sensitivity ofour present day munitions and have highlighted the need for protection of rocket motors, warheads, and otherexplosive devices.

The most spectacular accident involving the sensitivity of munitions was aboard the aircraft carrierForrestall when a rocket motor was accidntally ignited and flew into a group of parked aircraft which werearmed and serviced for takeoff. The subsequent fire which was fueled by spilled Jet fuel engulfed the air-planes, ultimately setting off additional rounds. This process continued and initiated additional rocketmotors, warheads, and aircraft fires. Damage and loss of life were heavy.

The second incident occurred in Germany with the Pershing I1 system on a cold winter day. The methodof initiation of one of the Pershing I1 rocket motor stages was determined to be by electrostatic discharge.The resulting damage to the facility and personnel loss was also very costly.

These two incidents emphasized to the Military services that a need existed to reduce munition hazardswherever possible. In essence, these considerations led to an Insensitive Munitions (IM) policy which was sub-sequeritly signed by the joint services and evolved into Military Standard 2105A. 1

The IM policy of the Department of Defense outlines and describes MIlitary Standard 2105A. MilitaryStandard 2105A defines tests which must be passed by munitions of all kinds which are put into productionafter 1995. The purpose of the regulation is to require the munitiorn industry to develop systems which areinsensitive to several initiating stimuli. This includes rocket motors (solid and liquid), warheads, and anyother devices which when heated or penetrated by projectiles can produce reactions more damaging than fire.

There are four tests which each new munition is required to pass with no result greater than a fire.These are illustrated in Fig. 1 through 4 and consist of fast cook-off, bullet impact, sympathetic detonation,and fragment impact.

In addition to these four tests, a vulnerability analysis must be performed on each missile system todetermine if in the fielded application the system will be exposed to additional stimuli. Additional testsmay be required, four of which are iliustrated in Fig. 5, 6, 7 and 8 and consist of slow cook-off, shapedcharge jet impact, shaped charge jet spall, and electrostatic discharge. Table I summarizes the requiredtesting. Other testing may be required depending on the particular system and its mission.

The IM requirements are aimed at insuring that all systems pass the Insensitive munition tests andthereby reduce the hazard to both personnel and equipment in the case of exposure to these stimuli. How-ever, the IM policy does allow for a waiver wherein systems which cannot meet the IM requirements mayapply for a waiver with the Army Acquisition Executive (AAE). The waiver application must be supported bydocumented evidence of the efforts to develop insensitive technology and why the regulation cannot be met.

*This work performed under Contract No. DAAHOI-89-C-0099 with the U.S. Army Missile Command,Redstone Arsenal, Alabama, 35807-3801.

Approved for public release; distribution is unlimited; August 1990.

55

Figure 1. Fast Cook-Off F ligure 2. Bullet ImpactUUsS

U s

A~.R

YMYy

Figure 3. Sympathetic Detonation Figure 4. Fragment Imipact

56

Figure 5. Hazard Analysis: Slow Cook-Off Figure 6. Hazard Analysis: Shaped Charge

Jet Impact

I( U

RA

Figure 7. Hazard Analysis: Shaped Charge Figure 8. Hazard Analysis: Electro-StatieJet Spell Discharge

57

TABLE I

INSENSITIVE MUNITIONS TEST REQUIREMENTS

"o Mandatory Tests For Insensitive Classification:

Past Cook-Off No Reaction More Severe Than Burning

Bullet Impact As Above

Sympathetic Detonation No Propagation

Fragment Impact No Reaction More Severe Than Burning

"o Hazard Analysis Tests:

Slow Cook-Off No Reaction More Severe Than Burning

Shaped Charge Jet No Detonation

Spell No Sustained Burn

Electro-Stat Discharge No Reaction More Severe Than Burning

APPLICATION TO SDC SYSTEMS

SDC systems, such as HEDI and GBI consist of solid rocket motors, liquid rocket motors, high pressuregas tanks, high explosive warheads, and initiation devices. All of these devices may be sensitive to thestimuli described in the IM tests. The hazards to which a system may be exposed vary with the basingmethod. in general, the mobile based systems are exposed to a wider variety of iazards. The fixed based,many times being installed in silos, are more difficult for saboteurs to approach. Nevertheless, hazardsstill exists for the fixed base system.

Regardless of the basing method, transportation of sensitive munitions always involves a hazard to bothpersonnel and equipment. While some systems may be manufactured on site, others are not and require trans-portation either on the public highways and/or by rail.

There are additionb, differences among the variety of systems founa in the Strategic Defense Command.In general, the exoatmosph'.ric interceptors are propelled by rocket motors which have long burning times andtherefore low propellant btrning rates. The low rate propellants are usually less sensitive thereby beingmore difficult to ignite anl less sensitive to impact and friction. Conversely, the endoatmospheric missilesnormally use high-buining .ate propellents in their rocket motors to provide short burning times. Thesehigher rate propellant.- ar.3 in general more easily ignited and therefore more sensitive to friction andimpact. Both endo- and exoatmospheric missile designs are using bipropellant and monopropellant liquid pro-pulsion systems in the upper stages. These may be pressurized by high pressure gaseous storage tanks or byliquid or solid propellant gas generators. All these devices have some sensitivity to projectile penetration,fire, and impact by jet spelling.

Design optimization to minimize the effect of projectile impact, fire and other IM hazards may have asignificant impact on rocket motor design, function and cost. In general the cost is expected to increase asdesigns are configured to meet the IM requirements. Vor example, the less sensitive propellants are usuallylower energy which results in larger rocket motors for the same size payload. Hence, the cost will increase.

There are some recognized methods of meeting the IM requirements. For example, insulation blanketscan be used during transportation to protect the munitions from the fast cook-off onvironmnent. Also, anarmor shield around motors and warheads during shipment may be effective in limiting the effect of pro-jectiles. Some rocket motor cases degrade significantly in pressure capability on penetration by a pro-jectile. The effect of igniting the propellant may therefore be significantly reduced uecause the chambercannot hold high pressure. The Strip Laminate case, for example, can be designed to unravel upon penetrationby a projectile or upon heating by degrading of the bonding material.

High pressure bottleo may be designed to simply bleed down to ambient pressure upon penetration or todevelop leaks when hested. Ideas such as these need to be studied further.

58

MEETING THE IM REQUIREMENTS FOR STRATEGIC SYSTEMS

Design is the key to insuring that SDC missile systems will meet the Insensitive Munitions requirement.The operational requirements and mission requirements must first be analyzed to determine the kinds of muni-tions which will be in the system and the type of basing which will be used. The technology level which willbe a part of the muitions is an important aspect of meeting the IM requirement. In some cases, new tech-nology will need to be developed. For example, propellant formulations which are insensitive to penetrationby projectiles may be possible. Rocket motor designs may be made with cases which resist heating.Transportation methods may be used which can limit the exposure to various forms of hazards. Early systemdefinition can define the testing which will be necessary to evaluate the sensitively of materials going intoa system design. However, the final testing must be at the system level; hense component testing may notbe useful in some cases.

Early evaluation of the requirements and application of these to the system basing mode can be amethod of reducing the IM testing required in a development program. For example, the GBI office has ini-tiated a position paper to look at the possible reduction of insensitive munition testing by design avoidance.Specifically, this position paper has suggested that the sympathetic detonation test will not be required if anon-detonating propellant (zero cards) is used. Since the sympathetic detonation testing must be performedfull scale, the cost avoidance of eliminating these tests is obvious.

Most of the testing must be carried out on the full scale unit because the size of the units tested havea significant effect on the results. However, there is some subseale testing which may possibly be used.For example, propellant sensitivity is one area in which subseale testing can have beneficial affects onrocket motor designs. Subseale testing of high pressure tanks can lead to designs which will not rupturewhen either punctured by projectiles or when heated in cook-off tests. The use of gelled liquid propellantsprovides a drastic reduction in the hazards associated with biprupellant liquid rocket motor systems.

The key to meeting the IM requirements for Strategic Systems is to examihe the system design and theoperational characteristics and use these inputs to guide the designers in coming up with configurations whichwill pass the IM tests. In those cases where little flexibility exists in the requirements, application for awaiver may be in order. The extent to which a waiver of the IM requirements will be granted has not yetbeen determined because very few systems have been designed with the IM requirements as part of the speci-fications. Meeting of the IM requirement is new and the munitions iodustry is just beginning to addrns.3 theproblems.

CONCLUSIONS

The USASDC is working to meet the IM requirements on All its systems. The propulsion anJ controlsdesigners are beginning to incorporate the requirements into the specifications, and these are resulting insome design changes to reduce the sensitivity of the explosive devices. These studies indicate that costsmay be increased and in some limited examples, the IM requirements may not be met and a waiver will berequired.

Both HEDI and GBI are sponsoring studies At the MICOM Propulsion Laboratory en rc:thods of reducingthe sensitivity of munitions. The Strategic Defense Command has the meeting of the V'I, requiremert a itsdesign goal and every effort will be made to insure that all design approaches are evalListed.

REFERENCE

1. Military Standard: MIL-STD-2105A (NAVY), "Hazard Assessment Tests for Non. Nuclear Munitions".19 January 1990. Prepared by Naval Sea Systems Command. DRAFT COPY - not to be used until approved.

FLIGHTWEIGHT DIV RT PROPULSION SYSTEM HOT-FIRB TEST

L. K. West, P. C. PhillipsenAerojeA Propulsion Division

Sacramento. CaliforniaPRA-SA-AF-SSD/I

ABSTRACT

On 8 June 1990 Aerojct Propulsion Division successfully hot-fire tested a flightweignt liquid bipropellant propulsionsystem designed to provide divert AV to a space-intercept kinetic kill vehicle This work is being performed on the SCIT-DACS contract with Martin Marietta Corporation, working in conjunction wsn the Air Force Space Systemx Division andSDIO. The propulsion system verified in this test is directly applicable to SBL, GBL and Brilliant Pebbles kill vehicles. In1991, Aerojet's propulsion will be integrated into Martin Marietta's kill vehicle and be hover tested at Edwards Air ForceBase.

The subject propulsion system was tested as an autonomous system wit iour divert thrusters, one each fuel and oxidizertank, two helium bottles, eight cold gas ACS thrusters, manifolds, and accessorw hardware.

The hot-fire test demonstrated the following operations in a 30 sc,;ond duty cycle which accumulated 267 divert enginefirings and 1,446 ACS thruster pulses:

• Waterharnmer is controlled during system activation.

* Thrust accuracy of divert and ACS thrusters is maiitained within ± 10%.

* System mixture ratio differs from single engine steady state mixtuie ratio.* The feed system remained stable during various rapid divert and ACS thruster firing cycles.

* The 90 in. 3, 9500 psia helium storage is adequate to supply the ACS, pressurize propellants, and operate bipropellant

valves.

Video, sound, and high speed film recordings of the test augment this written paper.

INTRODUCTION

On 8 June 1990 Aerojet Propulsion successfully hot-fire tested its bipropellant SCIT-DACS divert propulsion system(Fig. I) in a 30 second duty cycle. SCI'F-DACS (Systems Concepts and Integrated Technologies - Divert and Attitude

Fig. 1. SC.7". DACS Prupulsion Syrtcm in Test

Approved for public releas,!; distribution ir unilimited.

W-

Control System) is a 3-1/2 year contract with Martin Marietta sponso-ed by Air Force Space Systems Division and SDIO underthe SABIP SCIT Contract. 1be successb-,! inot-fire system test is a major milestone in tie contract. It verifies the propulsion7 ystmr. Is ready for integration intu Manii, Marietta's kill vehicle for hover testing in September 1991 at Edwards Air Force

Under fth SCIT-DACS contract, Aerrojet has:"(kcricrated a flight system design integrated with Martin Marietta's all composite structure.

"* Generated component designs."* Fabricated and verification tested all DACS components with particular emphasis on Aerojet's lightweight titanium

thruster at Atlantic Research's composite overwrapped positive -.xpulsion propellant tanks,"* Assembled a VTV (Verification Test Vehicle) DACS for system level verification testing."* Performed it pressurization subsystem verification test of the VTV DACS,"* Perfonw.4 a duty-cycle hot-fire test of the VTV DACS.

In the~ reniajadcr of the program Aerojet will assemble a second DACS for integration into Martin Marietta'; kill vehic-..and suppor hsver testing of the KV in September 1991.

This paper describes the result.s of the aforcmrnrtifi icd VTV-DACS hot-fire test including discussion or test objectives,TJACS hardware, test duty cyclt, and results.

0hI~JEYLV

Objectives of the VTV I)4CS hot-fire test wer,: derived from hov! rett requirements, KV deployment requirements, andl)ACS design requirements. Six primary objective, ef the test are:

I. Dcmoa Sygern • labih4~. This is the obvious objective to demonstrate the sy- tein functions as planned andremains intact over the test duration.

2. Contrl jjjb ]just y DUrng M~ltiple 'I apjostf Ejjng. Thrust accuracy is important for predictabte AVresponse to KV guidan~cecommands- 11w SCIT-DACS is required to deliver divert and ACS thrust within t 10% of nominalfor any ccnrbinatior, (if engine firinig (fun, a single jivert thruster to two diverts firing steady state with two ACS frnng a 50%duty cycle.

3. L0fMfLyVjJWiy When enginec are pulsing on and off and/or alternating, it is important that feedline. pressure oscillation~s not couple. Coupling can cause component rupture at the high pressure or can starve engines of .uffi-cient propellant at the low pressure erd. The objective is to show that system pressure oscillations converge rather thandiverge during duty cycle operation.-

4. CnuQI 5_kLW;W~cbmc Propellants are isol~ted in the propellant tanks for long term storage (10years). At system -itirtup, it iis quite important that the column of propellant filling the lines does not water hammer atcompletion of fill. Without the btart up control provided by SCIT DACS orificed pyro-valve, pressurest in the feed lines couldspike at 5,000 ptsi and cause compoinent failure, The objective is to demonstrate tihat pressure spikes are- controlled below2500 psi during startup.

5. I~LICmongrac 2ixiiMLL alio Control for Cceaj-rQLMAai~~UCii Propel laru ceniter -of-mass (CM) must bemaintained during a mission within a given band to minimize ACS impulse requirements. Our CM design goal required thesystem to draw proplclantit from the two tinks at an average MR of 1.00 ± .04. Our obiective i's to measure and show in thistest system MR is 'uaintuincdf within 4% of 1.0.

6. Demonuust at H um )ad Aftuwy.x Helium is stored in the DACS a. 9500 psia. It is used as propellant in thecold gas ACS thrusters,. to pres.%uriic and expel the propellant tanks, arid to actuate the bipropellant valves. This objective callsfor demonstrating 11.2 lbf-sec impulse delivered by the ACS while pressurizing propellants and operating bipropellant valves.

7. L~inmf~aitJIlirumtzj-Muo~n ~ ib~ikijtx. Th,: divert thruster mounts directly in contact with the graphite thermo-plastic. The objective is to demonstrate that heat soak to the thermoplastic does not raise its temperature above thc 250TF plas-tic trans~ition temperaturc.

SJS11hM OPERATION AND IJARD WARE DESCRIPTION

The SCHl -DACS provides divert propulsion with 3-axis slabilifdtiov, to a kinetic kill vehicle fi~ttd vwithi advanced sensorand guidance electronics. 'Mbe 1)ACS operdites autouioinously from tlrc electronics ._xcept for activation and thirusier command2ignals Divert AV is provided by thc four divert thrusters (thrust valuc is secret), %hown in the Fig. 2 schematiL, and the Fig. 3V IN l)ACS. A close-up view of the thruster v.~ given in Hig. 4. 3-axis stabilization is accomplished by orienting the diverts inthec pitch and yaw axis and aligning all to fire through th, vehicle's ccnier-of-mas% (CM).

0u2

He Ob ow .W P" own

ft..f a.,m0ý40 Yom E--

__di M3"lm Yub

_ mom voe -aACI ~ ~ ~ ~ ~ " POO%0 M- C lvuu

ACS ThrOusters COTrunT-OO

Feiuato 2.andWDC ShmtcadPrsueShdl

Oxidizer-r Tank&duc'rPot (8

ACSu BThuters (2)

Regu lato n

Syo-Vtv Sytem , Carbori/Epoxy and Carbon

yrVaeSytm Fig. 3. SCIT-VTV-DACS at Completion of Assembly Thro"'tcStutr

TotalWeight:lOg 1

hirpelnt Volvo',

buto

Fig. 4. SCIT-DACS Titanium Divert Thruster

63

Correction for minor CM and thruster mialipm•ents is provided by 2.5 lbf cold gas ACS thrusters in the pitch, yaw, androfl axes. Two axially oriented ACS thrsters are included to counter slight rearward AV predicted to occur in a hover testsituation where the KV is looking slightly above the horizontal when the divert thruster is fired.

The propellant tanks, made by ARC/LP, are composite overwrapped with a positive expulsion diaphragm that assurespropellant acquisition under all operating conditions. Additionally, the diaphragm maintains symmetrical propellant shapewithin the tak for CM control. The one fuei and one oxidizer tank are locatc- diametrically opposite each other and containequivalent propellant masses of 1.756 Ibm, all for CM balance. As suggested by the equivalent propellant masses, the systemoperating mixture ratio is 1.0. This is accomplished with a M-20 fuel blend (20% MMRH and 80% hydrazine) and N20 4 whichoperate near peak Isp at MR of 1.0.

The propellant tanks ar pressurized with regulated helium at 1450 psia nominal pressure. The helium is stored in twodiametrically opposite helium bottles at 9500 psia (at system activation). These bottles are also composite overwrapped formaximum strength-to-weight. Downstream of the regulator, helium is also used for ACS thruster propelant and bipropellantvalve actuation gas.

Isolation of helium and propellant prior to system activation is accomplished with pyro-technic isolation valvec. Dualpyro-.alves are utilized in the helium storage circuit to minimize propellant waterhammer at system activation. The orifice inthe upt.main pyro-valve retricts flow when the system is activated with the downstream isolation valve. The orificed pyro-val ! is ti ucd open after fuil system pressure is reached, providing full-flow of pressurant.

he. DACS is to be mfurbishable for multiple hover tests, uwerefore manifold connections are made with Swagelockfittings and fill/vent operations are performed through service valves. The VTV DACS (subject test article) was assembledwith CRES nmonifolds and fittings, while the HTV will have functionally equivaient manifolds of titanium to conne(t to dietypical titaniu.n interface connections of the components. Total mass of the propellant filled HTV DACS with structum, isexpected to be 6500 grams (14.4 Ibm) while the VTV was 9700 grams (21.4 Ibm) as tested.

The VTV DACS is instrumented with redundant Kulite pressure transducers in each of faur major fluid circuits as shownin the figure schematic. Thenrocouples we.e located on the external surface of each circuit, each thruster and the structurenear each thruster. To support structural dyn.mics ,valuation, accelerometers were located at key system features. The abovedata parameters were recorded along with valve command voltages and currents on FM Tape and/or high speed digitalequipment.

TETSEQUENCE AND DUTY CYCLE

The VTV DACS hot-fire test was performed at the A-Zone Rocket Technology Laboratory of Aerojet PropulsionsSacramento Facility. A pressurization subsystem test was performed to check-out the following system operations prior tocommiting propellants to the VTV:

"• helium fill procedure

"* system activation

"* regulator and ullage performance and stability

"* ACS operation

"• relief vwive operation

In this test, the propellant circuits werr disconnected and small pressure vessels installed to simulate the propellant tankpressurant cavities. System activation was successful with prescribed pyro-valve sequcncing. Other system operations weresuccessful while operating the ACS thuwters to a representative duty cycle.

The VIV DACS hot-fire system test commenced after the successful pressurization subsystem te3t. Following finalsystem checks, the hot-fire test sequence below was followed as a matter of safety and data production:

I. Fil fuel tank, record mass

2. Fill oxidizer tank, record mass

3. Pressurize helium storage to 9500 psig

4. Connect pyro-valve initiator cables

5. Activate system

6. Run system hot-fire duty cycle

7 Vent helium to 50 psia

8. Close propellant tank isolation valvc, then vent system to anibient

9. Remove tanks and weigh propellant

10. Decontaminate system and reassemble.

64

Th duty cycle fiuxd in the VTV DACS test provided data to meamure the system's performance against theaforementioned objectives. The 30 secrotd duty cycle is made up of 21 separatc sequermes of divest ard ACS thnsters tiringthat evaluaft the following thnrster operatiors.

"- sinigle divert engine, seady sate

"• single egine simtaalmng Hover test lift-off and sustain pulses"• two diverts, stady state

"• singl. ACS, steady staic and pulsing"* pulsing divert engin.es with offset and simultaneous firings

"* combined divert wnd AC3 pulsing.

The duty cycle wnalxtaed the following:

"* Total Duration: 30 sec

"* Four Diverts h•nisc-i Fired; 267 pu.ses

5.79 sec burn time

"• Fight ACS Thriu-ers Pulsed: 1446 pulses

6.925 sec cum. time

17.3 lbf-sec

Eight represntative duity cycle sequences are shown in Fig. S. Each sequence was initiated at a given time according to amnster duty cycle listing. The shaded portion of each sequence empresents "on" or firing time and blank portions indicate coast.If mome than one engine is involved, the left cilumn so indi.ae.ý, e.g., "A3" represents ACS thruster No. 3. and "1" representsdiven thruste No. 1.

Mma II I I' I I I I It II I I E J M

juma3

I I__ M - I I P

F- MWIS 111

emm

4- on 70uI 11111111 II-I IIk ý

1u6 W-u

em m mm

n-m .I 1M m m ý m mU -U mP 1, 0 M 1m II•mm

U4 Hom~ Im ! ____m

-- ... . -____________- .... . _u

Fig. 5. Representative Firing Sequences of Duty Cycle

M~uRladsThe VTV-DACS hot-fire tesm was extremely successful in producing data and in performance against the design and test

objectives. Key results are discussed below by objective:

I. LDnnsa Sy&stemOpMbjily. Every command given the DACS in the test sequence and Duty cycle wasperfornned. The pyro-technic isolation valves opened in sequence, pressure regulation was achicved, propellant expelled, allfour divert thrusters fired, all eight ACS thrusters operated. The entire duty cycle was performed with haidware intact.

2. Control Thrust Accursev During Multiple Thruster Firings. From our DACS simulation model we learned thatmaintaining divert propellant prssuw within a 500 psi bond (± 250) about nominal will deliver ± ! 9% thrust accuracy. Fig. (,test results display that divert thrust accuracy is maintained under maximum flow conditions of two diverts firing with threeACS pulsing on/off in a 50* duty cycle (5 msec on/5 msec off). Just prior to the 110 sec divert command propellant pressureis at regulator lock-up pressure of 1550. At the end of the pulse, the propellant pressure is reduced to 1360 psia, which is a 190psia fall (90 psia fall from nominal). By the ± 250 psi band definition given above, thnrst accuracy is maintained. This thnitmaccuracy was continued when the ithree ACS thrusters were cycled, sLarting at 70 msec after the two diverts were commanded(ACS on/off follows regulatmd helium pretssurm oscillations).

65

T-A1 1 1 1 1 11 1 1 1 1 1 1 11w p

;i I Thrus t Accuracy CwonatralldB

presue sclltinsexenece dnn te ut1c10 An fe exrdamle of Liepressuire oscwitinscnegneaesonicaes d Israi (ngtv or poIv prssr Wpks)cusdbyoeing or closin of the0 dsivr eof n giebp pvle

rT T II I 1 1,l1.

lig. . Fed Sytem Sabilty E idencd ' Pressures OsclltinCosege

presue Prcintepolatpressure 0-rp t4cc of 4~i 8. T4 tr'e shwterguae eim pes fe

pressure.~~~~~~~~~~~~~~Dl anaxd~rpeuefr20me fe ciali.Tepit of0 asem Withvation shownd WO h iirbneo h

TFicaly 6t Thrustim thcrac ContropllanWthianiod com0%i theirTw Divaertharnd ToASpAre Fiscpringe.I ti

3.sc Con firm ex rencd Syte Stw rabilit. Tir., theored s yste stabilt i deombnstrtJ bye~ th converg'¶ e iinch tfal an iflpressuresillations poticsexpeienifcaddntlrdue the pruts re2 ri few rxampes cofd theneIr h rcssure osiltoovrec rie rhote iancalculatIed bgased tr eacaed shopncorrlateto tlie divetenause fhisring sequlevelest air S.a 7,J 9, aheI In shiownd ihn pro5.Inllanclunases ditherbnesp(egative lineositied preisdacsue wpites cr ausdbyoenhimngor woi of uhedive'mrt eninm~el 'ipcrivalvhycalcuatin tatd converhetoamerd is otrleady bta ow pressur p.iinac tdsae

DietSqeceN.5Dvr Sqec o

Absence of Large Pressure Spike,Tlm4 of Indicates Control of Water Hammer

Activationp O Une Pressure

Accoleromeotr onS1O NC Pyro-Valve

SPrope~n I P...a m -a

AtnPaw Begin HRwsaur. 200oleb Fuel Urn Pressure

t ; , •• i 2 4 0 i s c:A • P ' I/-i~~He Prwesure~dn A ebv );I

, Regulated___"__i_;________ -f ... .p

,I .I ; I q

Fig. 8. Propellant Water Hammer Controlled at System Start-Up

5. Demonstrated Mixture Patio Control for CM Control. The VTV DACS hot-fire test improved our understanding ofpropellant utilization. We learned that boiloff of N 2 0 4 (N 20 4 vapor pressure much higher than fuel) significantly affectsmission average mixture ratio. The overall test MR was I. I1 with divert thrusters flow balanced at 1.00 ± .01. The 11% MRoffset exceeds our objective range of ±4%, but is a valuable result and will be used to modify our CM control plan.

From the correlation analysis with our single hot-fire tests of January 1990 where incremental flowrate versus timeis measured, it is clear the MR offset is caused by the combination of many short pulses (10 to 20 insec) followed bysignificant coasts of 100 msec or greawer. In each coast approaching 100 msec or greater, the N 2 0 4 remaining in the injector(dribble volume) evaporates while fuel remains. This loss accumulates over a duty cycle with many pulses to increase MR.To minimize the impact to our CM control plan, we may balance the thrusters at an MR below 1.0, where the MR value isbaed on a statistical evaluation of possible duty cycles.

6. Demonstrate Helium Load Adeouacy. The 9500 psi, 90 in. 3 helium load of the VTV DACS proved adequate toprovide the required 11.2 lbf-sec impulse through the ACS thrusters while maintaining ±10% thrust accuracy. This is verifiedby digital data output which quantifies the helium storage pressure to be 2880 psia at the point in time that 11.2 lbf.sec impulseis achieved. This 2880 psia source pressure was adequate to maintain the regulator outlet pressure at 1350 psia (within ±10%accuracy constraints) when two diverts and two ACS thrusters are firing at 50% duty.

7. Demonstrate 'thruster Mount Durability. The temperatures experienced in the DACS hot-fire test indicate that asignificant portion of stage impulse (greater than 50%) can be fired through one divert engine without fear of overheating thegraphite thermoplastic structure. During the 30 sec duty cycle, the temperature of the structure did not exceed 100°F. This iswell below the 250TF allowable. At a time 300 sec after test completion, a maximum soakout temperature of 168TF wasexperienced.

In conclusion, the SCIT-DACS hot-fire test was quite successful. It dcmonstrated system operability and compliance torequirements of thrust accuracy. feed system stability, waterhammer control, helium load adequacy, and divert thrusler mountdurability. With adjustments to syst-m MR, the DACS is ready to proceed to the Hover test portion of the SABIR-SCITProgram.

07

DEVELOPMENT OF THE ORBUSO I CARBON FIBER/EPOXYMOTOR CASE FOR THE STARS/STARBIRD PROGRAMS

Hugh M. Reynolds* and Manjeet Dhillon?United Technologies Corporation, Chemical Systems Division

San Jose, California

ABSTRACT

The Stiategic Target System (STARS) program required the development of a high-performance, thrust-vectorable space motor that would provide 25Z more performance than the existing fixed-nozzle version.One of key components that enabled the requirement to be met was the high-performance carbon fibermotor case. This paper describes the design requirements, design approach and analysis and demon-straced test data, along with manufacturing approach used by CSD to develop this motor case.Delivered fiber strength at burst was 715,500 psi. Demonstrated pressure vessel efficiency was1.85 x 106 in.

INTRODUCTION

In August of 1987, CSD was selected by Sandia National Laboratories to develop a third stage motor forthe STARS vehicle. The proposed design evolved from a sound motor case technology portfolio appli-cable to space motors. Nearly all of these technologies had been demonstrated on Government-fundedprograms such as the !mproved Performance Space Motor, SEP/CSD joint venture, KEW 10.2, Trident II(D5), Small ICBM pre-FSD or IR&D. CSD selected the AMOCO T-40 (Toray T-800) PAN carbon fiber and astandard aromatic amine-cured Bis-A epoxy EPON 828/ERL 420O/6DA resin system. This resin system wasdeveloped in 1972 and has been in use by CSD since 1976 on the IUS, Intelsat VI, TOS, Trident II (D5),and a variety of developmental filament-wound motor cases. The T-40 carbon fiber was fully charac-terized during the Small ICBM pre-FSU program and selected for the Trident II (D5) improved per-formance stage III motor. The selected internal insulation was Kirkhill Rubber Company KL (60)-269silica-filled EPDM. This insulation material is used on the IUS, PK, Trident I (C4), and Trident II(D5) rocket motor cases and represented the lowest risk insulation approach.

The design approach selected for the motor case was to balance the requirement for very high perfor-mance (PV/W) against the need for an ebsily produced product with very short development and qualifi-cation program. The more significant design parameter selections were stress ratio (SR - 0.64)eliminating the need for local port region reinforcement, a conservative hoop fiber design minimumstrength (SIGMA HO,.P - 580,000 psi), high fiber volume (VF - 0.60), film adhesive Y-joint shear ply(FM 123-5), and a unique oft skirt attach flange configuration allowing the stage to fit into anexisting envelope designed for a much smaller motor.

SUMMIARY

The Orbus I motor case design resulted in a wotor case 37.5% lighter than current SOTA productionunits made from Kcvlar-49, and 832 lighter than units made from titanium. Demonstrated pressure shellperformance an meagured by average pressure shell efficiency (PV/W) was 1.85 x 106 in. with hoop fiberstresses of 715.5 --i. Static motor firing verified insulation design with a demonstrated safety fac-tor >1.25 times worst case ablation.

Extensive design anelysis studies were undertaken to provide a minimum-weight design at low productioncost with!n the co,.straints of the motor devel,,pment program. All cases were designed and manufac-tured by CSD. Manufacturing reproducibility as measured by variability in insulated case weight is0.35 lb for a coefficient of variation of 0.82. The case and motor system, currently In production,has completed development and qualification phases.

DESIGN REQUIREMENTS

The key Orbus I case design req ilrements are shown in Table I. In addition, CSD imposed furtherdesign requirements based on the baseline motor design and CSD's previous development programs as wellas IRhD. These additional criteria are shown in Table I1.

This work was performed under contract No. P.O. 23-0957 for Sandia National Laboratories Albuquerque,New Mexico 87185-5800"Approved for public release; distribution is unlimited."* Orbus Insulated Case Design Engineert Orbus Insulated Case Manufacturing Supervisor

69

TABLE 1. ORBUS I DESIGN REQUIPEMENTS

Requirement Orbus i

GeometryLength, in. 32.9Interface bolt circle, in. 28.00Attachment 24 holes, 4 lift pointsForward envelope, in. 23.676 maximumDiameter, in. 28.58Space envelope drawing R08004Internal volume (propellant weight, lb) >905 lb

AccelerationsAxial 20Lateral 7

Vibration, ASD g 2/HzAxial 0.0156Lateral 0.C156

Shock, g 320 3 axis0.002 sec Haversine

Ultimate structural line loads, lb/in.Compression -320Tension +320Shear +_359

Electrical bondingGrounding of metal parts Resistance per MIL-13-5087

100 mohm maximumElectrostatic discharge _E 5000 ohms

EnvironmentsPressure, psia 0 to 16Temperature, (non-operating), *F 40 to 100Temperature, (operating), *F 350 case (motor 55-9.5)Humidity, (operating), % Up to 100Life, years 5

Reliability 0.9999826 (motor 0.998)

Note: ' Sandia Laboratories imposed requirementsPurchase Order No. 23-0957

DESIGN DESCRIPTION

The Orbus I motor case to a filament-wound pressure vessel with low length--to-diameter ratio (LID)),helically filament wound over an MeM Insulator with wound-in-place aluminum polar bosses machinedfrom Al 2124-T851 plate. Shear plies of uPtM rubbe: allow relative motion between the polar bossesand the carbon/epoxy structure. The Insulator Is made in a forward and aft dome cloned diecompression mold. The forward and aft polar bosses are molded Into each section using Chemlok 205primer and Chemlok 23h adheaive. The premolder dome insulaters are ihstalled on a segmented aluminummandrel and spliced together with uncured ePIM rubber overwrapped with dry Aramid roving and cured

prior ko winding. The presslure vessel consists of two helical and two hoop layers fully i.ntersperoedending with an external helical. The Orbus I case has a sing.l.e aft attach skirt. 1he skirt isintegrally wound in with a nitrile/epoxy film adhesive shear ply (FM 123-5) with intermediate shearmodulus In the okirt-to-case Y-joint. An aluminum flange Is machined from Al 2124-TO51 plate or6061-T651 roll and welded bar, and Is adhesively bonded and riveted to the wound skirt end for attach-ment to the vehicle. Two Al 1100-1-0 ground foils safldwicled between the aft skirt and flange provide

70)

TABLE 11. CSD-IMPOSED REQUIREMENTS

Requirement Orbus 1

Motor pressure at MEOR, psi 1090

Motor thrust at MEOP, lb 9380

Internal volume, in. 3 15,606

MaterialFiber T-40 carbon fiberResin EPON 828/EAL 4206/MDA

Factors of safetyProof 1.00FLT 1.25Captive FLT 1.25Ablation 1.5

Materials allowable (MIL-HDBK-5, -17, -23 or test)Metal A basisComposite B basis

Pressure shellBurst pressure minimum/nominal, psi 1363/1589Fiber volume fraction 0.55 to 0.65Fiber stress at minimum/nominal, burst, ksl 580/675

Allowable fiber strain at minimum/nominalburst, %/ 1.38/1.61

Pressure vessel efficiency

(PV/W) at minimumn/nominal, burst, x 106 in. 1.46/1.72

SkirtNominal margin on ultimate load 10%/ goalConstruction 0*/1±45*/90*Fiber volume fraction 0.50 to 0.60V-joint material FM 123-5Skirt-to-case design concept Integral

grloundi ng. lhe conductive fiber its also elect rically connected to the flange/foil using otiver-loadedepoxy to) stice electrostatic discharge tequirementsi. The OrLus I is suspended within the vehicle.

The Orbusj I care design is shown in Fig. 1, the finished came in Fig. 2, and the case design andstructurat haracteristirs in Table Ill. The helical-wound dome contours were developed for thrukt-

comperisated (aft dome), modified geo!(desic dome winding tising the CSI)-developed Composite Case Design

and AnAiysih (CIJAC) proyram. Vhit; design was then analyzed using CSU's Case Analysiis Program Sysie.m

(CAPS) computer code u'~irg the APGUS tinite element code.. This computer uses nonlinear large-

deflection geonetry as wull an mivnlinvar material properties to analy7c Oth! layered orrhotropic struc-

ture in order to) ye r I st ri~t ir..I nt(Y~r ity and prcm~u rtz ed enve I -w

pb - JI.-m ipa.! .J LA; u L ., I i.. j., oJm'n-iILI. Ira -d ,ufi j r I')rn.iisc, ain d na lt)ur ~LrIti. - 'A t hu mo~tor

i -A- -ift lI Sted ill FTblt'II I I tic 1,~n ic']L high! ight.. inc idt:

wor v'f t21r v /. /jI in. v -.tr hi 1, -It.M in. [I

* or.i s r ti. - lwo p'.'' s i~ .v,'ra;:-' 6 1, -. cI on.I ric

7'

AMOCO T-40 fiber/EPON 828/A 12-~ ~ahrnERL 4208/MDA resin 95

'to co posit

27.2 0028.000-diameter ground foil27.00b11crl-(ples

EEPOM case bolt CR2c39insulator FM.123-5 Y-joint Monet 9-2U

TRU-ARC (72 equally . 36 spacesNSOOO-375 Al 2124-TeS IIspcd

snapring bosses/closure .80iroe

(3 places) ~Note: Dimensions are 880daeein Inches

13435 dlameterI

32-47

Fig. I. orbus I Insulated Case Design

Fig. 2. ()rbus, I Flightweight. CAse

TABLE i11. CASE DFSICN AND STRUCTURAL CHARACTERISTICS

Requirement Orbus 1

Composite parameters (continued)Reinforcements

Aft dome and forward dome NoneSkirt-unidirecttonal hoop and axial, and 00 and ±4b" made from H'F-E-1948 AIN

:t45* interspersed reinforcementsY-joint Film adhesive FM123-5

Impregnation metliod In process wet for helical and hoopwindings, axial, and +-3° tapepreimpregnated

Insulation to case bonr- FM-73 with resin primerInsulation Silica-filled EPDM, KL (60) -26c

Resin content, % by weightHelical, hoop 31.5Axial, +451 40

Total weight including insulation, lbDesign 43.2Actual 43.23+0.35

PerformanceAverage burst strength, psi 1640 (3 tests)

Actual minimum burst, psi 1526PV/W, in. x 106 1.85

Minimum skirt structural strength,line load, lb/in.

Peak compression -370Peak tension +385Peak in-plane shear 680

Sending stiffness (El), lb-in.2 4.3 x 109Axial stiffness (EA), lb 4.7 x 107Shear stiffness (KAG), lb 5.2 x 106

GeometryDiameter, in. 27.23Length, in.

Boss face to boss face 32.87 (on mandrel)32.76 (finished/loaded)

Length-to-diameter ratio 1.21Port opening, % of case diameter

Aft 32.3Forward 12.6

Composite parametersHelical wind angle, deg 17.28Filament - helical,hoop axial, +45* Amoco T-40 12000 filamentResin EPON 828/ERL4206/MDAAllowable fiber strain, 0 1.61 nominal

1.38 B basis minimum

73

"* Skirt .-tructural strength - 370 lb!J•. compression, 385 lb/in, tension, and 680 lb/in, in-planeshear

"* Safey factor on ablet1ln - 1.5"* insulated case weight - 43.23 lb.

D%.'JLOPMENT HUrORY ANiD APPLICAbl.E DATAbASE

1he Orbus I de',elopment hivtory is based on .a solid foundation of basic material property charac-terization using establiebed materials and processes, verified by extensive testing, and culminatingin heavily instrumented and documented hydroburst and static motor testing. The validity of thisapproach Is supported by Lhe fact that all full-scale static motor and burst test case designs (inwhich Lhe fina' flightweight configuration re=ined unchanged) were 1002 successful.

Matertal Allowables. tabi2 IV suoamrizes allowables that ,re "B" ass from material tests run todate o,: the T-40/epoxy compusite and verify positive margins of safety for the full-scale design. TheKL(60)-269 silica-filled EPDM was Jnitially sized based on existLing data and verified by static motortest.

PFki-S~ale Tests. itill-scale tests centered on hydroburst and static motor tests. The test con-

tgi-rraticns are shown in Figs. 3 and 4.

Hvdrcburo. T"hSt6. The hydiburat test stano ts char.scterized by an aft-end up orientation with provi-.Aca * r aft dome t.hrust compensation and load takt"it on the aft skirt. The pressurizing medium iswater introduced i,.o Lhe thr.,st piston. H)drote~t pressurization is obtained by means of an electricpump. Hylr-t~sr iostrumentstion is shown in Table V. Dimensions, displacements at MEOP, and mass

TAhLE IV. 'MA'NRtAL ALL,)UABLES ("B" BASIS")

Para,,aeter Value

t:ull-scale filament-wco:,d prass'jre vessel

Fiber strain, % 1.38Fiber stress, psi 580,000

Tensile strength, psi0° 350,000

00/900 test 00 175,000

900 1900

Tensile strain, %

00 1.38

00/900 test 00 1.38

Compressive strength, psi90°/+20° 52,700

0d190/_+45' test JO (skirt laminate) 28,000

Bending/flexure, psi01 ultimate 142,000

Shear strength, psi0o ILS• 6500

001900/+450 ILS" (skirt laminate) 4000

Y-joint shear ply 4000

Rivet bearing, psi

0/b±450 i90° skirt mixture of plies edge-to-diameter (e/d = 3.0) 42,500

Note: " ILS - interlaminar shea,

74

Fig. 3. Orbus I Test Setup for Burst Testing

rig. 4. O)rblus I Stati Mo or lust Conti?,urat ton

75

TABLE V. SUMKARY OF TYPICAL RECOUDED DATA FOg HYDROBUk•T

Instrument Type Number of Instruments

Hydroburst test instrumentation 2Pressure gagePressuring rate, psi/sec 10Strain gage

Dome fiber 18Cylinder fiber 9Cylinder hoop 6

Displacement (removed for burst pressure cycle)Aft and forward boss axial 3 (0. 120', and 240)) Thrust vector angularlyAft boss radial 2 (00 and 900) and offsetDome/cylinder normal to surface Envelope determination, 12Acoustic 1 microphone in cylinderVideo Up to 2 camerasHigh speed (1000/3000 frarnes/sec) Up to 3 cameras

Note: Thrust relief on aft dome transmitted to the aft skirt is 8.60C x chamber pressure

properties of the finished insulated motor cases are suma,-rized in Fig. 5 and Table VI. The bursttest summery is shown in Table VII. Typical burst test configuration, failure mode, and strain beha-vior are shown in Figs. 6 and 7. Excellent Agr-ePnr was obtained between actual and predicted behA-vior, further verifyLog the performance and producibility of th. design.

Static Motor Teste. For the Insulated motor case, the objective was to verify inrulation thicknessand predicted ablation. The region of primary concern was the aft dome. Analysis and test measure-ments locations are shown in Fig. 8 and the static motor test resulte are sumrarlipd in Fig. 9. Asafety factor of >1.5 was maintained everywhere except ac the aft tangernt line wherr a safe.y factorof 1.35 was maintained. Insulator performance was very reproduc 4

ble, as indicated in Fig. 9.

Manufacturing. The aft and forward insulators are fabricated by compression moiling 'he ZPDl r,.bberdirectly onto the aft and forward end fittings as shown in Fig. 10. The bonding silrfaces of the endfittings are grit blasted and two coats of primer and adhesive are applied. These erud fittings arethen fastened onto the female insert and dropped into the mold cavity. The uncured EPEM rubber isthen cut into "donuts" and the correct amount is charged into the matchee die mold. "he EPDM rubbercure is staged from 100°F to 320*F and held at 320°F for a minimum of 2 hr at a mold p'essure of 500psi. The part is cooled down below 10F and removed with the insert.

The segmented mandrel has 36 segments and is made from 6061-T651 aluminum. Figure 11 shows themandrel canfiguration. Each segment is hard anodized, teflon coated and is small enough to come outof the aft opening. The segments are assembled from the forward dome upwards to the aft dome in awork stand. The assembly is then attached to the winding shaft and sleeve with the help ot adapters.The aft and forward insulators are installed onto their respective domes and an uncured EPDM centersplice section is bonded to the insulators with EPDM dispersion. Mylar shrink tape is then appliedonto the center splice, and dry Aramid helicals followed by Aramid hoops are wound over the inbulatorassembly. This procedure is shown in Fig. 12. This wound assembly is oven cured at 320*F. The nega-tive coefficient of expansion of the Aranid fibers and the positive coefficient of expansion of thealuminum mandrel consolidates and vulcanizes the bond between the insulators and centersplice as itcures.

Following removal of the overwrap and shrink tape, the insulator assembly is surface roughened, resinprimed, and film adhesive is applied to the whole surface of the insulator. T-40 carbon fiber/epoxyhoop and helical patterns are alternately wound. Figure 13 shows the completion of body wind. Thiswound assembly is removed frin thu filament winder and the skirt tool is mounted and fastened to theaft end as shown in Fig. 14.

76

Lee

14.--

OF DU DA

DCPS-m

DCCOCA

DCFD DCAD

No: All dlmesion are In Inche*

Fig. 5. Orbus I Insulated Case Dimensions and Displacements

Next the Y-joint seal plug, which uas pre-positioned on the skirt tool prior to mounting, is slid Intoposition betveen the aft dome And the skirt tool face. Peeforsed film adhesive wedges (Y-joint shearply) are cold pressed in a matched die sold and stored in a freezer until needed. These wedges withskived joints are placed over the dome butted and on top of the seal plug as shown in Fig. 15. Thisshear ply is compatted and adjusted to be flush and flat with respect to the cylindrical portion ofthe skirt tool.

After the compaction is removed and .he shear ply finished, an additional ply of film adheat'. isapplied followed by precut and stacked unidirectional prepreg sa'ored onto the skirt tool Lo form theskirt. This !)uJldup is completed by hoop overwinding .frh T-40 carbon fibet/epoxy as shovn in Fig.16. A porous tape compaction (wound under tension) layer is wound from the skirt end to the forwardtangent line completing the winding process (depicted in Fig. 17). The motor case is then S stagedand oven cured to the required parameters.

The cured motor case is mounted In a lathe. The okirt is ground to th- required length and diameter.An alumlnum attach flange io bonded to the skirt with the groundling foils in plice, and then sand-wiched between the skirt OD and attach flange ID as shown in Fig. 18. The skirt toal is then removed.The bondpd attach flange is drilled and riveted. The ground foils are bonded to the cylinder and domesurfaces with epoxy adhesive under vacuum pressure. The motor caue asescbly Is lolered inside andattached to an inverter stand as shown In Fig. I?. The wlni axis and mrndrel segments are carefullyremoved and the motor case is sent ti NDT for hydroproof and final empt), X--,.y testu and Inspections.The completed case is shown in Fig. 2.

CONCLUSIONS

In sumary, rig. 20 highlights the demonstrated Orbus I insulated motor ýase capability. P-r furlhercomparloon. an additional measure of per'orm.ance (shown at the battom o' the figu e) represents atotal chamber efficiency 'less insulatioa) as ompired to Lhe current .tate of the art. The Orbu- Iap~bility represento ip, oximately a 102 decrcasc in weight for given buret ttrengklh .I v Internal

volume over curre:it tarbon fiber motor casee, and a 372 decreast in weight over V,'vlsr-49 fiUte motorcases.

77

TABLE VI. ORBUS I DIMENSIONS/DISPLACENENTS MASS PROPERTIES

Parameter As Designed Actuals

" iBB, n. 32.87 32.80 to 32.87

LF, in. 9.50 9.48 to 9.53

DCFB a' MEOP, in. 0.204 0.193 to 0.263

DCFD at MEOP, in. 0.144 0.150 to 0.168

DCAB at MEOP, in. 0.264 0.266 to 0.330

O)CAD at MEOP, in. 0.130 0.012 to 0.040

DF, in. 27.21 27.20 to 27.23

DM, in. 27.21 27.19 to 27.26

DCCC a' MEOP, in. 0.140 0.136 to 0.146

DA, in. 27.26 27.21 to 27.27

Envelope in. 523.676 23.490 to 23.577

Insulatcd case weigfit, lb 43.2 (44.8 maximum) 43.23, CV -0.8%

TABLE VII. ()RBUS I tIYDRUBURST SUMMARY

Required PredictedParameter Minimum Failure Actual

Burst pressure. p.-,ygDB-1 CSD development 1326 1546 1526DB-2 CSD development 1363 1589 1751DB-3 CSD development 1363 1589 1642

Fiber ;trnss, kr.,D8-1 CSD development 552 644 632DB-2 eSlD development 579 682 738DB-3 CSO development 579 682 693

Pressure shell, PV/W x 10-6 in.DB-1 CSD development 1.34 1.56 1.biDB-2 CSD development 1.46 1.72 1.89DB-3 CSO development 1.46 1.72 1.81

•.77

ProiBurst

,At Burst Post Burst

7ig. 6. Tyical Orbu3 I Burst Test

20,000 - I - !

15,0001I

-10,000

.2

~5000

[Symbol -Descriptieon

C_) B-2 at 1302 pal0 B-3at 1302 psi

-| - Inner ply at 1302 psi (ARGUS orazed)- Outer ply at 1302 pal (ARGUS cnrzed)

- - . Ii I 10 5 10 15 20 25 30 35 40 45 s0

Arc fltqih, In.

Fly.. 'V1. nt. . il t. U,• I. } ' Fi ber Str • fa Iit's At I 1W)/ 1 i1

""- I i

levier~~~ Bl2 at 1302 psi(RGScrd,W ~ a• t I Mo rM (A R "U c m,, @M 0 B-3 at 1302 Pei

I' -- e-O ,plyat 1302pae(ARIS __ 0

SLPT04 W .- No 8*349

Fooo woo togr ._

IAft u Mm

0 120 5 10 Is 20 25 30 35 40 46

Arc length, In.

Fig. 7b. burnt Unit. Helical Fiber Strains at 1302 psi

r186

'16

14 thIckness

028.760

026.870

41727.000

3 4 [

M1i

P I -'J, (r u l k m ll l l • fn yi l i. l.l il. ill :l llI,.

Description14 [ o AS-A,ul burnlh'ough limnit s

Aa-bult SF . 1 25 limnli

12 I - b A.JIII SF wI5,mu I

* os-i ma::'nuinol 0s-3 meaznmum

ll o 2 A D C q u e n c h m o-C I ----I10 - - A ~~~~~01-4 alfu wst I-

La __-Ice

2 Sell Pot I

UI slot I -,.-u-0--1-.,, -,/

711.I. Orbus I Sagmtonced Mandrol

li . li. (soip ,,i I A.am 14 live,~ 1two~~g a,1 L,,.ja a

t I I'. '' . I~ tod I l .I '. I I j . 9 lC;

Fig. 15. orbus I Y-Joint fthear Ply Compactlon

St g * I * I I' I

I 0

TLmi. 17. ('i~um I C7AlnuJas C4Dipoautivii

plot ~ ji. , I, 0 10 4 11 pt a lles Ha.j I g10161m i ...

Fig. 19. Orbuta I Mandrel Removal USIAd

pgb4, vwm % 6O t I

*4.gte pasesure Sho4 tflecionCy. PY/W, litA i10

Phepfuducojimt ycooffitiom'y W9 viwaIboy b~aso on built pmeurlF sold fit-o ,I64th TO4

WOCghiI v8us81004, 4% OA

116,18, at..

4 " 14eu so III A,(.I ~ 4

"C 'u'. IV, de's .0, of.,111.

PP. ie 64 ,,".o 1 # 6-1 1~I~ "4,.1'i 4i.0h I.~ fi ,In' So '

owl

GLOSSARY

AEDC Arnold Engineering Development Center

CSD Chomical Systems Division

CTE coefficient of thermal expanasio

SPON aliphatic epoxide

RILL epoxy resin liquid

YE finite element

IUhD Independent Research and Development

IUS Inertial Upper Stage

L/D length-co-diameter

MDA aromatic amine hardener

NDT nondestructive testing

NSWC Naval Surface Weapons Center

QA quality assurance

ARGUS stress analysis of three-dimensional solids for motor cases (computer code)

SOTA stets of the art

Orbus Orbital Bois

CDAC composite case design and analysis code

PK asacokesper

F75 full-stale doval;)-mer-

Small I1M Small nt. .t Sen~ol 3e.1 ltic Missile

hi?

JA* AFS I IZA7TON OF TESION, OH, AND MM 1RTM 20 JETM MHANICAL IAIAL MPERw TEr

FOR Fitmw• waUH aoas smiciu

Terry L. VandiverU.S. Army Missile O2miandReds trne Arsenal, Alaba

ABST~AC

Ibis pae.r pamnta tte history and prgress to C.te of the Jolnt-Army-Navy-tMA-Air Force(3ANNI ) ( itae 9to Case Subactnittee (OCS) efforts to develop standard test techniques f-rthep renmnatirn of mechanical material properties of filamTet wouxn composite structures. Thewmk is a joint effort includ'sq both qovernment ard industry j ibere of the 7est and Inspection,Design and Analysis, and er=-msing Panels of the CPM. .bis paper provides a chronologicalseqimnce of e"-erts leaoi-g Lo the selection of the inierim test methods, specimen gew•try,sw&-.Aen fixture oalhution, testing apecifications, deeign of the 1U4D RBOIN Test (RRT),prooesaing spwiticaticns, mamafacture of the test specimens, and manufacture of the test fixtures.This paper also def-ribes the NorAdstructive Testing technique used to determine the quality of theapecImJns, anatys-i of the NDE data, strain gaging of the specL.ens, and plans for statisticalanmlysin of the data.

•IMMJCUION

Ibls effort was started due tn a need for valid mechanical material properties required todisign and analyze filament woud acmpoeite structures. The results fran a government and industryPurvey indicated there was a lack of standardization on the testing of filament wound com•r itestructures. Private industry and goverrnent orgarlzations primarily had to rely on sAnufacturer'sdata that is usually developed wer ideal laboratory condditins. The testing done an filamentwound utxuture oftentbw involves modificatIon of metals testing techniques. WAt of thetesting to determine mactanical material properties, with the exception of ultimate strength, ofomposita structures Is de with flat lamintes which are manufactured and curod using differentprocesses than are usually used with filament w structures.

The behavior of filament wound structures is typically different than the behavior of flatIumonated structures. noted differences result from the type of cure, tesin vid content,microcracking, and free edge constructitn. 1over, filament ound structures require themechanical property data as for the design and analysis of general flat laminated structures. Alarge portion of filment w",& strucaces Is used .n the ortjasite rocket motorcae ommm ity and,oormetmnly, the JPJ~I4A OCS Interim Wat Standards are in the form uf cylinders which moreclosely sim.Oate tJ' str.'.urq-n o b.W dosigned and analyzed. Initially, standards will bedeveloped to dsterclne uniaxia3 tmterlAl properties; hmwser, the owrp ite rccket motoreasermuminity also has a vital interest in th, biaxial and triaxial matoriaJ property resprisee.

("iYJ .er MW uhgI m SUJRMY

P JAMW CLY :"•ve" y w= 7ouc-='d in rlovwtbr 1983 at U, tstj Lvi-iqe uaw6 by qmrnrmtand inductry to deteurm•i, metet tal prtp rtiea of f ilamnt wuund oa"Ite otructires. 'It surveyoonuisited of 35 govarnmnt, idustry, and university WArticiPAntfs. '11W test surfvy asked whattests wt* usae and what. outpt parowbeteru were otCaMifd frcn each tat. `1w cutput desired by thetastr included informatJon tor design and valynin, quality *x;trol, rrreoning, jcocxmingeftect-, vwircromw tal ofteuts, a Otatiwtical dstAbmw, and co.har purpr's. 'tb curvey resulted in17 different taimile taste, 17 com~pcessirn testa, and 1( dJifr•rent shear tents. It became apparenftt•at tere were no sta.dard tqats toing urmid to (tetbrl'Ai. nachfnJicl materil propertirn offilamet vwiund structuras. Hydrciurst tottle testing w a clyw-i taiclnriq, 'joa by the MotorcAa0'anufiwcturero tn datermdrw ultinwtt atronglJo of va"Itt.u In a filnment 'aournd c.xfiguratir,.Ilkvuvsr, this taut wan nt utlizled to deu.ermine wuaslal noterial properties. Am monpanies useda teit cyliru-ri• A'aWi&tLrrj of a atztbination of lelic•l ajim 1fxri Aasera to hydriLurnt, aridarialytcitJ.)y Lwk u.,t the uLr|axia•. wateril proImrtiem. tw JNWIAF (Or-! dveired a jxuror form ofut)t.ai)iln) tim unioxu&l parprrt•ir imi a)*•teV a 90 duegee filament wurm) tube.

ASInl Id I f X I IAi rl-juej 41iti ibuL1wM ui n1 it' UP),

Jr/ I rMnG M Sa WT S(mvM

Mabers of the JRM~ C Test and Irnpection, and Design and Analysis Panels held a jointmeeting in Decmer 1985 in Magna, 1U1 to determine which properties were most needed for the designand analysis of -gwite rocket miotorcases. the test survey was screened test by test, and thosetests were identified that obtained the desired uniaxial material properties. Also at thismeeting, a panel of experts in the filament winding o unmity was selected to guide and makereoonxwdations to the r Working Group. The tests that provided the desired uniaxial mechanicalmaterial properties are sumw ized in Fig. 1. The joint panel organized a JJNW workshop topresent technical papers on each of the techniques selected in Fig. 1. Session chairmen wereselec-ted for tat ion, r-1[ression, and shear testing.

In February 1986, a JAMAP workshop was held in Lancaster, CA on the test methods selected oythe joint panel meeting in Decembr 198! to determine unlaxial xaterial properties. Each sessionchairman asked the presenters to address how tte test technique is cc oould be used to determinethe required outpit parameters. The technical paper presentations included six tension, threeoom:esaion, and six shear test methods. It was requested that the panel of exprts makereoomien ations to the OCS on test methods to determine uniaxial material properties. Ite resultsof the wokshop, including the recounendaticns of the panel of experts, is inc1luied in rTIAPublication 448 £LU.

S•L•EIM OF JWN DMM 729T EMR S

In April 1986, a meeting of the JANNAF IRT Working Group met for the purpose of seiectingJMNNA Interim Test Standards to determine uniaxial material properties for filament woundo•psitet structures. Each Ier on the panel of experts gave their recommexdations withs•uorting rationale. A group discussion was conucted and a vote for the JAWW Interim TestMetds followed. All major filament winders aM government representatives were allowed one voteeach. The selected test types, specimen type, test data required, and output properties is shaonin Fig. 2. AUl of the test types listed in Fig. 2 were redwkd to a single test, with theexoeption of the Zero Degree T7sion Test. There was a tie between a ninety degree filarent woundpressurized tube test and an NOL Ring type test. 'Ibis particular test type was also the focus ofthe group d..csicxn.

GOE* SLYPECR MEETIM

The (W-C Cnairmw pesewted the 1W Working Group's plan of action to develop the test methodsto the JANWF Executive O(mittee with a request for approval in June 1986. This request wasrequired because it was planned to rovide goverrment funding for selected portions of the effort.Awproval of t2e Executive Committee was received in February 1987. The Executive Crvmdttee ofJNW approved the plan and spaIfiei that the OCI would be responsible for providing the fundsrequired to oonduxt the M. In Deoeqrb,.r 1987, an all goerrnowt meeting was held in conjunctionwith the 197 3MNWF Propulsion M96tirg at San Diego, CA to determine the level of funding,in-homwe labor, and use of facilitie that axild be ommittaed by each government participant. Ite

" % ne- In the fonn of rwnndestructive testing efforts to determine the quality ot the testspecimens, and to provide Pah~r4 ahop stport to fabricate test fixtures required for the RJrEffcxt. h~t the same me.ting, it vah dcided to obtain a technical secretariat to assist in theon~vA:t of the FUC. The furaing frcon the gornment agencies wmuld be pooled for use by thetehnical secretariat to oc.dinaLe the 1tW. A four year program and oaumidmet was establishei.The government agencies that agraed to 4 years fundirq ur4'or labor, mid faciliti*5 included ArmyMaterials T:cW, Liboratory (AMI), Watertown, I9! Army Missile Command MICCOM), RedetioneArranal, ALt aM0 Astronautico Lab•o•a)ry, Edards Air 1,rne Bas, CAr; w,, Nval Weapons Center'MC), China [Axe, CA.

.he R1 EMfort 1P being ooordirtntei with MIL-IUIA -l78 (or "Fwymr hitrix CVc tuten,*Americaj rr-:lety frx. Tstetirj ar MeteriAls (AWfl), 24upplieuz n! Advwý'ed Composite MaterialsAMociation, Dspacti•st of rofaise &tb;UardiAtloi Pratr,?. 2'c i)TrV•it.es TS toiogy, t.CUomnunut.icon u, A" Crcr31rnation C nnJ'Ute4 for t!m ýitand)i,,lizjtion of 'est Mettli, and ot.|rcupumites Ural tn darUti.q, C(',;anlIat N, Te 7P,." pflanr of axf.im has bew riefue to each of

these organizations. The JANNAF Interim Test Stfandards are currently included in volume I,MII,-HK-17B, under the Filament Winding Section 6.6.9 (24. The ASTH D30 Committee on HighModulus Fibers and their cmposites has reviewed and provided comments on the JANNAF Interim TestSpecifications.

F40UW TES T KICOFF M1TIZjE

The JXNNAF JMCS WT Kickoff meetc c in oonjunction with the May 198Z @MCS meeting atthe Presidio in San Francisco, CA (31 c• seen in Fig. 2, four out of the six test typesinvolved cryiinders and the remaining t.. .ved flat laminates. The A.137M Test Standard D-3410 oncxpressicn tU-ting was determined adequate to obtain the desired properties for the Zero DegreeCompressicu Te!;t t43. The ASTM has o:nducted a RRT on the Iosipescu Test to determine transverseshear properties. The J4NNAF Wr Working Group recommencds the use of these ASTM Standards withfilament wound flat specimens. The RRT Working Group decided to develop the test types thatrequired ninety 49gree filament wound cylinders which included: (1) Zero Degree Tension;(2) Ninety Degree lxV-wian; (3) Ninety Degree Compression; and (4) In-Plane Shear. A blockdiagram of the entire JANAP PRT process from the kickoff through final reommendatlon for finalspecifications is presented int Figure 3.

Asisignments were given to each participating panel of the JARNAF CWS. The Test and InspectionPanej wa- as•signed the task of developing generic test specifications for each desired test. TheDesiqn and Analysis Panel was assigned the task of making recomendations for specimen geometry aidtest fixture design based on analysis in such areas as specimen wall thickness to diameter ratio,erd effects, and loading mciditieris. -,he Processing Panel was assigned the task of preparing ageceric processing specification to be used in fabrication of the test specimens. All assignmetswere to be completed and the results presented at the December 1988 RW Workshop in Huntsville, AL.

SEETION OF WT 1U3WICAL S EaRIAT

Materials Sciences Corporation (MSC) was selected as the JANNAr dr Technical Secretariat inAugust 1988. The MSC is also the Technical Secretariat for MIL-HM(-17B on "Polymer MatrixCcmposites" on an AMWL contract. The contract was modified to aceept the additional governmentfunding. The M3C had been active in the Filament Winding Working Group of MII,-HDBK-17B for twoprevious yearn. During this period, the primary focus of the Filament Winding Working Group wascordination of activities of MIL-HDBK-17B and the JANNLAF OC•S. The results of the JANNAF RRT willbe published in MIL-HLEK-17B as well as in CPIA JNNAF dzcuments. The MC is responsible forcoordination aimcg the various government and industry agencies throqv7'Kot the M process. TheM!9C is also responsible for interfacing with other ,wrpites technical groups such as ASTM,MIL-HBK-17B, and SNCMA. The MSC statistically designed the RRT in accordance with ASTM E691 (5]and will provide statirtical analysis of all data fram the F41T results.

R WCPKSHOP ON TEST ME'IWX) SPBCIFICATIONS

A JANN4F Rr TWorkshop was held at MICUIM. in Huntsville, AL in December 1988 (63. As a portior.of the Design and Analysis Panel Effort, government funding was provided to issue competitivecantracts to conduct an analysis of the JANNAF interim test specimens. Issues to be investigatedincluded specimen geometry, thickness, edge effects, loading conditions, tabs, etc. Each of theparticipating contractors were provide] a generic set of dita tc toe in the analyses of thespecimens and test fixturc-s. The results of the analyses "nd o. juny recommendaticns werepresented at this meeting. The pa-ticipating companies were: (1) Atlantic Research Corporation,Gainsville, VA; (2) Brunswick Corporation, Defense Division, Lincwoln, NEI (3) Hercules AerospeceCormany, Magna, LuT; and (4) Molrton Thiokol, Brigham City, UT.

At this meeting, a co•rensus was reached on three of the four tests: (I) Ninety DegreeTension; (2) Ninety Degree Conpression; and (3) In-Plarv. Shear. There were oppsing views as towhether the desired properties for Zero Degree Tension oculd be obtained from an NOL Ring type or a90 degre& filament wound pressurized tube test. It wvA decided that the Design and Analysis Panelwould resolve the issues, and reavorO a tebt methxd at a workshop to be held in May 1989. Draftsof test specificaticie, processing specifications, specimnw geaittry, processing fixtures, and testfixtures were establi]hed for the Ninety Degree Teneson, Ninety Degree Compression, and In-PlareShear tests. Figure 4 is a sketch of the M test tpeciman which includes geo•etry and materia]selection. Figure 5 is a sketch of the test fixture that will be used for the three differenttests. As oort be seen fram the sketch, the Ninety Degree 7wiion and Ninety Degree Coaixession

91

Teits use tne basic fixture, and the In-Plane Shear Test requires attachment of the torsionadapter.

DEIGN OF RrT 1ET PIAN

The JANOkF Ma Test Plan was designed by Materials Sciences Corporation as per ASTM E691Guidelines C53. Each test method will have six specimens, each from three diffei'ent batches, for atotal of 18 specimens. Each test method, consisting of 18 specimens, will be tested by sixdifferent testers. This results in 108 test data sets for each method. The six different testersinclude fair industry testers and two government labs. The industry testers were selected based oncompetitive bid. Figure 6 lists the participating testers in the order they will be testing.

WORMXSPS QN ZFO D M4SICN T

A 0-Degree Tersion Test Workshop was held in May 1989 to determine whether the NOL Ring or a90 degree pressurized tube should be selected for the JANNAF Zero Degree Tension Test. Sufficientinformation was not available at this meeting to make a decision an the Zero Degree Teroion Test.The fo•r contractors involved in the analysis WrR presented plans of action to aid in the decisionas to which test provides the desired properties.

Atlantic Research Corp. agreed to perform analysis on a 4-inch diameter, ninety degree filamentwound tube and design a fixture to perform the test. Brunswick Defense Corp. agreed to manufacturenine filament wound tubes oorxiisting of three sets of three different configuration 90 degreefilament wound cylinders. One set would be manufactured with all 90 degree windings, one set withVeilmat reinforcement, and one set with some 0 degree fibers in the cylinder wall. Hercules, Inc.agreed to investigate the NOL Ring to include a technique to instrument the specimen transverse tothe fioers, and a technique of obtaining the total tow length of the fibers in the test specimen.Thiokol, Inc. agreed to filament wind sone flat laminates and perform testing to obtain the desiredproperties. The results of the test efforts were distributed by MSC among all participants at theZero Degret! Tension Test at the November 1989 OCS and Structures and Mechanical BehaviorSubcommittee (9M&BS) Joint Meeting in Pasadena, CA. If a decision was not reached at this meetingon the Zero Degree Tension Test, scheduling would not allow inclusion in the current JANNAF FRTEffort.

Each oontractor presented their findings at the November 1989 (SM&BS) 0-Oegree Tension TestWorkshop. Each company had a different approach, but not one fully convinced the members of theC2.CS r&r Working Group to select tlhe NOL Ring type test or a 90 degree filament wound tube. Workin both test areas looked promising, but did not appear on solid enough ground to be included inthe RRr. The NOL Ring and the 90 degree filament wound cylinder were originally favored becausethe geometry more closely simulated a rocket motorcase than a flat laminate. Hoawever, until theproblems can be worked out with the two test techniques, the wr Working Group adopted the ASFMD-3039 Standard Test Method for Tensile Properties of Fiber-Resin Composites (74 with therecommendation that the test samples be filament wound.

MANUFAIURE CF rT SFZIMENS

A ocsetitive bid contract was utilized for manufacturing of the cylindrical test specimens tobe used in the JANNAF RRr Effort. Brunswick Corp., Defense Division, Lincoln, NE, was the winnerof the contract and manufactured 360 ninety degree cartom filament wound specimens as per theprocessing specification developed by the JANNAF N&T Working Group and M9C. The 360 test specimenswill be sufficient for each tester to perform the three selected tests with 18 specimens per test.Six extra specimens have been given to each tester.

MAMWAYRE OP RT TEST FIXTURES

Dimensioning and Lolerancing of the JANNAF iWT test fixtures were reviewed by NIC, MIOOM, andM9C. Detailed drawings were prepared by NWC and M9C. A total of 24 test Zixtures was manufacturedby three different government facilities. The NWC was responsible for 12 sets of fixtures, AMT.was responsible for 6 sets of fixtures, and MIOOM was responsible for six sets. The fixtures weredesigned so that they could be interchanged for all three tests. Each of the six testers will haveto perform their 54 tests and pass on the test fixtures to the next tester. The VSC is responsiblefor the test specimen and test fixture coordination, and shipping among the six testers. The NWCwas the first tester in the RRT and mixed test fixture parts for eech of the tests to assure thatall fixture parts were interchangeable.

92

NDE OF RRT SPECVINS

Nondestructive evaluation of the 360 filament wound carbon JANNAF •r specimens was performedby AMTL. The AMI'L and MIOOM worked jointly on the manufacturing of a test standard to inspect thecylinders. Standards, which oinsisted of cxmposite test tubes with known flaws, were manufacturedat MICOM with the assistance of AMTL personnel. The MIOCM used the sane carbon fiter, resinsystem, and winding tensions as used by Brunswick Corp. when manufacturing the 360 RRT 3pecimens.The flaws varied in size and material, and were placed at different levels through the wall1-hickness of the cylinders. Figure 7 is a photograph of the filament winding of the teststandards. A series of circular flaws are being placed in the wall thickiiess of the compoisiteetandard in this photograph.

The AMTL used the standard test cylinders to calibrate their NOE equipment. Ultrasonic DefectC-Scanning was the technique used by AMTL to inspect the cylinders. Figure 8 is a photograph ofthe siginal intensity output of one of the test standards with known flaws. Figure 9 is anultrasonic C-Scan of a typical RWT cylinder. The AMTL is preparing a Standard Operating Procedurefor Ultrasonic C-Scanning for filament wound composite cylinders.

ANALYSIS OF NDE RESLTS

An analysis of the NDE data was performed by MSC. The NDE tests were performed to deteminethe extent of manufacturing variability in the test specimens. Isolating the manufacturingcomponent of the vriability will increase the accuracy of the random error variability estimate.The procedure used xo isolate the variability associated with the manufacturing process requiresthe quantification of the C-Scan results. This quantification is performed using the signalintensity scale that accomplishes each C-Scan as can be seen in Figures 8 and 9. This scalerepresents the received signal intensities with colors that document different intensities. Inaddition, the scale includes the percentage of each range that appears on a scanned specimen. TheMSC used the percentages to quantify the relative quality of each specimen.

An Analysis of Variance (ANOVA) was also used by MSC to statistically examine the C-Scan data.This procedure helps identify the variation in the C-Scan results that can be attributed to randomerror and the variatico, that can be attributed to specimen quality. It is not known how specimenquality will effect the results of the tests performed on the specimens. Therefore, attempts willbe made to correlate deviations in test results with specimen quality using ANOVA.

STRAIN GAGING OF WT SPECIENS

It was decided by the JANNIFA RRT Working Group that the instrumentation should be applied by asingle source in order to maintain a higher degree of quality control. Strain gaging of the360 JANNAF RRT specimens was determined by oxompetitive bid. Tte contract was won by DynamicEnginr-ring, Inc. of Newport News, VA. The Ninety Degree Tension and Ninety Degree Coipressionspecimens were instrumented with strain gages centered with respect to the length of the cylindersand spaced circumferentially 120 degrees apart. The In-Plane Shear specimen required two straingages at 180 degrees apart and located at center length of the specimen. All strain gages wereMicro-Measurements 350 ohn 3 gage rosettes. Figure 10 is a photograph of a strain gaged In-Pla-eShear specimen.

RAN4OEIZATICN OF THE MT TEST SPIMED2S

Upon oompleticn of the strain gaging, the specimens were sent to M9C to perform a statisticalrandomization before being s-nt to the testing participants in the JANNAF PRT. The ran-lomizationproes was performed in accordance with both AS1M E691, "Practice for Conducting anfnterlaratory Test Program to Determine the Precision of Test Methods," and MIL-HDBK-17B on"Polymer Matrix CaTposites." The randomization process was reviewed by the Statistics WorkingGroup of MIL-HDBK-17B before being implemented.

DATFOtM DISCJSSICN

The Technical Secretariat and t'SC provided DA[TF4M to all participants in the RRT. A computerprogram, DATF(I4M is designed to format data for transfe: between laboratories by producing astandard output file. The computer program, DATFCR4, is maintained by MSC as an agent for AMI¶.Tie DATFURM onists of two modules: (1) The Materials Descriptors Module and (2) The Test

93

Descriptors Module. The Materials Descriptors Module is used to format information pertaining tothe materials and mnufacturing prooedures wued in the. preparation of specimns for a testingprogrm. The Test Descriptors Module serves the ftnctio of formatting information about theprooedures used and the results of a testing program. Upon completion of testing, the RRTparticipants will provide the IYF4 information to M9C for data analysis.

ROND REIN TEST

The actual JANNAW WT began in March 1990. Eac of the six testers will be allowedapproximtely six weeks for conducting 18 tests on three different test methods for a total of54 tests. The MW is responsible for the ouordination to include test specimen and fixtureshipping to the six testers. Timely testing is critical to the success of the testing effortbemuse it is necessary to share the 24 test fixtures • the six testers. The M9C has sent testspecimens and OIR to each of the testers so that they can enter the physical data on each ofthe test specimens. As the D&I•h 4 data files 1eo1-e available, the MW will begin data analysis.At the time of this update, tOC has completed testing and the fixtures are in route to AMRL. Thecompletion of the actual testing portion of the JA.NF Fr is scheduled for January 1991. The MSCplans to have the last two testers test in tandem. Approximately 9 months is scheduled for MSC toperform the analysis and coordinate the results with AS2I, MIL-HIFK-7B, and the JANNAF •r WorkingGrOup. 11e final recoenwidations for test standards is scheduled for the last quarter in FY 91.

CONCLUSION

The history and progress, to date, of the JkNWAF CMS I•T Effort to develop test standards todetermine mechanical material properties for filaiment wc1 c~z-4(xisites has been presented. Theschedule is aggressive and cmpletiti depends an the cooperation and coordination of numerousorganizations in government and private induztry. The effort by the JA14AF CCS will providestandardized test methods for government and industry to obtain mechanical material property datato use in the design and analysis of filanent wound cmpsites.

1. "Test Methods for the Mechanical Character ization of Filament Wound Cmfosites," Lancaster, CA,Chemical Propulsion Information Agency Publication 448, Johns Hopkins University, Applied PhysicsLab, Laurel, MD, February 1986.

2. KUM-HI(K-17B, *Military Handbook., Polymer Matrix CA'%xiites, Volume I: Guidelines," NavalPublications, Philadelphia, PA, February 1988.

3. "1988 Composite Motorcase Submittee Meeting," Presidio of San Francisco, CA, ChemicalPropulsion Information Agency Publication 489, Johns Hopkins University, Applied PhysJcs Lab,Laurel, MD, May 1988.

4. American Society of Testing and Materials D3410, "Compressive Properties of Unidirectional orCrossply Fiber-Resin Czcmposites," AS'IM Annual Book of Standards, 1987.

5. American Society of Testing and Materials E69i, "Practice for Cnnducting an InterlatoratoryTest Program to Determlin Precision of Test Methods," Volume 8.03, 1979.

6. 0 pite Motorcase Subkointtee Rouxnd Robin Test Workshop," Huntsville, AL, ChemicalPropulsion Information Agency Publication 524, Johns Hopkins University, Applied Physics Lab,Laurel, MD, Deomiber 1988.

7. American Society of Testing and Materials D3039, "Starnard Test Method for Tensile Propertiesof Fiber-Resin Cci•qxites," ABIM Anual Book of Standards, 1987.

94

Test Specimen Type Test Data Properties

0O Tension Tests

Flat Laminate (00 Layup) oil. ElI Ej1 . 2Elongated Ring qI, E1l E11 . Ju2

Pressurized NOL Ring Oil, Ell Ell JI12

Prepsurized Tube (90o wound) a,,, ell El I .u12

901 Tension Tests

Flat Laminate (90" Layup) Ok22, E22 E72. u2,

Tube (90' Wound) %. -' 22 E22. JJ2 1

0 Compression

External Pressurized Tube 0i I11 E 1, I . 12

Flat Laminate ( 0O Layup) I. Ell ElI. jJ12

90' Compression

Flat Laminate (90" Layup) 22, 2 E22. M21

Tube (90* Wound) '222 E22, Y21In-Plane Shear Tests

t459 Layup Tension Laminate ""12, 1'12 G12±45" Helical Filament Wound Tube 7-12, Y1'2 G12

900 Filament Wound Thin Tube 7'12, Y12 G12Rod (Torsion) - -G12

losipescu "'12, Y12 G12

Asymmetric 4-Point Bend T"12- Y12 G12

4-Point Ring Twist - -G12

Transverse Shear

losipescu ?23- Y23 G23

Asymmetric 4-Point Bend 7-23, Y23 G23Torsion Rod - -23

+451 Compression 1'73- Y23 G23

FIGURE 1. VM4RY OF VEM REULTS

95

Test Type Specimen Type Test Otoa Properties

0* Tension NOL Ring

ASTM D 2291

Pressurized Tube 0j I , 1 * E IU 2

90g Tension Tube (22 E2 2 E.22. -02,

0 Compression Flat Laminate

ASTM 0 3410 Oi I1 , E1 1 E,1 1 , Jul 2

900 Compression Tube C22' E22 E7.2 JJ21

In-Plane Shear Torsion Tube "1i? 2 ' 1 2 G12Transverse Shear Flat Laminate

losipescu ?23- Y23 (;23

3

00a00

000000 0

1~

A) FLAT LAMINATE PROPERTIES ORIENTATION

3

12

8) CYLINDRICAL PROPERTIES ORIENTATION

FIGURE 2. JANNAF INTMIN D4 METH W

Phase I

Phave III

FIGURYE 3. JANNAF ROLM) ROBIN 1 WT MU..flNE

97

MATERIAL:

AMOCO T650/42 FIBERFIBER VOLUME - 60±3%WET FILAMENT WINDING PROCESSVOID CONTENT < 5%12 PLIES OR 6 LAYERSBAND ADVANCE - 0.106FIBER TENSION - 4±0.5 J/TOWFIBER WIND ANGLE - 90

MANUFACTURER:

BRUNSWICK DEFENSE, LINCOLN. NE4. 4 I.D.

~0.0804

5.5"

FIOUZt 4. JAN ? ROUhD R' TEST • spEC

98

TORSION ADAPTER

SPHERICAL WASHERSFOR ALIGNMENT

04.00 GAGE LENGTH 04.00

EA934NA OR EQUALPOTTING MATERIAL

FIGURE 5. JANNAF RI•AND ROBIN TEST FIXUR

99V

1. NAVAL WEAPONS CENTERChina Lake, CA

2. ARMY MATERIALS TECHNOLOGY LABWatertown, MA

3. ADVANCED COMPOSITE TECHNOLOGIES, INC.Sparks, NV

4. HERCULES INC.Salt Lake City, UT

5. THIOKOL INC.Brigham City, UT

6. BRUNSWICK DEFENSELincoln, NE

.•.,•, *. *?:• .e•e,*1** * te s

20..

IA 9.9-

67-5 39.6

P5.2 34P5

- ,39.1 3:.G

13.3 ?¶

Figure 9. Utirasonic C-Scan of a Typicca JANMAF RR Specimen

(IM

-g~ura i0. 53traiin Gage& tr-ý-Pjanr, Shgnr t3i

102

SOLID ROCKET PROPULSION APPLICATIONSFOR ADVANCED POLYMERS

JAMES S.B.CHEW and JOHN RUSEKAstronautics Laboratory (AFSG)Edwards AFB, CA 93523-5000

ABSTRACT

The demands to lower the cost and increase the reliability of solid rocket propulsion systems have forcedthe aerospace industry to investigate new materials and fabrication techniques. The performance specifications forstiffness, strength and temperature resistance make the process of finding better materials and processes quitechallenging. Advanced structural plastics, such as liquid crystalline polymers, may meet these requirements. TheAstronautics Laboratory (AL) has initiated an in-house program to determine if advanced polymers could meet thecosvperformancelreliability requirements. Components identified for polymer appiicalion includes motor cases,ignitor housings and nozzles. Injection molded parts, fabricated from VECTRA(R), XYDAR(R) and RYTON(R) weredesigned and initial testing has started. In addition, compression molded CELAZOLE(R) was also evaluated. Anoverview of this program and the progress to date is presented.

INTROUICTIJO

The design impetus for most solid rocket propulsion components is performance. Performance criteriadrove the industry to develop extremely strong and lightweight components. When the inert component weight isreduced, propellant weight can be increased thus improving overall performance. For rocket motor cases, DCA6steel and graphite epoxy were used because they yielded the desired higher strengths and lower weights, however,material and fabrication costs remain high.

The complex nature of the DCA6 steel and graphite-epoxy fabrication processes can lead to high componentrejection rate or components which vary significantly in terms of mechanical properties and structural integrity.By comparison, the bulk cost of liquid crystalline polymers is quite low. Material properties can be tailored to theapplication by heat treatment or mold designs. A variety of low cost fabrication techniques including injectionmolding, pultrusion, compression molding and resin transfer molding can be used to make net shape parts in one step.

This paper presents a brief description of liquid crystalline behavior, the rocket propulsion applicationsidentified for these materials and a summary of relevant progress; these programs are combined Air Force effortsbetween the Air Force Logistics Command at McClellan and Hill AFB, the Air Force Armament Laboratory (AFATL),the Air Force Institute of Technology (AFIT) and the AL.

Liquid crystalline polymers (LCP's) can be subdivided into two classes; thermotropes and lyotropes.Thermotropic LCP's exhibit liquid crystalline behavior in the melt, while lyotropic LCP's are liquid crystalline insolution. Examples of thermotropes include VECTRA (R), HX4000(R) and GRANLAR(R), all of which exhibit anisotropic/nematic phase tr.' ,3ition above 300 C. Typical lyotropes include KEVLAR (R) and polybenzthiazoles.polybenzoxamides and polybj.izimidazoles. KEVLAR (R) exhibits liquid crystalline behavior in sulfuric acid while theother three polymers are drawn from polyphosphoric acid, where they behave as nematic LCP's.

In general, the molecular architecture is the prime determinant in liquid crystalline behavior. LCP's arerigid rod polymers, usually polyesters which have a molecular "aspect ratio" of 30:1. This implies an average

Approved for public release; distribution :1- unlimited

103

length of 90 A and an average degree of polymerization of 10. Lyotropes generally have many barriers to rotationdue to molecular gcometry. steric barriers and buried polar moeities. Polybenzoxazole, a typical lyotrope, isdepicted in Figure 1.

FIGURE 1. LYOTROPIC STRUCTURE

Thermotropes generally contain large pendant groups, no buried polar species, and are more free to rotatearound the backbone centerline. A typical thermoropa is depicted in Figure 2.

/0 1

/ n

FIGURE 2. THERMO TROPIC STRUCTURE

An interesting phenomena that has been observed in recent years within the class of thermotropes isannealing. A given theimotrope is injection molded into a cylindrically symmetric part as an unfilled resin. This partis then subjected to a long durat!on temperature cycle after which the part will behave as a thermoset. Thisphenomena is not well understood and i. under intense invesligation by the Air Force laboratories. Materials whichv neal° clearly address the application area for high temperature solid rocket case and nozzle materials as well as

liquid rocket engine component materials.

APPLICATIONS IDANTIFSC" TON

The purpose of the Advanced Polymer Component (APC) program is to utilize the benefits of these polymersin the development of rocket components. Figure 3 presertt. the interrelationship between this program to other ALin-house research programs. We feel that the application potentia' for those polymers is enormous.

T.iis paper focuses primarily on solid propu.:ion motor applications . Figure 4 presents a generic solidrocket motor. The inert components, such as the motor case and nozzle make up the majority of the total weight andcoat of the motor. When the case akirti and inteTrtages are added, the woight and cost of tho inert motorcomponents Increases. When these components are made from aovanced composites such as graphile-epoxy, the

!i 4

NASPTECHNOLOGY

.*ccPlastic

'Combustor.CWc

-Plastic

FIGURE 3. ASTRONAUTICS LABORATORY PROPULSION IN.HOUSE INITIATIVE RELATIONSHIP

labor for fabrica•tion represent the majority of the costs. By using the LCPs low cost fabrication techniques, such asinjection molding, these costs are greatly reduced.

Gtlarl W ItSNI lI'flt/ hlllnhl.Ol C...• re41 ea f~I" cavi"y) {" W, (1y--,) ("• o ,

FIGURE 4. GENERIC SOLID ROCKET MOTOR SCHEMATIC

Figure 5 compares the LCPs strength vs. stiffnec.s relative to other care materials. As shown in the figure,LCPs meet and: in some cases exceed current state of the art case materials. For this reason, a majority of the solidrocket work using LCPs concentrate on motor case applications.

TEST ARTICLES

Because the AL is a relative newcomer in working with thermally processed polymers, the most logical planwas to initially develop cylindrical motor cases. Depending on the success of these, more complex polymeric caoeswould be developed.

06-0 , •, I

10-I-0RAI

ARAM103 I H STF ENGTHC

W to 8 --- 4/ -

S; HIGHMO LUSoS

/ X6ýORQERED.4_P O L4ER

UZ n. ,r ', . - - -

0 V T3ojS L/QE

TEELE RLI1T~!' ALUMIN11'M

2 6 8 10 12SPI CiýiC W.ODUWL,'S

(h.'J. X 10--8)

FIGURE S. MATERIAL STRENGTH VS. STIFFN&SS COMPARISON

Figures 6-9 show the current AL motors case configurations. Figure 6 I, a 2x4 motor, used for obtainingpropellant ballistic property data. The operating conditions are listed with the figure. These cases are compressionmolded or injection molded from the various aforementioned LCPs.

ii _ _, ,: - -

MAX CHAMBER PRESSURE: 2000 PSII -- I , (HIGH PRESSURE TESTING, 12,000 PSI

7 ,MAX CHAMBER TEMP: 3700 K, " MAX BURN TIME: 1 SEC

___ -.-. ...-, -_. 1 _

FIGURE 6. 2X4 TEST MOTOR

Figure 7 shows the Air Force Academy Motor and it's operating conditicns. Notice the in-.reasing complexityof the motor cases. The LCP design for the Academy Motor case would be a two-part design vs. the current throepart design. For this motor case, the head-end closure and the motor case would be incorporated into one piece.

~ I

,. CHAMBER PRESSURE: 1000 PSICHAMBFH TEMPERATURE: 3300 KS' '" - 7 M A X V E RfT IC A L LO A D S : 15 ( S

iU : ------7[-.r---- ! so.BURN TIME: 2.1 SEC

I • . ... . .. . i. . . . . . . . ,. . . ,I '

FIGURE 7. AIR FORCE ACADEMY MOTOR

L 06

DFLSIC, N APPLICA TION.¶

Figures 8 and 9 present more ambitious potential applications for these materials. The AIM-gL Sidewindershort range air-to iir missile is shown in Figure 8. The design of this motor case presents many challenges. Notonly does the rlotor c;.se have to survive the flight environment (up to Mach 8 with a 35 g turn capability), but ithas to survive captive carry and handling loads. The wing tip location of an F-16 was selected as the designcondition (Fig. 9). Thermal conditions of -45 F to 145 F will be considered, as well as the associated transmittedmoments and forces. The AL will attempt to design the launch lugs *into" the motor case. This means eliminating thelaunch lug "bands" that are currently used on the AIM-gs. The information gained from this one change has thepotential of producing a new generation of short range air-to-air missile designs.

ton--.... X'"_._.J_ . 5 O.D., 71" LONG (MOTOR)

"""" - •7' -. " U.,.'L-! •-45F TO 145F TEMP SURVIVABILITY,,'*, A;,, *,,, -;o AEROHEAT UP TO 800F

FIGURE 8. SIDEWINDER MISSILE

FIGURE 9. F-16 STORES LOCATION

Thu most attractive feature of the LCPs for tactical motor applications is the potential "insensitivemunitions" application. The mechanical properties of these materials degrade at significantly lower temperaturesthan current tactical mto!or c.ase materials. At 1000 F, the mechanical properties of the annealing LCPs degrade to apoint that a motor case made from these materials will lose significant structural integrity. This attribute, coupledwith the LCP's inherent high strength and stiffness and low cost fabrication methods, have a high potential of beingan ideal material for tactical motor case applications.

Th'- Astronautics Laboratory has recently started a joint motor case design program with the Air Forcelnstttute of Tecnnology, at Wright-Patterson Air Force Dase. A group of Master's candidates will be designing ashort range air-to-air motor case usinlg LCP materials. The design requirements are for the wing tip of an F-16. Thesurvivability requirements for this rrotor case are the same as for the current systems.

A,. shown in Figure 3, the AL is currently developing, c•emonstrating. and integrating, advanced low costsolid propulsion technologies for a new generation intercontinental ballistic missile (ICBM). A first look at applyingLCPs to ICBMs produced several ideas (Fig. 10). Possibly the most promising areas are the polar boss. ignitorhousing and interstage. The most interesting application is for external protection of the various stages. The goal isto demonstrate that tow-cost molded sheets of LCPs are viable materials for the debris impact and external thermalproiection '•nvironments.

FIGURE 10. INTERCONTINENTAL BALLISTIC MISSILE MOTOR

PROCESSING CONSIDERATIONS

Polymeric components have been fabricated by a variety of low-cost methods; these methods have beenextensively developed by the toy and automotive industry. American automakers were surveyed on polymercomponents fabrication and design techniques. The automakers requirements for low cost structural composites arequite similar to the aerospace Industry's desires, with one exception: they are interested in a fabrication processthat can produce consistent quality parts very quickly ( on the order minutes). We are also using as a resource anAir Force Armament Laboratory (AFATL) contracted effort with McDonnell Douglas in St. Louis to investigate avariety of low cost composite fabrication methods for tactical weapon applications.

We are focused on two fabrication processes, injection molding and structural resin transfer molding. Mostof the polymer component fabrication techniques, including compression molding and pultrusion have beeninvestigated; it was assessed that injection and resin transfer molding held the most promise for fabricating partsto meet AL requirements. We are teaming with the Air Force Logistics Command to use their expertise in theseprocesses to assure the timeliness of this program.

PROGRESS TO DATE

Our approach to date encompassed simple feasibility demonstrations and design data generation. The 2x4motor case and the Air Force Academy motor nozzle were targots of opportunity and it is necessary to ascertainthe mechanical properties of the various polymers over wide conditions to establish a data base for design purposes.

Figure 11 presents the molded tensile strength specimens that were evaluated at the AL. These articlesyielded three tensile test specimens in the longitudinal direction and three in the transverse direction. Thesadirections refer to the polymer molecular orientation, which is determined by the direction of the polymer flow inthe mold.

Tensile specimens were molded with VECTRA A625, VECTRA C130 and RYTON. The nomenclature on theVECTRA(R) refers to the type of filler (letters signify polymer type; numbers signify type and concentration offiller). Advertised tensile strengths of these materials are 24,000 psi, 31,000 psi and 12,000 psi, respectively.Fabrication of the tensile specimens permitted exploration of injection molding methods and mold design. Theresultanta specimens were used to establish tensile lest procedures and provide an early indication of materialproperties.

108

FIGURE 11. POLYMERIC TENSILE STR ". T.' TST PARTS

The scrap pieces from the tensile specimens were used for propellant bonding tests. Eight samples of eachmaterial were used for this testing. Uncured propellant (listed in Table 1) was applied to the surface of thesesamples. The propellant was then cured at 140 F for one week. For these tests, the propellant was peeled from thesamples after cure to qualitatively assess the bending behavior A good bond was deemed if the propellant failedcohesively. A bad bond was when the propellant neatly peeled from the polymer substrate.

TABLE 1.AIR FORCE ACADEMY LOW ALUMINUM PROPELLANT FORMULATION

R-45 M 9.5% 0M 2.00%Al 3.0% AP(400/200/50/10) 83.0%DOI 2.24% TEPANOL 0.15%

The surfaces of the samples were treated in the following manner:

- no surface treatment- surface roughened- surface washed with N-100 Ieocyanate- surface roughed and washed with N-100 isocyanate

The majority of the test samples yielded good results. Only the samples which did not have any surfacetreatment yielded bad bonds. The results of this testing added confidence at this critical phase of the program, whichwas needed prior to casting the 2x4 motor cases to be molded from the LCP materials. It was felt that the propellantcould now be cast into these cases with a minimum amount of surface treatment.

Figure 12 shows the variety of 2x4 motor cases machined or molded by the AL. On the left is a standard2x4 metal case, followed by one molded from VECTRA C130, one molded from VECTRA A625, and finally one thatwas machined from a billet of polybenzimiazole, known commercially as CELAZOLE. This material, being a lyotrope,has to be compression molded and machined to form parts. For our application, existing tube stock was bought andthe interior machined to the required dimensions.

The wall thi, kness of the CELAZOLE 2x4 was .250 inches, while the wall thicknesses of the molded partswere .125 inches. The CELAZOLE part was designed to withstand an internal pressure of 7000 psi. The molded 2x4swere designed for the 'low pressure* 2x4 test conditions of 2300 psi. The CELAZOLE motor survived a chamber of

109

pressure of 1800 psi for 1 second. The molded 2x4 cases failed at chamber pressures of 1000 psi. Analyzing thesetest results and comparing them with our in-house tensile results revealed almost a factor of two discrepancy inmechanical properties between the published values and the tested values. For this reason, overdesigning by 65% isa good rule of thumb for first cut designs.

FIGURE 12. 2X4 MOTOR CASES

Nozzle abalation tests were performed using CELAZOLE nozzles. The Air Force Academy motors exposedthese nozzles to 650-800 psi for 4.5 seconds. The recessive behavior of this polymer is similar to graphite;0.050' nozzle enlargement was noted. This material is stronger than graphite; the CELAZOLE material seems to holdsome promise as a tactical motor nozzle material.

SUMMAR Y/CONCLUSlON

The test results to date have shown that the Liquid Crystalline Polymers have great potential for solidrocket applications. Initial testing shows that there are some design techniques which need to be used to realize thefull potential of these materials; these techniques are being learned with each test design exercise. Industry andgovernment agency comments on this program are strongly encouraged.

The amount and quality of the work presented in this paper could not be possible without the help of thefollowing people:

Chris Frank, Rich Griffen, Heiu Nguyen, Pete Huisveld, Jim Trout, Shirl Breitling, Janet Shelley, Tom Duffy, DaveRobinson, Jason Baird, the AL LCP team and the Automotive Composite Consortium.

This paper would not have been possible without their significant contributions.

110

SOLID ROCKET BOOSTER INTEGRATIONWITH THE AQM-37C MISSILE TARPGET

F. M. CumboNaval Ordnance StationIndian Head, Maryland

ABSTRACT

"Thi paper presen the work performed to imegwe a solid rocket booier (SRB) with the AQM-37C Miusile Target. The SRB petformaumand interface requirements were derved based on AQM-37C operational requirements. Tradeo's of propellant and imulation/atress reliefsysam were made to define a quality design that meets system safety and performance requiremens. Boosa materals ,ad manufacturingprocesses were carefully selectd in an e to minimize cost. Propellant and igniter characterization tests were performed and rocketmotor case test articles mamnfactured for future testing.

INTRODUCTION

The AQM-37C Missile Target is an air-laumched, supersonic, guided, unrecoverable aerial target system employed by the Navy forweapon systems evaluation and fler training exercises (Fig. I). be AQM-37C flies two different mission profiles: Mach 3 and 80,000 ftand Mach 4 and 100.0D0 ft. 11,! AQA-37C is Lainched subsonically at Mach 0.8 and 36,000 ft from A-4 and A-7 aircraft and supersonicallyat Mach 1.5 and 50.000 ft from F-4 aircraft. The AQM-37C employs a liquid bipropellant rocket engine which operates in a boost phaseto bring the target so the desired operating conditions and a sustain phase to maintain the target at those conditions. Launch from theF-4 is required for all Mch 4. O0,0000-ft missions and is preferred for the Mach 3 missions because of the greater range obtained whenlaunched supersonically.

FIGURE I. AQM-37C TARGET WITH SOLID ROCKET BOOSTER

Obsolescence of F-4 aircraft will leave the AQM-37C without a Navy-operated supersonic launch platform. Options to maintain theMach 4 mission capability and the current range of the Mach 3 missions include the following:

(I) Perform thp necessary aircraft and target system modifications required to obtain F-14 launch certification(2) Contract for F-4 launch services(3) Integrate a solid rocket booster.

Integrating a solid rocket booster (SRB) is the preferred option since it could he used with either of the other t.wa options and significantlyincrease the range of the target system. This is very desirable because of the current trend toward.s long-range standoff weapon systems.This paper dicusses SRB design requiremeit. the SRB design, and supporting component characterization tests.

Approved for pubic release; distribution is unlimited.

III

BACKGROUND

This projecs wu conducted from fucal years 1988 to 1990 in a series of studies leading to design definition.

DEOIGN S DY

A SRB design study was performed by the Naval Ordnance Station, Indian Head, MD (NAVORDSTA), and the Naval Air DevelopmentCeter, Wrainnta, PA (NADC). fron February to Mardt 1988. The investigation evaluated SRB/AQM-37C/aindft interface and launchaoditions to derive SRB requiremaeis. The JuD Rocket Motor M-k 7 Mod 4 was identified as a candidate for integration with the AQM-37C.

INTEGRATION STUDY

NADC coordinated the intepation study between NAVORDSTA and Beech Aircraft Company (BAC), the AQM-37C manufacturer.The study was performed from September 198 to June 1989. BAC evaluated the Mk 7 Mod 4 jato design and determined it was notsuitable for integration. BAC continued the study by evaluating booster installation, launch compatibility, structures load, weight andbalance, aerodynamics, and flight controls to define SRB requirements.

Evaluation of the resulting SRI requirements did not identify any technological roadblocks that would preclude SRB integration withthe AQM-37C. A search for off-the-shelf SRBs was conducted and the Lapwing Rocket Motor, u . on the British Sea Petrel AerialTarpe System, was identified as a possible candidate based on ballistic performance. Though LiAp ,lg ballistic performance was satisfactory,a new propellant (for incrase mechanical properties), case (attachinet), and nozzle (15* cma) were required to accommodate the AQM-37C.

ADVANCED TARGET BOOSTER (ATB) DEFINITION

The ATB definition project started in November 1989 and will culminate in beptemoer 1990. An oMitrkum AQM-37C SRB designwas defined to facilitate a cost versus performance tradeoff study of performing a product improvement program of a qualified off-the-shelf SR8 versus developing a new SRB. Component characterization tests were performed to verify the ability to meet optimum designrequirements and to support a future product improvement program or lull-scale AQM-37C SRB development of a new design.

BOOSTER REQUIREMENTS

The SRR requirements are summarized below:

Total impulse 50,000 to 3,300Bum tim Wj 30 to 32Average Ltm-' Ir', 1,665Length (imnh,) 100 to 103Diameter (inrh,-' 6 to 8Total weight (ibn) 250 to 290Nozzle can (degrees) 15Thrust misalignment (df.0 ees) 0.5

A short account of the requirements evolution is gi%,.', in the following paragraphs.

BALLISTIC PERFORMANCE

Ballistic performance requirements were derived using siml,l, Ms wD,.servaion calculations during the booster design study to determineif SRO integration was feasible and to serve as a starting point fr,, th. subsequent integration study. During tire integration study, flighttrajectory simulations and load analyses were performed to deterr,, ;.- fo-Ae, performance requirements 1vr v.itsequent hoosier definitionand development.

Initial Sizing. Initial sizing calculations were performed to determine it a hgnt is s-,.A e, ougi SWB could he integrated with theAQM-37C and launch aircraft. The first iteration of cal-ulations was based on increasing the target range by 85 nautical miles for theMach 3 mission when Launched at Mach 1.5, 50,000 ft (The requirement at this time was to only increase the range of the target; losingsupersonic launch capability was nor an issue.) The amount of liquid propellant that would he required to fly an extra 85 miles for theMach 3 and 4 missions was calculated to he 90.7 and 49.5 Ibm respectively. Next the change in vehicle velocity during boost was calculatedto determine the amount of solid propellant required during the boost phase to save the aforementioned quantities of liquid fuel for thesustan phase. It was calculated that 212- and 167-Ibm hcuscr .re required for the Mach 3 and 4 missins respectively. These cak-ulationsassumed a propellant-to-total.weight mass fractio'i of 0.75, spec.:fic impulse of 250 lbf-s/lbm.

112

Once it was determined that a fairly lightweight SRB could be used, NADC performed several two degree-of-freedom flight simulaionsto verify the hand calculations and to more closely define the booster pe'formance requirements. The resulting requirements were as follows:

Minimum impulse: 27,000 hbf-sAverage thrust: 5,000 lbfMaximum diameter: 10.5 inMaximum weight: 230 lbNozzle cant: 30 degrees

A search of several off-the-shelf boosters for compliance with these requirements resulted in identifying the Jato Rocket Motor Mk 7Mod 4 as a candidate fi;r integration.

Performance Parametric Study. BAC performed six degree-of-freedom flight trajectory simulations using the Mk 7jato performance.They determined early on in the study that Mk 7 average thrust would exceed the AQM-37C airframe load capability, and that the physicaldimensions fell out of the envelope required for safe takeoff. (Basically the canted nozzle would strike the runway during rotation.) BACproceeded with a parametric study of SRB impuLe, range, and burn time to determine the optimum SRB performance (Fig. 2).

30

(40 Z

cc

J/55100I50,000

45,000 IMPULSE (Ibf-s)3 4 5 6

TIME ON STATION (min)FIGURE 2. BOOSTrER PARAMETRIC PERFORMANCE STUDY

[Std. launch - Mach 1.5, 50,000 ftl

Load Analysis. A preliminary evaluation of captive flight conditions was conducted in accordance with MIL-A-8591C for wing-mounted stores. The load analysis was based on a design gross weight and center of gravity location of the combined AQM-37C target,booster, and 3dapter weights. Fuselage design loads indicate the primary area of concern is in the adapter attachments at fuselage stations160.255 and 224.135. The aft lug vertical load would be 11,391 pounds, and it has a design load limit of 10,400 pounds which is inadequatedue to the extri weight of the booster. The existing airframe, manufactured of cast magnesium, would require structural reinforcementto sustain the vertical and side loads induced by the bcoster.t

INTERFACE

Propulsion System Sequencing. Several flight simulations were performed to determine how to sequence the liquid and solid rocketpropulsion systems. Simultaneous versus consecutive solid then liquid operation was considered. Simultaneous operation causes two basicproblems: target pitch-up is unacceptable, and liquid propellant unporting from the tank outlets would cause thrust interrupt. Ignitingthe solid booster first followed by liquid operation allows less liquid propellant to be used during climb, decreasing the likelihood ofpropellant unporting. Consecutive operation provides a smooth climb profile and increases the mission time at speed and altitude. I

Physical Envelope. The booster will he attached to a sway brace/adapter, and the Launch Adapter Unit (LAU) 24B or LAU 96 willbe used for integration on the A-4, A-6, F-4. and F- 14 aircraft. The envelope for safe separation distances with aircraft stuts fully compressedand flat tires dictates the booster size. All aircraft except the A-4 had sufficient clearance for conventionally sized boosters. Booster sizerestrictions to approximately 9 inches in diameter and 100 inches long are required to prevent the booster from contacting the runwayduring takeoff. The A-4 aicraft centerline station would be used to obtain the maximum amount of clearance. t

113

Thrust Alignment. A canted nozzle is required to align the solid rocket booster thrust close to the combined center of gravity of thetarget and the booster. Flight simulations using nozzle cant angles of 5. 10, and 15 degrees were performed to determine the best cantangle. Nozzle angles of 5 and I0 degrees caused the target to pitch up too high for the canards to compensate, causing the vehicle tobecome ballistic and overshoot the desired altitude. The best angle for both missions was 15 degrees. A slight improvement in time onstation was also predicted with the 15 degree nozzle angle. t

BOOSTER DESIGN

An illustratien of the general ATB assembly is presented as Fig. 3. A composite case, passive nozzle retention system with removableigniter/exit cone and igniter thermal isolation features will be employed for safety considerations. An end-burning propellant grain wasselected to obtain the volumemic loading required to meet ballistic performance requirements. Given this grain design, a compusite-nubber.typepropellant and a booted insulator/stress relief system were selected to meet operational temperature requirements.

GRAPHIrTEEPOXY ALUMINUMCAS•E EIqE'.NT IO

FAIRING ADAPTER RINGINSULATION

-- ~..-----.-. STEEL--. DAPTIER

BULKHEAD STRESS NODZLE BODYRELIEF - W•ACowF.•

ENTRANCEINSULATIIN

Jfr?ý__` -Is.tTV TYPE 11PELLETS

FIGURE 3. SOLID ROCKET BOOSTER DESIGN CONCEPT

CASE

A graphite/epoxy case was selected to meet desired performance and operational requirements. The composite structure is manufactured

by braiding graphite fibers wetted with epoxy resin over an aluminum mandrel. The plies comprising the composite laminate possessthe structural integrity required under normal operating environments. Stainless steel closures with special thread forms are employedto complete the pressure vessel and facilitate attachment of the nozzle and forward fairing. A layer of stainless steel foil is used to provideelectrical continuity between the closures in order to dissipate electrical charge buildup in the case. The aluminum mandrel has the samethread torm as the closures. After cure the mandrel is pulled and the closures are bonded to the composite structure.

PROPELLANT

Two propellant formulations were considered during early propellant definition. Both formulations utilize an ammonium perchlorateoxidizer, aluminum fuel, and a binder synthesized from the hydroxyl terminated pnlybutadiene (HTPB) polymer. The more conventionalof the two formulations uses iron oxide and extra fine ammonium perchlorate to achieve the required burn rate. The other formulationuses a new chemical called butacene. Butaccie is the basic HTPB polymer with ferrocenyl groups grafted at the carbon-carbon double bonds.

Basic propellant requirements of a 2.5- to 2.6-in/s burn rate at 1800 psi, 25% minimum strain, 75-p1i minimum stress at 140 *F,and 1.3 hazard classification were derived from booster requirements. Several one-gallon propellant mixes of both formulations weremade to characterize the safety, processibility, and physical and mechanical properties of the formulation. Based on the one-gallon mixesthe butacene was selected for further development because it was easier to ,-rocess and the burn rate was easier to tailor. A blend of74% butacene and 26% R-45M HTPB polymer is used to achieve the desire.' burn rates.

INSULATOR/STRESS RELIEF BOOT

An insulator/stress refief boot design was defined to protect the case from hot combustion gases and to relieve stresses in the propellantcaused by thermal expansion, motor ignition, and vehicle acceleration. An ethylene propylene diene monomer (EPDM) rubber with aramidfiller will be laid up in thickness proportional to the time each part of the case su'cture will be exposed to hot gases. The inner insulator,which will sprve as a stress relief boot, is laid on a mandrel to a thickness of 0.060 inch. Next the five strips of Teflon tape are laidon the insulator prior to laying the remaining EPDM to the required thickness. Teflon tape acts as a release agent which facilitates formationof five stress relief boots running longitudinally down the case. Next the insulator is vulcanized and secondarily bonded in the motorcase everywhere except the head end which allows relief of stresses in the longitudinal direction.

114

IGNITER

The igniter design consists of a MIL-1-23659 qualified (I -amp, I-watt no-fire, electrostatic discharge-safe) filtered electric initiatorand pelletized Type II magnesium-Teflon-Viton (MTV) ignition material enclosed in a cellulose acetate capsule. The capsule will be bondedto the area just inside the nozzle throat. The initiator (or qualified safe and arm device if required) is threaded into the center of a nozzleweather seal and is readily accessible for installation and removal.

CHARACTERIZATION TESTS

PROPELLANT

Propellant burn rate. Shore A hardness, mechanical properties, and pot life test results are presented in Figs. 4 through 7. Table Ipresents propellant safety data.

I e24.9 m

0.P m

S25-, .

1 , ............ .......... . i-. . .; ' . . . . . . . . . .

20-.. 4.9Am

4 C C 7 8BuA.cene (% of wI.)

FIGURE 4. WEIGHT BUTACENE VERSUS PROPELLANT BURN RATE

188° Sohds; NCO/OH = 1.00; 70/30 Ratio of 200/30 M m API

50% Butacene

" - 100% ButaceneS/,l •• 58% Butacene (9 ois

S• 58% 81".acene (90% solids)

flays

FIGURE 5. PROPELLANT SHORE A HARDNESS

to

• , - Strain 25 05

SStress -

Bulacer,e (% of wI)

I IGURL 6. PROPELI.ANT MECHANICAL. PROPERTIES

I1s

80-

70 - + 50% Butecene

60 o 75% ButaceneA 100% Dutacene

50 50S 40 .............................. .......................

A 40 0300

>~ A 0

10201 0 0 + + + + +

10 + +

0 1 2 3 4 5 6 7 8 9 10

Pot life (hr)

FIGURE 7. PROPELLANT POT LIFE

rABLE 1. PROPELLANT SAFETY DATA

Butacene levelTest 50% 75% 100%

325 200 150235 315 315

ge (J) 0.625I 0.500 0.050" g(.Medium Medium Medium

IGNI'Ek

Prototype igniter tests were performed to select the optimum amount of MTV to achieve the desired dispersion of particles. Even,semil-spherical dispersion is desired to ignite the flat end-burning propellant grain. A minimum of 10 grams of MTV was calculated tobe required for propellant ignition. Three each igniters were loaded with 10, 15, and 20 grams of MTV and fired in the open air to assessdispersion. Review of high-speed films indicate that igniters loaded with 15 grams of MTV possessed the most desirable dispersioncharacteristics. Subsequent tests will be performed in a closed bomb to acquire pressure versus time data. A rubber witness pad representingthe propellant grain will be installed in the bomb to verify the ignition material is evenly dispersed.

PERFORMANCE PREDICTIONS

A prediction of rocket motor booster bal'istic performance is presented in Fig. 8. AQM-37C Target performance based on this boosterdesign is presented in Figs. 9 and 10. A comparison of range for mission profiles of Mach 3, 80,000 ft and Mach 4, 100,000 ft versoslaunch option is presented in Fig. I1.

THRUST

2400; ..-- PRESSURE 11600

,a t IS- -.8- '\180

S1200 ................. ..... ... .. ... '800

6001 2 3 3 41400

0 00 6 12 t8 24 30 36 42

Time (s)

FIGURE 8. "RFORMANCE PREDICTIONITempera.•, O °F; altitude = 50,000 ftl

116

4.00 O SUBSONIC LAUNCH W/O BOOSTER

+ SUPERSONIC LAUNCH W/ BOOSTEP3.00 SUBSONIC LAUNCH W/ BOOSTER

z2.00

:•1.00/

0.00 ,,-0 200 400 600

Time (t)

FIGURE 9. MACH 3. 80,000 FEET MISSION PERFORMANCE

5.00

4.00 •0 SUBSONIC LAUNCH W1O BOOSTER+- SUPERSONIC LAUNCH W/ BOOSTER

3.00 - A SUBSONIC LAUNCH W/ BOOSTEREz 2.00

1.00

0.00 ý -__L _ _J

0 200 400 600Time (s)

FIGURE 10. MACH 4, 100,000 FEET MISSION PERFORMANCE

250

200-

150

,100'

0 .. .....

A-4 "/-4BOOSTER F-14 F-14/fOOSTEn AQM-XXLaunch Optlons

1 Mach30h 3 0. S t 1 Mach 4.0. 100ODf R

alane as defined a$ ditaflce traveled On station.

FIGURE I. AQM-37C IAUNCH OPTIONS PERFORMANCE COMPARISON

CONCLUSIONS

A SRB can he integrated with the AQM-37C to maintain its current performance when launched from suhsonic aircraft and greatlyincrease its performance when launched from supersonic aircraft. Results of preliminary analyses and tests of the SRB deiign have beenfavorable and demonstrate that the SRB could be fully developed and integrated with the AQM-37C Target at low technical risk.

REFERENCE

1. D. R. Arab, "Booster Integration Study for the AQM-37C/AQM-37(EP)". Beech Aircraft Corporation, Wichita. KS, April 29, ;989.

117

STAMIZJTY TESTINGOF FULL SCAI2 TACTICAL DKeYIYRS*

F.S. Ulomih~eidNavat weapons Center, China Lake, CA 97555

J.E.Crump a" N5.B.athe-sCcomsrco, Inc., Ridgacreet, CA 93555

H.W. SecksteadBrigham Younig UniAversity, Provo, UT 84602

The Navel Weapn Center is participating in an ongoing program to develop an Improved understanding oflinear and nonlinear combustion instability in solid propellant rocket motors. One goal of this progra is to

devlopa sstentc date bern. of motor and stability dats. Earlier paprs have reported on previous WUC workvn this program. ' This paper describes the linear aspect%of the motor firings and analysis. A compaonionpaper describes the nonlinear aspects of the motor firings.

7.&. motors being used by NUC are- 5 inches in diameter a" 67 inches in length. The majority were loaded..th an Wt solids reduced smkok AP/INPI propellant with a nominal burning rate of 0.24.0 in/sec at 1000 pal.In addit:con, motors have been f ired which contain 1 percent 8 micron aluminum oxide, 90 micron aluminwi oxide,eond 3 micron zirconiun carbide as stability additives in place of one percent ammnonium perchlorate. Motorpressures ranged from 500 to 1500 psi. various grain geometries have been tested. Pressure coupled combustionresponse measurenmprnts were made at the nominal motor operating pressures.

Motor perfornience and stability chiculastionoi were made using the Air Force Solid Performance Program (Vand the Standard Stability Prediction Program (SSP) ior the several motor configurations that were fired.OThe stability predictions were compared to the data obtained from the motor firings.

~INTRDUCION

Motor Cori1faurallons; Several motor configurationr. and propellant variations were incttxied in the program.A total of 23 motors were fired. Thc motor configurations are shown In Figure 1.

The baseline grain geometry was a six-point star in the aft two-thirds of the motor and a cylindricalsection In thc. forward en. Vtesr motors were 67 inches long with a diameter of 5 inches, giving themn a Lengthto diameter ratio of over 11. Yen of these motors were fired using the basel ine propellarn NWR-11, and threewere fired using pr-opellants containing stability additives. Three motors with star-forwa-d grains, one motorwith a full star perforation, two motors with circular perforations, and four half length motors were alsofired.

The noomenclature used to -dentify the motors describes the geometry of the motors. The four graingeometries ar. star-aft (SAFT). star-forward (SFiD), full star (SFUL),and cylinder (Cr1). The length of themotor is designated nos 3 or 6. The nuriber 3 r~fers to the frequency of the first longitudinal Rnode for the futllength motors (300 Hz) and the number 6 to the first mode frequency for the half length motors (600 Hz). Formotors that have propellant stability additives, the propellant forwmuition nuriber Is included. Thus, SAFT3-12Is a star-forward, full tength oo~tor using NWR-12 propellaent. table I shows the test nmtrix and whien the motorswere L.ist and fired.

Er.j,11iantjip The baseline propellant (WI.I-11) was an88%-sotids AP/HTPS propellant. The burning rate was 0.24.0in/sec at 1000 psi; the burning rate exponent was 0.491. The propellant combustion response funaction wasueasured using the pulsed-during / pulsed-after burning 1-burner te.chnique. The response curves for 500 and1000, and 1500 psi are shown in Figure 2.

It was planned to use the sane propellant (NVR-1i) in Motors 11-14 and 19-21. Hoiever, the burning rateof the propellant mix for those motors was considerabty lower than desired The rate was 0.213 in/sec at 1000ppsi. Since the moto-* nozzte throats toad been machined, all of these motors operated at a tower pressure thanoriginally planned. Also, It was decided that the propellant response funrct ions at !o00 psi and 1000 psi shouldbe determined. The response function for both the lower rate propellant (Mix 10196) and the original (Mix 9293)are shown In F'igure 3 (500 psi) and Figure 1. (1000 psi). Th2 huge increast. In the response fun -ions of thelow rate propellant at 500 pri Is not explainable at this time. A careful review of 1-burner response data hasryv. revealed any reason for the differences.

* his effort was sponsored by the Air Launchred weaponry Technmology Block Program Oft ice under the authorityof Tcmi Loftus, Code 372, kaval Weapons Lenter, China Lake, CA 93555.

I~pproved 'or Public. Release; Distribution is Unlimited.

O STAR-AFT(BASEL I NE)

A-A B-BOSTAR-'FORWARDA8-8

A WA-A B-B

____ ____FULL STAR

A-A

0• FULL CYLINDER

FAmORARD :A= A-A

FIGU E 1. Motor Configurattons

TABLE 1. Motor Test Matrix

TEST GEOETRI PROP PRESSURE FREQ MIX/TEST NOZZLE CASTING FIRINGNo NW,- psi NZ GROUP inches DATE DATE

I1 S-AFT 11 1000 300 1 1.668 JUN 85 AUG 862 U N U U U N NOV 563 N 0 U U N N U

4 S-FAl " M " MAY 875 S-AFT 4 U U , N N

6 H 500 0 " 1.850 N a

7 FULL " 1000 " 2 1.680 DEC 87 JUL 888& S-AFT N 1000 " ' 1.680 •8b 500 • 1.850 "a " N 1500 " "4 1.460 " N

8d " 1000 N " 1.668 MAR 899 CYL " 1000 ,4 1.575 JUL88

10 U 50O0 " " 1.710 " "

11 S-AFU l1b 1000 600 7 1.200 DEC 88 MAR 8912 • N 1500 " " 1.130 "13 CYL " 500 " 1.040 [14 $-FWD " 1000 1.1!0 I; ...

15 S-AFT 12 1000 300 3 1.668 MAY 88 MAR8916 " 13 4 ". ".17 17 " 5 , AuG8818 18 " 6 1." centered

S-AFT Il 2000 - 30 E 8 KR820 S-FW 1000 , , 1.668 LI _

z

z 1.20

0.8

02.0.4 -

O.Z IX

0 200 400 600 0oo 1000 1200 1400 1600 1800 2000

FREQUENCY - Hz

FIGLUE 2- Pres.;,.-op '.4jpted Combustin Response of MYR-11L- - propetlant at 500, 1000 and 1500 psi

3- --- ----

2.5 - -- , " "-

w z

00

coo

0 200 400 600 800 1000 1200 1400 1b00 1800 2000

FREQUENCY - Hz

FIGURE 3. Cimparison of Pressure Coupted Combustion Response ofDifferent Baseline Propetlant mixes at 500 psi

:21

3

p ap-Il - IMl. I"I

Z.5

--- -- --- --

oU

0u2 1.5

0.5 ,"

0 200 400 600 800 1000 1201, 1400 1600 1800 2000PREQUENCY - Hz

FIUMAE 4. Comparison of Pressute Coupled Combustion Response ofDifferent seetline Propellant Mixes at 1000 psl

Tnree motors were loaded with variations of the baseline propellant. one percent of the 5S micron wIlueperchtorate in the baseline propellant was replaced with a stabi tity additive. The NW-12 propellant containedone percent of 8 micron alumiurs oxide, NWR-13 contained one percent of 3.5 micron zirconium carbide, and NV-17contained one percent of 90 micron altminum oxide. The response function for the three propellents weredetermined from T-burner testu. They are ahown in Figure 5 along with the baseline propellant, NW-11. Theadditives also caused a slight shift in the propellant burning rate. Pr lnttent properties have been smmrizedin Table II. From Figure 5, it can be seen that each of the additives lowered the response function by 25-35percent below the baseline propellant over the frequency range of interest (300-1000 iNz).

TABLE II. Additive Propellants Properties

Propellant Additive Size Rate* Response Exponent Motor(micron) (in/sec)

WR.-11 (none) .-. 0.238 1.25 0.491 1,2,3,5NW,-12 AtO 3 8.0 0.248 0.75 0.493 15NWR-13 ZrC 3.5 0.225 0.82 0.493 16NWR-17 I A t 203 90.0 0.230 1 0.7; 0.413 17

B lurning Rate at 1000 psi** Response Function at 1000 psi and 300 Hz

MOTOR PREDICTIONS

The xmin thrust (. the overal( program was to develop an Improved understanding of nonlinear (pulsed)combustion instability. ;he nonlinear stab-lity of a motor is dependent on the linear stability. The linearstability is tht threohold stability that must be exceeded In order to produce the nonlinear behavior. Thelinear growth (or decay) of a pressure oscittllation is given by

Spoe° (t1 )

122

3 .

A me-12 Ova L2O3... m.... ?.... ... ........ _ __ __ _ __ _

2.5 -2-ON- A._______

2.

Z0C/)2 1.5..... .. +

0'

0 200 400 800 800 1000 1200 1400 1600 1900 2000

FREQUENCY - Hz

FIGURE 5. Pressure Coup~led lesponae at 1000 psi ofPropellants Containing Stability Additives

The rote of growth (or decay) is expressed In termn of the alpha in the exponential expression of Ecrqt~nt~ 1.If a pressure disturbance In the motor, e.g. pulse. is such that it produces a pressure decay alpha tess tnanthe linear decay alpha, the pressure oscillations will decay; If the alpha produced Is greater than the lintearmotor stability alpha, the pressure oscillatiorws will grow. Since knowledge of the linear stability wasrequired, this facet of the program presented an opportunity to predict the linear stability of several motorsand to make comarisons with linear experimental data which might be obtained uinder some test conditions.

The Air Force developed Solid Pr~toieant Performencea comuter program (SPP) and the Standard StabilityPrediction coippter program (SSP) were used to predict the motor performence and linear stability of thebaseline motor and all variations.

Batgi.s.Iicse. One aspect of this study was developing the capability to predict the ballistic performience of arocket motor using SPP. A comprison of the measured pressure-time curve with the predicted curve for Motor5 is shown in Figure 6. The Initial prediction utilized a propellant strand burning rate determination overthe pressure range of 200 to 2000 psi. the burning rate was 0.240 in/sec at 1000 psi with a pressure exponentof 0.491. After the motor firing thew burning rate was adjusted to give the correlation shown in Figure 6. Theburning rate used in the correlation was 0.238 at 1000 psi. The large ignition spike and large tail off areboth indicative of erosive burning in the motor. Although the a and a erosive burning parameters were adjustedto give the agreement shown, the Inputs to SPP could have been further optimized by adjusting the erosiveburning parameeters further. However, this level of refinement was considered beyond the scope of the project.

Ark exempt* of the mean pressure for a motor that was pulsed un'stable Is shown in Figure 7 along withpredicted steady state pressure. The pressure trace is from motor #4 which was a full length star forwardgemAwtry operating at 900 psi. The high pressure during the oscillating portion of burning was due to theacoustic erosivity associated with oscillatory combujstion. It should be noted that the motor was pulsed at 0.97seconds at an amplitude of 43 psi, and the motor was stable to that pulse for that geometry. The second pulsewas at 1.96 seconds at an amrplitude of 25 psi, which was sufficient to trigger the observed non-lintearcombustion Instability.

Stability Prediction: The Standard Stabilt~t Prediction Me#thod (SSP) was used to predict the linear stabilityof the baseline motor a" all variations. ' The linear stability elements (expressed as growth end decayalpha's) for the baseline motor ore shown In Figure 8. The algebraic sum of the pressure -coupl ad (PC) alphp,the velocity-coutled (VC) alpha, the nozzle doiping (NOV), and the flow turning darping (FT) yields the !wtoralpha (MOTOR). If the sum Is negative, the motor is said to be linearly stable, e.g., it is stable to smallpressure disturbances. If the sum Is positive, the motor is said to be spontaneously unstable, e.g., minutePressure oscillations will grow.

1800 - -..

STAR AFW PREOUTW-

1400 -. _-

1200s!_.. __ _ e__0__c__ _ _o_

1800 0

iz Goo

400

2800 - _ _....__ _ .,, • _ _- __

2600

200 051 .5 2

00 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

TIME - SEC

FIGURE 6. Comaprison of Predicted and Measured Pressure-Time Curves for Star Aft 300 Nz Motor S

1800 --

"- PREICT IO

1600 '

1400___

. 1200

) 1000

r. 800 ...... '""

600 -,

400 __ "

0o - - --- - -

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6

TIME - SEC

FIGURE 7. Comparison of Predicted and Measured Pressure-Tim Curves for Star Forward 300 Hz Motor 4 (Unstable)

124

2000 FULtEm "Mift]MG)oC •pi - p - 1o • moUe~l

cn) -- - - - - - - - -

0

- 200

0 0.2 0.4 0.6 0.8 1 1.2 1.4WEB - INCHES

FIBROE a. Stabitity Prediction for Star-Aft Configuration (300 NZ) at 1000 psi

0

--1001

-200

-250

-3000 0.2 0.4 0.6 0.8 1 1.2 1.4

WEB - INCHES

FIGURE 9. StabiLt ity Prediction for Star-Aft Configuration (3 00 z)at 1000 psi for First Three AxiaL Accustic Had**

125

Since iwnoinear stability involves stesp-frented pressure waves (Oitiple acoustic mode frequencies). theinser stability of the higher tongitudinat acoustic modes of the baseine motor is of interest. The motor

eighas for the first three modes of the baseline motor at 1000 pal are show •n Figure 9. The higher modes havelower tinmer stability limits (emitler volums of the motor alpha) than the first longitudinal made (300 Hz).This is the result of the propellant response ftuction Increasing with Increasing frequency over the frequencyrange of 300 to 1000 Iz (Figure 2). This is also somewhat responsible for the susceptibility of the motor topulirli. The pulse contains relatively high frequency components oaich tend to excite the higher frequencymodes of the motor. The fact that the higher frequency modes are less stable Linearly, tends to promote thenon-linear respose of the motor to a pulse.

MOTOR TEST DATA

All the motors in this program were instrumnted with at least one ballistic pressure transducor and twoctosely-coipted high frequency response pressure tronsducers located in the forward Wid of the motor. Detailsof the Instrumentation are found in Reference 2.

The motors tested, the pltnnd pulsing, und the qualitative test resultc are indicated in Table 11. Thepulsing of Motors 11-14 and Motors 19-21 was intended to get three pulse decays, which would provide data formore correlations with linear stability predictions. The Intent with Motors 15-16 was to pulse them in the sammvner as the baseline motors to see if the aditives had an effect on nonlinear instability.

Doat from motors •hich were pulsed but rmmined stable was analyzed to determine the linear motor stabilitymargin. This was done by measuring the decay rate of the pressure oscillations (motor alpha) fotlowing thepulse. The technique is described in Referenc 4. Table III also lists the pulse times, pulse amplitudet, anddecoy alphas. As shown in the table, se of the motors were stable to the pulses. Sowe of those pulse" werevery weak and resulted in situations where the pulst was too small to detect in the date or too small to givemeaningful decay elphat due to the poor signal-to-noise ratio. Pnan is the nominal pressure designation forthe motor; Paove is the approximate man pressure when the pulses were fired.

TAILE I11. Pulse Time-AwpLitude-Alphs Comparison

TETGCTY Prom Psv# PULSE 61 PULSE 112 PULSE 0TEST GW(IETRY ....e *

NO (psi) (psi) sea psi Paloi a sec pail pail a sec psi' poll 0

I UAFT3 1000 800 nrd SO -..- 2.016 20 36 +209 *ac 102 0 W a 0.703 50 12 -159 1.706 20 12 +207 sec 10 . .3 a • 0.805 50 14 -112 1.790 20 50 *'61 osc 10--4 S2FD3 900 0.970 50 38 -190 1.962 20 22 .56 osc 10 .. ..5 SAFT3 800 0.965 50 100 -176 npf .. .. .. rpf ..6 500 S40 0.760 50 18 -214 2.005 20 7 -102 3.010 10 6 -75

7 SFUL3 1.000 900 O.748 50 38 .639 0ec 20 . . esc 10 --Be SAFT3 1000 800 rpf npf .- 3.020 10 10 .1348b SAFT3 500 540 0.950 50 50 -210 2.216 20 20 -89 3.419 10 32 +366& C 1500 1330 0.617 50 20 *151 Inc 20-- . . oc 10 .8d 1000 8a0 (mistake; no pulses fired)9 CYL3 1000 830 0.745 50 20 '232 oec 20 "- " osc 10.. ..

10 " 500 610 0.887 50 36 '244 osc 20- ;-- e0c 10-.

11 SAFT6 1000 480 2.067 50 40 -143 3.6W6 10 6 -75 5.218 5 4 -6012 a 1500 600 1.374 10 12 -182 2.S43 10 6 -247a nrd 5 -.13 CTL6 500 580 1.261 10 8 -165 2.429 S Is nrd 3.445 S 4 -7114 iFtD6 1000 660 1.945 50 50 -201 3.65 10 8 -133 4.896 5 2 -33a

15 UAF3-12 1000 940 0.946 75 174 -122 1.764 20 40 -84 nrd S16 11F3-13 7 7 0.824 75 52 -232 1.505 20 26 -137 nrd 5 --17 SA13-17 f 830 0.866 75 70 -188 1.66 20 46 -75 2.572 5 2 -110oa16 GAF3-18 ' (canceled)

19 IAFT3 2000 1460 nrd '10 .- . 1.81? 5 2 -26. 2.767 S r2 -64a20 SfiW3 1000 730 0.964 40 16 -193 2.167 10 4 -81 rpf .21 SFi)3 " 1.012 40 42 -251 2.195 10 10 -143 3.506 5 nrd

rpf - no pulse fired nrd - not reducible osc - oscillating a - weak pulse

0 planned puLse mplitude

actual pulse alitude by extrapolated beck to zero time (See Raef. 6)

126

LINEAR CORRELATIONS

Mean Pressure Ccarlson: Motors 1-6 were fired in the first test series. A total of three pulses decsyed InMotors 2, 3, and 5. The pulse decay rates are plotted in Figure 10 along with the linear stabitity predictionfor the furdamntal frequency for that configuration at 1000 psi. A totai of five pulse decays were alsomeasured in Motors 6 and 3b. These are also plotted in Figure 10 along with the linear prediction for S00 psi.The agreement between the masuremnts and the predictions is very good. It Is particularly significant thatthe motor becoes Less stable during the firing and both the calcutaticna and dats show this trend. it Is alsoworth noting that the Lower pressure motor was more stable then the higher pre-ssure motor. Table IV ehowsindividual alphas end total alphas predicted by the SP program for these two motor pressures.

TABLE IV. Individual Alphas for 500 and 1000 psi Star Aft Geometries

web Total Alpha PC Response VC Response I wozzle Orp. FLow Turning

SOpi 00 500 1000 S00 1000 500 1000 50 10

0.02 -319 -237 104 123 *23 -20 -261 -222 -140 -1!90.10 -250 -178 88 I10 -10 -8 -214 -17 10.20 -197 -138 74 85 -4. -3 -i17 -141 -92 -750.30 -157 -109 63 72 -1 -1 -145 -119 -7 5 -610.40 -131 -89 55 63 1 1 -124 -102 -630.60 -97 -67 37 47 1 -88 -72 -47 .380.80 -81 -56 31 36 2 , -75 -61 -39 -32

1.00 -77 -51 35 40 6 5 -75 -64 -39 -311.20 -30 -18 4 4 10 9 -55 - I -19

1.41o " 1 .1 5 1 70 3 3 -7 .7 -.1 -•

0

PRflEICTION - 1O_

,- -. -PC-C9 IS t . I.U ...... ...... . . . . . -

Q -1~0 -- ----- ----------------- __

T -zoo

C. 20 ___-,---- __

-Z50 I1

- 3 0 0 0 0. 0.4 0.6 0.8 1 1.2 1.4

WEB - INCHESFIGURE 10. Measured and Predicted Star Alt 300 Nz Mctor Alpha& for motors 2, 3, 5 (1000 psi)

and Motors 6 and 8b (O0V psi)

127

f.~rjjfnj Motor 11 was a half length motor Intended to sho66 the of fects of f reqUOMnY, but I t wastoo" wi th the low rate propellent mix (*Aj-ilb) that had a burning rat* lower than plarmed. This motor hada fuwdmotat freqvency of 600 Ht. Decay rot** frcm all three pulses were measured and ptotted in Figure 1ifor comarison with the half tength linear stability calculations (solid line). The predicted linear stabliktymargin lor the baseline motor is shown for comprison (dashed trise). This is the full length baseline withnormal burn rat* propellant. The trend of the dots Is in agreent with the prediction, but the magnitude isvery different. It Is not certain whether the observed effects are dues to the change In burning rate, frequencyor propellant responi... As mentionied previiously, the V-burner data for the low rate propellant was wunuuallyhigh and no rational explanation for the high response could be detected. If on assumption were made that theresporae of the low rate proosliant were the Sav as the normal rate, then the predicted stability (solid line)would be such closer to the measured data. Table V shows the Individual alph~as and the total alphas predictedby the SSP program for these two motor length*. Un~fortunately, this table and the plot in Figure 11 arecoa~ring both propellant burning rate and frequmncy effects and it is difficult to distinguish them aport.

TABLE V. Individual Atphas for 600 Hz Half Length and 300 Hz Full Length Motorsat 1000 psi with tho. star Aft Geometry

Web Total Alpha PC Response VC Response Nozzle Dam. Flow Turning(in) -- I- -

6~0 iiz 300 600 300 600 300 600 1300 600 300

0.02 *149 -237 159 123 *9 -20 -175 -222 -124 -1190.20 -17 -724 1 2 Z 0~ -8 -146 -177 -13 90.20 -102 -138 126 e10 -3 -116 -177 -105 *750.30 -50 -109 108 72 6 -1 -95 -119 -68 -610.40 -37 -89 96 63 7 1 -82 -102 -58 -520.60 -39 -67 55 42 5 1 -55 -72 -4 .380.80 -XI -56 47 36 4 1 -47 -61 -35 -321.00 -25 -51 42 40 6 S -4?2 - -31 -311.20 0 -18 48 48 is 9 -39 -57 -25 -191.40 .5 1 a 7 5 3 -. .7 .7 -1I

0 RITO - Iwo$-

-- Wa.:'C RtlCTION

-50

100

0 -20

-3000 0.2 0.4 0.6 0.8 1 1.21.

WEB -. INCH-ES

FIGURý )I. Measured aiid Prodicted Half Lerqth 600 Hz Notor Alphas for Motof 11

128

Additive Propellant*: Motors 15, 16, and 17 were loaded with propellants containing stability additives. Thepropellants with additives had burning rates that were somewhat different than the baseline propellant (, -11,mix 9293). The previously shown Table 1! shows some of the properties of the propellants. Due to thedifferences in burning rates (and the use of standard nozzle*), the bettiscic behavior of the motors wasdifferent than originally planned. These differences are reflected in the pressure-time curves shown in Figure12. The results of the linear stability calculations are shown in Figure 13. The motor stability margins forall the motors with additives are virtually the s and are significantly greater than the stability marginfor the baseline motor (NUR-11). The increased stability is due to the lower response functions of the additivecontaining propellant* (refer to Figure 5).

If one looks at the stability margins for the additive propellants in greater detail, Figure 14, it isseen that the motor uwing NR-13 propellant had the largest stability margin. One might expect that thepropellant with the smallest value of the propellant response function would have the greatest margin (NUR-12,Table 11). However, in this comparison, the ballistics of the three additive motors were different due to theslight burning rate differences (Figure 12). Hence, the changes (increases) in the other factors that make upthe motor stability margin (mainly, the changes in the mean flow velocity) were greater than the change(decrease) in the pressure-coupt ed alpha due to the decrease in the response function for the additivepropellant motor firings. The point to be made here is that in the motor geometry-propel Iant system, a decreasein the propellant response function does not necessarily result in an increase in the motor stability margin.

Pulse decay rates for the first end second pulses in the three additive motors were measured and comparedto the linear stabitity prediction. These comparisons are shown in Figure 15. The trend of the data is, again,in agreement with the prediction. The predicted linear stability margin for the baseline motor is shown forconparison. The stability margin for the baseline propellant is considerably less, primarily because it doeshave a significantly higher pressure coupled response, see Table if or cigure 5. Table VI shows the individualalphas and the total alphas predicted by the SSP program representi e of the additive motors using NWR-12propellant and the baseline propellant motor.

1900i IMOTOR 5 8aeeLL•S I I MOTOR 15 8&. R1203

1600 MOTOR 16 ZCi RMOTOR 17 90..n AL203

J400 I -..... I _

U) 1200

1000

S600

400__ _

0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6TIME - SEC

FIGURE 12. Pressure-time Curves for Star Aft 300 Hz motors with Propellant Additives

129

0

(12m

-150

-3000 0.2 0.4 0.8 0.8 1 1.2 1.4

WEB - INCHES

FICIMtE 13. Stability Predictions for Star Aft 300 Mz Motors with Additives at 1000 psi

-50

* NWt-12 Oue R 203

-60

-70

-80

0

90

0.4 0.5 0.6 0.7 0.8

WEB -INCHES

FIGLEE 14. Enlargemd View of Stability Predictions for motors with Additives

130

The principle intent with Motors 15-17 wea to pulse them in the sw minner as the basetine motors to soeif the additives had an effec- on nontinear instability. Unfortunately, the strengths of the first pulses (174,52 and 70 psi) were not, in general, comparable to the first pulses of the baseline propellent (0, 12 and 14),see Table I11. However, the fact that these motors remained stable to more severe pulses white weaker pulsescaused the baseline motor to go unstable is significant. This suggests the powerful rote stability additivescan play in combustion instability suppression. For the second pulses, if one considers the pulse aplitudeson the additive motors 15, 16 and 17 (40, 26 and 46 psi) to be nearly comparable to the second pulse in motors1, 2 and 3 (36, 12 and 50 psi), one could conclude that the additives did increase the margin of stability sincethe pulses in the additive motors decayed whiLe the pulses in motors 1, 2 and 3 grew. The lower value of theresponse function for the higher modes for the additive propel tants also supports this conclusion, since a pulsecould not as easily excite these modes into a non-linear instability. The particle damping coqapted for thefirst longitudinal mode is very tow as can be seen in Table VI. The particle damping will be more effectivefor the higher modes and wilt help demp out this form of non-tinear pulsed combustion instability. However,more d•ta is needed to reach a positive conclusion. For the third pulse there was no comperable data.

TABLE VI. Individual Alphas Representative of the Three Additive Motors using NWR-12 Propelt-%nt andthe Baseline "-opettant motor at 1000 psi and at 300 Hz with the Star Aft Geometry

Web Total Alpha PC Response VC Response Nozzle Damp. Flow Turning Part Damping(in)- -

Add. Base Add. Base Add. Base Add. Base Add. Base Add. Base

0.02 -263 -237 66 123 -17 -20 -202 -222 -110 -119 -2.3 1 00.10 -210 -178 58 102 -8 -8 -167 -177 -89 -94 -2.4 00.20 -164 -138 48 85 -3 -3 -135 -14 -72 -75 -2.4 00.30 -132 -109 41 72 0 -1 -111 -119 -58 -61 -2.4 00.40 -110 -89 35 63 1 1 -95 -102 -49 -52 -2.4 00.60 -82 -67 24 42 1 1 -68 -72 -37 -38 -2.4 00.80 -68 -56 20 36 1 1 -57 -61 -30 -32 -2.5 01.00 -65 -51 22 40 4 5 -59 -64 -29 -31 -2.5 0-1.20 -41 -18 29 48 9 9 -57 -57 -19 -19 -2.4 01

1.40 -13 1 4 7 3 3 -7 -7 -1 -1 -2.4j 0

This Particle dainping assumes 1% 89m A( 2 03

PRCtUICTION - IlOps,

* MOTO 15 8u. RL203

o nOTOR 16 Z-C0 a MOTOR 17 9O.* RL203--- W Q..ClNE PREDIrTION - . .......... _

S--100 6

<0

S-150

0 -200

-300

0 0.' 0.4 0.6 0.8 1 1 2 1.4

WEB - INCHES

fIGURE 15. Measured and Predicted Motor Alphas for Additive Motors 15, 16 and 17 (1000 psi)

and Prediction for Baseline Motor

M1I

92Mtry Variations: the primery purpose of Motors 4. 20 and 21 was to obtain edditionalt linear stability datefor correlation with the linear stability predictions and examine goetric effects. These motors were castwith the star-forward grain configuration. Unfortunatety, motors 20 and 21 were cast with the low ratepropellant mix, MOR-11S. Although geometric effects cannot be directly compared for these two motors, the pulsedecay data can be compared with the SSP predictions. Decay rates for these three motors were measured and areplotted in Figure 16 for comerison with the linear stability calculations. The open triangle on the plot isthe decay alpha measured for motor 4. It is to be copred to the solid line prediction. The solid date pointsare for motors 20 and 21 and are to be copered to the dotted tinP. Also shown in Figure 16 is the predictedstability for the baseline star-aft geometry, dashed line. The data in Figure 16 is in fairly good agreemntwith the prediction. It is again particularly significmnt that these motors become less stable during thefiring and both the calculations and data show this trend. The increase in stability for star forwardgeometries is due to an increase in the nozzle dosping because of the smaller port areas at the aft end of themotor. Table VII shows the individual alphas mnd the total alphas predicted by the SSP program for these twomotor lengths using the baseline normal rate propellant. The other two geometries fired in the program (fullstar motor nLm.er 7 and full cylinder motor number 9) yielded no stable pulses so no linear comparisone couldbe m•de. However, the fact that both these constant cross section geometries wer.t unstable, my indicate thataxial symmetric geometries are more susceptible to pulsed instability.

TABLE VII. Individual Alphas for Star-Forward motor V and Star AftBaseline Geometry at 1000 psi and at 300 Hz

Wb lTotal Alpha PC Response VC Response Nozzle Damp. Flow Turning

Star-Fwd Star-Aft StAr-Fwd Star-Aft Star-Fwd Star-Aft Star-Fwd Star-Aft Star-Fed Star-Aft

0.02 -311 Z37 108 123 -2? -20 -272 -222 -125 -1`190.10 -249 -178 93 102 -15 -8 -230 -177 -97 -940.20 -199 -138 78 85 -10 .3 -191 -14•. .77 -730.30 -159 -109 66 72 .7 -1 -157 -119 -61 -610.4,0 -132 -89 57 63 -5 1 -133 -102 -51 -520.60 .99 -67 39 42 -4 1 '96 -72 -39 -380.BO -80 -56 33 36 -2 1 -78 -61 -32 -321.00 -60 -51 25 40 -4 5 -57 -64 -24 -311.20 -24 -18 19 48 -5 9 -29 -57 -9 -19

1.0 -5 1 6 7 .3 3 -7 7 -1 -1

I- L -------

PR•Ol'ION M, WR-I 6A roM a4, - I]

-50 I "TOR 21 - mm-e - - - - -Z..II. ......TJio---

100

- .( .-____L_jcL__ -. .. "/

0 150

-//II --0 g0 ... L_ ___. _ _ [ _ _" _ _-250 --

0 0.2 0.4 0.6 0.8 1 1.2 1.4

WEB -- INCHES

FIGURE 16. Measured arKJ Predicted Star Forward 300 Hz Motor Alphas for Motors 4, 20 and 21

112

Summary of Linear Stabiltf.ty Comaarisons: Figure 17 shows all the measured decay alphas versus the theoreticallypredicted decay alphas. The plot is marked according to motor pressure, Length, propellant and geometry. 25percent error bands are also shown. Most all of the comparisons fall within the error bands. The exceptionis the half length motors 11 through 14. The stability of these motors is consistently under predicted, ie.they are predicted to have a tower stability margin than the data indicates. Reasons for these discrepanciesneed to be addeeszed in future work.

CONCLUSIONS

The ballistic pressure-time curves for most of the motors in this program was satisfactorily predictedby the Solid Performance Program (SPP). In many cases the actual values of measured stability were correctlypredicted. In all cases the trends in measured stability were confirmed by the Standard Stability Program.Perhaps the most important input for the program (and one of the most difficult to determine) is the correctpressure coupled response function for the propellant. Also, critical for accurate stability analysis is thecorrect burning rate of the propellant in the motor.

Effects of pressure, motor Length or frequency, additives and geometry were all examined. The pressureeffect which was predicted and validated with data showed that increasing pressure reduces the margin ofstability. Conclusions reached in varying motor length were difficult to determine due to the change inbaseline propellant properties. The effect of stability additives on nonlinear stability was significant. Thedata indicated that stability additives were effective in increasing the nonlinear stability margin. However,differences between stability additives were not evident. Finally, motor geometry comparisons were made thatindicated that a star forward geometry was more stable than a star aft geometry. This conclusion was validatedboth experimentally and theoretically. In addition, the constant cross section geometries (full star and fullcylinder) both were observed to be very sensitive to pulsing and resulting instabilities.

Future plans included pulsing motors at higher pressures and evaluating newer formulations of reduced smokepropellants. This mutti-year program has generated an extensive amount of nonlinear combustion instability testdata. This data is currently and wilt be used for present and future use in understanding nonlinear instabilityand in validating nonlinear computer models.

-250 - 0 SWrT 300Hz SO~puL ,' /* 5&rT 300Hz IO00pL+ SRrT 300Hz IS0(0,.L* SbWO 300Hz 1000pL* SAFT Bus RL203

SV SqfT ZrC -,B S*FT srur AL203 0

-i 2000 SRFT 600SzOOpaL 00 SF 600Hz 500paL " "3 C'L 600HzSOp. ""-- "

PERrFtT FGREEDENTS+5Z ERROR O • , x-150 - --- 3.ý-'ERROR -tDO- -- "

D A

•-1 s-o -00,

100

E - 5

0

0 -50 -100 -150 -200 -250

EXPERIMENTAL ALPHA - I/SEC

FIGURE 17. Measured Versus Theoretical Decay Alphas for all aotors

133

REFERENCES

I. F.S. ilomhietd, J.E. Crump, H.S. MatheS, R.L. Derr and M.L. Seckstead, "TTCP Won-Linear InstabilityCollaborative Progrmn--NVC Participation," 22nd JANNAF Combustion Meeting, CPIA Pub. No. 432, Vol. 2, pp155-164, 1985.

2. H.S. Mathes, "TTCP Non-Linear Instability Progra: U.S. Motor Data," 23rd JANNAF Combustion Meeting, CPIAPLt. No. 457, Vol. 1, pp. 59-64, 1986.

3. U.H. Clark, J.E. Crump, H.8. Math", and M.L. Beckstead, "Motor Test Data Comparisons with Laboratory Dataand Stability Predictions," ?4th JANNAF Combustion Meeting, CPIA Pub. 474, Vol. 1, pp. 101-114, 1987.

4. UI.H. Clark, J.E. Crum and Ht.S. Mathes, "Results of Tests for Non-Linear Stability of Full Scale RocketMotors," 25th JANNAF CQmbustion Meeting, NASA Marshall Space Flight Center, Huntsville, At, 1988.

S. 1H.S. Mathes, J.E. Crump, F.S. Blomhield and M.W. Beckatead, "Non-Linear Test Results of Full Scald RocketMotors," 26th JANNAF Combustion Meeting, CPIA Pub. 529, Vol. IV, pp. 165-176, 1989.

6. f.S. *lonshield, C. BRlter, H.S. Mathes, J.E. Crump and M.U. Seckstead, "Nonlinear Stability Testing andPulsing of Full Scale Tactical Motors," Proceedings of the 27th JANNAF Propulsion Meeting, Anaheim,California, October 1990.

7. R.W. Hermuen, J.T. Lafberty and R.E. McCormick, "A Computer Program for the Prediction of Soltid PropellantRocket Motor Performance (SPP)," Volurme V., AFRPL-TR-84-036, September 1984.

8. G.R. Nickerson, F.E.C. Cutlick end L.0. Dang, "Standard Stability Prediction Program for Solid RocketMotors,- AFRPL TR-83-017, September 1983.

134

NONLINEAR STABILITY 7 ING AND PULSING OF FULL SCALE TACTICAL MJT1DS

F. S. Blomahield id C. A. BeitrResearch Department

Naval Weapons Center. China Lake. Calif.H. B. Mathes maI J. E. Crump

COMARCO, Inc., Ridgecrest. Calif.

M. W. Beckstead

Brigham Young University, Brigham City, Utah

ABSTRACT

This paper reports results of an ongoing effort to develop an improved understanding of linear and nonlinearcombustion instability in solid propellant rocket motors. One goal of the program is to develop a systematic datbaseof motor stability data. Earlier papers have reponetd previous NWC work on this program. This paper describes thenonlinear aspects of the motor firings and analysis. Subjects discussed include motor test parameters and methods usedfor acoustic pulsing of motors. Nonlinear motor resposme to pulsing. pulse amplitudes, and characteristics of sustainedlongitudinal chamber pressure oscillations are summarized.

IWITRODUCION

An experimental test program is under way at the Naval Weapons Center (NWC), China Lake, Calif, that involvesresearch into linear and nonlinear acoustic instability.'-6 The term "nonlinear" as used in the context of combustioninstability can describe several phenomena including: processes that lead to limit amplitudes, presence of highamplitude shock-fronted pressure waves, triggered instabilities, ad mean pressure deviations associated with acousticwaves in a motor. Considerable attention has been paid to this form of instability in recent years because ofpotentially severe consequences to motor operation when it occurs.

The possibility of occurrence of a "triggercd instability" is of particular concern to missile designers and usersalike. To the user in the field, the consequences of a aiggered instability can mean failure of that motor's mission. Atriggered instability can occur in a motor that. without the triggering disturbance, is stable. Once triggered, the motorremains violently unstable until the end of bum. Triggering can occur from nozzle ejecta or from chamber pressurepulses induced by other means. In nonlinear instability research, ejects or pressure pulses are deliberately introduced atpredetermined times and at desired initial disturbance amplitudes as the motor bums to test the motor's response. 7-4l Inthe present program, pyro pulsers rather than ejects were utilized to create pressure pulses in the motors.

Triggered instabilities have been under intense study and analytical techniques are available for ad hoc assessing ofvarious aspects of this form of nonlinear motor behavior.tZ IS In the following, we examine some experimental resultsof triggered instabilities obtained in the present program and discuss methods used in pulsing the mourt.

APPROACH

The motor testing portion of the program involves ruing 5-inch-diameter motors loaded with a reduced-smoke AP-HTPB baseline propellant formulation. A number of internal motor configurations, several minor variations of thebaeline propellant formulation, two motor lengths, arn several mean chamber pressure; have been used in the testprogram. The program utilized readily available hardware, but a specially designed forward closure, as well as somecustom nozzle sizes, were required. Mot! .. c.,,ors were acoustically pulsed three times during burn with a variety ofpulse amplitudes. The pulses were genPrk :.- by "pyro" or "blowdown" pulsers of conventional design. All motorswere equipped with high frequency response pressure transducers in addition to standard ballistic pressure transducers.

The objective of pulsing motors was to test motor stability to acoustic waves. Motor response to pulsing fallsinto two categories: (1) the motor is stable to a pulse (i.e., acoustic waves generated by the pulse decay with time), or(2) the motor becomes violently unstable following a pulse. Stable motor response is disuassed in a previouspapers.'. 6 Unstable response to pulses and methods of pulsing are the subject of this paper. Also included are pulsingdata for ten motors that have been tested in the past year.

*This effort was sponsored by the Air Launched Weaponry Technology Block Program Office under the authority ofTom Loftus, Code 372, Naval Weapons Center. China Lake, CA 93555-6001.

Approved for Public Rcleasc; distribution is unlimited.

135

COPPER WASHER -----. I PIEZOELECTRIC

WRENCH FLATSURE_

RECESS FOR

POWDER CAVITY-- ELECTRIC SQUIB

1-2o THRAD• ,PULSER BODY

BURST DIAPHRAM(UPSTREAM)

,,--ORIFICE PLATE

THERMAL BARRI ER•-\•

BURST DIAPHRAGM10TRD:(DOWNSTREAM)

1- 20 THREAD ,--

3/4-20 THREAD-_' ------ MOUNTI NG ADAPTER3/4-0 THEAD-MATES WITH

FORWARD CLOSURE

Fig. 1. Pyro Pulser Components.

PULSER DESIGN, PERFORMANC.. AND APPULCATION

PULLER DESIGN AND COMVPO•NNE

The pyTo pulser used to introduce pressure pulses into motors is shown in Fig. 1. A small oe-ivy-walled stainlesssteel body is bored through and counter-bored to provide for a powder chamber, an electric squib. and a method formounting it on a motor. A squib is potted in the pulser body with room-temperature vulcanizing (RTV) rubber. Thesquib alone may provide sufficient energy for pulsing a motir or. should additional energy be needed, the squib mayignite a charge of propellant placed in the bore of the pulser body. The end of the body opposite the squib is threadedfor a mounting adaptcr. The adapter clamps an orifice plate and one or two burst diaphragms between itself and thepulser body. The other end of the adapter threads into a hole in the forward closure of the motor. A thermal barrier isinserted into the adapter to protect the downstream diaphragm from adverse effects of exposure to hot gases in themotor. Some of the parts in the pulser were varied to suit the needs of a particular pulsing situation as explainedbelow.

Pulser performance varied de-pending on the powder charge weight used, the orifice diameter, and the placement andthickness of burst diaphragms. The powoer charge, when used, was cannister grade "Red Dot", a doub!e-base flakepropellant commonly used in the loading of shotgun shells. Orifice plates were made of 0.125-inch-thick stainlesssteel. They were bored and stamped to indicate the hole diameter. Diaphragms were made of commercial half-hardbrass shim stock. Tests of the shim stock, run separately from the tests of the pulsers, showed that samples of shimstock from the various packages had similar shear moduli. Thus the burst pressure was a linear function of thicknessfor a given orifice diarnetey. A "downstream" diaphragm was always used in conjunction with a charge of Red Dot tosCal the pulscr interior from invasion by hot gases from the motor. As the program progiessed, a thermal protectorwas placed between the downstream diaphragm and the motor to reduce the possibility of diaphragm softening due tocontact with motor gas. Several thermal protection schemes were investigated. The method finally chosen was a 0.1-inch-thick disk of white polystyrene foam. The disk was a light :Tess fit in the adapter port- It was pushed againstthe diaphragm as the ruial step in pulser assembly. The disks werc of norninal 0.5-inch diameter. Each disk weighedabout 22 milligrams.

i 136•

The initiator was a modified Navy Mark [1 electric squib.which consists of a coined copper case. a base assemblywith two external lead Wit" anid an internal bridge wire. The source of squib energy is a 90-milligram charge of 4F-Gblack powder mixed with BKNO% to provide a specified brisance. (The standard Mark UI Contains no BKNO3.) Outerdimensions of the squib's body are 0.28-inch diameter and 0.43-inch length.

PULSER TS7N

Prior to use in motor tests, each pulser load was evaluated in a laboratory facility using steel tubes pressurizedwith helium. The tubes were instrumented with pressure transducers to measure the pressure wave amplitudes generatedby the pulsers. A variety of pulser load combinations were tested. Each combination of powder charge, diaphragmthickness and arrTangement was given an alphabetical load ID as shown in Table I.

Table 1. Summary of Pulser Loads aLoad ID Charge Weight Diaphragm ickness (in) Orifice Diameter

0X Upstream Downstream (in)A 0.0 None None 0.355B 0.1 None 0.002 0.099C 0.1 None 0.008 0.144D 0.2 0.008 0.008 0.144E 0.1 None 0.008 0.199Fe 0.1 0.008 0.008 0.199G 0.1 None 0.008 0.295HO 0.1 0.008 0.008 0.29516 0.0 None 0.004 0.1441 0.1 0.004 0.004 0.144K 0.1 0.008 0.008 0. 1 dL 0.0 None 0.002 0.0,S 0.0 None None 0.099T 0.0 None None 0.144

These pulser types were not fired in motors

The pulser laboratory tests were analyzed using theories developed by Hercules, Inc. and Aerojet Tactical SystemCo. 16.17 A brief description of each theory wil be followed by a comparison with experimental data.

Hercules Analysis. In the lIercules analysis. the pulser chamber immediately aft of the rupture diaphragm/orificeassembly is treated as a classical shock tube. The shock forms at the orifice plate. travels through the motor mountingadapter (Fig. 1). and expands in a quasi-spherical manner into the test chamber. Figure 2 illustrates a simplifiedschematic of the pulser test setup and motor cavity.

pressure transducer

"C' a c

s- hock tbPC

)'ch tc P ~ jppulser v 3 motor chamber

burst diaphragm

Fig. 2. Simplified Schematic ,f Pulser Test Setup.

y. a.. and D are the specific heats, speed of sound. and diameters of thr. pulser, subscript ch, and motor chamnber.subcript c, respectively. At the instant that the diaphragm bursts, the relarion between the rupture pressure p, andshock amplitude p, is given by the classic shock tube equation:

I 137

PC PC L 2 Tc+(y.+1)(p./ptl)

The magpitude of the shock wave decreases as it expands in a quasi-spherical manner to fill the test chamber. Thepulse amplitude is defined as the shock wave amplitude once it has expanded to fill the test chamber. The wave then ischamcterized as a traveling planar shock wave. The effect of the quasi-spherical shock wave expansion is given by

ýp.ý'-) - 1 2 (2)

I +l.75()

Thus, given the pulser and test chamber dimneters, the rupnure and test chamber pmrssures. an the necessary ratio ofspecific heats and speeds of sound, the effective pulser amplitude 6p can be computed from eqrw. (1) and (2).

In reality, the expansion of the shock wave to fill the test chamber is a complex. three-dimeiuionalevent marked by a chaotic inflow of hot pulser gases. The organization of Ote shock wave int.' Ine travelingwave does not occur until several chamber diameters downstream. Due to this complex proces itial pressureamplitude measured by the transducer is substantiaUy less than the first reflection of the Waveli.6..- .ve. This initiamplitude can be thought of as the attenuated effective pulse amplitude. Dacay of the frust and subsequent reflections ofthe traveling wave occurs approximately in an exponential manner. Extrapolation of the exponential envelope to theinitial pulse time gives an approximate value of the effective iatuenuated pulse amplitude. Figure 3 shows thisexponential decay and Fig. 4 shows the exponential curve fit beck to zero time and effective pulse amplitude.

I\I \

Effective Approximate exponential decay of reflected traveling waves

unsitenuated

amp tude

Fig. 3. Example of Test Record for Pulser Tests. Transducer at head end of test chamber.

5.2

0o In p - 5.1837 - 49.481t, RA2 - 0.9895.0 In p(O) - 5.1837

4.8C.

- 4.6

4.4

4.2

4.00.00 0.01 0.02 0.03

time (sec)Fig. 4. Example of Data Reduction for Pulser Tests.

138

AmkLaba~. In the aralysis developed at Aarojec.17 ma mid morgy balaws eoquations us1 applied to diecombusimm chamber. These take the form of two orditwy differential equations with pressure aid temperatue asdependent variables. and when solved smuhsasouly. yield the amplitude of the pulise inooduced into the gast chamber.The -ul amplitude~ &p. is reiated to the nam of praiser gases bsjechidntos the wea chamber. &m r via:

x .I A. L (ii) '

where 0 is a ountant of propoutiomality, s an md y is the speed of sond mid qpaefic heat in the pulser. A, mad L anthe teat chamber cross-sectiotmal area end length. t, is the praiser vast time. mid k is the tinie period of the bastchamber's first longleudbial mode. It should be noted thw in the Aerojet wialysis the convuuAilo of defunin the pulseuniplitds is the extrapolated value (as previously described) dividedf by two. This facmo of two wises from a smideicyof a prease tranithacer mounted normal to the direction of tho traveling wave to onewue ewaly twice the senplitude ofthe reflected wave (see discusseon pp 39-42 of Ref. 17).

Comnatimo WMit Emoeriments. Table U lists the physical coristarts utilized in both analyses. Pulser loads A. S. andT (see Table 1) wen omitted from the anaysis due in their lack of bumt diaphragnis. making thern inappliable ioeither theory. The determination of fth proportionality coristar. 0 (Acrojet analysis), was determine empirically. Acomparison of the experimental mid predicted pulse amplitudes for both theories ms shown in Figls. 5 and 6 midtabulated in Table M. In general, the Hercules theory gave better agreement with experimental results. Except forloads C. 1. J. B. and part of the wtets with load . the majority of Wrms were Less than 120%.

Table 11. Physical Constants Used in the Pulser Performance Analysis.

Chre I Charse 2 Charge 390 mgsquib 90mgnSsquib 90Omg squib

_______100 __ _ _ _ Omg Red Dot 200 mlRed DotRuespective loads L L B. C, E.F. 0.H.J. K D

&.h,. pulser gas sonic 27769 37607 39723velocity (in/s&")

R. pulser &a; constan 420 537 573

y. putserSas ratio of 1,1697 1.1925 1.1945specific heatw______ _______

139

300° .

0

250W)

500

0 50 100 150 200 250 300EXPERIMENTAL PULSE psi

MSg. S. Comparison of Espcrimw"aua d Predicte Pule Amplitudes. Hercules Theory, 201i ETTor Banids.

200 " "

APO

I 150 - -," -- _ _ __ __ __0 ii

100

Oz., ~ ~ ~ f L l,',olin

ao 050

00 50 100 150 200

EXPERIMENTAL PULSE - psi

FiS. 6. Comparison of Experimental mad Predicted Pulse Amplitudes. Acrojet Theory, 209 Error Bands.

140

Table M. Compn *son of Hemules mnd Aemroet Anal wis on Pulser Bench Tests.Hmuiles Aerojet

Test Pulse (psis) Predicted pulse (psi&) O.SxPulse (psis) Predicted pulse (psia)82C 106.4 253.5 53.2 113.383C 113.8 266.1 56.9 120.084C 98.5 259.4 49.3 116.485D 189.2 214.1 94.6 94.486D 184.9 200.7 92.4 88.187D 177.5 161.8 88.8 73.188 148.5 121.9 74.3 89.589E 128.4 109.6 64.2 89.4"90E 136.0 114.9 68.0 87.091F 176.3 162.2 88.2 106.892F 174.0 164.2 87.0 107.693F 171.0 162.2 85.5 106.894G 187.6 169.0 93.8 191.495G 184.0 164.6 92.0 188.796G 197.1 173.4 98.. 192.097H 159.7 157.4 79.8 183.898H 151.3 161.5 75.7 186.1"99H 169.7 167.6 84.8 189.21011 68.4 30.6 34.2 26.21021 63.2 25.6 31.6 26.81031 63.0 22.5 31.5 28.01071 97.6 45.0 48.8 16.01081 75.6 46.8 37.8 16.31091 71.7 46.8 35.8 16.3I1OK 112.9 91.1 56.5 24.9IIIK 110.7 91.4 55.3 25.0112K 122.5 90.7 61.3 24.8113D 138.4 103.6 69.2 28.3114D 135.6 104.3 67.8 28.411SD 126.1 102.2 63.0 27.9116H 106.9 99.0 53.5 69.1117H 106.0 105.9 53.0 71.7IISH 112.0 103.2 56.0 70.7119D 74.6 78.6 37.3 21.5120D 77.8 79.7 38.9 21.8121D 77.5 79.5 38.8 21.8125C 55.7 38.1 27.9 10.9126C 53.9 37.7 27.0 10.8127C 50.4 38.5 25.1 11.0128D 60.6 69.3 30.3 18.8129D 55.7 67.1 27.9 18.1130D 76.4 69.8 38.2 18.9131E 63.8 53.1 31.9 20.9132E 54.1 51.9 27.0 20.6133E 53.2 53.1 26.6 20.9134F 70.3 77.0 35.3 28.613SF 73.6 78.7 36.8 29.2136F 71.1 79.5 35.5 29.5137G 39.1 46.3 78.2 79.11380G 72.5 80.5 36.3 46.71390 71.1 74.6 35.5 44.4140H 81.3 92.2 40.7 52.2141H 68.2 96.7 34.1 53.9147H 75.2 93.7 37.6 52.8143C 44.1 17.9 22.0 14.2144C 45.7 17.9 22.9 14.2145C 48.5 20.0 24.3 13.9146C 48.9 20.3 24.5 13.9147E 91.4 62.6 45.7 30.71488 117.0 66.7 58.5 31.6149E 90.8 60.3 45.4 30.215OF 115.6 95.6 57.8 38.815IF 97.1 105.2 48.5 41.4

141

Table M11. Comvwi... of Hercul an Acre y ao Puber Bench Team (Cont'd.).Hercules ASmia

Test Pu.se (paia) Predid pulse (*is) 0.SxPumae (vsia) Predicted vule (pais)152F 112.1 98.1 56.0 39.51530 111.9 88.6 56.0 66.31540 123.3 97.3 61.7 69.01553 112.3 98.7 56.2 68.9159E 130.0 70.0 65.0 41.4160P 112.5 65.7 56.3 40.7161E 123.9 69.6 62,0 41.5162F 49.4 59.1 24.7 15.0163F 50.7 55.0 25.4 14.3164F 54.9 62.4 27.5 15.61650 58.2 57.3 29.1 26.0166G 54.7 56.7 27.4 25.91670 54.7 57.7 27.4 26.3168D 77.4 63.5 38.7 10.9169D 68.4 62.8 34.2 10.8170D 66.8 64.8 33.4 11.1171B 10.5 13.2 5.3 2.6172B 10.4 11.5 5,2 2.6173B 9.7 14.5 4.8 2.7174B 8.1 13.9 4 " 1.8175B 7.0 14.0 3.5 1.8176B 9.6 13.9 4.8 1.8177F 39.0 46.4 19.5 10.8178F 37.9 45.8 19.0 10.7179F 41.9 50.3 21.0 11.6

ADAPTING TO MOTOR CONDITION

The test hardware for evaluating puiser load performaiwe provided a basis for evaluating the pulse pressureamplitude in a test chamber of known cross-sctional area. Conditions in motors, however, involved pulstr pressurewaves in chambers (defined by the motor's interior) with cross-sectiona areas different from the laboratory tests. Theprocedure for determining which pulser load was best suited to give the desired pressure amplitude in a motor at a givenpoint in burn was as follows:

1. Pulsing times were initially chosen in the baseline motor (Star aft. 1000 psi) for 1.00. 2.00. and3.00 seconds.

2. Pulsing times in other motors were based on the fractions of web bum established by the baseline motorpulsing Linmes.

3. Pulse amplitudes were determinod for each motor based on prior experience regarding susceptibility of a motorbeing triggered unstable. As a rule, pulse amplitudes were tailored to be progressively weaker as burn progrested.

4. A given web burn defines a cross-sectional ar in the motor chamber. That wes typically differed from theares of the laboratory test apparatus in which pulsers had been initially tested. In order to account for the effect ofcross-sectional area on pulse strength, the pulse strength desired in the motor was scaled back to the test apprawua.The scaling rule was that the pulse amplitude was inversely proportional to the cross-sectional wree of the chamber.

5. When the motor pulse amplitude was saed to an equivalent test chamber amplitude, the test data were reviewedand a pulser load chosen that came closest to the calculaed test chamber amplitude. That load was used for the motortest.

The process of choosin& a pulser is illustrated in Fig. 7. Of the twenty pulser load combinations tested in thelaboratory only about half was used in motors. A listing of pilse loads used on motors is shown in Table 1. Note thatsome of the loids consist of the initiating squib alone in association with an orifice plate. Such combinations werenecessary to achieve a low pulse amplitude for some motors.

142

________ ret pu___ a lse IVLMto bTs Matru __ _ _

1st ong. Nomial Sar Sar Fll Cl indePressre At Fw. Str _Type _

Pu00 2otor

Fig.in 7.pelau PusHSletonPrces

TbeI.MotrR TestMtrx

folsw itheet L mo tor. Noemiadrl wSe:atu lll t Star a Fprt ll taadaCylinder. w oo eghwee Farca end. Mo rosc sur int 66Ficwogcae siltd. aSafnaetar ogtdnlfeunyo

approxmat0l 332LTeewr eintda 0 zmtr.Ma ftetsswr odce nte30Hmotors. Fou tes0s use (30)c~lu 1oo t oeaiebhvo thihrfeuny h 3-nhmtr

A0ln-lds oosw th og m1c oMi- flith motor.90 h10.ardwe waZ dpe roIvial sucsCh

9.The closrex rpeln was bcaed and thede oac-ommdted aicnto al plachnede ister, mowe csstad. ballsting wessr

traUsduerL~ twohig fres uento r es. nse prlsuwre: tansducl-enghsta, aprtlsa. and ahe cuyrslet ntindertto. sp mto e ngtly

aotrrage Four theete sts, 3included pulse fiing et chaminel andvio muti-ighnel frocuessncy ofThe high frequencys

reurmnsfrmlpepligwdfrhg rqec epnepressure transducers rsle intedsignals.

143

A ** *:~ * *STAR-AFTA A

(eBASEL INE)

.. ~ * *STAR-FORWARDA-A B-B

@ FULL STARA" A-A

itl•Ai0 FULL CYLINDER

FORWAR AE AFT EM A-A

Fig. 8. Motor Geometry Variations.

i

F IRE CONTROL CHANNELS

IGNITINPULSE FIRING

PULSER ioPtA.SER 29PtLSER 3_-

IONITIONO[OTQR PRESSI.JRE TRNM•JOMESPC- I*

PK- 1020

PILLSER PRESSRE TRANSUCEWRSpP- i *-

PP-20PP-30

FORVARD CLOSURE (SCI-EMAT IC

Fig. 9. Schematic of Forwud Closure and Test lns'Lunentation.

144

Pulser Performance in Motors. Results of motor testing showed a variety of pulsing behaviors. Few pulse

pressures measured in motors were " desired. There were a variety of causes: (1) not all planned pulse amplitudes couldbe matched with an appropriate load and (2) pulse timing wu often different from the times that were w.L The lauerproblem occurred mainly due to variations in motor igniter performance. In one cae (motor 13) pulse times wereincorrectly entered into the timer and for motors 12 and 13 the pulsers were inadvertently interchanged. Pulser types. acomparison of pulse amplitudes, pulse times and resulting motor stability alphas are presented in Table V.

Table V. Pulser Performance in Motors.

Motor Info. Pulser Pulse Amplitude in Pulse AlphaMotor Chamber (psi) Fire Times (1/se)

No. Geom/Press N. Type Planned Measured Planned Actual Growth DecayI SAFT3/ 1 B 50 NPF 1.000 *..-

1000 2 C 20 36 2.000 2.016 2093 B 10 OSC 3.000 -...

2 SAFM3/ 1 B 50 12 1.000 0.703 -1591000 2 A 20 12 2.000 1.708 207

3 B 10 OSC 3.000 ...

3 SAFT3/ I B 50 14 1.000 0.805 -1121000 2 C 20 50 2.000 1.790 161

3 B 10 OSC 3.000 .-. -.

4 SFWD3/ I D 50 38 1.000 0.970 -19)1000 2 C 20 22 2.000 1.962 36

3 B 10 OSC 3.000 L. ...5 SAFr3/ I D 50 100 1.000 0.9o5 .i"6

1000 2 -- 0 NPF ..-..3 -- 0 NPF ..

6 SAFT3! I D 50 18 O.G W500 2 B 20 7 2.250 :.%o05 -I2

3 B 10 6 3.250 3.010 -757 SJU.L3/ I E 50 38 1.000 0.748 6(9

1000 2 G 20 054% 2.000 ..3 A 10 OSC 3.000 .. - -

81 SAFM3/ 1 - 0 NiT 1.000 ... .1000 2 -. 0 NPI 2.000 ... .

3 A 10 10 3.000 3.020 1348b SAFr3/ I D 50 50 1.160 0.950 - -210

500 2 J 20 20 2.400 2.216 -893 G 10 32 3.550 3.419 366

8c SAFr3/ I C 50 20 0.850 0.611 1511500 2 B 20 OGC 1.160 .. .

3 A 10 OSC 2.350 ... .

8d .... PULSES DID NOT FIRE ----

9 CYL3/ I K 50 20 1.000 0.745 232 -1000 2 G 20 OSC 2-0OU-.0

3 B 10 - OSC 3.00010 CYL3/ I D s0 36 - 1.14 V " 244

500 2 E 21) OSC 2. ; -

3 G 10 OSC 3.50 ._I1 SAFT66/ I J 50 40 1.24 r 2J6, - -143 -

1000 2 B 10 6 2 .2 3 "19& -753 S 5 4 3.1 1 -60

12 SAF"6/ I T 10 12 1.0 .374 1 ýIz1500 2 B 10 6 1.85 2,543 *-47I

3 S 5 NRD 2.64 - "-"- - L -

13 CYL61 I B 10 8 1.03 1.261500 2 B 5 8 1.85 2.429 NRD NRD

3 S 5 4 2.64 3.445 -7114 SFWD6/ I J 50 50 1.24 1.945 -201

1000 2 B 10 8 2.22 3.465 -1333 S 5 2 3.13 4.898 -33a

NPF - no pulse fired NRD - not reducible OSC - oscillating a - poor data

!45

Table V. Pulser Performarne in Motor. (Con.d.)Motor Into. Pulser Pulse Amplitude in Pulse Alpha

Motor Chambr (,psi) Fire Times 0!ue)No. Geom/Press No. Type Planned Measred Plarted Actual Growth Decay15 SAFi'3/ I K 75 174 1.00 0.941 -122

l% 8 P 1000 2 C 20 40 2.0o 1.764 .- 4S3 S 5 NRD 3.00 ...

16 SAFT3/ 1 K 75 52 I.0o 0.824 -2321% ZrC 1000 2 C 20 26 2.00 1.505 -137

• 3 S 5 NRD 3.00 ... .-17 SAFT/ I K 75 70 1.00 0.866 -188

1% 90 1 n 1000 2 C 20 46 2.00 1.669 -75A]• 3 S 5 2 3.00 2.572 -liOa

19 SAFT3/ I S 10 NRD 0.75 .2000 2 S 5 2 1.41 1.817 -26s

3 S I 2 2.09 2.767 -64a20 SA-IT3/ i C 40 16 1.00 0.984 -193

1000 2 B 10 4 2.00 2.167 -813 L .- NPF 3.00 -. .. -

21 SAFT3/ I C 40 41 1.C0 1.012 -2511000 2 B 10 10 2.00 2.195 .143

3 L 5 4 3.00 3.506 NRD NRDNPF - no pulse fired NRD - not reducible OSC - oscillating a - poor data

Summary of Motors Pulsed Unstable. All but oie of the 23 motors tested were pulsed at least once during burn.Most were pulsed three times during burn. A wiiring effel on motor Sd resulted in no pulses being fired although threehad been planned. Pulses on a few motors were n3t found or it, the case of motor 20 a pulse (the third) did not file.Ten cr the motors were triggered into high amplitude unstable operation following a pulse. Ir. each case theoscillatory waveform assumed a jagged shape, the oscillations persisted throughout the remainder of burn. and asignificant DC chamber pressure shift was associated with the oscillations. A pressure-time curve from a motor number4 that was pulsed unstable is shown in Fig. 10 indicating the behavior described.

1800 -

1600

I 1400

W 1200

•) 1(00PULSE I

600

400 -_

0-0 0.5 1 1.5 2 2.5 3 3.5 4

TIME - sec

Fig. 10. Pressure-Tirn li tory of a Motor Pulsed Unstable (Motor Nr. 4).

146

A summary of parameters related to nonlinear instability for mobis that were pulsm untasble is presented in

Table VI. The terms used ui that table a describ-d in nows in the tole and are graphicially depicted in Fig. 11.

Table V1. Oscillation Amplitudes and DC Pressure Shifts in Motors Pulsed into Nenlineair nstability.

Motor Information 1 ri)•Lring ilulse and Alpha Oscillation Parameters TimeNominal Ph Pd. Pk of'

Motor Coufig. Chamber Pulse Time Pulse a P. Pta P P p PP Meas.No. Pressure No. (see) (psi) (I) (psi) (psi) (psi) (___)I SAFr3 1000 2 2.016 36 209 830 750 510 0.90 0.61 0.56 2.282 SAFr3 1000 2 1.708 12 207 770 670 455 0.87 0.59 0.55 2.403 SAFT3 1000 2 1.790 50 161 FlO 820 410 1.01 0.51 0.60 2.204 SFWD3 1000 - 1.962 22 56 830 580 370 0.70 0.45 0.80 2.607 SFUL 1000 1 0.748 38 639 1053 550 267 0.52 0.28 0.42 1.50Is SAFF3 1000 3 3.020 10 134 835 825 583 0.99 0.70 0.58 3.25SbO SAFT3 500 D 3.149 32 366 556 133 102 0.24 0.18 0.20 4.008b" SAFT3 500 3 3 149 32 366 556 58, 338 1.05 0.61 0.65 4.209c SAFT3 1500 1 0.617 20 151 1250 1129 890 0.90 0.71 0.53 1.259 CYL3 1000 1 0.745 20 232 765 4SO 272 0.59 0 36 0.42 1.50

10 CYL3 500 1 0.887 36 244 544 289 233 0.53 0.43 0.37 1.75" Motor 8b oscillated mildly and then more strongly as indicated by two sets of resu.jP,. = Motor chamber pressure at the tiure of the initialing pulse (psi)P., Limit amplitud, of the acoustic oscillations (psi p-p)

. Shift of motor ch,-.,oer pressure associated wit acoustic oscillations (psi)

DC PRESSUfnE SHIFTPULSE 1

PULSE 2

NORMAL PRESSURE~

Id

TIME - -

Fi,. 11. lilt sitation of Terns Used in Table VI.

Mention was made previously regarding the presence of sharp- or shock-fronted pressure waves in a motor ks asymptom of nonlinear instability. An esimple is provided ir, Fig. 12 to illustrate the point. The data is from motorno. 4 but is typical for triggered iista'nility sen in this type- of testing. The triggering pulse 's indicated in Fig. 12.The pressure amplitudes between 1.95 and '..05 seconds ate mplified and shown in. greater deLai! in the upper left hand-of. -r of Fig. 12. The pulscr ail high frequcncy pressurc tr.nsducer are located side by side. Therefore, the bursting

of the pulser ini'.ally generates an outflow of gas Pial rushes dowp the motor bore toward the aft end. The rapidity ofLis discharge from the nulser generates a high frequency acoustic oscillation in the region of the pressure trtasduce-lital gai flow from the pulscr is dirccted down .. c a8Xs of t-e motor. Thus. diiz pressure transducer docs not "sce awel organized pre;sure wave uri; the initial diswtubanrc travcls the length of me motor and r -.- Rs to crei .e whit iscalled the first reflection. The growth of the finri and suhsequent refc•.tons of the traveling wave occursapproximately in an exponential mannei. Extrapolati-in o." the -xponcnfial eniclopr tn Ohw ume of the initicdmeasured amplitude gives an approxtmn:e value of the e,.cctive unsuenuated pulse amplitude. A similar sequence ofevents is seen in laboratory testing of pilsers with sid.: by side mountin-- of pulser and tranmducer (see Fig. 3).

1600

1400 1I H' IJ

12P0

Cl)

800 PL) AA& I\A I 1h1.94 1.(G 1.98 2 2.02 2.04 2.06 2.08 2.10

TIME -- sec

Fig. 12. Details of Pressure Waves in a Motor Pulsed Unstable (Motor No. 9).

SUMMARY

A total of 23 motors have been fired. Most were pulsed three times during burn. Ten were pulsed and becameunst3ble AS a result of the pulse. In deciding how to pulse the motors, the guiding rule was to choose pulse amplitudesc;as- to the point that would trigger nonlinear instability. Experience with pulsing motore indicated that the motors:tsted would be more resistant to pulse triggering early in burn and les.s resistant later in burn. Thercfore. plans forpulsing generally resulted in relatively strong pulses early in burn and progressively weaker pulses as burn proceeded.As -night be expected. motor response to pulsing was difficult to bracket for many of the test conditions since onlyone or two motors were fired per condition. Obtaining such data typically requires multiple tests under a given set ofmotor conditions in which the pulse amplitude is systematically varied. We refer here to go no. .-. sensitivity, orBruceton data. which result from up-and-down testing.'8

Motor test data obtained in the present program substantiated earlier etperience. which indicated that mnotors wouldbe relatively re~sistant to being triggered unstable early in burn as compared to susceptibility to triggering late in burn.The trends observed in these tests seem to follow the linear stability predictions discussed in our previous papers. 3.4.6

All .he motors are stable to a strong pulse when the lInear stability alpha is lasg.- but unstable to a similar pulseamplitude later in bum whet, motor stability is lower. As noted above, test results obtained to date do no' providemuch quantitative information concerning amplitudes that bracket conditions for triggering nonlinear instability. Anexception is the data for the 300-Hz star aft (SAFT3) rnotors at 500 psi (motors 6 and 8b) in which a 6-psi pulse 6idnot trigger instability but a 32-psi pulse did.

The principle intent with Motors 15.17 was to pulse them in the samne marnner as the baseline motors to see if theadditi',es had an effect on nonlinear in.;tability. Unfortunately, the strengths of the first pulses (174, 52, and 70 psi)were not. in general, comparable to the first pulses of the baseline propellant (0, 12, and 14 psi), see Table V.However, the fact that these motors remained stable to more severe pulses while weaker pulses caused the baselinemctor to go unstable is significant. This documents the powerful role stability additives can play in combustionitistability suppression. For the second pulses, if one considers the pulse amplitudes or the additive motors 15. 16.and 17 (40, 26. and 46 psi) to be nearly comparable to the second pulse in motors 1. 2. and 3 (36. 12, and 50 psi),ore coutld c )rclude that th, additves did increase the. stability margin sinrc the pulses in the additive motors decayedwhile .he puises in motors 1. 2, tnd 3 grew. The lower value of the response function for higher modes for theadditive propellants also supports thii oncluslon. since a pulse could not as eusily excite these modes into anonlinear ;r.•.tiility. 6 The particle damping computed for the first longitudinal mode is very iow. 6 The particledamping will be more effective for the higher modes and will help damp out this form of nonlinear pulsed combustioninstability. For the third pulse there was no comparable data.

i-.i,

The data in this program suggest that susceptibility to triggering is influenced by interior geometry of the Metor.For example, tests on the 300-Hz full star (motor 7) and on the motors with cylindrical perforaiions (motors 9 and 10)showed those motors to be more susceptible to triggering than any of the other 300-Hz configurations. Both thesetypes of motors had constant cross sections and were symmetric front to back. Motors 7. 9. and 10 all went unstableon the first pulse with 38. 20, and 36 psi pulses. Other 300-Hz configurations successfully withstood first pulseamplitudes in C range of 40 to 70 psi.

Unfortunately, none of the 600-Hz motors (11-14) were pulsed unstable nor were any of the motors with stabilityadditives pulsed unstable (motors 15-17). If an opportunity should arise to fire more of those motors, stronger pulsesthan were used in the tests documented herein will be required to produce nonlinear instability.

This program also allowed the opportunity to evaluate pulsers and various design tools to estimate pulseamplitudes. Both Hercules and Acrojet theories were applied to laboratory pulse daa. The results indicated goodagreer.ent. usually within 20 percent, of the measured pulse amplitudes. The Hercules theory was slightly better forNWC's type of pulser.

The DC shifts in the chamber pressure, caused by acoustic wave action on propellant burning rate, are stronglyinfluenced by the acoustic wave amplitude. Thai influence is seen in Fig. 13 where the DC pressure shift (Pd,)associated with the oscillations is plotvd against the limit amplitudc (Pthm).

Data for Fig. 13 were taken from Table VI. Motor no. 8b provided an interesting case. It oscillated rathermildly for a brief time following the triggering pulse. Oscillations then grew rapidly to a higher level, which wassusmained briefly until burnout was reached. Data are presented in Table VI from measurements in each of the twoportions of bum for that motor and the two data points are plotted in Fig. 13. The line in the plot is a linear leastsquares fit to the data with forcing thiough zero. The plot indicates that an approximately linear relationship existsbetween acoustic limiting amplitude and burning rate increase.

Data of the type piesented here carry a strong message both to motor designers and to motor users. Triggeredinstabilities arc of more thar, academic interest. Standard ballistic testing will seldom reveal how susceptible s motoris to being pulscd into nonlinear instability. The phenomenon is capable of ruining motor performance when itoccurs. The results obtained in the present test series reemphasize the impotaance of testing motor stability bypulsing in addition to performing the usual stability calculations. Pulsing should be an integral part of motorqualification programs. Otherwise, the range of motor conditions (including debris discharged through the nozzle)under which a motor will rcmaini stable is difflicult to asscss.

1000

) 800 .. .......

-" 600

En) 400 771

0

0 L0

0 200 400 600 800 1000 1200

LIMITING AMPITUDE; -- psi

Fig. 13. DC Presure Shift in Nonlincarly Unstable Motors Versus Limit Amplitude (Data from Table IV).

14)

Although pulsing is an excellent tool for testing motor stability. many motors are in service which have not bealdeliberately pulsed. The reasons we too numerous to relate here but aost is one factor. Another is the argument that ifa motor is normally stable it will perform properly in service and a pulsing program will not improve a motor thatalready behaves well.

REFERECES

1. F. S. Blomshield. J. E. Crump. H. B. Mathes. R. L. Den'. and M. W. Beckstead. "ICP Non-Linear InstabilityCollaborative Progrmn--NWC Participation". 22nd JANNAF Combustion Meeting. CPIA Pub. No. 432, Vol. 2 (1985).pp. 155-64.

2. H. B. Mathes. "TCP Non-Linear Instability Program: U.S. Motor Data'. 23rd JANNAF Combustion Meeting.CPIA Pub. No. 457. Vol. 1 (1986). pp. 59-64.

3. W. H. Clark. J. E. Crump. H. B. Mathes. and M. W. Beckstead. "Motor Test Data Comparisons with LaboratoryData and Stability Predictions". 24th JANNAF Combustion Meeting. CPIA Pub. No. 474. Vol. 1 (1987). pp. 101-114.

4. W. H. Clark. J. E. Crump. and H. B. Mathes. "Results of Tests for Nonlinear Stability of Full Scale RocketMotors". 25th JANNAF Combustion Meeting. 1988.

5. H. B. Mathes. J. E. Crump. F. S. Blomshield, and M. W. Beckstead. "Nonlinear Test Results of Full Scale RocketMotors." 26th JANNAF Combustion Meeting. CPIA Pub. No. 529. Vol. IV. Oct. 1989.

6. F. S. Blomshield, J. E. Crump. H. B. Mathes. and M. W. Beckstead. "Stability Testing of Full Scale TacticalMotors,- Proceedings of the 27th JANNAF Prcpulsion Meeting. Anaheim. California. October 1990.

7. W. H. Beck and W. H. Jolley. 'Harmonic Analysis of Piston and Pyrotechnic Pulsers for T.Burners". in J.Propulsion and Power. Vol. 4. No. 3 (May-June 1988). pp. 283-85.

8. R. S. Brown. "Blowdown Pulser Criteria for Solid Propellant Rockets'. in J. Propulsion and Power. Vol. 2, No. 2(March-April 1986). pp. 110-116.

9. R. L. Lovine. J. P Baum. and 1. N. Levine. "Ejects Pulsing of Subscale Solid Propellant Rocket Motors", inAIAA Journal. Vol. 2.). No. 3 (March 1985). pp. 416-23.

10. J. A. Murray. J. A. Condon. and D. E. Krusch. "Pulsing Criteria for Solid Rocket Motors Volume fl: MotorPulsing Design Manual". AFRL-TR-79-45. Air Force Rocket Propulsion Laboratory. Edwards AFB. Edwards. Calif.March 1981.

11. R. L. Lovine. "Nonlinear Stability for Tactical Motors, Vnlume IV--Pulsing Considerations Parts I and 2".AFRPL-TR.84-017. Air Force Rocket Propulsion Labortt ry,. Edwads AFB. Edwards. Calif. August 1985.

12. J. D. Baum, J. N. Levine, and R. L. Lovine. "Pulsed Instability in Rocket Motors: A Comparison BetweenPredictions and Experiments". in J. Propulsion and Power, Vol. 4. No. 4 (July-August 1988). pp. 308-316.

13. R. M. Hackett. "Analytical Model of Pulsing of Solid Propellant Rocket Motors". in J. Spacecraft and Rockets.Vol. 22. No. 2 (March-April 1985). pp. 201-210.

14. J. D. Baum. 1. N. Levine. and R. L. Lovine. 'Pulse-Triggered Instability in Solid Rocket Motors". in AIAAJournal. Vol. 22. No.10 (October 1984). pp. 1413-19.

15. J. D. Baum and J. N. Levine. "Modeling of Nonlinear Longitudinal Instabilities in Solid Rocket Motors". in ActaAstronamaica.Vol. 13. No. 6-7 (June-July 1984). pp. 339-48.

16. J. A. Murray. ].A. Condon. and D.E. Krusch. "Pulsing Criteria for Solid Rocket Motors. Volume [I: Moutr PulsingDesign Manual," Hercules Aerospace Division. Hercules Inc.. Allegany Ballistics Laboratory. AFRPL-TR-79-45.Edwards AFB. CA 93523.

17. R. I.. Lovine. "Nonl:near Stability for Tactical Motors; Volume IV - Pulsing Considerations," Aerojet TacticalSystems Co.. AFRPL-TR-84.017. Edwards AFB, CA 93523.

18. E. L. Crow. F. A. Maxwell, and M. W. M.xfield. Statistics Manual. Dover Publications, Inc.. New York. 1960.

SHEAR STRESS TRANSDUCER CONCEPTS*

UNITED TECHNOLOGIESChemical Systems Division

E. C. FrancisR. E. ThompsonS. W. Heerena

The measurement of shear stress in a solid propellant rocket motor can beaccomplished by discreet transducers embedded in the propellant-insulationinterface. However, shear stress transducers and application techniques I -' nctbeen generally available. CSD has had experience in the past with thin .4 oftransducer, and is currently involved in a program that is evaluating m-. sheartransducer concepts. This paper presents the different types of shear 3itresstransducers available, and shows laboratory data on stability and sensitivity. Thepotential usefulness and drawbacks of each type are presented.

Current Shear Stress Transducer Confiauratjlon

There were five models of shear stress transducers concidered on the project.Four different vendors were involved in the manufacturing of these sheartransducers. A summary of the shear design concepts are illustvated in Figure 1.

The first transducer design is the Entran dual beam type with compression balias shown in Figure 2. In this concept, two instrumonted beams are aligned andattached to a rigid force summing diaphragm. The diaphragm, or moveable portion ofthe unit, is mounted to the beams with an epoxy bond. A Ailicon cerbide ballsupports the diaphragm to eliminate sensitivity to compressive loads durinn ignitionconditions. As shear forces are applied, the two beass bend and produce electricaloutput from the bonded semi-conductor strain gages. An elastomeric seal is placedbetween the diaphragm and remainder of the transducer body to keep moisture andother corrosive products from reaching the stra;n ga~as.

Figure 3 shows a Micron integral dual beau transducer. The transducer body ismanufactured as one solid unit rathar than an assembly. This design concepteliminates an epoxy bond in the load structure, but may also be sensitive to tensileor compressive loads. Senseo-Metrics. Ini. makes a shear stress transducer thatemploys a diaphragm concept. Figure 4 shows the Senso-Metrics, Inc. combinedshear/normal stress transducer. This transducer is strictly a standard diaphragm-type, but wired (see figure 5) such that forces in both normal and shear directionscan be measured simultaneously.

Figure 6 shows the Micron potted T-beaa transducer. This unit employs, a sirqleinstrumented T-beam. The gaged area is then potted over with Veraamid/Epon-typeepoxy or silicone to protert the critical components. The last shear gagetransducer is an instrumented wedge shown in figure 7. This device has beenmanufactured by Konigsberg and Senso-Metrics, Inc. Both foil and semi-conductorstrain gages have been utilized. The wedge itself has been made from bOth soft andhard polymeric material and even from solid propellant. CSD has evaluated each ofthese types of gages for electrical stability, shear stress sensitivity, and thermaloffset sensitivity in the laboratory. A summary of the data is presented below.

Each transducer type was evaluated for the requirements listed in Table I.Stability of zero offset is critical because once this type of transducer isembedded in the rocket motor any electrical Phifts would cause erroneous stressreadings. Viscoelastic transducer effects must be eliminated because thesetransducers are intended to measure the nonlinear viscoelastic response of the solidpropellant. Repeatability is essential especially during motor thermal loading.Solid propellants are nonlinear viscoelastic and will not generate repeatablestresses during thermal excursions. The electrical output requirements are

FiThis work sponsored by the Astronautics Laboratory under contract No. F04611-87-C-0062; Approved for public release; Distribution unlimited.

151m

generally >10 millivolts for full scale readings. If the shear transducer issensitive to hydrostatic loads, then the corrections should be linear so thataccurate shear stress can be calculated. Embedded transducers must be of small size(especially height) or they may significantly change the stress that is beingmeasured. if the stress is perturbed sufficiently it could cause premature motorfailure. The stress gage will measure the altered stress on the sensing surface, sominimizing disturbance is essential.

Table I. Minimal Shear Transducer Requirements

1. Stable zero offetr during thermal testing2. No viscoelasti.:; 6frects3. Repeatability4. Adequate electrical sensitivity for accurate stress measurements5. Insensitivw to hydrostatic loads or subject to linear correctLon6. Minimal propallant stress disturbance and interaction (minimal height)

The initial tests conducted with each transducer were measurement of shearload, and hydrostatic load repeatability and viscoelastic effects. The instrumentedwedge and potted T-beam transducers exhibited large viscoelastic effects andsignificant electrical shifts or lack of repeatability. Both of these transducerconcepts :cz-e dropped from serious consideration because of these problems. Typicaldata is shown i-i Figure 8. Use of polymeric materials in the loading structuresappears to - nonlinear aio viscoelastic transducer response. A summary of theshear gage prot,4_s using the wedge and T-beam designs is presented n Table II.Both of these designs are not considered acceptable for solid propellant rocketmotor use.

Table II. Potted Shear Gage Characteristicsfor Wedge and T-.Eas Designs with Polymeric Materials

(1) Large calibration errors(2) Time-dependent viscoelastic response(3) Nonlinear load-displacement cmrve potentially caused by load transmission

througn the potting material(4) Gage sensitivity decreases with lower temperature tests(5) Zero shift during calibration produces large error(6) Gages are very sensitive to handling loads; nay incur permanent shifts

These factors show that shear gages using polymeric materials are not suitablefor accurately measuring the nonlinear viscoelastic shear stresses generated in asolid propellant rocket meotor.

Zraducer Stabillty

A transducer screening test was devised by CSD to approximate the long-timeoffset stability of new units after they are received from the vendor. Experiencehas shown that coaparinq th-4 no-load output of a transducer at ambient temperaturebefore and after thetw ,y•cling will reveal any units that are inherently unstableThe test consists of a- ':,jnt of the no-load output at a conditioned 70.F,thermal cycking of the . •nree times between nominally 140-F and -65.F, andreneasureaent of the r %J .1 output at 70"?. If the before and after ambientoutputs are within 100 L!crovolte of each other, the transducer is considered to bestable. Delta values higher than this indicate an electrically unstable unit thatexperience has shown will very likely continue to drift electrically during use andwould generate data with a large uncertainty bond. Table I below summarizesscreening test 4ita for the various shear stress transducers.

Table III. Initial Screening Test Results

Early Unite

Sonso-Metrics SK703 0.180Shear/tension; 8K704 0.269601887 8K705 0.091

8K706 0.1)38

152

Table III. Initial Screening Test Results(continued)

Micron 150022 0.133Potted T-beamVersamid/Epon

Micron 5401708 0.208Potted T-beam 540170C 3.196RTV compound 540170D 0.050

Entran, Inc. 9U6U-V4-4 0.441Dual beam; 9U6U-Vl-1 0.350

ESL-100 9U6U-V5-5 0.980

Micron 9H440 0.278Dual beam; 9H441 0.369601911 9H442 0.062

9H443 0.2399H444 0.1089H445 0.1209H646 0.0849H647 0.594

Entran 89G9gG28-V03 8.67389G89F19-V05 4.53589G89F19-V02 1.77789G89F19-V06 3.20389G89G28-V04 3.43289H89H09-VOl 0.539

Based on the above screening data, most of the earlier units were consideredunstable. However, the technology for shear stress transducera is not yet to thepoint of that for normal stress transducers. Manufacturers have not been makingthis type of gage for as long as they have been making normal stress transducers,plus the geometry is somewhat more complex. Many of the later units by Micron werestable and were the best shear transducers received. The later Entran unitsexhibited large temperature-induced shifts which were attributed to the manufacturerusing unfilled epoxy for bonding t 'rain gages and no stabilizing the metal partsafter machining.

Transduc_ .deQar Sensitivity

The shear sensitivity calibration test measures the output sensitivity inmillivolts per unit of applied shear stress. The transducers ware installed on avertical fixture by attaching the base of the gage to the fixture with double backtape, Figure 9. The direction of shear measurement was in the vertical direction.A small stainless steel hook was attached to the transducer sensing surface. Promthis hook a series of weiqhts were hung by a thin wire to apply a shear stress tothe gage (Figure 9). The 11 point shear stress calibration was performed with databeing measured at specific times for each weight increment. Gages which exhibitedviscoelastic response would give different output if the weight times were varied.

All of the early and later shear transducers had adequate shear sensitivityexcept for the Senso-Metrics, Inc. shear/tension transducer which was the diaphragmtype. Units with polymeric material exhibited nonlinear viscoelastic behavior. Asummary of the shear stress sensitivities are presented in Table IV.

Transducer Thermal Offset Sennitivity

In addition to the pressure sensitivity, one other pnrameter is required toconvert the total sensor output to pressure units. That parameter is the thermaloffset versus temperature curve which is the sensor no-load output versustemperature over the temperature range desired. The transducers are thermallycompensated at the factory, but always have some sensitivity to temperature. Thisthermally-induced offset must be subti' ",i from the output of the transducer whenit is mounted in propellant to produce tn,, net output due to the propellant induced

, , i I I i I I I I 53

load. C used a programable temperature controlled Thermatron oven to astp cyclethe transducers and asure the temperature offset response of each unit. Thetransducer output measurements were taken approximately every 20* between 150*Fand -50.F. The cycle was repeated 3 times so data reproducibility could beverified. Two typical sets of data are presented in Figure 10. The thermal offsetIs nonlinear but repeatable and is used to correct the electrical output for alldata generated during later transducer use.

Table IV. Transducer Shear Stress Sensitivity Values

A~rlX nitsmV/psiSense Metrics RK703 0.0067Shear/tension SK703 0.0066601867

Micron, Inc. 150022 13.36Potted T-beanversauid/spon

Micron, Inc. 540170B 91.64Potted T-beam 540170C 74,41RTV compound 540170D 113.96

Entran, Inc. 9U6U-V4-4 0.213Dual beam 9U6U-V4-1 0.365ESL-100 9U6U-V5-5 0.299

Later UnitaMicron 9H445 0.268

9H440 0.- .79H442 0.2339H646 0.2809H443 0.2859H444 0.289

Entran 89C89G28-VOl 0.48389G89G28-V03 0.40289G89F19-V06 0.25889G89F19-V05 0.29789G89F19-V02 0.296

The zero offset curves for each transducer have unique shapes and must becarefully measured. Each set of data im fit to a fourth order polynomial as shownin Table V. Fourth order fits have been adequate to represent all of thetransducers tested at CSD.

Compressive Load Effects

When shear stresses are being measured duL-ing solid propellant motorcombustion, a large compressive load also exists. Ideally we would like a sheartransducer that does not sense this compressive load. In reality, all shear sensorsexperience some electrical readout ii, a hydrostatic or compressive load environment.Both the wedge and potted T-beam designs exhibited sizable hydrostatic loadelectrical outputs.

The two transducer designs which were considered the best were the Micronintegral beam and the Entran dual beam with compression ball. Both designs weretested in a hydrostatic pressure environment and showed no output. On furtherinspection, we found that pressure leaked Into the transducer through the lead

154

MICRON - 9H444 Shear Transducer*

% ResidualsSums of about mean Polynom.

g2rM Sausres ZAM coexhcD

0 0.74408E-006 0 0.001084431 2.52489E-006 71 -3.25052E-0072 8.25187E-007 91 -8.02303E-0083 2.6777E-007 97 4.47436E-0104 2.55134E-007 97 -7.43136E-013

MICRON - 9H443 Shear Transducer*

% Residuals

Sums of about mean Polynom.Dere gUMara Z ind Coeffic.

0 2.817982-006 0 0.0004472021 7.88064E-007 72 6.01933E-0072 1.8075E-007 94 -1.35641E-0083 1.80748E-007 94 -1.95873E-0104 1.59985E-007 94 9.53761E-013

Total points - 130Fitting Interval Limits: -59 to 165Points in Pit Interval - 130

wires and seals to equilibrate the system. We isolated the compressive loads byusing dead weights on top of the sensor plattens and found both units providedsignificant outputs as shown in Table VI.

Table V. Polynominal Fit Statistics for Two MICRON Shear Transducers

Hydrostatic TestMICRON Sensitivity

23 0.009224 0.002737 0.031396 0.000232 0.000691 0.0084

Dead Weioht Compressive TestsMICRONIntegral Sensitivity

9H445 0.0619H440 0.0139H442 0.0609H646 0.0319H443 0.0259H444 0.049

Entran Dual Beanwith Compression Ball

Senvitivity

Vol 0.062V03 0.036V05 0.004V02 0.029

The Entran compressive ball did not eliminate the compressive load effects asexpected. The manufacturer thought that metal tolerances and metal seatingunder load may have reduced the potential design benefit of the compressive ball.

155

The compressive load sensitivity is much less than the shear sensitivity, but thefinal motor application condition has a much higher compressive stress than shearstress.

Summary of ResultsThree shear design concepts were discontinued midway through the program

because of numerous response problems which are listed below.

Table VII. Transducer Concepts Discontinued Midway Through Program

1.0 Wedge Type Transducers"* Large calibration errors"* Large hydrostatic sensitivity" Sensitivity changes with

temperature"* Easily damaged during handling"* Viscoelastic effects"* Large electrical shifts

2.0 T-beam Transducer"* Large calibration error"* Hydrostatic sensitivity"* Large electrical zero shifts"* Sensitivity changes with

temperature"* Viscoelastic effects"* Easily damaged during handling

3.0 Diaphragm Shear Transducer"* Low shear sensitivity"* Hydrostatic sensitivity much

larger than shear sensitivity

The other two transducer concepts were selected for further improvement later in theprogram (Table VIII).

Table VIII. Shear Transducers Selected for Further Improvements

MICRON - Integral Dual BeamENTRAN - Dual Beam with Compression Ball

Goals:

* Improve manufacturing to achieve acceptablh thermal stability* Design for 5 mhA constant current excitation* Low and stable hydrostatic sensitivity

CSD reviewed transducer design and manufacturing with both Micron and Entranengineers and suggested manufacturing improvements that would improve their yield ofsuccessful transducers. Both companies sent us their best shear transducers forfinal evaluation. Transducer screening results showed that the Micron transducerhad electrical instability in the low microvolt range and would be useful. TheEntran transducers exhibited large electrical shifts during thermal screening in themillivolt range and would be useless for measuring small shear stress in a rocketmotor environment.

Final verification of these shear transducers is being conducted now. Thesetransducer designs are being evaluated in solid propellant motors being subjected tothermal and pressurization conditions.

156

1.0 Duel boom with oa~resioin boll

- flScOfL-ductor q90e0/17

-4

IIS

2.0 integral dual boom- semi-conductor qsqe0/1

7-4

SB

3.0 Diaphiragm - vired for shear readout

- Seat-conductor 902OV/ll-4 69

~ 4.0 Instiumeanted ?-beau

- patted in polymeric waý.erillsemi-conJucter strain 94qeS

- zpoxy-vorsasld/oilic~ns

5. nstrumented wedgea

replymer

Ins~ulationl- solid p'ropellanlt

Figure 1. Shear Stress Transducer Concepts Considered

Strain $ages

Silicon carbide boll

_________- -0.30 In. diernele.

Figure 2. Entran Dual Beam with Compression Ball Shear Transducer

Ribbon Cable

(ClDS~uppiI160) SiliCOne Of. Wl`V 030

S~n~o' bale (17.4 CREG) (1age CR.000340AIES

Epoy 00inar No. 10)1335 ---- 00

Ep130 6203 20

Figure 3. Micron ~~InterlBa.SerSno

1J',

A

- 0.440 In.

0.130 In.SA ction A-A

Figure 4. Senso-Ietrics, Inc. Combination Tension/shear Gage

SHEAR

I

f

Figure 5. Senso-Metrics, Inc. Combined Shear/tension Gage Wiring Schematic

158

Us sleki... oftel bm

N Ca w e bo bbahd".

Figure 6. Micron Potted T-beam Shear TransducerSemi-conductor Strain Gages

Figure 7. Instrumented Wedge Shear Stress Transducer

So

Figure 8. Typical Step Load and Unloaded Response forWedge and Potted T-beam Sensor

159

I,

10

7-

4

1 2 3 4 9 6 7 0 6 10Tinw, Irne mlentfsU

Figure 9. Shear Calibration Loading Sequence11-point Shear Stress Transducer CaJibration Procedure

0 000100 1

10000400-vý " 4 b 0

0.000 -

IQ P MICRON IO43I. 3192-9 MICRON 9H444

-0.000400 1 ., , ... - , . I I

-SO -00 40 00 140'l0Iplr~turw. "1F

Figure 10. Zero Offset vs Temperature for Two Micron Shear Transducers

160

NORMAL STRESS TRANSDU~CER BEHAVIOR.

R. E. ThompsonE. C. Fr~ancis

UNITED TECHNOLOGIES/C!herical Systems DivisionSan Jose, CA 95161-9028

Btru.:tural analysis procedures are routinely used on composite caies andcomplex ge, atry propellant grains to predict structurtal margins. Hotever, theseanalyseR Pmmt he txperimentally verified in order to instill confidence in theirpreoictions. C5D ha. lundcrtýLken a program to demons!-rate the application of stresstransducers under real motor conditi..ns and assess their ability to accuratelymeassure the notot response to thermal and pressure-induced loads. Severe.l differentType& of norz'^) ptrews transhiicere were evaluated under conditions of constantstress, slowly changing (low grhad~ent stress), and termination stresses -heredramatic changes occur within a small loc~tion This paper presents a summary of thetype of trAnr~ducers evaluateA Ar.d their 1-otential suitability to measure solidroch'i.. motor stresL.La.

Pipt Normal jlrjE_.IXfpsducer Histvor

ý'~ use of iiurmal ._itrp-as trý nsduc..ers in tP.e rolid propellant rocket motor~ndustry has been going on for apprreki~aely :5 y&;trs and can rougs-.Iy be broken downinto four time phases. In the uerly phare. %1.%roximately 1965 to 19.'5, the 11-Jitedtypea of transducers avai]7'ble su.ffered from several problems. AinorJ those weregage internial vacuum leakage duu to inadequate lead wire scair, chemical corrosionof the semiconductor qages and lead1 wire cor.tacts, unstable output, high structuralinteraction wit?. M~le prupellant d to A hin diaphragms, and low accuracy acquisition'clactronics. Overall, thesa tzaii:_Qrýers were aaequate if used for measurellent oflurge ignition-type prossuros, bct wert use~ess for the measurement of the muchlower thermally-indur-ed stresses. The secoril phare of transducer use occurredbetween approximatýeiy 19)5 and 1983. During this time, the stability of thetr,,neducers and their assucia't9d acquisition olcctronics great .y improved. Themachined transducer bodies wer~e subjected to an elahorate metal conditioning processto eliminate residual stresses that c-intrJibuted to retal instability. The epoxyused to bond the semit.',rductor strain gagur ti

tite metal dliaphragm was first filtered to elirinatA "large" aqglor,.erates which wouldaffect tha qgags, and thenapplied with a masktng techniqie so i, %inimuri amount o' thc. diaphragm was covered.The epoxy thickness was kept to 0.0005-liachpa or less to minimize creep potential.However, the size of the transducern was too cum~bersome for use in actual motorsand, in fact, the best transducer was a large throu.~h-the-case unit that wirs quitesuitable for laboratory st~ructural test vehicles. but nolt in-situ uso in realnotors. The third phaseL of transducer tome wan kqtw~een 148.I and 1987. During thisperiod, the size of certain. normal. strnss transducerG ras reduced to that compatiblewith real notors. Prinn concern with usiw in real. Aotors was that the transducercould n~ot disgvade the safety or reliability of the motors, and should be configuredto minimize stress disturbarco. S',7.a of these trunsdiictrs were actually used inflight, motors. Manuf.%';turing procenses were looked at sore closely, and pre-usescreening techniques designed to "weed -Nut" unstable gages were developed. Thefourth phase of the work is currently .znder-way. Ir. this phase, the continuedrefAn."ment of normal stress transducers is be mci put-sued, along with the developmentof 'ery small transducers tLuita;)!e for teraination lo-:ations and the tLevelopment ofstable, accu'rate shear stress tran~zduce.rs. A termination location is defined as abondline aroa of the motor. such as near grain ?;lot tips or boot tips, which have alarqe and sudden stress riser or qradi'.sr-t.- 53iico t?~ese stresses occur over very

This. work eponsored by the Atti-onauties L.aboratory under contract No. F04611-J7-C-0U62; Approved for Public Relpe~e: DistriL'ution is Unlimited.

161

small areas, a very small transducer is required for their measurement. This paperwill .,ncentrat. on the normal stress transducers being tetudied in ULhi fourth phaseof strOat transc ac.r development.

C It Transducer Development

Aftnr the third phase of transducer development had been completed, one primaryvendor of stable transducers had been identified. That vendor was Sens'i-Metrics,Inc. of Simi Valley, California. When the current development effort began, CSDsurveyed the industry to see if other vendors could be competitive with SMI inquality, price and stability. Three other vendors were selected in addition to SKI,to munufacture traneducers to evaluation on the new program. These three aeditionLlvendors were Kulite and Entran of Leonia anA Fairfield, New Jersey, respectively,and Senao-Tec of Columbus, Ohio. A total of six transducer models were evaluatedfor stability, pressure sensitivity, thermal offset sensitivity on the bench, andmeasurement accuracy after embeddment in propellant blocks which wero subjected totensile, copressive and st.ear loads. The designs of these size transducers arepresented in figures 1 through 6. Figurs 3. shows the through-the-case model useablefor laboratory structural test vehicles with metal cases, but not useable for eitherstnictural test vehicles with composite cases or actual flight motors. Each of theother five units could be used, with a proper lead-wire management system, in flightmctors. The iead-wire management system is defined as the system used to bring thew-res alor4 the insulation surface and then outside the motor (or between the caseand insulation) terminating in a connector which can then attach to the powersupply/data acquisition system.

Current Transducer Parformance

The most important electrical characteristic of any transducer is it's long-term stability, or output, under no-load conditions. once a transducer is installedin a rocket motor, its no-load output cannot be remeasured nor can any otherelectrical calibration values be remeasured. Electrical instability can be causedby several factors. Perhaps the metal. stress-relief conditioning may be inadequateleaving residual metal stresses causing metal creep. The glass-to-metal header usedto hermetically seal the lead-wire exit from the transducer body could actually havea minute leak causing a small, constantly chanqing internal pressure (from vacuumtoward positive pressure) which directly affects the semiconductor strain gageoutput. Another possibility is that the epoxy bonding the strain gages to the metaldiaphragm is too thick, or under-cured, and is creeping, causing semiconductorstrain gage output change. Whatever the reason, transducer instability will producean erroneous output that cannot be identified if the qaqe Js in a motor. Thiserroneous output would cause inaccurate stress measurement which could be eitherhigher or lower than actual stresses. Gage stability must be assured beforeinstallation in a so:id propellant rocket motor.

A transducer stability rapid screening test was devised by CSD to evaluate thestability of the new units after they were received from the vendor. The test wasdevised from experience wherein comparing the no-load output of a transducer atambient temperature before and after thermal cycling would reveal any units thatwere inherently unstable. The test consists ot measurement of the no-load output at20.F, thermal cycling of the gage three times between nominally 140.F and -65., andremoasurement of the no-load output at 70.P. If the before and after ambientoutputs are within 100 microvolts of each other, the transducer is considered to bestable. Delta values higher than this usually indicate an electrically unstableunit that experience has shown will very likely continue to drift electrically andproduce data with a auch larger uncertainty band. Since transducers cannot berecalibrated once they are installed in a motor, these electrically unstable unitsmust be avoided. Table 1 suolarizes the screening test data for the better normalstress transducers tested during this current program. In continuing the evaluationof the transducers, each of them were subjected to pressure sensitivity and thermaloffset repeatability testing.

Transducer Pressure SJ.L-AXi,

The pressure sensitivity calibration test is a key test as it measures theoutput sensitivity in millivolts per unit pressure. The gages were installed in apressure vessel and pressurized from 0 to 250 poig and back in 20% increments, or 11steps using an Anetek Dead Weight Tester. Linearity and hysteresis are measured for

162

each transducer. A linear curve fit is applied, the slope of which is thetransducer pressure sensitivity in mV/psig. A summary of the measured pressuresensitivities for several transducers is presented in Table 2.

The pressure sensitivity value, unlike the previous screening values, has nounacceptable limits. It is a constant, but unique number for each transducer. Thisconstant is used to convert the raw output to pressure units by dividing the netoutput (see text below describing the thermal offset sensitivity for net outputdefinition) by this constant to achieve net stress. A structural analysis of the

raw data, shown in figure 7, calculates the pressure sensitivity based on a leastsquares curve fit of the raw data. It also calculates the linearity and hysteresisof the data. The data for the Senseo-Metrics, Inc. transducers generally showed lessthan 0.1% error in the linearity and hysteresis, while other manufacturers generallyshowed errors much greater, often in excess of 1% error.

Transducer Thermal Offset Repeatability

In addition to the pressure sensitivity, one other parameter is required toconvert the gross sensor output to pressure units. That parameter is the thermal<Afset versus temperature curve which is the sensor no-load output versustam:.erature over th3 temperature range desired. The transducers are thermallycompensated at the factory with the appropriate temperature compensation resistors,vut still have some sensitivity to temperature due to the nature of thesemiconductor strain gages. This thermally-induced offset must be subtracted fromthe gross output of the transducer at each temperature when it is mounted inpropellant to produce the net output due to the propellant loading. CSD uses aprogrammable temperature controlled Thermatron oven to provide the thermal historyused to measure the temperature versus zero offset of the transducers. Themeasurements are taken approximately every 20-F between 150"F and -50.F. The cycleis repeated 3 times so that data repeatability can be verified. At each temperatureincrement, the temperature is ailowed to stabilize for a minimum of one hour whilethe output readings are taken before changing to the next temperature. The datapresented in Table 3 lists the maximum delta output due to temperature between thetwo temperatvre extremes.

As can be seen from Table 3, some the transducers had a smaller delta outputover the temperature range than others. The transducers with the smallest zerooffset shift during screening show the best repeatability during the zero offsetversus temperature testing. However, the absolute zero offset curve shape andamplitude is not critical as long as it in repeatable and accounted for in dataanalysis. Figures 8 and 9 show typical offset versus temperature plots for veryrepeatable transducers. Figure 10 shows data from a transducer that has very poorrepeatability. The importance of this thermal repeatability cannot be emphasizedenough. Once gages are installed in propellant, they cannot be recalibrated.Therefore, all transducer behavior measurements taken before installation intomotors is final. If the installed gages electrically drift at this point, it wouldbe interpreted as a change in propellant-induced stress and be a part of theexperimental error. When very accurate experimental measurements are needed, onlyvery stable transducers should be used. CSD has routinely rejected unstable gagesand requested that the manufacturer rework them until they are stable.

Future Transducer Wor

The transducers referred to in this paper were evaluated in propellant blocktests after the bench calibration tests described above. This propellant block datawill be summarized in a separate paper. Following the block tests and dataanalysis, several of the better gages will be installed in two different compositecase subscale motors. Some gages will be mounted in a Kevlar case motor to be castwith a PEG/NG propellant, while others will go in a Graphite case motor to be castwith an HTPB propellant. Gages will be located in areas of relatively constantstress and also in areas of high stress gradients (termination areas). This willprovide a critical evaluation of the gages under actual motor loading conditions.The evaluation of the gages will incorporate two finite element analyses of themotor. The3e analyses will calculate the stress et the gage locations, both withand without the transducer present. The inclusion of the transducer in the motorchanges the stress it is intended to measure (otherwise known as the disturbancefactor). For a well designed unit, this disturbance factor is usually only a fewpercent, but should be accounted for.

16 1

Conclu±2le

CSD has evaluated transducers from four separate vendors to determine theirability to measure stresses in solid propellant. These gages are both small sizeunits suitable for areas of constant stress, and very small units suitable tortermination areas or areas of high stress gradients.

Of the four vendors who supplied transducers, Senso-Ketrics, Inc. producedgages of superior stability. The reasons for this are several. Firstly, SKI usesthe specific metal conditioning process mentioned earlier to reduce metal stressesinduced by machining and other thermal processes. This reduction of residual stressin the final product reduces paceible metal creep and epoxy creep which effectsoutput stability. SKI developel and uses an epoxy filtering technique whicheliminates large filler particle sizes which can contribute to thicker epoxy and mayinduce creep. concurrently, SKI devised an epoxy masking procedure mentioned earlierso only the area beneath the stain gages are epoxied. No other diaphragm areas are"smeared" with epoxy. The othar vendors were supplied with the metal conditioningprocedures, but they chose not to nise it. They felt that their own metalconditioning processes, uaed in their routine manufacturing, were sufficient. TheSKI epoxy filtering/maskiing techniques were proprietary information, and could notbe supplied to other transducor vendorN. As a consequence, the moet stabletransducer came from SKI.

CSD believes that normal stress transducers, including sensorb small enou,'. fortermination regiona, has reached a new state-of-the-art level. At the conclusion ofthe current research effort, gages able to measure relatively constant and gradientstresses, and produce data that is stable and accurate, should be available forroutine use in the solid propellant rocket motor industry.

Table 1. Screening Test ResultG

senso-metrios, Inc.

Through-Case

Serial Nuzbe Delta mY. Screenina

3A020 0.0334G408 0.0504E105 0.0144G412 0.1224G405 0.0214E495 0.089

aenso-Metri.s, Ins.Vormal/In-situ

9G334 0.0609G336 0.0639G331 0.0089G330 0.1019G333 0.0539G335 0.0329G345 0.0120G344 0.0059G341 0.0649G340 0.0279G343 0.0109G322 0.0339G339 0.0259G342 0.0069G337 0.0419G338 0.0557M453 0.0287h454 0.1027M4455 0.015

164

Table 1. Screening Test Results(continued)

Benso-Toe, Inc.Termination

187133 0.580187134 0.179 Large187135 0.550 Offset188461 0.270 Shifts188460 1.258

BntranTermination

16X3M4-H4-3 0.77316X3M4-H5-1 0.91016X3M4-H2-4 1.300 Large16X3M4-H3-2 0.114 Offset88F88E16-05 1.230 Shifts88F88E16-06 1.77088F88E16-02 2.01088FS8E16-03 4.280

KuliteTermination

5057-1-13 0.2095057-1-14 0.1145057-1-15 0.0825057-1-16 0.2595057-1-17 0.6765057-1-19 0.2925057-1-22 0.9555057-1-25 0.2081208-4-227 0.0601208-4-232 0.0201208-4-233 0.079

Table 2. Transducer Pressure Sensitivity Values

Sensa-Motrias, Inc.

Normel/In-situ Pressure Sensitivity, mV/psi

Serial Number 0 Deo_ 70 Deg. F 140 Deg. F

9G330 0.1209 0.1234 0.12229G331 0.1305 0.1326 0.13149G332 0.1246 0.1268 0.12669G333 0.1392 C.1415 0.14019G334 0.1336 0.1366 0.13609G335 0.1522 0.1544 0.15269G336 0.1463 0.1497 0.148796337 0.1472 0.1488 0.14719G338 0.1330 0.1338 0.13299G339 O.109q 0.1118 0.11099G340 0.1395 0.1415 0.14029G341 0.1229 0.1260 0.12519G342 0.1587 0.1598 0.15919G343 0.1498 0.1506 0.15069G344 0.0944 0.0965 0.01)69G345 0.J249 0.1?77 0.2(.,

Table 2. Transducer Pressure Sensitivity Values(continued)

RuliteTermi•ttion

5057-1-13 0.3290 0.3326 0.33825057-1-14 0.3490 0.3503 0.37655057-1-15 0.3900 0.4005 0.42965057-1-16 0.3330 0.3240 0.32886057-1-17 0.2990 0.3300 0.36455057-1-19 0.3440 0.3473 0.36205057-1-22 0.3540 0.3400 0.33425057-1-25 0.3230 0.2800 0.2817

Table 3. Maximum Transducer Thermal Offset

SxjiL Number aximuz Delta O.P.. mV

gazo-Metrics, Ino.Through-case

4G408 0.7004G412 0.5004E105 0.4004E495 1.800

senss-Metrics, Inc.Normal/In-situ

7M453 0.3007M454 0.2007M455 0.600

Seac-TeocTermination

187133 4.000187135 1.700187466 0.600

EuliteYurination

5057-1-13 0.5505057-1-14 0.7305057-1-15 0.5005057-1-16 0.8005057-1-17 0.9605057-1-19 0.4505057-1-22 0.9005057-1-25 0.900

166

SA

Figure 1. Senso-Metrics, Inc. Through-case Pressure Transducer

Figure 2. Senso-Metrics, Inc. florual SensitivityStroes 'Transducer

I(67

P0,014 aeewISy

Nout owls~leod 34-9ags

ftoe Dlium.nlons are k In UAo. o~ Wods.i 41 -in SoIg

Figure 3. Senso-Tec Termination Transducer

Po"aplvsgm 0.0035 Mock etche to

350 to400 WJMlcsGUla at khA-sCae.

0. O3 00. In 4 r~.lOn Coated Ribbon wi,.

OMM

Figure 4. iKulite Termination Transducer

felerso

Figuro S. Entran Norsmnal Stress Transducer

Now.: DbmeehtoEmt we bL bum..

Figure 6. Entran Termination Transducer

Irv; MIA

144I'Psifils] 1 2 3 to S ~ gto 4.370 4.134 11211 17-26S 16.214 21.20I 16.231 mm25 81tt 4.32 63

IM 1.116 4.114 6.196 02.261 l4,231 24.214 14.214 02.266 9.11 4.1211 3.161

1.661 23.666 41.111 61.130 g1,g1. 111.644 11.111 &1.664 41.464 ll.446 IASI

5/4 IND S I 5 13

0N . 4.31411 0,111117 1.066686 f.45516 0.46531

I I$ rldwsI*4: It 111 461 lot oil 1111

OF lSw~Iii.111 (..-l) 04 .344 4.W2 0.2V 2. 1?.? .231 31.11?

31 S(1 - 0.04 .11 Voz .424 6.633 6.663 6.014

Figure 7. li-point Pressure Calibration Data Printout forSKI Transducer SIN 7M454

2

100')

1.OO0E-003

,- 1.084E-019

o) [CKI~.Ct [m PCTABLII

-1.000E-003

S-2.OOOE-0030~

NI -3.000O-003 4 9G343QII 9G342

UNMOUNTED ZEFU , .. DATA

-4.OOOE-003 ...... . .. - , ..... ... .... ....-60.00 -10.00 40.00 90.00 140.00

TEMPERATURE F

Fiqure 9. Transducer ZOFF Data

7 -B

• N B

4

2

2 "I

Figure 10. Entran Transducer ZOFF Data

170

PiOCBSS AND QOALITY INPROWVITS FOR N7 PROPELLANT

R. A. GottHercules Incorporated

Aerospace and Ordnance GroupRedford Army Ameunition Plant

Redford, Virginia

ABSTRACT

A modified MY propellant has been processed using the solventless propellant method ofmanufacture. Significant improvements in overall propellant quality have been made by usingsolventless manufacturing techniques. The improvements included reduced chamical variability by281, improved dimensional uniformity by 80%. and reduction of internal and external voids anddefects by 801.

INTRODUCTION

Hercules Incorporated. operating contractor of the Redford Army Ammunition Plant (RAAP),has adopted the managemaent technique of Total Quality Nanagement (TQN). Using this managementphilosophy, improvmonts in the quality of current production items are being made. The basicmethodology is to use the cost of non conformance to identify those products where quality costsare greatest. Quality action teams are then formed to study both short-term and long-termsolutions to reducing these quality costs. Statistical tools are used to determine the criticalprocess steps that contribute most to quality problems and quality costs. Teams working onshort-term solutions have been successful in making minor changes in process parameters tominimize failures, except those inherent to the process. Those responsible for long-termsolutions to inherent quality problems have proposed that alternative methods of manufacturing beevaluated in sow cases.

This paper outlines the approach to solving basic quality problems and reducing qualitycosts within the fraumvork of Total Quality Management. It describes a case where, by using analternative process for propellant manufacture, a solution to an inherent long-term qualityproblem which effects many end-items was developed.

PROCESS KVALOATIOU

An inspection of cost information on production items currently manufactured at RAAPindicated that products which utilize MY propellant have one of the highest quality costs. Also.most of these quality failures are inherent to the solvent propellant manufacturing process.Products such as TOW launch. Subcol LAW, and 1180, manufactured with NM propellant, areconsistently among the top five failure rate products. The quality cost analysis indicated thatthe average percentage of grains rejected is approximately 131 for TOW launch, 24% for SubcalLAW. and 281 for N1SO.

Pareto analysis was used to determine where in the process the majority of qualityfailures/defects were occurring. For UY propellant, quality failure could not always beidentified during the UY propellant manufacturing process shown in Fig. 1. easy of these qualityfailures/defects were found late in the end-item finishing process shown in Fig. 2 for TOW launch.

Approved for public release; distribuLorn is unlimited.

171

DE TECT

~a -- aFAD PRESS

SEND TO END-ITEM FINISHING

ITOW launch. MISO. Sub-Cal!

Fig. 1. X? Solvent Process

so MACHINE mEMAY ED

MAGAZINE RCKING MOTOR X-AAI

STOR*E ASSEMBLY

TEST

Fig. 2. TOW Launch End-item Finishing

Table I lists the typical quality failures/defects for W? propellant that were detectedduring the end-item finishing process steps. In this specific case, finishing steps, machining,and X-raying required for TOW launch were analyzed. Using this analysis end analysis of otherprocessing steps, it was found that most quality failures could be classified into threecategories: internal defects such as voids (making up over 90% of the total defects at i-ray),external defects such as blisters end pin holes (making up over 601 of the defects at sawing),and processibility such as rheology problems (which currently cause the propellant to be mixedapproximately 2.25 times in order to produce a usable strand). It is estimated that these threecategories make up 90% of the quality costs and appear to be inherent to the solvent propellantmanufacturing process.

172

Table I. Pareto knaljysis of M Solvent Propellantat TOYi Launch Finisbing Stops

Machining X-rayDefect Type I Df Total Defect Type 7 of Total

Blisters 44.9 Voids 92.9Pin Holes 16.3 Foreign Matl. 4.7Nashed 10.3 Other 2.4Dents 8.9Crooked 6.SLength 4.9Other 8.2

The solvent propellant process currently used to manutacluve M7 it a vatca process designedfor bhe a'nufacture of large quentitiws of qun prooell~r~t. During the f'nal process ster, 'hepropellanL (composed of many different lots/mites) ii blended 'o meet befls!tic perfortincerequirements. Traditionally, these soivent-type pvopcllants nave been used in artillery and tankapplications where several hundred, or even thousands, of granules made up a singlw propellingcharge. Performance of this charge would than be based on the averaAe burning chara:turixtics ofthe propelling charge. li a few "bad" grains were in the charge, performance of the total chargewould suffer very little. ine same would not be true for shoulder-fired rockets that use the M7propellant, which have a small number of propellant grains in each motor. In the case of theshoulder-fired rockets, performance of the motor is extremely dependent on the performance ofeach propellant grain.

After reviewing quality costs, pareto analysis of key process steps, and the intent forwhich the solvent process was designed, it appears reasonable to consider an alternative processfor the manufacture of M; propellant to reduce or eliminate inherent quality failure.

SL,,VMTLBSS PROCISS DIVWLOPRINT

Previous attempts to produce M7 propellant ujing a solventless process have bed onlymoderate success. Vork on this stlbject ass perio.m'ed in the mid-fifties by R. I. Buell of BadgerOranance Workd. Mr. Btell chose to modify the intredients in the M1 formulation, reducing the.,cunt of carbon black and Aubstitutin! dietbylphthalato fcr the resioal solvents(alcohol-acetone) used in the processing )f solvert-type M? propellant. The project wasterm;nsted after two press shots that involved this AaLerial. Investigation into these incidentsindicated fe ty die design as the cause of initiation, not the propellant. The program appearsto have lost support sf¢er these incidents.

In the development of the M? solventless propellant at RAAP the goals were (1) to make asfew changes to the pL-opellant formulation as possible, (2) to most all the requirements ofMIL-P-14731, the basic specification for M7 propellant, and (3) to show that the M7 propellant,when processed in a specific end-item configuration, would met the end-item ballisticrequirements. TOW launch was chosen as a test vehicle for proving out the solventleas M7 in aspecif|- end-item.

bvaluation of the solventless process began by studying the effect of changing the solidsto plasticizer ratio on processing, within the limits of the current M7 specification. Table IIsummarizeF the three modified M7 propellaht formulations and the current M7 specification. Thesethree propellants were formulated with the aid of a thermochemical computer program in an effortto maintain a constant calorific value. In all cases the total of potassium perchlorate andnitroglycerin remained constant at 43% of the total composition.

, i ii i i ;;1;3

Table 11. 17 Solvent and Solventlets CgWO W22

X7 Solventless m7Ingredients Solvent.% IPD-2. RPDo-13. _D-14Nitrocellulose (13.lSIN) 54.6 nos 55.0 55.0 55.0Nitroglyerin 35.5 nom 40.0 36.0 43.0Potassium Perchlorate 8.0 max 3.0 7.0 0.0Carbon black 1.2 nos 1.1 1.1 1.1Ethyl Centralite 0.8 max 0.6 0.8 0.8Candelilla Wax 0.1 0.1 0 1

The type of nitrocellulose (NC) was also considered to be a key variable in theprocessibility of these propellants. Based on past experience with processing NC that had bnennitrated directly to fairly high levels (13.151U), it was decideS to also consider using a bleluof 201 low-grade and 801. haih-grade to achieve the desired nitration level of 13.151. This VC iothe same type that is currently being used in the 120m solventless propellanta.

The process fnr manufacturing solventloss 17 is described below and shown in Pig. 3. Twoiterations of each propellant were made using this process. One contained the standerd MYsolvent-type NC and one iteration of the NC blend. In both cases, NC at 28r water wet was mixedfor 10 minutes with the ethTl centralito and nitroglycerin. This formed a light colored pastewhich was dumped into cotton bags. The paste was allowed to age for five days minimum at ambienttemperature. At the end of this rest period the paste had a moisture level of approximately161. The paste was then t310on to a blending barrel where the remaining chemicals were added,The total blend time for the paste and chemicals was 15 minutes. After blending, the past* had amoisture level of 13%. The paste appeared to be finely divided and no problems with lumping orballing were observed. The paste wae then transferred to the rolling operations.

/

SCRADER ABOW.. AGE 9LENO P-RIEALLMIX bM/NG/ECI FASTE AWIE

aRO PRESS 9 UR.IL I EVINWHEOTAX E ,•Y RIL L

SENO TO END-;TEN FNISHING

1(7.4 Ibuncti, MISC. Sub-can

Fig. 3. Solventless 17 Process Flow

Optimat differential speed rolling (preroll) conditions were found to be 100 seconds on a2j8"F roll using a 4-lb paste charge. In most cases this produced in etreomely smooth, jet-blackshiet be&ve,n 0.035 and 0.040-in. thickness. It was observed that ell sheets produced withstandard hi solvent-type cotton were extremoly brittle as they cooled down. It was alse foundthat stiffness/ brittleness correlated extremely well with level of solids (i.e.. pntecsiumperchlorate) in the composition.

174

Only the N7 propellant containing the NC blend was processed on the evenspeed. It was foundthat most of the other material wiich contained NC directly nitrated to 13.1SIJ level was toobrittle to be processed any further safely. Optimal conditions for evenspeeding were identified;a roll temperature of 153"F with 1 marriage, 4 books, and 1 long-fold, using 20% rework. Theseconditions produced a smooth, black sheet 0.085 to 0.090-in. thick. Even with the NC blend, itwas found that the solids content dramatically increased the brittleness of the sheet. Theseobservations are supported by physical property data shown in the propellant description sheetswhich are attached. Significant splitting and cracking occurred along the edges of those sheetscontaining the highest potassium perchlorate level. These evenspeed sheets were then slit intofour-inch strips and rolled into 4-inch cobs for extrusion.

The RPD-13 propellant, which contained 71. perchlorate, could not be extruded. Extrusions ofthe two remaining propellant formulations RPD-12 and RPD-14 were successfully performed in a4-inch horizontal solventless press equipped with & single-lane airvey takeaway system. Theprocess conditions for extrusion are as follows; 150"f carpet roll temperature. 160-170"F dietemperature, 15SF basket temperature at an extrusion rate ot 1.05 in./min and 2800 psigpressure. Both RPD-12 and RPD-14 were extruded into solid _71indrical Franules and the TOWlaunch configuration. The solid cylindrical granules were used in ciosed bomb testing to obtainburn-rate information whi'i is provided in the attached description sheets. RPD-12 propellantstrands extruded into tie TOW launch configuration were then processed through the TOW finishingoperations as shown iý, Fig. 2. This propellant was ballistically tested in TOW launch motors,the results of which are discussed in the section oi solvent vs solventless H7 comparisons. AdimenE onal summary of the comparison of all. RPD-12 propellant lots produced and solvent-type M7TOW is shown propellant in Table III.

Table III. RPD-12 Dimensional Comearisou and Sumeary

Dimensional Su- & Solvent RPD-12 (Solventless H7)n7 Lot-I Lot-2 Lot-3

Outer Diam.. in. 0.780 0.759 0.771 0.781Ynner Diem., in. 0.363 0.336 0.337 0.356'deb, in. 0.208 0.212 0.217 0 213Surface Area, jn2 charge 191.9 185.5 183.3 191.0Web range, in. t 0.060 1 0.010 a 0.010 a 0.010Web Std. Dev.. in. ± 0.011 1 0.010 1 0.002 0.002

I0PD-12 Solventlesu 017 Lots 1 & 2 were processed and tested at the same time.

RPD-14 propellant was not processed any furtter due to limited resources and the beliefthat a replacement which contained no potassium perchlorate would not be considered a replacementbut a new propellant formulation.

PRODUCT INPROV•ENIITS

During the processing of the solventless N7 propellant through the TOW launch finishingoperations, significant improvements in dimensional uniformity and overall propellant qualitywere ubeerved with respect to the current solvent-type M7 propellant. Table IV sumarizes theimprovements that were realized as a result of this process change. All economic and qualityimprovements were made relative to 1989 production figures.

Table IV.Product Imorovesents for Solventless N7

Solvent SolventlessImproved Catexory NM7 m

Internal tefe.ts, % 9.5 <1.5Surface defects, IL 5.5 <O.sQuality Costs, 8/yr >l.lN <lOOKExtruded Cost, 8/lb 19.88 13.00Finished Cost, $/stick 20.00 12.00HOg variability, cal/g 4.4 3.4Web Variabilitj, in. 0.011 0.002

175

After radiographic inspection, approzimately 9.51 of all solvent-type K7 T launch sticksare rejected and subsequently burnAd because of internal voids that are inherent the solventp-ocesa. Inspection of over 300 solventless P7 propellent sticks indicates the reject rate forinternal defects would drop to less than 11.. The risulting cost saving would be greater than300,000 $/yr. This number is based on 16,396 stickt being rejected at an average cost of 20S/stick.

At the sawing and machining operations. approximately S.S% of all solvent-type $7prop,!lent is rejected due to blisters, pin holes, and other rurface defects. By using thesolventless orocess, surface defects could be reduce. to less than 0.51. This would result in asavings of over 200.000 S/yr. This savings Is based on over 14.500 sticks/yr that are rejectedfor surface defects. TOW launch sticks cost 10.33 S/stick after sawing and 13.88 S/stick aftermachining.

Curreatly, solvent-type X7 propellant is mixed and extruleu an average of 2.25 times beforei usablt strand is produced. This remizing of green dough propellant results in i averagekqlity cost of 500,000 S/yr and contributes to the final propellant cost to be elevated to

nearly 20 $/'b. It is estimated that, by using a multi-port solventless extrusion, thepropellant could be processed for approximately 13 S/lb. Based on the cost differential of 7$/lb and an iverage production rate of 68.000 lb/yr, 475,000 $/yr in savings could be achievedusinr the so ventless r.ethod of pro~essint.

By using the solvnntlear process, dimensional variability was reduced by 801.. The.Jistributions of prc vllant web for both the solvent and solventless M7 propellant is

shown in Fig. 4 to illustrat• the reduction in dimensional variebiity. The solventless M7propellant is Pkewel slightly to the right because of the propellan- not having an optimizedweb. This greater t-nformar,ce should translate into improved ballistic uniformity.

ax -

ITI \

oi - .... -/0. 7 0.18 V.,J 0.2 021 0"2 023 024 025

VM 1)

-- 9)Mthms - 97int

Fig. 4. Distribution of Solvent and Solventless N?VsO. Variation

Improvements :n chescal uniformity is also possible since blending, in the solventlessprocess, is dons prior to miing anc extrusion. If heat of explosion is utsd as a relativemeasure of chemical variability, then a 281. reduction in chemical variability was achieved withthe soiventless process. The normal variation for double and triple-base solvent typepropellants is +/- 4.4 cal/g whereas the normal variation for solventless propellants is #/- 3.4cal/g. This improvement in chami ,l uniform,mty should also improve end-item ballisticconformance.

176

SOLVEINT VS. SOLVKnTLES M7CONPARISON

Although the solventless N7 propellent appears to provide a significant reduction ofquality coats and substantial improvements in dimensional and chemical uniformity, the propellantwas tound to have a considerably lower tensile strength than its silvent counterpart. Thesolventless MY propellant in the TOW launch configuration was found to have a tensile strength of400 Ibf/in2 whereas the minimum requirement is 600 1bf/in 2 and standard solvent-type propellantaverages around 720 lbf/in2 . It is also observed that the standard solvent MY is much morebrittle than its solventless counterpart. Based on this limited data, it cannot be determined ifthe change in physical/mechanical properties will improve or be detrimental to motorperformance. The need exists to better understand the differences in physical/mechanicalproperties between solvent and solventless Ny and to determine what effect they may have onend-item ballistic performance.

Using the TOW launch finishing operations (Pig. 2), RPD-12 MY solventless strands wereprocessed into sticks ani then loaded into motors for static ballistic testing. A summery ofthis ballistic testing is provided in Table V for the two solventless test firings that wereconducted. Table V also provides a sumary of ballistic test requirements end firing date for astandard solvent M7 TOW launch lot (RAD89HO07-0ll).

Table V. Ballistic Summery

Solvent Solventless MNN7 Lot 1 Lot 2 Lot 3

Ballistic Sumary (-2S*F):Charge weight, g 570.0 546.0 573.0 575.4Delay time, msec 10.4 11.1 21.3 11.1Action timo. m,sec 59.2 70.5 71.8 61.7Press max. psig 5541.0 4353.0 4615.0 5238.1Thrust max, lbf 6097.0 5015.0 5313.0 5671.9Impulse, lbf/sec 270.0 241.0 262.0 269.8

Ballistic Sumary (+125"F):Charge weight, g 570.0 546.0 573.0Delay time. msec 6.4 7.4 7.3Action time, msec 29.4 29.6 28.9Press max, psig 10907.0 10293.0 11223.0Thrust ma:, lbf 12402.0 11614.0 12725.0Impulse, lbf/sec 278.9 262.2 272.0

* Lost sticks on first motor; no other motors were fired at +125"P.

The first solventless MY TOW firings were made from RPD-12 propellant, lots 1 and 2. Twodifferent web sizes were produced and evaluated for ballistic performance. sizing of the dide,to produce these lots, was based on the best shrlntag*/swnlling date of other solventleaspropellant and on rocket motor performance predictions. Predictions were based on the closedbomb burning-rate data as Rhown in the burning rate analysis summary of thy attached descriptionsheets. Dimensional analysis (Table III) Indicated that the solventless M7 had substantiallyleos surface area then its solvent counterpart. The effect of this was to lower the averageoperating pressure in the motor, causing the propellant to burn at a lower rate. A slowerburning propellant combined with a large web resulted in long action times and low Impulses at-7?S* for both lots I and 2. Analysis of ballistic date indicated that solventless N? andsolvent N? have similar burning rates in the presmure range of TOW launch; therefore, It appearstl ý TOW launch end-item ballistics could be achieved with RPD-12 solventlese Ni propellant intbt standard TOW iaunch configuration. Due to the difference in burning-rate between the Clowa,1bomb data and the moto- firing data, the need esists to generato burning-rate data over the 2-9Kpsi pressure rant- W ich could be used to better model the system at estreme temperaturo.

177

Balsed an the first firings, the th'.rd lot of RPD-12 solventless NY TOW launch wasproduced. titrusion dies for lot 3 were design based on shrinkage data from the previous twolots. Dimeasional analysis of this propellant Indicated the propellant to have about the rightamount of surface area; however, the web was larger than desired. from this analysis It wasknown that the t pallant would stIll be slightly long on action time and low on imrnulse at-25*F. Problems also developed during the Vin assembly step. I new polyuretbano adhesive. usedto pie the sticks, had foamed, opening up channels for flawe penetration into the pin stick.interface. Since the predicted performance of this lot would be subttantielly bqtter than theother two previous lots. it was decided to test this propellant even at the risk of losing sticksdue to the adhesive. During the firing. Use sixth motor cold (-2S7F) and the firut motor hot(*12S*F) lost sticks; however, ballistic data from the first six cold firings indi xted aprogression in performance toward the specification. Based on this firing, it is 1t34sible thatballistic performance parameters for TOW launch could be met with RPD-12 propellant using a0.209-in, web and greater than 192.0 in.2 of initial propellant surface area in the motor,

Although ballistic performance for solventless NY propellant has not been completelyoptimized, substantial improvement in performance was made during the last test series. Thisprogression in ballistic performance, combined with product improvements such as; roduction ofinternal and external defects, improv@C chemical and dimensional uniformity, and redicedprocessing costs warrant further optimization and development of a solventless my propellant.3f(forts will be made to generate Interest at RICOH to complete 4evelopment and implementation ofthis propellant in new or existing shoulder-fired rocket systems.

CONCLUSIONS

1. It is feasible to produce a modified MY propellant, that meets the requirements ofMIL-P-14737 using the solventless method of propellant manufacture shown in Fig. 3.

2. By using the solventless process for manufacturing NY propellant, an 601 -'e.-i'ttou ininternal and external defevtts to the end-item can be achieved.

3. By producing NY propellant using the solventleas method of ma~nufacture. quality cost&can be reduced by IN S/jr given i989 production rates. This cost saving Is due inpart to a lower propellant reject rate end ecoaamies-of-scale that. are possible inusing a multi1-dis press facility.

4. NY soiventless propellent extruded ialco the TOW launch conflguratios will show an SotImprovement In dimensional uniformity over the standard solvent-type NY prnpollatitcurrently being used for TOW launch.

S. By using the solvenUost method of blending prior to miziag, to process thesolventloss NY propullant, a 257. improvemesnt in chemIcs al ~forelty con b-0 schievadover the solvent technique of misiog and then blending.

6. The tensile strength of NY solventleps pitpolle'bt extruded int-* the TOW launchconfiluration was 3Acv-!r. ur'der low-ttaLk lostiol tests, to it? prpopllmnt insanutocturedusing the solvent pro ess. It has nct boso ea'ablisimed frovi the liaitod mtatorfirings it the lower tensil, strength of tbhs A? solventloss propellent is Importantto moto4r parfoemanc*.

7. Cnd-item (Y*W launch) ballistic #,erformance roquaroemmto wvte hot met. vit~h MY7solvootles. pr~pellovif. toctupe the *9propviate Sanoatry was viot .chioiol. liowetat,based on tiat bol'.Igtir toest fitiedg' o'. forped to datv, It tcposer& that by floolvmingpropellant $9eAtry. talli,'tic Vvrformtrc* for TOV )*%i~ch bod #,thel rackeLs WAtahutilize W~ v~old b4 achievcO.

I iN

REFERENCES

1. R. R. Buell, "Multiple strand Extrusion of Solventless Propellant: A Study ofModified M7 Propellant Extruded As 3/8" and Stick Powder (N28)", Project5S-304-BK246, Badger Ordnance Works, Baraboo, Wisconsin, January 1956.

2. W. G. Clark, "Evaluation of the Solvent Solventless Process", Dept. of Army Project517-06-003, Picatinny Arsenal, Dover. NJ, August 1959.

3. J. N. Juran and F. N. Gryna, "Quality Planning and Analysis", 2nd Edition,McGraw-Hill Book Company, New York, 1980.

PBOPgoUANrT DUSCRIPTION SHET

RPD-12 (SOLVENTLESS M7)

Chemical Properties: Stability Properties:

Composition Nominal, Analysis,

Nitrocellulose 55.0 54.54 100C Talian'i TestNitrogylcerin 40.0 39.91 Slope at loom0 Hg 0.50Potassium Perchlorate 3.0 2.67 minutes to lOom Xg 213Carbon Black 1.1 1.05 Slope at 100 minutes 0.45Ethyl Centralite 0.8 0.89 120"C Heat TestCandelille Wet 0.1 0.09 No color change. minutes 60.Moisture 0.84 No fumes. minutes 60

Thermochemical Properties: Physical Properties:

Calculated Measured Absolute Density, &/cc 1.64HOE. cal/g 1267 1268 Dot bone tensiles at 70'? (lengthwise)

Thickness Width Stress Strain Tension(in.) (in.) (in.) (i.)• (lb)0.085 0.381 1038.8 0.094 33.S

TOW Launch grain tensile 400strength. lbf/In2

Hazards Sensitivity Data:

Dry WetPaste Paste Preroll Evenspeed

Impact. cm 11 64 41 64Friction, lb@ 8Upa 120 217 275 345ESo. . 0.125 0.262 > 9.5 > 9.5Moisture. 1 0.01 12.2 3.2 0.5Sample Thicknessmail 31 24 49 87

Strand Burning Rate Data:

Pressure. pair ,25*F ,70"F 41256F

late Date (in./sec)

1000 0.56 0.68 0.772000 1.00 1.12 1.203000 1.28 1.50 1.68

Closed Bomb Burnrate Data:

r - spn where r - rate in (in./sec) and P . Pressure in psi&

Temperature, F a n

-25 2.88x10-3 0.84+70 2.56z10-3 0.86+125 2.69nl0"3 0.87

Obtained by calculating linear burn rate from closed bomb data reduction at a 0.15loading density in a 700 cc bomb. Solid cylindrical grains with a diameter of 0.4inches and LID of 1.0 were used for this purpose.

Best Available Copy179

110.10 IOOLVINTIISS 1771

gbjIMII2L Frer+el Stab~l~tY too P olor .

Nttrooe11o5 $0.0 S5.10 100*C T.llet TestNltOeoplcorIO 36.0 30,51 Slop. at LOOm "IN 0.0

Potassium Perchlorato 7e 0 0.07 mLnutes to loom hg 250Carbon Black 1.1 1.14 Slope at 100 mLnato. 0.52

1'llL Contralito 0 0.8 0.88 1200C Mest ToolC."i l l .e 0.1 0.01 No cor cbang.. .inot.5 00..let.r. 0.41 No tome, . e. 30

Tbeo0rcb tell ProoprtiLe: physical PropertLes:

Calculated Ro orod I tO Density. 1/ct 1.62NO!,. oe1'0 1270 1270.0 Dos bone .. le51.. at 70-F liengthree,

Thickness Width Stress Strain Teolon(in.) ýJ.,)j. i .]•__ I_ A Il. bL0.008 0.383 1010., 0.001 07.2

Nluards SoMlItitity Ot.1:

Dry Wtg

Fast* Plate Prorl ... n 11

Impacet 13 $1 21 00FrLclLoo. lb. oft. 100 217 200 3i1[S0, J 0.13 0.202 1.0 9..MOistore. o 0 10.0 2.30 0.0Sea.T11 Thi ...ck .. 1 30 "I 01 00

Utreed sursing late Data:

Pressure. OIlt -25-F T.? *2s"rRit* Data (Ift./Ift)

1000 0.03 0.00 0.70

2000 1.10 1.13 1.223000 1.31 1.52 1.67

closed so" I300n0t0 pets:

Closed bogb data asl 000 obtaliOd for Wll lot.

IPO-16 lSOLVNTh1E$S MI)

Cheicalto Proporties: Soablilc Prootrtlo.:

Cowsiotloo !o!1is.L. &.slopisA

MItrocelloloso 0W.0 00.29 l0OC Thilsol Testcierololti. 03.0 42.70 Slop. *1 loOm Mg 0.0?

pgaesllt PTrch.or.t. 0.0 1.00 min.utes to lOMe NM 1isClrbOo 111 e.1 1.10 Slope :t 100 leont.l 0.S010071 C+enr.1i00 O.8 0.00 120'C Nlot TestCsod.11115 0.1 0.1 0.09 No color chanso. minutes 60.mOisture 0.17 NO (i.. inutes a 0

The.sochelmicl Propwrtln: rko.0cal Prooortl.:

C0&l11 Msure d blots Density. 1/! c 1.02001. c01/1 1 201 1209.0 Do, boos 000.l1.. t0 '.* FloegtO.il

Thiclll. 1idth Stte.. Strain Tenslio(I-) (in.•2 1,-I. •tsL ..) OhL.

0.013 0.37S 1001.3 0.100 13.7

otrolt1. l0

Notatd Sensltivity Data:

Dry wotPast- Fast- prstoll fooooo..d

"m'oct' .3 11 04 20 01PrIct.,,. 1t.1.rps 100 27 200 2.1110, 2 0.209 0.202 9.4 1.,NOlatot.. • 0 10.0 1.7 0.1Semple ThloOtl~llll 31 20 01 80

Pressoro. oaio , 20I .11 !12 5"F(1.15 1. De5 (0./sd)

1000 0.07 0.7 072000 0.91 1.14 1,201000 1.25 1.00 1.00

C ...d Boush Worstt. Date:

r - ap whbere r - rate in lb/sl . 0) and P Pres.sretin pily

T.loersturV. -A oS

-2s 0.01" :.S

,?0~~ 1.•Ol- 0.9

.120 2.S112103 0.08

Obt.1n.0 b. ololttlol 110..r bore redl rog olomo 0.05 rc:ucontI 0 0.11oed1ol denei•y In o I 00 to bogO. -.171 cyllodris l .t.in. ilth . dl.i tor of 0.0

ob.e and L.'0 ot 1.0 ere u." for this Pourpos.

Sest Available Copy

1 80

MECHANICAL DESIGN OF SURFACE LAUNCHED TACTICAL MISSILES

Jon J. YaglaProtection Systems DepartmentNaval Surface Warfare Center

Dahlgren, Virginia 22448

ABSTRACT

The paper discusses the mechanical design of ship launched missiles and launchers. There are manyrequirements of the missile besides being a flight vehicle. The secondary requirements drive much of themechanical design of the weapon and launcher. The paper highlights the secondary requirements, points outwhere they are critical in design, and indicates the present state-of-the-art.

INTRODUCTION

Ship launched AAW weapons pose design challenges beyond meeting flight performance requirements.Interceptor missiles must minimize weight in order to attain maximum speed in flight and maximum kinematicperformance at intercept. The most advanced interceptors use the rocket motor chamber as the airframe, andemploy the strongest steels available. There are many requirements of the airframe besides being a flightvehicle, e.g. stowage environment, launch environment, shock, vibration, and insensitive munitions. Thesesecondary requirements drive much of the design of the weapon and launcher.

Ship launched cruise missiles and anti-submarine missiles also must meet the same basic secondaryrequirements as AAW missiles. However, these missiles are slower, heavier, and not designed forintercepting high speed, maneuvering targets. The performance penalties for designing these missiles forthe secondary requirements are not as severe as for AAW missiles. The thrust of this paper is to highlightthe secondary requirements, to point out where the secondary requirements are critical in design, and toindicate the present state-of-the-art for design of AAW missiles and launchers.

The design of the launcher can no longer be divorced from the design of the missile. In the pre-launchcondition, the missile and launcher and/or magazine are a coupled mechanical system when under excitationfrom the secondary environments such as underwater shock. The design of the latest interceptors hasrequired very sophisticated mathematical models of the coupled system in order to concurrently meet weight,space and performance requirements. The modeling effort interplays with experimentation in the laboratoryto overcome specific design issues, and to verify the final mechanical designs.

Futuristic ship designs project even greater capabilities than present designs. These ships will attainunprecedented speed, maneuverability, and sea keeping qualities through innovations in mechanical design.At this level of sophistication, the design of the platform, launcher, and weapon will be coupled in themathematical sense.

The paper traces the evolution of the design of the ship launched AAW missile and its launcher. Theemphasis in the paper is on the secondary environments. The present state-of-the-art is presented throughexamples of the current design approach. Projections for analytical and experimental design and methods forfuture AAW ships are xe~ented.

THE MISSILES

Figure 1 shows the three current AAW missiles under development in the US Navy. The missiles aredeveloped and improved through an evolutionary process. Each missile was first deployed in the 1960 decade.The present Standard Missile SM-2 has its ancestry in the Tartar and Terrier missiles. The weight andexternal appearance of the present generation is closely similar to the original missiles. However greatlyimproved performance has been attained through gradual improvements in propulsion' For example, -Tartar and------...-Improved Tarter employed the MK 27 rocket motor. A new milestone in performance was attained throughsubstitution of the much hotter MK 56 Dual Thrust Rocket Motor for SM-l. The present SM-2 missile employs alonger and more powerful MK 104 Dual Thrust Rocket Motor. The MK 104 rocket motor has been produced inthree "MODS," the MOD 2 version having improved capability for surviving underwater shock, and the M3D 3version to be the upper stage propulsion of the SM-2 Block IV missile which employs a separate boostermotor. Boosters were already employed in the Terrier, which has recently been superseded by SM-2 (ER). Theearly Terrier booster was the MK 12, which employed nitrocellulose propellant, and has been replaced by themuch more powerful MK 70, which employs ammonium perchlorate as the oxiditer and aluminum as the fuel.

The basic outlines of the missiles have not changed much, and are somewhat limited in variety by the factthat the missiles have to be fired interchangeably in many different launchers. The launchers and shipsimpose weight and volume limitations on the missiles. So the design challenge has been to attain improvedperformance within the bounds of older pre-existing weight and volume constraints. The SM-2 flight vehiclehas been 13.5 inches in diameter since Tartar, missile weights have increased modestly, but the limitsimposed by the thips and the launchers have essentially been reached. Future increases in missile weightswill have to be offset by weight reductions in the launching system.


Best Available CopyIs1

Figure 1 shows a sketch of the NATO Sea Sparrow missile. The missile has evolved through three decades.The missile has always been launched from ships from trainable and elevatable box type launchers. The

launchers were originally adaptations of a MK 25 gun mount to the missile launching requirement. The system

evolved from the Basic Point Defense Missile System to the present NATO Sea Sparrow Missile System, with

much improved missiles, launchers, aad fire control. A special "Dual Pack" canister, that can fire two Sea

Sparrow missiles from a single 1K 41 VLS cell in under development. The basic 8 inch diameter of themissile has not changed.

Finally, figure 1 shows the Rolling Air From Missile (RAM). Again, the system is evolutionary. Themissile is an adaptation of the air launched Sidewinder missile to the ship launched role. The 5 inchdiameter sidewinder propulsion and warhead have been used, with vastly different seeker and guidancesystems. The launcher is an adaptation of the missile box launcher concept to replace the cannon in the

Phalanx Close-in Weapon System mount.

In conclusion, the three main AAW missiles in the US fleet are the products of an evolutionary process.There are three basic diameters of missiles employed. It is the writers opinion that, barring radicaltechnological advances in propulsion, these basic weight, space, and envelopes will continue well into thetwenty first century. The mechanical design challenge will be to get more performance out of each of thesesy-tems, rather than a wholly new design.

THE LAUNCHERS

The basic launching systems are shown in figures 2,3.4, and 5. The Standard Missile is launchedinterchangeably from various rotating arm guided missile launching systems, the MK 26 shown in figure 2being typical. The missiles are stowed in a magazine below deck. To launch a given missile, the magazinecycles the missiles around the interior of the magazine until the desired missile is at the hoist station.The missile is hoisted up onto the launch rail and retained on the launch rail by a latch until it is fired.The rail is ttained and elevated until it points along the intercept trajectory, then the missile is fired.

The Vertical Launching System is shown in figure 3. The missiles are transported to the ship in theirlaunching canister, which also serves as a shipping container. The complete canister is installed in thelauncher. The missiles are never handled or exposed to the outside world until they are launched.

The launcher for the NATO Sea Sparrow is shown in figure A. The launcher shown is installed on theaircraft carrier USS Roosevelt. The RAM launcher, figure 5, is shown installed on a small German warship.

THE SHIPS

There is a great variety of ships that fire AAW missiles. The large Standard missiles are only launchedfrom large surface combatants. These range in size from the 3600 ton USS Perry frigate to the 10,000 tonUSS Virginia class cruiser. The smaller ships employ the single arm MK 13 GCLS. The cruisers employ the MK26 GMRLS with larger magazines and twin arms. The most recent Aegis ships, beginning with USS Bunker Hill(CG-52), employ the MK 41 VLS. It is a continuing challenge to improve the perforngance of the missileswithin the weight, space, envelope, and interchangeability constraints imnosed by the such a diverse set ofplatforms.

The NATO Sea Sparrow Missile System is deployed on an even broader set of platforms. The system isdeployed on aircraft carriers up to 90,000 tons, and down through smaller patrol size ships, and on a widevariety of combatants, auxiliaries, and amphibious assault ships. With carriers and a-Tphibious assaultships, the systems are frequently installed on sponsons at the level of, or immediately below the flightdeck. On cruisers, destroyers, and frigates, the system is usually the main AAW battery and installed onthe center line on or above main deck. Figure 6 is a typical sponson installation on the forward starboardquarter of USS Midway.

The RAM system, as installed for OPEVAL on USS David Ray, DD-971, is shown in figure 7. 1he RAM system,small and compact, can be the main AAW battery of a smaller ship, or a secondary point defense battery for alarger ship.

In summary, there is a tremendous variety of ships deploying AAW missiles. As the platform size varies,and the locations of the systems aboard these ships varies, the mechanical environment varies. As the Navystrives for commonalty and interchangeability across this wide spectrum of platforms, the price for this ispaid in performance and difficulty of design.

UNDERWATER SHOCK

Figure 8 shows an underwater explosion test of USS Mobile Bay. New ships are routinely tested fortolerance to underwater shock. Figure 9 shows a cross section of a ship in an underwater shock situation.The shock is nearly spherical and is shown at several times. The explosive energy is proportional to thecharge weight, W. The surface area of the wave front is proportional to the square of the slant range, R.The energy flux density at the ship is therefore proportional to W/R Z. Early experiments showed targetdamage scales as the square root of this quantity, the so-called keel shock factor WI/2/R.

8 Best Available Copy

As the shock wave diffracts around the ship, large surface pressures are created, on the order of athousand pounds per square inch. The only significant counterpressure is the atmosphere, so the largestmotion is vertically upward. There is also significant motion in the athwartships and longitudinal

directions. The motion is characterized by an impulsive change in velocity. The velocity changes typically

stand in the ratio 1:0.5:0.3 for the vertical, sthwartship, and longitudinal directions respectively.

Looking at the ship cross section in figure 9, the velocity change would be most sudden at the lowest and

stiffest part of the ship. Equipment in the center of the ship is connected to the hull and keel andthrough elastic structure, so the accelerations are not so large. Equipment installed in the superstructureexperiences even lower accelerations. For a given shock factor, smaller ships experience larger velocity

changes and accelerations.

During underwater shock testing, impulsive velocity changes of the order 10 feet per second areexperienced. Accelerations as high as 400 g are measured in the deep inner bottom of the ship.Accelerations in the range 40 to 100 g can be expected for equipment in central spaces below main deck.Accelerations on the order of 20 g are measured in the superstructure.

For ships with rotating arm launchers, the most direct shock load path to a stowed missile is from theinner bottom of the ship, to the foundation of the launcher, to the ready service ring support structure, tohanger rail rollers and hanger rail, aad finally through a rattle space in the aft shoe latch into themissile.

The rotating arm missile launchers were designed before the finite element method and sufficiently largecomputers were available as engineering tools. The development of these launchers was troubled by lack ofdata on the design environment, and the unavailability of a sufficiently large floating platform forconducting underwater shock tests of a complete launcher in a laboratory environment. The prototypelaunching systems were delivered to operational ships and evaluated in the fleet. It took many years oftesting and engineering changes through the ORDALT process to develop shock capable launchers.

The MK 13 and MI 26 launchers are deployed in USS Perry class frigates, USS California and USS Virginiaclass cruisers, USS Kidd class destroyers, and the first five USS Ticonderoga Class cruisers. As new St-2type AAW missiles are developed, they will have to meet the requirements and carry the loads imposed bythese ships, well into the next century.

The most direct load path in VLS ships in from the bottom of the ship, to the magazine foundation, to themodule plenum, to the canister launch rail, to the safe and arm device and into the missile. The path ismore direct and through stiffer elements in the VLS case. This leads to higher inputs to the missile.

The VLS with AAW missiles is deployed in the Ticonderoga class starting with USS Bunker Hill (CG-52). Theentire USS Arleigh Burke class is being built with VLS. The VLS weight, volume, shape, and accelerationconstraints will dictate much of the mechanical design of future AA3 missiles and launching canisters.

The finite element method was used extensf.vely in the design of the VLS. The analysis employed verydetailed structural models of the launcher plenum, uptake, lattice, and dock, canister, and attached partsand panels. However, the early calculations employed veo .imple missile models, which modeled the missileas a "stick" with three segments.

Figure 9 shows that the shock design acceleration levels are highest for the launchers with magazineslocated below main deck. From the standpoint of underwater shock, launchers located in the superstructurewould be most desirable. This is possible for the smaller missiles, but for heavy missiles the ships wouldbecome unstable.

The right side of figure 9 shows accelerations experienced by a typical missile in flight. By comparingto the figure on the left, it can be seen that weight and strength penalties attendant to the design ofmissiles and launchers for the shock enviro.znent will set in at and below main deck. A very large part ofthe mechanical design effort for the SM-2 Missile was in the area of underwater shock. The major mechanicaldesign activity for the SM-2 Block IV missile and canister was underwater shock.

The requirements for underwater shock are enforced through application of MIL-STD-901D, which prescribestest procedures for shipboard equipment. The Navy also conducts shock testing of every new ship upondelivery. A floating platform and test tlxture for underwater shock testing encanistered missiles is underdevelopment.

GUN BLAST AND NUCLEAR BLAST

Figure 7 shows RAM and NATO Sea Sparrow systems installed in close proximity to the 5-inch gun in USSDavid Ray. Air blast environments are very severe for exposed equipment close to guns. Guns as large as16-inch are deployed on na-ial ships, however the largest guns deployed in proximity to AAW missiles at thepresent time are 5-inch. The air blast is an instantaneous jump in pressu.e with an exponential ortriangular decay of pressure back to ambient. The design level is 7 pounds per square inch incidentoverpressure. The gun blast pulse has a positive phase duration of 5 to 10 milliseconds. The nuclear blast

Best Available Cop-183

pulse is the aomi sbape, but the positive Vihase durationsa re in tho rage of 350 amilliseconds for tacticalsized wespena * to three soocouds for megaton miss weapons.

Figure 10 shaw, bleat waves impacting a ship mad launcher. Typical values of the accolerations impartedto equipment are sho-a. The acceleration levels are typical of what KAYEC engineers have measured innrmeroua testse an expr~oimets with missiles,* launcheors, and fire control equipment. As one mtakesmeaurements Lower In the ship, below main deck, the values drop very quickly to insignificant levels.

The NATO Sea Sparrow and RAN sysatis were carefully evaluated for Sun blast survivability through tests atthe Dahlgrmi Laboratory. The test. simpoyed S-inchi guns and fully operational launchers and missiles.

Again, the right hand side of the figuLe shows typical flight acoeleration levels for coparison toacceleration& imparted by air blast. The blast accelerations axe more severe than for flight levels fortopsidet liocations on end above main deck. To survive the blast enviroret. above main deck, either thelauncher mast protect and isolate the mtiesile. or the mtissile mist be designed with strength well beyoudwhat Is required in flight. There will be a weight penalty in either the launchier or the missile.

The requiromeint for Sun blest survivability is enforced through RAMSAZNIST 9110.1 'Sttuotur&l Test firingas Surface Ships.' - ach nw ship, modernised ship, or ship receiving new ordnance has to conduct Sun blastteats. The tests are conducted by firing the guns at the angles of train and elevations that Impose thedesign Level loads on the new equipment.

FUMAING, UANLINO, TRAWSPMtATIOU., AND SYCRAaE

The AAM weapon mest be delivered to the chip after it hao been asembled. The current practice In the USNavy is todesign the shipping containers to mitigate the transpourtation environments, The main mechanical concernsare for vibration during transportation, WAd shock that would occur If the cootainer were dropped. Theusual requirement is for the systemi to survive and operate after being dropped oen foot. The item must also,survive, with out exploding, a drop from a forty foot tower. Figure 11 is a sketch of a missile container.'Lbo requivememe for transportation vibration is enforced through application of KZL-STD-6101). Rather thancompare with flight envicromets as before, it. is suggested that the rounad should be protected or Insulatedfrom theoe mechanical environsents by its shipping container to levels below the shipboard and flighteniviroemownts. There soumld be no reason to accept a weight or perforimancet penalty for a non operetingemvirormont.

"IM15AM VI3RATIOU

Hissiles deployd on surface ships aust survive years of vibration, without being teeted for flightrsediness, then be reedy to fire at a target, Ships vibrate continuously while underway. Under certaineanditions of speed at so-% state. the excitation frequency will coincide with oaae of the nornal nodes ofvibration of the belA . Ad the vibrstiosis became very large. Most. cruisers and desotroyers have largeregeoneeso around 2C~ beit6. Auy *16oile must survive the resonant vibrations of the ship. The requirement..fer viberetiso testing of shipboard equipment are speelfied in MIL-STP167-1(Slihips). The specificationrequireis the systems to be vibrated in a variobtle frequency test in a range of 4 beirts to "O harts indiserete froequeny intervals of ans haft$. At each integral frmoquaiy the vibration is maintained tog fiveminutes. Tn. variable frequency test reveals any, resonances In the syatem being tested. The second test Isan enduramsee test in wtaish thte eq'iipont Is vibrated for a total period of two hours at its resonantfiequenoles, or a speaified upper frequency if no resonaticea erao fo~tud.

The asiduranee test loads to a loris numer of oyslea of loading. For examle, after 2 hours at 25 hertz asystem hee been cycled 160,000 timess. When one sonsidera the uesul allowable stroee-ntembor of cyclos curvesfer struetural material., or aleowable stress detratinge Go recoemmawded by the Auerican Welding Society endthe Amerisan Institute of Steel Construction, the test becomet stevetre.

nhe %$0. is all the mere severe It sesonanwoe in the test Item occur. For simple anettanitail eyetows insteadY sLats vibration, th astual dieploomewt, stress, a"i acceleration sen be mcain' times tMe value thatwould 4,41ut W1der the influenee of the &m forse applied quasistatiaally. The 46tmal stagnitieationattained depend on the daminit present. For moet structural materials the dmaing is In the zrange of 21 to69 of eritisal damping, Yes rubber type materials, as used tit vibration dempre "n shock isolators, theIaimplg sam be as large aso M, The megifieations attained at resonene, are 25, 10, nW 3.11 got damping ofS1, i,. OWd MI, reapeotively. At fr6equelces eeey from resonance the systeM response drops off veryrapidiy, nd Misese is ne design peiralty.

Ihe boy to. evoidiatg v41stlouo pueblomo in missilesensd laustabhes is to design the eysimai suftficietlyWt(i W. reasemeto et flawietteelo above thooo. W. toe ontuwatered in the ship and required In the test.

The shipboard viboetis., peoltlem for an hAN olecle tos shetlsed in figure 12 rhe drawing on the leftahaws a missile In its sanister As with olio'b, ship oultiet Imparted to the vaoister arm thrseediesassusl. Pop the VL4Oo tlr~put.. to ilie 4oiletei are whosea it. meetso the upoper deck of the lsunches, tielptsiface Loatosn the w~dis t ti!s t.f:t t:219 -it .H. tit.esee0. e4A at 11a A,-@ down, i&Labsa a short,

1014

distance above the plenum. Ship vibrations drive the canister in a general motion that can be resolvedalong the three orthogonal directions as shown. The vibrations are passed to the missile through the strong

hack (canister launch rail), to the safe and arm device which clasps the missile by its forward launchshoes. In the box type and rotating arm type launcheos, the attachment to the missile is closely similar.The boxe* ire supported by a mount with trumions and train and elevation drives, and the rotating arm typelaunchoer support the missile on a magazine rail (hanger rail) with a shoe latch.

Figure 13 shows handbook curves for of the steady state vibration of a system as a function of excitation

frequency. Damping in the range of 0.02 to 0.15 of critical should be expected in practice. The response

at resonance is drmtic. no matter what the damping. The missile, launcher, and ship vibrate as a coupled

mechanical system. The frequency response curve of a coupled system is different from the frequencyresponse of any of its individual elements. For example, an empty 3000 pound canister would resonate at ahigher frequency than the same canister loaded with a 3000 pound missile.

Analysis of the vibration problem is just now becoming possible with the finite element method. The

problems are so large that very powerful mainframe and super computers are required, and solutions areexpensive. The payoff in computer analysis is a vastly improved understanding of the vibration

environments, anticipation of problems, and correction of deficiencies at the earliest possible time.

The right hand side of figure 12 shows a qualitative c,wg-ilsons of the vibration environments for flightvibration, transportation vibration. and ýiWioard vib: .:'. The comparison shows the vibrationrequirements aro essentially similar. rlo%..s -he syste.ý r .- _e.med to resonate in the shipboard vibration

environment. If the missile/launcher system is allowem.La v :-onate, there may be a significant, design

penalty eototndnt to the higher stress and acceleration at. resonance.

CURRENT DESIGN PRACTICE

The current focus in the Navy on AAN missiles and launchers is on the VLS, Sea Sparrow canisters, for VLS,

and the S4-2 Block IV missile, booster, and canister. The 1i80 decade has seen the finite element, methodmature from infancy, skepticism, and crude rudimwntary applications, to a sophisticated analytical tool that

produces data conaidered to be the baseline for design and comparisun with experimental results. As

explained below, the SI-Z Block IV and its canister have been designed entirely based on results from the

finite element method.

Experimental date from Tomahwmk cells in the US$ Hobile Bay shock trials, figure 8, were used to guide the

design of the S4-2 Block IV missile and its canister. The Tomahawk missile mad its canister are

approximately the ame weight as the Block IV. Fortunately, date for the Tomahawk missile were available toguide the design of the Block IV round. The motions at the interface between the canister and the VLS were

measured in the shock trials. Those motions were used as the driving force for design of Block IV.

The design process was very involved. The finite element method was first used to predict the response ofthe launchers and stowed missiles in USS Hobile Day, figure 8. The experimental date obtained during the

shock trials were then used to refine the finite element model. The missile response in each cell was then

recomputed with the finel "correleted" model. The refined models were used to predict cell-to-coll.

variations In the response of the launcher. Based an the response in the model, the experimental data wore

scaled upward to represent the worst case loadings to be expected in all points of the launchor, The scale

factors for the worst case loadings were then iised as multipliers for actual experimental accelerationversus time data from the shock trials. These motions were then applied as Lime-dependent boundary

coniditions In finite element models. The process is as sketched In figures 12 and 14. The acceleration;time histories from underwater shock measurements were applied in a transient finite element analysis to

comqute the it po•he of the missile.

The model then predicted bending moments and sheer forces in the missile at each axial location. Thesevalues were used to design the sections and Joints. Time predicted values were also used to specify force*and moments for static qualification tests of the missile sections and 'oints.

Figure 14 is a , -grm of the Nastran finite elemerit model of the Block IV missile in its canister. The

model represents a t testune in the application of the finite element method in that the missile and

canister are anslysdo together as a coupled system. The sie level of detail is used in the missile as used

in the canister. At .ht rnsti level oa sophistication in the devel.omaert of ship AAW weapons and launchers,the rocket motor of the missile would be modeled as a shell filled with viscoelastic propellant. In the

final level of sophisticaticoi, the ship will be incorporated into the coupled system model.

The vibration pr blem is being analyzed in & similar way. The sams finite element model as was used forshock Js being used. The vibsaLiorns specified in MIL SID 167 are being applied as boundary conditions tothe finiteo @lament tdul. The analysis hies been ccarplicatd by several factors. First, the damping

coefficients and material properties are uncertain for this prqblom. Secnnd, the problem is nonlinear. The

uonllnearitits arise in the first level of detail because if gaps and rattle spaces. For large amplitude

I I.',

motionmi, a second set of nonlinearities in the viscoeLastic properties of the propellant may have animportant affect on the vibration response. In a strict sense. the powerful methods of uodal analysiscuiot be used for nonlinear problems. Calculations with a transient inta44atlon scheme that does not relyon model analyeis, and still accounts for the nonlinearities in the finite element models are beingattmted. Results from linearised models ate believed to be qualitatively correct, end sufficientlyaccurate for the design of section level and colete round vibration tests and apparatus.

The above analysis and design effort took several years of work by a combined government and Industrytem. The original finite element models of the VLS were constructed by Martin Mariette. The coupledanalysis of the Block IV missile and canister was carried out by 1MC. The original S1-2 missile model wasprovided by General Dynmics, with alterations for Block IV done by Raytheon and FM•. The HK 104 rocketmotor was deveLoped by Thiokol. Detailed mathematical models of the MKi04 rocket motor, booster motor, andfine were provided by EAVSWC. The Naval Sea Systet Comand, Naval Weapons Center, and The Applied physicsLaboratory provided management, technical oversight. and independent celculations.

IJIIARY AND COPMUSICIPS

The mechanical design of AAM mlscilas for the secondary enviromments has been as challenging and difficultas for the operational flight environment. For many cases the secondary enviroment is the most criticalfor stress and acceleration. Waen the critical stress in an AM missile is due to a secondary environment,extra strength and weight are required. The extra weight leads imodiataely to a perforo¢ce penalty inflight. e.g. loss of maximum range, lower operational coiling, less acceleration available at intercept.

There have been recent major yoffs in undorstanding and performance as a result of the finite elementmethod. The past decade has seen rapid progress in the development and widespread acceptance of thisanalytical tool. Finite element analysis has been used extensively in the design of the vertical launchingsystem and its canisters. The method has been used extensively in the design of the Block IV missile. Forthe first time, a coupled analysis of the missile and its launching canister has been used to arrive atdesigns for each. Ta4 payoff will be superior performance of the AAt weapon due to mitigation ofsignificant secondary loads by the canister.

Future designs for AAt missiles will be based on finite element calculations of the ship, Launcher, andweapon as a coupled mechanical system. Barring unforeseen quantum stop improvements in the energy orimpulse of propellants, future improvements in thc flight performance of AAW weapons will be the result ofattaining the goal of fully stressed designs of the flight vehicle, optimized for the mission and flightenvironment. All the secondary environoents will be mitigated down tr equivalent flight Levels by thecanister, launcher, end ship.

1861

SEA SPARROW

r *OKXMOUNTID SOX*OUAL PACK'

TARTAR. TERRIERSM-t. SM-2

* ROTAlUE AIM LAUNCHERS mMRN* MOSILE14AGA4ES-1 h BELHAT

2AI% = WTS MISSILE hAIC: ;LTEY OF SHIP

RAMT 1

~~ *MA MK 3I LAUNCHER r(APATION OF CIWS MOUNT)~r~

FIGAURE I. THE THREE MAIN TYPIS OF AftO4PfNSE MISSILES OWNED ViRo,

NAVALSNIPS. ~d

SAIN SpACE /STrIEEOOmWI WLL

FIGUJRE 2. THE MK-20 GUIDED PASSILE LAUNCHING SYSTEM. TV[. SYSTEM

$TOWS THE MISSIES SIN THE MAGAZINE BELOW DECK. THE MISSILES ARECYCLED IN* THE MAGAZINE T0 THE LOADING POSITION WHERTEY A14EVEHOISTED ONTO THE LAUNCH RAUS THE LAUNCH RAILS TRAMN ANDELEVATE THE MISSR.ES TO THE INTERCEO" TRAPECTORY.

OUTBOARD

INITEEOSATE4STRUCTIORI

CANISTER LATCH

BASESTUCTUR

(PLEHUM)

FIGURE 3. MODULE Of THE VERTICAL LAIUNCHING SYSTEM_ A COMPLETEL.AUNCHER CONSISTS Of SE VERAL 440OOEJLS. THE MODUL.E CONSISTS OF FIGURE 4. NATO SEA SPARROW MISSIL LAUNCHER. THE 'BOX IYOA FRAMEWORK TO SUPPORT ENCANKISTERD MISSILES, A PLEUUM BELOW LAUNCHER CONTAINS ?HE &NISSILES IN A RINADV.IO4AUUCH CONDITION.THE FRAME WORK TO RECEIVE THE MISSILE EXHAUST. AND A CENTRAL THE BOXIS TRAINED AND ELEVATED TO POE THE hUSSIES AT THEUPTAKE DUCT tOODISCHARGE THE EXHNAUST. TARGET.

187

FIURE S. RAM LtAUNCHER ON FANTAIL Of GERMAN ShkP. THE RAMLAUNCHER IS AN ADAPTATION OF THE BOX LAUNCHER CONCEPT TO THEPHALANX CLOSE-IN WEAPON SYSTEM GUM MOULNT.

FIGU[E6 SPONSON INSTALLATION Of NATO SEA S*AI6O* LAUNCH4EON STARBOARt SOW OF USS FORtARSTAL THM INSTALLATION IS TYPICALOF THE iNSTALLATONS ON OTHERP AIRCIIAF T CARBIE AS AN) AMPHISIOUS

ASSAULT SHIPS.

. t.

FIGURE 7. RAM AND NATO SEA SPARROW LAUNCHERS ON US% DAVID

RAY. THE LAUNCHERS ARE INSTA.LED IN CLOSE PROXIMITY TO A S.INCH

GUN-

FIGURE 0. USS MiOlDI IAY, AN AIGIS SHIP WITH THE VZi, It SHOWN DURING ANUNDERWATiR SHOCK TEST. SHOCK TESTS ARE ROUTINELY CONDUCTED SIIORTLY

AFTER DELIVERY ON NEW NAVAL SHIPS

NUCL IAft10.20, &LAST

41,0 ILAST 7 ou

MI T) INOPINALT'Y 'p 420 TYPICAL MIS VLA

MILN DECKW~ 40-2009 P N L Y ACCIL E RATIONS

40. MAI OCK204 PENLT LAUNCH: 209

1.-29STAGE: 24TYPICNAL M~ISSILE log ~ MANEUVER. Il

YPICAL SHIP / /NOFNAtACrCIUEtAI1ONsI EXPLOSION

-,,.LAUNCH JogSTAGE 25g0MANEUVER, I~

FUIGUE 9. CROSS SECTION SKIETCH OF A SHIP IN SHOCK TEST. THE SHOCK FIGURE 10. GUN BLAST AND NUCLEAR BLAST WAVES IMAPACTINIGACAUSES AN IMPULSIVE MOTION Of THE 514W UP AND AWAY FROM THE MISSILE SHIP. MISSILES AND LAUNCHERS MUST WITHSTAND NUMEROUSCHARGE. ACCELERATIONS TYPICALLY MEASURED IN THE SHIP All GUN BLAST WAVES. NAVAL SHIPS AND IEQUIWVENT ARE ALSO DESIGNEDSHOWN. THE INSET FIGURE ON THE RIGHT SHOWS THE ACCELERATIONS TO WITHSTAND NUCLEAR BLAST. THE ENVIRONMENT MAY Bt MORETYPICLLY ATTAINED IN MISSILE FLIGHT. THERE IS A DES"G PENALTY FOR SEVERE THAN FLIGHT FOR LAUNCHERS ON OR ABOVE MAIN DECK.MISSILE AND LAUNCHER FOR THIS ENVIRONMENT WHEN THE MISSKISARE STOWED BELOW MAIN DECK.

THE ROUNDOSHOULDBIt 001P20TECTIOANISULATEO 1 10 100IN.FROM THIS ENVIRONMENT I,

BY THlE CONTAINERFPllET

IIODOILI I.E ISNINI01011

VIIIIATIOII 0-1 ,.-~NNEENTI

"MILS0.16611wo'OIN? Io 16 ';0 10000SHUIOARO VIGA~TON 04110 INVINONMENTS COO.1AIRID

FIGURE 12Z. SHIPBOARD VIBRATION. THE MISSILIE 6A LAUNCHER RESPrMOSTOSHI VIBRATIONS, WHICH CAN aE SEVERE 0 CERTAIN MANIEUVERS. SEACONDITIONS. OR ENGINE SPEEDS. THE VIBRATION LEVELS AREQUALITATIVELY Of THE SAME LEVEL AS FOR FLIGHT AND PHISIT. EXCEPT INA RESONANT CONDITION, IN WHICH THE SHIPIORAD VISRATIONMAYBEfMUCH MORE SEVERE1.

FIURE 11. PACKAGING. HANDLING, STOWAGE, AND TRANSPORTATION.THME MISSILES ARE PROTECTED DURING STOWAGE AND TRANSPORTATIONBY SPECIALLY DESIGNED SHIPPING CONTAINERS. IN SOME CASES THESHIPPIJNGCONTAINER IS ALSO THE LAUNCHER.

0.10

2 ~-0-as*MAGNIFICATION .

FACTOR 0.371 0 5 0C / C

a0a0 1 2 1

FREQUENCY RATIO ("-c.j

FIGURE 13. HANDBOOK CURVES OF THE VIBRIATION RESPONSE OF SAIPLEMECHANICAL SYSTEMS NEAR RESONANCE. THE RESPONSE INCRIASISDRAMATICALLY AS THE RESOHANT FREQUENCY. iv., IS APPROACHED.PARAMETRIC CURVES ARE SHOWN FOR VARIOUS VALUES OF DAMPING INTHEI SYSTEM.

StE CANISTER IX21 SM-2 KOCK WCANISTER NISSU

ACEIIP0-Zwol wools

NIS 7,07 - HIMOIuoACGBWUT MS 70? stow,100

Flu REPRESENTATIONS

RADOM

SH

us 22.07 N142 RADIAL AND TRANSEVEIRSE NMIGUJIDE CHANNEL K

TRANSVERSE K a SO LBDI&INEAR aCANDANCIRADIAL K a I,0AM LBAM&IAEAAN

MIS 37.S4 - N13N300tDORSAL FIN

LONGITUDINAL-RADIAL

ORNDANCE- RIGIDBAR

Mus 60.71 Ni"W04U40BAR

APSU -S3A

VERTICALZ---Y ATHWARTSHW --

XFOREAFT - --MK 104 --

SHIP-

N17S --

N100 ~FINflRimj DORSAL

-ms I".00 Milo *A CHANNELFICP

- MS170.6s "II FlINSHEARCENTERNM

STEE RING CONTROL

/ UNMI

POTEIOAL RIS RINT -MS 164.00 N7001

SADDLE INTERFACE0.072 ON GAP

BOOSTER-

Nils RIG. NIAR

ALA,.TERAL PIMG 111700

LOWER K.ZI7LMN N1 7009

-MIS 2".S - UTOISo

MS2So42J STAK PA"NEFC 7011t ISHIP 1 NtiS K *4891,000 LIME

ACCE INPUT- .

SHP SHIP

fI*URE114. FINfILI ELMINT MODELING APPROACH TO DE SIGN OF SM-2 BlOCK IV MISSILI AND CANISTE n THE MODEL CONSISTS OF A MISSILE MODEL, CANISTERMODEL, AND CONNECTING ELEMENTS. THE $ASIC MODEL WAS USED FOR ANALYSIS OP UNDERWATER SHOCK AND SHIPBOARD VIBRATIONE.

190

COOLING SYSTEM AND INSULATION CONCEPTFOR A MACH 5 TURBO-RAMJET AIRCRAFT

S. C. JonesLockheed Engineering & Sciences Company

Hampton, Virginia

D. H. PetleyNASA Langley Research Center

Hampton, Virginia

Abstract

A cooling system and insulation concept for a Mach 5 cruise aircraft, using non-cryogenicfuel is presented. Catalytic endothermic reaction of petroleum fuel is used as the heat sink forengine cooling. A secondary closed-loop coolant circuit removes heat from the engine andtransfers this heat to the catalytic reactor. Insulation on the engine flow path surfaces reducesthe cooling requirements. A high temperature insulation system, which is capable of a surfacetemperature of 4,0000 F, is used for the combustor and nozzle.

A complete closed-loop cooling system design is shown in detail. Main features of thissystem include a fuel preheater. a catalytic fuel reactor and engine wall cooling panels. Asilicone-based liquid polymer, designed for extended use at 7500 F. is used as the coolant. Thepreheater and reactor design are based on the results of recent experimental work sponsored bythe USAF. The cooling panels are designed using a thermal fluid analysis computer programwhich was originally developed for the Nation-il Aero-Space Plane (NASP). Major componentsare analyzed structurally as well as thermally and weights are presented.

Introduction

In a cooperative effort by the NASA and the US Navy, a conceptual design study was madeduring 1989 of a Mach 5 cruise aircraft [I]. Design specifications for :his aircraft included arange in excess of 1,200 n.m. and a cruise altitude above 90,000 ft. It was also required thatthe aircraft be capable of carrier operations. These specifications led to some distinct designrestrictions in terms of aircraft size, weight, takeoff and landing characteristics, safety andlogistics. Top and front views of the aircraft resulting from this study are shown in Fig. 1. Therequirement for hypersonic flight, coupled with the safety and logistics restrictions, created aunique challenge in the areas of fuels and thermal management. This paper presents a detailedthermal management concept for such an aircraft.

Propulsion

The combined cycle propulsion concept incorporated in this aircraft was derived from aMach 5 penetrator study done in the mid-1980s as a joint effort between NASA, LockheedCalifornia, and Pratt & Whitney [2,3.41. That system consisted of an over/ under turbo/ramjetdesign which utilized a F-100 class turbo fan, integrated with a two dimensional ramjet. Thatsystem was designed for liquid methane fuel. The syster" designed for the current concept issimilar to the Lockheed/P&W concept in that both use a 2D mixed compression inlet design witha variable geometry tamp system scheduled to position the shocks for efficient operation.Because of the extreme temperature of the exhaust gases, the current concept uses a thermalcnoke in the ramjet nozzle. This airplane uses two turbo/ramjet units.

UNCLASSIFIED : Approved for public release; distribution is unlimited.191

The modes of operation for the current concept are shown in Fig. 2. Subsonic propulsion isprovided by the turbojet with afterburner, while the ramjet is cold flowing. The ramjet isignited transonically and the system operates in dual mode until the turbojet begins to spooldown near Mach 2.5. The ramjet has been shown to be of assistance transonically, primarily fordrag reduction, even though its low speed efficiency is rather poor. As the aircraft acceleratesfrom the speed of sound to Mach 3, the ramjet becomes more efficient and provides increasingthrust. At the same time, aerodynamic heating becomes an increasing problem for the turbojet.The turbojet remains in full afterburner until it is finally shut down at Mach 3. The ramjetthen provides all of the thrust up to Mach 5.

Fuels

In the high Mach number region of the flight envelope (M>4), aerodynamic heating is toogreat for conventional structural materials to survive without active cooling. The fuel must notonly have a good heat-of-combustion, but must also provide the necessary heat sink for thecoolirg system. Three different fuels are compared in Table I. Both liquid methane (CH4) andliquid hydrogen (H2), are cryogenic. The heat sink and heat-of-combustLon figures for the fuelsare based on heating to a given temperature prior to injection, labeled "when heated to" inTable 1. The cryogenic fuels would be heated directly in a cooling system whereas thenoncryogenic fuel would be used with a secondary coolant loop. The given cost figures do notinclude the cost of energy necessary to maintain the fuel at cryogenic temperatures. Because ofthe extreme cold involved, cryogenic fuels pose many design, logistical and safety problems. Inthe cooling system design, extremely low temperatures cause high thermal gradients which leadto thermal stress problems. Materials tend to lose toughness at low temperatures and use of H2can cause problems with hydrogen embrittlement of the coolant passages. The insulationrequired for cryogenic fuel tanks is extra weight for the aircraft. In tnis particular design. USNavy specifications do not allow the use of cryogenic fuels in aircraft carrier operations.

An alternative to cryogenics is provided by the chemical heat sink available in anendothermic petroleum fuel such as methylcyclohexane (MCH) (C 6 H I I CH 3 ).

Methylcyclohexane is a cycloalkane formed by exothermic catalytic hydrogenation of highpurity toluene (C 6 H5 CH3). MCH is a clear liquid with approximately the same density andheat-of-combustion as JP-7. The viscosity is less than that of JP-4. On a volume basis, it hasmore than four times the heat-of-combustion of liquid H2. The flashpoint for MCH is higherthan that for conventional jet fuel (JP.4), making MCH safer to handle (5).

Note rern.te Itermal Choke

Mach 0 to 0 9 cold flo.Mflg

SDual node

Mach I to 2.9

2 1.1 Mach 3 to 5

Figure I. Mach 5, Carrier Based Hypersonic Aircraft Figure 2. Operating Modes for Turbo/Ramjet Engine

192

The heat sink is provided by a two-stage preheater-reactor system. Such a system has beenbuilt and tested in experiments by Garrett Engine Division, sponsored by the USAF (6]. A fuelpump pressurizes the fuel beyond the critical point to avoid boiling. The preheatcr heats thefuel to the proper reaction temperature while removing heat from a secondary coolant. Tomaximize efficiency, the preheater is designed as a counterflow heat exchanger. The flow pathis shown in Fig. 3. Experiments have shown that the preheater provides about one-third of thetotal heat sink of the system. After preheating. the fuel passes through the catalytic heatexchanger/reactor (HE/R). The endothermic catalytic reaction of the MCH is basically thereverse of the process by which it was formed from toluene and hydrogen. At 100% efficiency

the reaction will absorb 1,000 BTU/Ibm of MCH at 925 0 F and 500 PSIA. In this design, theendothermic reaction efficiency is estimated to be 58%, with the reaction absorbing 580BTU/Ibm. Catalytic efficiency decreases with temperature and is limited by the operatingtemperature of the silicone-based coolant. Preheating of the fuel, prior to reaction absorbsanother 313 BTU/Ibm from the coolant.

The air conditioning system for the pilot and avionics uses the fuel directly as a heat sink.

Prior to takeoff, the fuel is chilled to 20°F. By the time the aircraft has accelerated to Mach 5,

the fuel will have been heated to 701F, due to acrothcrmal heating. Thc fuel is chilled toachieve maximum range at Mach 5 speeds. If unchilled fuel is used, the cruise airspeed orrange would be reduced. Launch delays for an aircraft with chilled fuel arc not a problem.Because the thermal mass of the fuel is so great and the tank insulation is very effective, the

fuel bulk temperature in the tank only increases at a rate of 1.60 F/hr. under the most stringentambient weather conditions. This means fully fueled aircraft could be left on deck for longperiods before having the fuel rcchilled. Only the surfaces in and near the ramjet and nozzleare actively cooled. A basic coolant routing diagram is provided in Fig. 4. Coming out of thecoolant pump near the rear of the airplane. the coolant is used to cool the airframe nozzle areaand is then piped forward to the turbojet inlet closeoff doors, where it is used to cool the seals.The coolant is then pumped downr the ramjet inlet sidewalls and then back down through coolingpanels which surround the ramjet, forming a cooling jacket. The coolant is then returned to theheat sink.

Liquid tue;Table I. Fuels for Hypersonic Aircraft (MCH) in

(Indit (Direct (Directc•olng) cooling) cooling) PHEATER

Cryogeni no yes ye[--

Heat of Cml_ _Cln anCoolant(TUlb) 191184 21,46S 52,170 in- Cooant

(BUCa•12"t,09 $3,475 30A614 out

When Heated to (F) I 0 so 1340 1340 _______

Heat Sink 18'1lb) RE-ACTORChemilcal (S111%) Sa 0 0Physical .I212 M1 MZTotal 12 0

Boiling Point (F) 213 -259 -423

Cost ($/Gal) 0.90 21tod4 1.30 Gaseous fuel(H 2 +Tol) out

Figure 3. Fuel Heat Exchanger Reactor System

193

Heat load and heat sink are shown in Fig. 5. The fuel tankage is allowed to absorb heat bothfrom the airframe hot structure and from the air conditioning system. The ramjet area heatloads are as follows: engine nozzle (46.0%). ramjet combustor (35%), ramjet inlet (13%),airframe nozzle (2.6%), and turbojet inlet close-off door seals (2.6%). Approximately 1/3 of thebeat sink for active cooling comes from the preheater, the rest from the reactor.

Survey of Coolants

The chosen coolant is a commercial heat transfer liquid "Syltherm 800" a product of DowComing Corporation of Midland, Michigan [ 7). This commercial polydimethylsiloxane fluid hasexcellent heat transfer characteristics and is odorless, non-toxic, non-fouling and non-coking.It has a useful life of at least six months at a temperature of 830°F in a stainless steel systemas used in this design. The fluid was chosen primarily for its low freezing point and highboiling point (-400 and 390°F at I atm, respectively). "Syltherm 800" and two other possiblecoolants are compared in Table II. Many other coolants were considered including both liquidsand gases. First considered was a liquid metal alloy of sodium and potassium (NaK). AlthoughNaK has superior thernial characteristics, it was rejected because of flammability. NaK willburn violently when exposed to water and will also ignite when exposed to air at temperaturesabove 239 0 F [8]. A leak of thi.; coolant in flight could be catastrophic. Water was consideredbecause it is an excellent coolant but, to avoid boiling, liquid water would requirepressurization to 3000 PSIA at 6950 F. A decision not to allow boiling in the cooling panels wasmade early in the design process. Gaseous water (i.e., steam) was considered but would requirea separate boiler in the system and special start-up procedures. Gaseous hydrogen hasexcellent heat transfer properties but was rejected because of low density and extremeflammability. Gaseous helium has good heat transfer properties yet is inert. A completesystem using helium was designed with a maximum temperature of 851°F and a maximumpressure of 1050 PSIA at 4 Ibm/sec. flow. This system required a 628 HP, 800 CFM (inlet)compressor for each engine. The weight and size of these compressors and the associated APUwas judged prohibitive. Characteristics of cooling systems for this aircraft using differentcoolants are shown in Table IlW.

Molten salts were considered because they are excellent heat transfer liquids and arc amongthe safest high temperature fluids available [91. A eutectic mixture of potassium nitrate,sodium nitrate, and sodium nitrite (known as "heat transfer salt" or HTS) is useable up to8400 F. Molten salts were not chosen because they have the disadvantage of a high melting point;HTS melts at 288°F. , TO FUEL

lo 1TO COOLANT

COO6MM COAW10

HIM 120

co cool coo 004 €OMMUSTOR NOZL 30

NOZZLE HEAT LOAD HEAT IBNK

Figure 4. Coolant Routing Diagram Figure 5. Heat Load and Available Ileat Sink at Mach 6,for On. Engine

194

Cooling System Design

The cooling system design was driven by Mach 5 cruise conditions, where the heating loads

are greatest. Operating conditions for Mach 5 flight are shown in Fig. 6. The maximum

pressure is 424 PSIA at the cooling panels' inlet; the maximum temperature is 838°F at thecooling panels' outlet. The cooling panels are arranged in series with the coolant flowingforward through the airframe nozzle, then downward to cool the turbofan inlet close-off doorseals and back through the ramjet. The 77 Ibm/sec flow splits after the last cooling panel outletwith 40 Ibm/sec flowing through the heat exchanger/reactor and the rest flowing to thepreheater. The flow is remixed at the inlet to the expansion tank. The pump requires 143 HP at242 PSIA. and 80% efficiency. A separate power unit using LOX and MCH drives these pumps.

Syltherm 800 has a high coefficient of thermal expansion, making an expansion tank

necessary. Between room temperature and 838 0 F. Slytherm 800 expands 91% in volume.According to the manufacturer, this tank should be designed to be 1/4 full of coolant when coldand 3/4 full when hot [10). The remaining volume is filled with a nitrogen gas blanket. Thetank is located at the lowest pressure point in the system, and the constant flow of coolantliquid through the tank ensure that vapors will be isolated in the tank. The physicalarrangement of the components is shown in the scale inboard profile of Fig. 10.

TO OTHER ENGINE COO.AN it

FUEL SYLTHERM 900MC" 812F WL01 F

PRHETE N2 14H

6W PSIAAIRFRAMES' EXP. TN]NOZZL.E

"AP TURBOJET

I COM13USTOR

BYPASS E N NOZZLE

803 F tNG 6

NOTE: ALL PUMPS 586 __& COMBUSTION0.8 EFFICIENT L0 HEIR

A/C USES FUEL " 40 LOBM/STANKAGE AS r 791 FHEAT SINK 242 PSIA

H2 AND TOLUENE MIX OUT762 F, 564 PSIA

Figure 6. Cooling System Schematic, for One Engine (Conditions at Mach 5

Flight Shown)

195

A unique feature of this cooling system is that it is self-compensating, i.e., an increase inheat load will lead to an increase in the temperature of the coolant entering the heat rejectiondevices. The increase in temperature will lead to an increase in the efficiency of the heatexchanger/reactor, increasing the temperature drop in the coolant. Similarly the preheater willalso transfer more heat to the fuel. The effect of the increase in heat load on the temperature ofthe coolant will have been diminished and a new equilibrium will be established. In this off-design case, the coolant and catalyst may be degraded but the system will operate until the fueltemperature reaches 9250F; this is the limiting condition for both the catalyst and coolant.

Syltherm can operate for five days at a temperature of 9320F, but the pressure would have to beincreased to avoid boiling. Note thut the heat capacity of the coolant is so great that a 15%

increase in heat load will only cause an increase of 60 F in the coolant.

Heat Exchanger/Reactor

The heat exchanger/reactor (HE/R) (Fig. 7) is designed to remove heat from the coolant bycatalytic endothermic reaction of the fuel. It consists of a five pass shell-and-tube heatexchanger and is made of 316 stainless steel. It is designed to operate at a coolant pressure of350 PSIA at 1000°F and a fuel pressure of 600 PSIA at 8500F. The 2% platinum-on-aluminacatalyst is packed to 70 mesh size in 3709 1/4 in. ID tubes. The fuel, in the form of asupercritical fluid, flows through the tubes at a design liquid hourly space velocity (LHSV) of100. The shell pressure vessel is designed to a factor of safety of 3 on yield strength; the vesselwalls are 1/10 in. thick. The coolant inlet pipe has an ID of 2 in. for an inlet velocity of 60ft/sec. The HE/R is rectangular in cross-section and approximately 100 in. long. It isdesigned to handle a maximum fuel flow of 1.8 Ibm/sec. The device weighs approximately 964lbs, empty of coolant and fuel.

Insulation

Properties of insulation considered for use in this aircraft are shown in Table IV. Theinsulation with the highest temperature capability consists of 1/8 inch O.D. zirconium oxidetubes filled with a foam of the same material. The zirconium oxide has 8% yttria stabilizeradded to it, increasing the amount of cubic crystaline structure for increased strength. Thereis no known substitute for this insulation which is designed to withstand exposure to gases attemperatures above 4000OF fill. Except for its temperature capability, the zirconia is not verygood insulation. A lower temperature system, such as carbon-carbon sheets backed by saffil-aluminia fiber, might be used if the cooling capacity of the fuel were greater.

Each of the two metallic TPS systems, mulitwall and honeycomb, is stiff enough to supportits own weight. The multiwall is made of cobalt L605 sup•cralloy. Honeycomb is usually used asload bearing structure and is so used in this design. In addition to its structural usefulness,this honeycomb has good resistance to thermal shock. This honeycomb consists of Ti 6242facesheets and a Ti 6-4 core, with a core solidity of 1.65% and a core height of 1. inch. Saffilalumina fibrous insulation is a good candidate as fuel tank insulation. It is not used in theactively cooled areas of this design mainly because it is not self-supporting.

Wall Construction

The wall construction for the ramjet inlet and airframe nozzle consists of multiwall ThermalProtection Systcm (TPS). the cooling panels, and a structure of honeycomb supported byrectangular rings. A section is shown in Fig. 8. The multiwall is designed to withstandexposure to the inlet air or expanded exhaust gas while the aircraft is cruising at Mach 5. TheTPS is 0.68 in. thick in both places and weighs 36 lbs on the nozzle; it weighs 18 lbs on the

196

inlet. The backside of the TPS will be maintained at 818°F by the cooling panel. made of

Titanium 6-4. The land thickness is 0.03125 in. while all other thicknesses are 0.020 in. The

inlet and nozzle panel weigh 148 lbs each. The honeycomb is supported by rectangular rilgs

made of the primary structural material and designed to operate at 7500F. The primary

structural material is assumed to be a NASP derived advanced metal matrix composite.(AMMC 1).

The wall construction for the ramjet combustor and engine nozzle consists of zirconiainsulation, with the backside temperature maintained by the cooling panel. Behind the cooiing

panel is a structure of honeycomb supported by rectangular rings. A section of thisconstruction is shown in Fig. 9.

Table II. Coolants for Closed Loop Table III. Effect of Coolant on System Design,for One Engine

FSptlW' O00h He N&K FSy 1tww11m 800*1 H% NaK (78%K)

IP1O5o psi (78%K) Max. pessure 424 10so 35State liuid OWs liquid (peru)

Highy fammable no no yes Pump power (HP) 143 u28 1.4

Toxic no no yes

Max. ustenmp. (F) 92S none 1400 Massflow(bsec) 77 4 112COOlOa'nIw~ight (b) 1415 2 1700

Oensity (lb/cu•t) 30.5 0.21 53.1 C (

Fxparmion tank yes no YeDensity X Spec. heat 1 7.0 0.26 11.9(STL~icu It F)

Thermal Conduct 0.034 0.16 13.8(8TUtw ft F)

Freezing point (F) -40 -459 14

Syltherm (out) -•I- -" Syltherm (out)

Fuel (in) --a- ul ot

Pre-heater Heat exchanger/reactor

6 in. M 26 in.

12 in. ,81in. - -21 in.

K - - 100 in.3709 1/4 Tubes

,116 Stainless steel

Figure 7. Preheater and Heat Exchanger/Reactor (to scale)

1% )7

Rectangular Rings

818 F(Shee 116242, Core T16-4)

'--Mu~wall TPS

'SYLTHERM 800"

NHOT GAS (IOW0F)

Figure 8. Ramjet and Aircraft Nozzle Wall Construction (not to scale)

Tube mounting shaft,

HOT GAS (4200 E) S, Tube mounting plate

Z= Cofneyo Panel

(Sheet T16242. Qor TI -49

Figure 9. Ramnjet Combustor and Engine Nozzle Wall Construction (not to scale)

198

The insulation in the combustor weighs 416 lbs and is 0.61 in. thick. The insulation in thenozzle weighs 611 Ibs; and is 0.63 in. thick. The tubes are mounted on a plate which isattached to a strain isolater. A design for the tube mounting and strain isloation is shown infigure 9 in the expanded view. The plate underneath the tubes is made of carbon-carbon. Thebottom plate has mounting shafts welded to it and is attached to the cooling panel. The bottomplate and shafts are made of Haynes 242, chosen for its high temperature strength, and acoefficient of thermal expansion which matcaes that of the cooling panel material. The coolingpanel is made of Ti 6-4. The land thicknes:: is 0.03125 in. in both panels. The combustor wallis 0.020 in. thick, while the nozzle wall is 0.026 in. thick. The combustor cooling panel and thenozzle panel weigh 161 lbs each. The honeycomb acts as insulation with a 88OF temperaturedrop through-the-thickness. The honeycomb is supported by rectangular rings made of theprimary structural material and designed to operate at 750 0 F.

The cooling system was designed using a thermal/fluid analysis computer program calledNASP/SINDA. It was originally developed for the National Aero-Space Plane (NASP) and usedhydrogen as the coolant. Special versions of the program were written, first using helium andthen using "Syltherm 800". Structurally the panels were designed to have a minimum factor ofsafety of 1.5 on yield against internal pressure. The relatively thick lands are designed toresist buckling by through-the-thickness compressive loads. The results of the thermalanalysis of the combustor and engine nozzle panels are shown in Fig. 11. A thermal stressanalysis of the panels was not done but the temperature gradients do not appear to pose aproblem. The airframe nozzle and inlet panels were almost isothermal through-the-thickness at

8030 and 8100 F, respectively.

The available heat sink is insufficient to allow active cooling of those parts of the airframeaway from the engines; fortunately, such cooling is likely unnecessary. Wing leading edge andnose temperatures are expected to be below 2000 OF.

Conclusions

An operational military Mach 5 aircraft is possible without the use of cryogenic fuel. Thisaircraft can have a range in excess of 1200 n.m. and cruise above 90,000 ft.. This can beachieved by taking advantage of the chemical heat sink of the endothermic fuel MCH. Thecommercial heat transfer fluid "Sylthcrm 800" is used as the coolant in a secondary loop. Theweights and volumes of various major cooling system components for one engine are presented inTable V. The heaviest single part of the system is the required 183 gallons of coolant. All ofthe primary insulation combined weighs 1180 lbs. Because they are made of titanium, both thecooling panels and expcrsion tank are relatively lightweight. The HE/R, made of stainless steel,is heavier. All combirf.,, the cooling system and engine insu!ation amounts to about 13% of theaircraft 80,000 lbs take-off-gross-weight.

The weight of these components is very sensitive to the heating load which is a very strongfunction of top speed. The maximum speed attainable with this concept, with the given rangeand weight requirements is most likely not much beyond Mach 5. The zirconia insulation is theonly part of the design which is beyond the current state-of-the-art. Research, sponsored byNASA-Langley, has resulted in the fabrication of several specimens but more development workis necessary. Catalytic reaction of MCH has been accomplished in the laboratory OLt a flightweight system has yet to be built.

Table IV. Insulation for Hypersonic Aircraft

Zlrc. Oxide Cobalt L605 Titanium Safill AluminaTubes Multlwall Honeycomb Fibrous

Self supporting no yes yes" no

4 Max. use temp. (F) 4000 1900 1000 3000

*Thermal cond. 0.71 0.126 0.706 0.0771(BTU/hr ft F)

€ Density (lb/cu it) 185 14.9 4.84 3.5

*Density X Cond. 131 1.87 3.42 0.27

4Therm. diffuslvity 0.032 0.094 0.811 0.073

(sq ft/hr)

Used as load bearing structure

Table V. Cooling System Weights and Volumes, forone Engine

COMPONENT WEIGHT (Ib) VOLUME (0f3 )covdrt 33 0.75Coolalant ..

expn~jr!Preheater 33 2.5.

LOX ank tank Fum hftt HEIR 14 25.

exchangwi/ree':tors Coolant 1415 24.5Cooling Panels 639

Expansion Tank 160 30.4

0INSULATION WEIGHT (Ib) THICKNESS (In)

(I inlet 18 0.68N2 Tank (2 Ramjet Combustor 456 0.67

Figure 10. Carrier Based Hypersonic Aircraft, Inboard (2 Ramlet Nozzle 670 0.69Profile (to scale) (1) Airframe Nozzle 36 0.68

Total for Insulation 1180

(1) Multiwall (2) Zirconium Oxide Tubes

Cokl*JSTOR ENCM NOZMT TA0 =t as T (MAX) - a17 ALL PAIELS MADE OF 116-4

/ , I MAX1USE 1EMP. •. 1001j _T M .. _)= __ ._ -- . .•- ...

G(Q02D86 OL026 900 910 900 C SSLE::M8()R

85 -~ 90~COOLAN ISSYLTHEF04 S

"Ct 821 281 Q01312 8 031

82~J 1 1 -amTEIERA.TURrS iN

DEG:EES F

Figure 11. Cooling Panel Dimensions and Temperatures at Mach 5 CruiseConditions (not to scale)

2WX)

References

I. Pegg, k.J., et.al.: Conceptual Design of a Mach 5-Carrier Based Aircraft. NASA TechnicalMemoranduim 102634. March 1990.

2. Watts. J.D.. et.al.: Mach 5 Cruif. Aircraft Research. NASA Conference Publicatio' 2398.April 1985.

3. Perkins. E.W.,Rosc, W.C. and Horie. G.: Design of a Mach 5 Inlet System Model. NASAContractor Report 3830. August 1984.

4. Cassidy, M. D.: Performance Sensitivities of a High Altitude Mach 5 Penetrator AircraftConcept (U). NASA Contractr Report 3932, September 1985.

5. Anon: Handbook of Aviation Fuel Properties. CRC Re.rt 530 Coordinating ResearchCouncil. 1983.

o. Lipiihski. J. ct. al: Design. Fabrication and Testing of a Boi!erplats. Endothermic Methyl-cyclohexane Fue, Heat/Reactor System. WRDC-TR-89-2045, Jun,. 1989

7. Anon.: informat~on about Syltherm 800 Heat Transfer Liquid. "arm No. 22-761G-86, DowCwrning Corporation, 1985.

8. Jackson, C. B.: Liquid Metals Handbook, Sodium (NaK) Supplement, Atomic EnergyCommission and tht; Bureau of Ships. Department of the Navy., 1955.

9. Singh, Jasbir: Heat Iranmler Fluids and Systems for Process and Energy Applications.Marcel Dekker, Inc., New York and Bascl, 1985.

10. Anon: Syliherm 800 Heat Transfcr Liquid Design Gu~dc Form No. 22-964C-88, DowCorning Corporation, June 1988.

I I. Deane, Charles W. and tlaught. Alan F.: Thermal Shock Testing of Zirconia-TubeInsulation UTRC Letter Rcport R88-957798-1, 1986.

.(1oI[

A SUMMARY OF EUROPEAN AND JAPANESE"HYPERSONIC FACILITY ACTIVITIES*

0. G. DeCoursinFluiDyne Engineering Corporation

Minneapolis, MN

ABSTRACT

The interest in hypersnnic vehicles In Europe and Japan has created a need for test facili-ties. This need is being filled by upgrading old facilities and building new facilities. Thefacilities are designed for research in hypersonics and for the development of vehicles rangingfrom commercial transports to aerospace planes. Vehicle development programs include the FrenchHermes, German Sanger and Japanese HOPE.

It appears that both Europe and Japan have well defined programs for the acquisition of com-prehensive hypersonic test capabilities. The status of this activity is presented through a briefdescription of modifications to existing facilites and of new facilities being planned or designedand constructed. A comparison is made with the major U.S. facilities.

INTRODUCTION

Major programs are underrway in Europe and Japan to develop commercial high spetd transportsand aerospace planes for traveling to and fror low earth orbit. The aerospace planes requireflight at hypersonic conditions ind create the need for suitable wind tunnel facilities for aero-dynamics, propulsion and thermal protection system (IPS) development.

In Eurcpe. the European Space Agency (ESA) provides an umbrella organization through which thevarious countries participate in the cooperative programs. A vehicle to be launched from theAriane V. Hermes, originated by France. is now in the design stage. The initial flight isscheduled for 1998. Facilities are being designed and built to support the design and developmentof Hermes. In West Germany. a two-s'age-to-orbit vehicle. Sarojer, is being studied.

In Japan. the National Aerospace Development Agency (%IASDA) leads a program designed to giveJapan the capability to travel to and from low earth orbit with the shuttle-type vehicle HOPE(kefs. 6 - 8). The initial flight of HOPt is scheduled for about 1997. This will be an unmannedvehicle of 10 metric tons gross weight launched by the H-2 bioster. A manned version of 20 tonsweight is being studied for launch early In the 21st century. Development of HOPE will be sup-parted by a combination of ground and flight tests. Ground testing will be carried out inJapanese, U.S. and French facilities (Per. 7). The Thermal Protection System (TPS) will be testedin the Orbiting Reentry Experiment (OPtX), to be launched on the first flight of the H-2 booiter in1993. Aerodynamics and guidance witn be studied with a quartercale mo'el of HOPE to bp launched in1994.

It is reasonable to expect Is that over tt~e next ten years both Eu-ope and Japan will place inservice hypersonic ground test capabilities that are each sufficient for the development of vehi-cles for traveling to and from ow earth orbit. A brief look at the specific facilities that arenow being planned, designed, or :onstructed, is the subject of this paper. All of the Informationis taken from the open literattre and most of the figures are from published papers and bro-chures. Surveys of facilities arv given in Rofs. I - 5.

For purposes of discussion and to simpligv the comparison with U.S.. facilities are dividedinto the following categories:

1. Hypersonic Wind Tunnels: Blowdown tunnels of nominally one minute flow duration with airheated sufficiently to eliminate ati condensation. Either air from a storage heater or viti-ated air.

2. TPS Turinels: Arc tunnels used primarily for TPS development. Produce enthalpy level and heatflux equal to flight conditions.

3. Ft•lse Tunnels: Various kinds of shock tunnels with run times up to a few milliseconds. Usedfar aerodynamic measurements and supersonic combustion research.

This work wds performed by FluiDyne as an internally funded project. Approved for public release;distribi-tion is unlimited

2(1

4. Scramjet Tesi Stands: Used for the development testinM o•• ,cr•ip.Jet engine modules and

eng i nes.

EUROPE

FRANCE - UNEVA Si

This is a bIoidown wind tunnel, very similar to the 3.5 foot Viy•ersonic gunnel ait NASA AmesIt is located at #1odne and is the oremier hypersonic tunnel In Europe. Figure 1 is a sketch ofthe layout and a phOtit cp.-.king upstream from the end of the tesL. chamber is Shown in fig. 2 (Ref.9). The tunnel was piaced into service in 1970 with a Mach 6 nozzle, just as h)personic workdeclined. It was subsequently used for various purposes, but especially for ramjet testing (Fig.6). It is now used extensively on the Hermes program ani several iai~rovements have been made. Thepebble heater was renovateOj to reduce dust level in ths air, ii dust lilter (F.g. 5) was developedand successully operated, at-i r Mach 10-12 no7ýIe (figs. 3 and 4) with one meter exit diiineter wasadded.

Some characteristics of the present configura'.ion are:

Alue.'ra pebble heater with potential for a'r tempeature to 1850 K (333t, R). o4ximuvu stagqa-tion pressure 15 MPa (2175 psia). Mach 6 'izzle with diam&eter C,6W m (2.2 ft). Mec;, !0 - '2nozzle (interchaiqeable throats) with diaweter I m. Curre.rtly r.iining witl- full sch(-cule atone shift per day. Current dust filttr operated at 4.5 NWa (650psia) and 110C X (1980 F).Filter being developed for higher tevperature.

FRANCE - ONERA 14 HOlSiIOT

This facility (rig. 7) was begun in 14d8 to meet the needs oý the Hermes pro~raw for aero-dynamic and heat transfer mieasurements for design purposes and CFD vllidati'.n. Energy stcred ir. alarge flywheei is used to generate a pulse of electricity that is dis':harg•e in the arc ceahiber,producing high pressure, high erthalpy flow that expands through the mzzle. Various "zzle. areolanned with diameterc. from 0.4 to I m, and Mach rwraber from 7 to 113. The ma•xlm~ -tagnat.on con-ditions are estimated Lt 200 MPa (29000 psia) and 16 MJ/kW (6900 B/ib).

Hotshot tunnels are no longer used in the U.S. becanse the test gas was significantlycontaminated by material lost from the electrodes. The F4 results will be especially interestingin this regard. Intlil operation and calibration Is planned for the latter part of 1990.

FRANCE - AEROSPATIAtE SIMOUN

Simoun is an arc tunrel located at Aerospatia)e near Bordeaux (Figs. 8 and 9). The arc poweris 5 MW and the tunnel is used for IPS development, currently for the Hermes. It is the same siueas the NASA Johnson arc turnels but seualler !han the 20 MW tr'1 60 MK tunnels at NASA Ames, all ofwhi•h have been used ifor TPS testing. It has semi-circular nozzle which is similar to the semi-elliptic no'zles used at Ames.

A.erospatiale is currently developing a lerg.ir arc teal;!-, of about 20 .W inlet power.

WEST GERMANY - DLR HYPERSONIC FUNNEL H2K

H2K (Fig. 10) :s a blowdown wind tunnel located at Porz, near Cologne. It was cropletei in1958 and operated for' only a short time. It was reactivated 4bout 1987 and is now used for theHer.mes and Sanger progr.,ms. Hot air is supplied by an elettric heater at up to 5 MPa (?25 p.sia'dnd 1300 K (2340 R). Tr.ere are five nozzles, from Mach 4.8 to 11.2, with diameter 0.6 m (? ft).Wuns of about it sec d-iration can be made -very ?0 min. The electric heater art' hot velve (betweenthe heater and the nozwle) are being upgraded.

WEST GERMANY - DLR FREE PISTON SHOCK TUNNFL

This is a ,iei, pulse-type facility (fig. 11) being bwilt ait CoLttingen. Initial operatd'on isscheduled for mid 1991. The oyerall design is similar to that of other free piston sho.•k tun-rfels. In physical size, it is iarger thu.n T4 at the University of Queensland, iiriseanp, Australijand 15 now unler construction at GAICII, Caltect and smaller than RYFLF, beinrg constructed byRocketdyne. The piston diameter is 0.56 m (I.q ft) and the overall facility lerngth is 61 m (200ft). Stagnation conditions in exce~s of 200 MPa (30,000 pst) and 11000 K (20,000 H) a-eexpected.

Aerodynamic measuremerts cf the HiIPMES conf 4guration wilP be inade in the first test program.

2(u.

TALY - CIRA PLASMA WIND TUNNEL

Italy has created a national agency for conducting aerospace research. the Centro ItalianoRicerche Aerospaziali (CIRA). It was started in 1984, current employment is nearly 200. and astaff of 500 is planned by the end of 1993. Offices and facilities are being built at a site (Fig.12) near the town of Capua, about 50 km north of Naples . Three large wind tunnels now beingdesigned, a low speed, a high Reynolds number transonic and a Plasma Wind Tunnel (PWT) for IPSdevelopment.

PWT (Fig- 13) will have a power input of 70 MW making it slighitly larger than the 60 MW tun-nel at Ames. Intial tests of the Hermes TPS are scheduled for early 1993. Models will be testedat the flight conditions of enthalpy. surface pressure and heat transfer corresponding to theHermes trajectory. The design point stagnation conditions are 17 atm and 20 MJ/kg. Two nozzleswill te used: a conical rozzle with exit diameter of 1.5 a (4.9 ft) and a semi-elliptical nozzlewith maxiinum width of 1 m~ f3.28 ft). Large models will be tested: nose tips up to 0.6 m (2 ft)diameter and flat plate% 0.6 x 0.6 m (2 Y~ 2 ft).

This tunnel will be the largest of its type in the world. Although designed for TPS testing,future plans include its use for aerodynamic measurements.

JAPAN

JAIAAN_-_HVPERSLNIC TUIAI4L

As mentioned inthe IntrodUttion, Japan's aerospace pine program is directed toward launching

ars utiannel -iersion of :IO0OE (Fiý. 14) In 1997. Because o' the limited wind tunnel facilities inJapan. tests are planned in Frinc%' Lod the United States. Fhe only hypersonic tunnel in Japan of aMgnificant size. Is locarqd at the~ National Aerospace La)oratory (NAL) test site at Chofu (Fig.15). which is n'ear Tokyo. This tunniel (Fig. 16) is similar tc NASA's 3.5 ft tunnel at Ames andOKLRA'i. tunnel ý4 at V~xiane. It has an altuuPm pebble b~ed hea~er which operates at temperatures upto 130C K. bu.. has the potev..ilc) for rl~her temperatures. The maximum stagnation pressure is 9.8MP. (1400 psi). Existing nozz~es a-e Mach 5, 7,. 9 and il with a diameter of 0.5 na (1.6 ft).

A major additior. to the tcrnne, Is now being planned. A .,.*coný., parallel. Itg will be addedthat %ill tian'.*Ie nozzles of 1.0 m (3.28 ft) diameter. The fiirst nozzle willI he a tiach 10 to oper-ate at 9.8 MPa (1400 psia) and 1300 K on a schedule of five ttows per single shift day

JAPAN - PULSE TUNNELS

Japan his completed studies of two pulse tunnels: a conventscial shock tunnel for aerodynamictesting and a free piston shock tunnel for supersonic combustion testing. These facilities wouldprosbably be built at Kakuda. funding for design and construction hiss not yet been aspproved by theyovernment,

.JAPAV - R\JV4JTISCRAM ET TES" CELL

1041 l1'is a ma,'or 'test site (Fic. 17) at Vakuda, "but 200 im north cf Tokyo. An existingroket altitude test cell (tig. 18) Is being converted to a ra'PJet/scr40jet ellI. s-quare nozzles(C 51 x 0.51 mi) wicii Aach numbers 3.4, 5.3 anM 6./ will te installed. Th'ý%e w!)! provide the testconditions for ii-mi~indire~t testing at flight Nacii r~ueiers corre~porid~ng to 4, 6 and 8. A ceramicstorage heater ond a vitiation heater wý'll be isea to give clean air cr~nditfuns and to compareclear. air w'th vitiated ~iir. The' modifications are beingj made Ini two phases, with the comp~letefacility t~o Lop ready in 1993.

JAK"AN - lEST SCHEUDif

An ovrervit-m of the! Japanese progyam was given at the H~ypersonic Flight Conference held at theUniversity of North Ddkcta in 1988 (Ref. 6). A schedule of wind tunnel tests was, presented and Ismreproduced iti Fig. 19. It ,hrvs a tota; of !0.000 tests during the period 1988-1996, whichincludes over 3000 hyoersoric tests.

CC'MPA2I1SON WITH UNIIFF STAILS

A comparison of the [uropean arul Japanese facilities with their tmo'.t simlilar counterparts Inthe U.S. Is. giver in tne Table I. In general, the J.S. hA' musltiple. and larger, facilitie,: ofeach type. On the other hand, most of the U.S. facilities 3-e vzry old whereas most of the foreignfacilities di-e. or v4lij be, new and therefore will have modern con~trol systems and, one wouldexiiect. higher productivity.

2015

TABLE I

COM4PARISON WITH U.S. FACILITIES

United States Europe Japan

Hypersonic Wind TunnelsAEDC Tunnel C ONERA S4 Upgrade NAL Hypersonic UpgradeAmes 3.5 ft OLR H2K UpgradeNSWC Channel 9

TPS TunnelsAmes 20 1MWAmes 60 MW Aerospatiale S MSW NAL 30 MWJohnson 5 11W CIRA 70 MW (1993) being studiedJohnson 10 MW

Pulse lunnelsCalspan Shock HAL Shock Tunnels andGALCIT Free Piston DLR Free Piston (1991) Free PistonRocketdyne Free Piston being studied

ScramJet lest StandsGASL Direct Connect Unknown HAL Direct ConnectMarquardt Cell 2 HAL Kakuda (1993)Aerojet E[f

CONCLUSIONS

The scope of hypersonic facility activity in Japan and Europe indicates that each will, earlyin this decade, put into place the critical facilities needed for aerospace plane development.This capability, combined with selective use of out-of-country facilities, will enable them toindependently develop aerospace planes and other hypersonic vehicles.

REFERENCES

1. F. E. Penaranda and M. S. Freda, Aeronautical Facilities Catalogue. Vol. 1, Wind Tunnels, NASARF I!3, 1985.

2. F. E. Penaranda and M. S. Freda, Aeronautical FacilitiesCatalogue, Vol. 2, Airbreathing Pro-pulsion and Flight Simulators, NASA 1133. 1965.

3. AeronautliCl Facilities Assessment, NASA RP 1146, Nov. 1985.

4. J. F. Wendt, European Hypersonic Wind Tunnels, in Aerodynamics of Hypersonic Lifting Vehicles,AGARO rP 428, 1937.

S. Aerospace Technoiogy. fechnical Data and Information on Foreign Test Facilities, U. S. GeneralAccounting Office,

6. T. Yamanaka, SAL's Research for Hypers9nic Flight, Proceedings First International Conference

Hypersonic Flight in the 21st Century, University of North Dakota, September 20-23, 1988.

7. Japan Forging Aggressivw Space Development Pace, Aviation Weel, August 13, 1990.

8. Japan's Space Progreon, Part 2, Aviation Week. August 20, 1990.

9. Large lesting Facilities, ONERA. 1909.

206

Alunn pwbbi4 bed t~mr,-

T.1 1h-be 0 JOK

I - -- 29-1 400t..'

Fig 3.ONEA Mch 0-1 NozleFig. 4. ONERA Myprch i Tu1nozzle S4

with Cooling Hater Manifold One Meter Exit Diameter

1t.S. ONEPA Filter Operated 3t 1101) K

207

Me. s• Mcx, 4•nato, Ac Winld vtune

Fig. 7. ONERA HotShot Tunnel F4

Fig. 6. ONERA S4 with Raamjet Engine

Fig. R. Aerospatiale Arc Tunnel Simoun Fig. 9. Aerosvatiale Arc Tunnel Simoun Nozzle

7.0,1l6 *.l'| 0(X am I l]W~nI

aelr hor- leo

/ t I I I l - : l -- -

F in. 10. OILP. lyprsoni. Tunnel IWK

209

. i • I.-

Fig. 11. DLR Free Piston Shock Tunnel HEG fig. 12. (.IRA Site at Capua

7-

A

-"'

- -''-,•€V•I, 'iJ*L .-

Fr tT g. r3. CI1. (rR Site t Cp

.. Ii

!r I:

-I -i -r

' : ff' II " '-

w•,,.,- , :;" - p',;Fig. 13. CIRA Arc Tunnel PAT"

I Iq. 14. .-*plrn sv, Aer'- p.(p P lantn !Iro, ht r

2(f)

Fig. 15. NAL Site at Chofu Fig. 16. NAL Hypersonic Tunnel

Fig. 17. %Mi. -i'.e at Kmkut.a Fig. 18. NAL Rocket Engine Altitude Test Cell

* : . .... _- -- ai 7$ . L _

I&D

I ' I Ti Iu,. t * f I S: I * a ,-I . ; u' IS,9f :•_ .

*t4- n I - "

' i '. iG l..'

Fig. 19. AL Wind Tunnel Test Schedule

210

MIGH PRESSURE MACH 10 TO 20 ELECTROTHERKALHYPERSONIC WIND TUNNEL

R.L. Burton and F.D. WitherspoonGT-Devices, Inc.Alexandria, VA

0. Riakalla and W. Chinitz

General Applied Science LaboratoriesRonkonkoma., NY

ABSTRACT

A new technique in described for generating high enthalpy, long pulse flows for testingscraumjet combustors at equivalent flight Mach numbers of 10-20. The approach derives from highpressure discharge technology developed for electrothermal guns and space thrusters. Cryogenicliquid air is heated with a high power electric discharge (500 MW for a 15 kg/sec test flow race)in a capillary tube, generating a quasi-steady pressure of up to 6800 atm at 20,000 K. The heatedair is mixed with additional cryogenic air and is expanded in a conventional supersonic nozzle to0.5 to I atm at 3000-6000 m/sec. Combustion and nozzle expansion test times of -300 amsecs arecontemplated. Facility nozzle chemical kinetics calculations show that above Mach 16, increasingconcentrations of argon and oxygen replenishment are required to obtain suitable test conditions.Initial experiments on a subscale test rig are described for 0.5 kg/sec and 300 atm pressure.

INTRODUCTION

Scramjet supersonic combustion requires typical test conditions of 2000 K at close to 1 atm.In order to simulate Mach 10-20 flight, stagnation conditions of 10.000 K at several thousandatmospheres are required. This is beyond the capability of existing ground test facilities, whosestagration pressures presently limit steady-state testing to Mach 8 flight conditions. Therequ:si -e conditions for higher Mach numbers can be achieved in shock tunnels for test times ofabout L usec, but many test objectives require test times of 0.1-1.0 sec, only achievable with adifferent approach. Among the test objectives which require test times of hundreds of msecs are:

I. Testing of scramjet engine dynamics, including thermal choking and combustioninstabilities.

2. Investigation of wall temperature effects.3. Use of certain types of non-intrusive diagnostics.4. Calibration of results from CFD codes and impulse facilities.

5. Influence of facility starting transients upon model flow establishment (eg.

establishment of separated flow regions)

The objective of the present research is achievement of Mach 10-20 free-jet and combustor

entrance (direct-connect) conditions in a wind tunnel utilizing a high pressure, electrothermaldischarge. This type of discharge, also called the capillary discharge

1, has been demonstrated

on the electrothermal gun2 and on the pulsed electrothermal (PET) rocket thruster3. The principle

is to confine a 10-1000 KW resistive discharge in an insulating tubular capillary with endelectrodes, and to supply the discharge with sufficient mass to hold the temperature to 15,000-

20,000 K. The discharge radiation is efficiently captured by the injected mass, providing massevaporation rates measured in kg/sec, all of which must exit through the open end of the capillarytube (Fig. 1). By increasing the power level and decreasing the capillary diameter, arbitrarilyhigh pressures can be generated, limited only by the material strength of the chamber walls.

Pressures of 6800 atm in the electrothermal gun, and 100-1000 atm in the PET thruster are routinely

generated by this method.

For screm et combustion tests at Mach 10-20. an air fiow rate of 10-15 kg/sec is desired(based on 2 ft facility nozzle exit area). Based on the enthalpy at Mach 20, the required

electrical power is 500 -W, 50% of which is assumed to be lost to the walls before reaching thenozzle exit plane. For a 300 msec test time the total mass throughput is 3-4.5 kg, and 150 1J ofenergy are required. Because of the small mass throughput, a 10 m diameter vacuum sphere is

sufficient to hold the exhaust back pressure to <0.01 atm during the pulse.

This work was pcrformcd undcr Contract NJAS 1-18450 sponIsoieJ by NASA Lbiigley ResterOh CentLer-. Mr.

Ernest Mackle-; was the Task Technical Manager.

Approved for public release; distrihution Is unlimit' '

211

The basic principle of the wind tunnel is to heat air electriotherinally to high pressure andtemperature, and then to expand it through a aupersonic nozz.e_ to achieve the desired testconditions. At the stagnation conditions required the air is partially ionized and highlydissociated, while at the nozzle exit and combustor entrance the air must contain the properspecies concentration of molecular oxygen, with a minimum of monatomic oxygen and oxides ofnitrogen. It is therefore necessary to perform kinetics calculations on the nozzle flow to predictchemical and gas dynamic performance.

CHEMICAL KINETIC CALCULATIONS

One of the major technical issues which must be addressed in tne design of any hypersonictest facility which adds energy to a stagnant fluid is the performance of the facility nozzle.Expansion of the gas through a nozzle to achieve the desired test conditions requires properconsideration of the chemical and gas dynamic processes occurring in the nozzle, the ultimate goalbeing to assure a test flow which is as consistent as possible with that encountered in flight.As mentioned above, the stagnation conditions required for Mach 16-20 simulation result in a highlydissociated test gas, making nozzle chemical kinetics calculations necessary to predict the stateof the gas at the nozzle exit plane. The stagnation conditions for the free-jet and direct connecttest modes were calculated using an in-house equilibrium chemistry code (EQSTATE) which utilizesthe CREK chemistry package developed by D. Pratt. 4 The desired nozzle exit conditions for a q-1000psf flight path and the required equilibrium stagnation conditions are presented in Tables I andII. The free-jet test conditions are those conditions under a vehicle forebody at eight degreesto the flight direction. Note that the mole fractions of oxygen, nitric oxide, and atomic oxygenthat exist in the various stagnation states are listed only to illustrate the initial degree ofdissociation that exists. The actual gas compositions include other compounds as well as ionizedspecies.

Table I shows that predicted stagnation pressures for the Mach 16 and 20 free-jet conditionsare well above the facility limits of 6000 - 6800 atm. Since the elevated pressures (beyond thosefor perfect gases) result mainly from dissociation and the formation of NO, a mixture whichreplaces some of the nitrogen with a monatomic inert gas would reduce the stagnation pressures totolerable levels. Figures 2 and 3 illustrate the effects of increased concentrations of argon onthe stagnation pressure and temperature required to simulate flight Mach 10-20 free-jet conditions.For the purpose of comparison, both equilibrium chemistry and perfect gas stagnation conditionsare plotted. It is seen that the Mach 16 and 20 free-jet conditions require about 18 and 65percent argon in the test gas, respectively, to reduce the stagnation pressures to 6000 atm.

The chemical kinetic nozzle processes which determine the chemical and thermodynamic stateof the gas at the exit plane can be adequately defined (initially) by one-dimensional conicalnozzle calculations. The calculations were performed using the NASA-Lewis LSENS general chemicalkinetics code, which is a revision of the GCKP84 code written by D. Bittker and N. Scullin. 5 Thecode was modified at CASL to accept high temperature (>5 0 0 0 K) curve fits for the thermodynamicproperties of the constituents of air, including ionized species. The neutral species chemicalkinetic mechanism used in the calculations was compiled by the NASP High Speed PropulsionTechnology Team6 , while the reactions and rate constants involving ionized species were obtainedfrom reference 7. Initial (sonic) conditions were assumed to exist in equilibrium and werecalculated by the EQSTATE code.

FREE-JET CALCULATIONS

A series of initial calculations were made for the free-jet mode of operation. Conicalnozzle expansions (half-angles) of 2.5 to 10 degrees were assumed for flight Mach 10, 12, 16, and20 free-jet conditions. In all of the calculationp, a negligible percentage of the total enthalpywaa consumed by ionized species. In addition, their rates of depletion through the nozzles werenoted to be extremely high such that they seemed to be in equilibrium with the flow conditions.These facts coupled with the large range of uncertainty associated with the reaction rate constantsof the ionization reactions led to their exclusion from subsequent calculations.

The results of' the calculations are shown in Figures 4 to 13. Figure 4 is a plot of the molefraction of 01 fit the facility nozzle exit as a function of the conical nozzle half-angle for thefoir flight coridltIons considered. It is sqen that, firstly, a significant fraction of there.qi ird to ,e frartion of 02 has recombilned nt the nozzle exit for the Mach 10 and 12 conditionseffi rand // '# to gI I , r!;perctively). Socoiidly, the amount of recombf ted oxygen for these twoconrgdFt lI,:; It; fauirly iFndep.rident: of tham cone h- lmigle however, delptlldoilCee 0l the cello hal[-agI1 rf IF,' rn',',.; for thr Mach 16 iind 20 rood ItI:f oil, whi FI , th,' amountt of recollb Itin d oxygen decreases11W, 1r10':' FV IF ,'ri}• FIrrdII.d Ir Vxp;1frrr F 0 thIle 2 tt

2 rorlirr-d r ,xit rena In pleottl d I| Fl guiru 5 versiko ;

tt, 1 '.r ', F r, , 1;lF I - ,t1)oplf., Not v Ihi' - rrm 1g o, (,tii ,irr c , (1t1 rOr'V -hr1 1 11 11) , 1 V 11110 1 n l k' I h l' F )(r'ploden'c 0of 1 hefIi ".I ,I f ,,,'d orl F' 1I r,,' '; 6 t 1 '0 !;how pro I I I -! 1 Il f ac I I I ty tI o , .I o (I tin mII o I Irm I c :1 on of I .f i o ' I, I r w irm ii M;irm ifm ) r2 ( s d IP I F t ir;r . I I I a i I r ', I , !w FIllo!; I Fl I ri, I wI e 'F rr!I I 11m( mr0 oe ' .'m; ov r, l

Best Available Cob.

a relatively short distance. However, oxygen recombination in the Mach 16 and 20 nozzles freezesat much lower levels. This result is expected since the higher total enthalpies tend to dampenthe rate at which reactions occur and mak•. chemical freezing a stronger function of the expansionangle.

The effects of chemical freezing on the nozzle exit flow properties are of equal concern,since they diverge from the desired, or equilibrium properties as the flow chemistry becomesincreasingly frozen. The Mach 10 and 12 nozzle exit conditions are close to the desired conditionsin Table I. The Mach 16 and 20 exit conditions, however, diverge from the equilibrium conditionsin Table I with both increasing flight Mach number and nozzle expansion angle, as shown in Figures10 and 11. Naturally, this trend is an expected consequence of the chemical effects discussedpreviously. Thus, additional effort was focused on the flight Mach 16 and 20 free-jet conditionsbecause of the significant effects of chemistry on the nozzle flow.

Initially, it wau thought that increasing the amount of argon in the mixture would resultin additional formation of oxygen. That is, increasing the amount of argon in the test gas fora & "v total pressure and temperature would result in increased 02 recombination over pure air.This occurs as a result of the reduction in the total amount of nitrogen present in the mixture(and therefore the amount of NO and extent of dissociation present), which in turn liberates moreoxygen and increases the net rate of formation of 02. However, the total pressure required tosimulate a test condition decreases rapidly with increased amounts of argon in the test gas, whilethe total temperature increases (the energy of dissociation is converted to translationaltemperature). The tendency therefcre is to reduce significantly the initial concentration ofmolecular oxygen and the reaction rate constant which governs the rate of formation of 02 (sincethe rate of formation of Oz from atomic oxygen is a function of a negative temperature exponent,k-Al- 1 ). The net result is a decrease in the concentration of oxygen at a given area ratio throughthe nozzle for increased argon in the test gas. To illustrate this point, two gases composed ofdifferent concentrations of argon (64% and 83% by mass), each containing 21% mole fraction of 02.

were equilibrated at Mach 20 free-jet sonic conditions and then expanded through identical nozzles.The stagnation conditions for each gas are tabulated below, and the resulting molar oxygen profilesare shown in Figure 12.

GAS (by mass) Pt (atm) Tt (K) (X0 2 )t

a) 64% Ar, 19% 02, 17% N2 5720 13,300 0.4326-2

b) 82% Ar, 18% 02 1580 19,500 0.4842-3

Note the large difference in both the stagnation conditions and the initial mole fractionof oxygen which the presence of nitrogen creates. Also note in Figure 12 the difference in slopes,or the rate of production of 02. and the final mole fractions. For the purpose of comparison, gas"a' was replaced with the equilibrium composition of pure air at iden-ical stagnation conditionsand expanded through the same nozzle. The 0 profile for this case is labeled "air (@ 641 Arconditions" in Figure 12. Note the substantial suppressing effect of iitrogen on the productionof 02. Figure 13 shows the rate of depletion of atomic oxygen for th= same cases.

Oxygen Replenishment. A set of calculations were 1a2de to evaliate oxygen replenishment asa method for increasing the concentration of 02 at the facilii noz:.le exit for the Mach 16 and20 conditions. In this regard, the initial undissociated concentratin of 02 was increased until21% mole ftacLion was observed at the nozzle exit.

A summary of the oxygen replenishment calculations is shown in Figures 14 and 15. Figure14 shows the mole fraction profiles of 02 0, and NO for a 2.5 degree flight Mach 16 facilitynozzle. The uxidissociated mass percentages of the constizuents are rhown, but more importantly,the undissociated mole fraction of 02 required to reach 0.21 at the exit is 0.289, which is anincrease of 38% over the nominal mole fraction. Note that at the nozzle exit the mole fractionsof NO and 0 are relatively low, indicating that oxygen replenishment for this condition is areasonable method for obtaining 21 percent mole fraction of 02 in the test gas. The ratios ofactual to desired pressure, temperature, and Mach number are 0.86. 0.82, and 1.0, respectively.

The result of the Mav., 20 oxygen replenishment calculation is shown in Figure 15, andpresents very different mole fraction profiles from the Mach 16 calculation. Here, almost 80% ofthe nominal mole fraction of 02 must be added to obtain 21% 02 by volume at the nozzle exit. Inaddition, the concentration of atomic oxygen has reached a level equal to that of molecular oxygenand almost z.ero N2 is present in the. Zn Conseqtientiy. the exit flow conditions arc far from thedesired test conditions (p./p., . - 0.18, T./T*., - 0.40, M./M. q -1.40). Given the sianificanteffects of atomic oxygen on the scramJet combustion process, it is clear from the results of thisand previous calculations that obtaining proper test conditions for the Mach 20 free-jet conditionwill require a different approach.

213

In the previous section it was determined that increasing Lhe amount of argon in the testgas leads to poorer facility nozzle performance by decreasing the total pressure and increasingthe total temperature requirement for simulating a given flight condition. Moreover, even underconditions of equilibrium chemistry, the test gas mixture properties become increasingly dissimilarto those of air as the concentration of argon is increased. However, if the latter situation canbe tolerated and accounted !or, then a test gas consisting solely of argon and oxygen can be usedat the flight Mach 20 total enthalpy, but at a (much higher) total pressure which would suppresssome of the initial disjociation and promote additional recombination to 02 through the nozzle.Admittedly. the elevated pressuras would lead to higher temperatures in the nozzle which would workin the opi.o.,Lni direction (toivard dissociation and early chemical freezing). This, however, canbe corrrctad for by replenishlug the mixture with enough oxygen to obtain the appropriate molefraction at the nozzle exit. The results of these calculations are shown below for a 2.5 degreeexpansion. (The nozzle exit area was corrected to reflect the maximum facility mass flowcapability, but resulved in only a 3% decrease).

Gas (by mass) Pt(atm) TI(K)

27% 02, 73% Ar 6000 18,000

P./P...q T./Te...q Mo/M,.q (Xo2 ) (Xo)

1.07 0.761 1.11 0.212 0.157

The molecular weight and ratio of specific heats at the exit are 34.5 and 1.55, respectively.Although the atomic oxygen concentration is almost 16% by volume, it represents only 7.3 percentof the total mass of the mixture, making it a reasonably workable mixture.

A final set of calculations were carried out to evaluate the performance of the facilitynozzles for the direct connect mode of operation of the facility, and to suggest possible methodsfor improving nozzle performance.

D.UCT CONNECT CALCULATIONS

The direct connect mode of operation implies that the facility must duplicate post-inlet,or combustor entrance conditions which feed directly into a combustor model. In this connection,inlet area ratios and kinetic energy efficiencies for a low contraction ratio, adiabatic inlet werechosene. The inlet entrance conditions for these preliminary calculations were taken to be equalto free stream conditions. The combustor entrance conditions for the two cases examined wereobtained from reference 8 and are listed in Table II. As seen in Table 1I, the total pressure lossthrough the inlet for the Mach 20 case is substantial, indicating that air can be used as a testgas, but that initial concentrations of 02 will be low.

As in the free-jet calculations, the direct connect nozzle calculations were performed usingthe high temperature version of the LSENS cjde. However, only the flight Mach 10 and 20 conditionswere examined to represent the range of conditions encountered. The results, plotted in Figures16 and 17, show the mole fraction profiles, oE 0, 02, and NO through a 2.5 degree expansion angle

nozzle. The oxygen profiles for both the 1$ch 10 and 20 cases are very similb- to those of thefree-jet nozzle calculations. Recombination to 02 for the Mach 10 nozzle is Tinirtal but resultsIn an exit mole fraction of 02 of 0.18, while the Mach 20 nozzle exit flow I .,mposed of 0.088mole fraction of 02. The normalized exit flow conditions, which indicate the e tent of deviationof the exit flow from equilibrium, are tabulated below.

M. P./Pl T./Tct M./M.,

10 0.98 0.95 1.0

20 0.72 0.65 1.1

The oxygen recombination problem in the direct connect facility nozzle is analogous to thedecrease in formation of 02 as the concentration of argon in the test gas is increased. Thestagnation temperatures remain high since the total enthalpy is constant. However, the totalpressure drops significantly through the inlet compression process, allowing the gas to furtherdissociate in the stagnation state. As in the free-jet mode, this process is similarly exaggeratedwith increasing total enthalpy. However, the desired test flow static conditions are more easilyobtained since the aree ratio for a direct connect facility nozzle is substantially smaller than

214

for a free-jet nozzle. Hence, calculations similar to those made for the free-jet moO'

made to optimize the flow quality at the higher flight conditions.

Finally, it is worth mentioning that the overall lengths of the nozzles may be re

expanding at the lower cone half-angles until most of the oxygen has recombined and then ii

the expansion rate to the desired exit area, i.e., expanding through a horn nozzle. Fora Mach 10 nozzle may be constructed by expanding at a 2.5' half-angle to a length of 0..

(where the mole fraction of 02 freezes at its maximum of 0.18), then expanding at a much 1a•

half-angle to t&e desired exit area. Obviously, this type of expansion would te2ult in a highi

nonuniform exit flow, so that a contoured (2-D) section would have to be added to obtain a uniform,parall,:! flowfield.

To summarize, the conclusions reached in the chemical kinetic calculations are:

1) The flight Mach 10 and 12 free-jet and direct connect conditions can be simulated

using air with small amounts of oxygen replenishment to obtain 21% 02 by volume. Exit

flow conditions are close to the desired test conditions.

2) Flight Mach 16 free-jet simulation requires less than 20% argon by mass in the test

gas to contain the pressure generated in the mixing chamber. A suitable test gas can

be obtained by oxygen replenishment. Direct connect test conditions can be obtained

using pure air since typical combustor total pressures are substantially lower than

those required for free-jet simulation. Oxygen replenishment is needed here as well.

Static nozzle exit conditions for both free jet and direct connect simulation areclose to the desired test conditions.

3) Simulating Mach 20 free-jet total conditions requires about 65% argon by mass in thetest gas. Oxygen recombination is minimal, however, due to the high total enthalpy.of the gas. Hence, the same holds true for simulating direct, connect conditions aswell. Suitable test conditions can be obtained by using a test gas mixture composedof argon and oxygen at elevated stagnation pressures.

4) In all of the above cases, very low nozzle expansion angles (< 3 deg) are assumed tomaximize the formation of 02. This also maximizes the heat flux to the nozzle wallsand hence increases cooling requirements beyond that encountered by similar facilitynozzles.

Recommended further investigation includes:

1) Contoured (2-D) nozzle chemistry2) Use of efficient third bodies, e.g. H.O, to promote additional 02 recombination.3) Investigation of nozzle wall heat transfer effects.

WIND TUNNEL SYSTEM

A 1 / 5 0 tb mass flow-srale version of the wind tunnel is being tested to verify the designapproach. The system (Fig. 18) consists of three subsystems: a discharge cartrid-e with nozzle;a high pressure intensifier pump; and a cryogenic dewar for supplying air test fluid. Exhaustgases are dumped into a 1 m3 vacuum tank.

With a vacuum in the exhaust tank, tunnel operation is initiated by admitting cryogenic fluidfrom the dewar through the high pressure end of the intensifier pump and through the capillary.The cryogenic fluid is allowed to flow until the pump is chilled and vapor-free liquid air isobserved to enter the exhaust tank. The pump piston with area ratio of 10:1 is driven by a highpressure helium bottle (not shown) through a I cm diameter diaphragm to a velocity of a few metersper second, pressurizing the liquid air. A check valve forces the pressurized liquid through athrottling orifice in the ca.hode and into the capillary.

Before pressurizing the liquid, the power source is armed. Several possibilities exist forsupplying power to the tunnel, which requires megawatt levels at an impedance of 0.1- 0.5-f..Typical requirements at 500 MW are 32 kiloamperes at 16 kilovolts. Capacitive sources can be usedfor pulse lengths of tens of milliseconds, but resistive losses prevent their use for hundreds of

milliseconds. Battery sources are possible but 50,000 lead acid batteries would be required at500 MW, leading to high maintenance costs. Flywheel generators are a good option, as units of the

required size have been successfully used to power fusion Tokamak magnet coils. Finally, one cancouple directly to the commercial grid, provided that the power is available and the problems of

;witching transients can be overcome.

"3est Available Copy215

%e present experiments a capacitive pulse forming nervork is used to supply constant2 usees to the capillary. Power during the pulse is 14 KW, roquirin the storage ofof anstgy, 20% of the bank capacity. The capillary resistance is 0.1fl, and the poverA by 12 kiloamps at 1.2 kilovolts. Typical current and voltage traces are shown in Fig.

* the pumw Is initially pressurized a 20 acsc transient occurs before a constant pressurei Id achieved. The triggering of the power source is therefore delayed 30-40 fhecs from

reeking of thlewlL•m diaphragm tn insitre constant flow conditions during the run. A liquid-slurs trace is~showt in Figure 20.

A'LVi, .Lthough provisiop is made in the design for fluid injection through the anode and the nozzle,hv.t for cooling (Fig. I), this was not attempted during initial tests. For longer pulses the-At transfer rate of -300 kW/c&2 will require transpiration cooling to prevent mass loss from the

electrodes aoa throat, with associated contamination of the air streas.

The electrothermal wind tunnal is early In this development phase. Potentially it can playa role in acranjet ground testing, particularly at high Mach numbers for hundreds of milliseconds.Larger size turnnas then the present one will be required in the future to evaluate perforianceand flow quality at the 15 kg/eec, 6800 eta, 300 macc level.

ACKYO'JIZDGEM.D4TS

The authors wish to thank D. vaii Doren, T. GoIcher and R. Lampman for their expert design,fabrication and technical assistance on this project. We also wish to thank E. Mackley and G.Anderson of N".•A lAngley for nu.erous enlightening discussions on the engineering, chemistry andphysics of higb enthalpy. pulsed flows.

R.EFfl.EfCES

1. Ibrahim, E.Z., 'The Ablation Dominated Pol~yathyLsi~thacrylate Arc," J. Phys. 0: AppI. Phys.,13, 1980. pp. 2045-2065.

2. Winsor, !.. ., at A1.. *Facility Descr'•pion: A. ?0 m E.tt rothersal Gun," GT-Devices, Inc.,39th Meoting, Aaroballictic Range Association, Tokyo, Oct. 1987.

3. Burton. R.L. at al. . "xperi-ents on a Repetitively Pulsed Electrothermal Thruster., J. ofrropulsion a•d Power, J:2, March-April 1990. pp. 139-144.

4. Pratt, D.T.., 'Calculatitn of Chemically Reacting Flows with Complex Chea stry*. in Suiesin CanvecLico, Volume II, B.F. Launder, ed., Academie. Press, NJew York, 1976.

5. Bittker, D.A., Scullin, V.J. , *GCKP14 - General Cheicacl Kinetics Code for Gas-Phase Flowand Batch Process.3 Including Heat Transfer Effects', NASA IP 2320, 1984.

6. Oldoiiborg, R., et. al., 'Ste.us Report of the Rate Constant Comittee, NASP High SpeedPropulsion Technology Team', Dpcegaber, 1989.

7. Bortner, N.H., 'Reaction i..te Handbook', D&SA 1949, General Electric Company, Chapter 19,October, 1967.

8. Billig. F.S., at. a), "Pxoposed Supplement to Propulsion System Management Support Plan',JHU/APL-NASP-PS-l. The Johns Hopkins University Applked Physics Laboratory, Laurel, Maryland,Jury, 1986, Unclassified, Controlled Need to Know.

TABLE I. FREE-JET NOZZLE EXIT AND STAGNATION CONDITIONS

(q - 1000 PSi)

in. - f

M. P.(N/mW: T,(K) V (m/s) M Pxatm) T(K) X, X0 Xo

10 1610 420 3000 7 11 441 3710 0.1569 n.01649 008442

12 3200 492 3660 8.26 1340 5010 0.1034 0.06205 0.135•5

16 2760 565 5010 9.80 8630 8240 0.03280 0.1558 0.158020 2560 1872 6380 11.0 38200 1800 0.01776 0.1689 0.1471

- -- "- -' - - -

z l(,

TABLE II. DIRECT COfJNLCT NOZZLE EXIT AND STAGNATION CONDITIONS

me P.(atm) T,(K) V.lm/s) M. PI/Pl P,(atm) T,(K) Xo, Xo Xwo

10 0.701 1300 2660 3.80 0.183 123 3700 0.1528 0.02760 0.0801

12 0.607 1500 3350 4.46 0.150 372 4870 0.09316 0.09273 0.1204

16 0.550 2050 4700 5.38 0.0898 2160 7860 0.02084 0.2067 0.1181

20 0.524 2720 5960 5.90 0.0391 6240 10,700 .,.86192 0.2223 0.09233= =, .... ____

RADIATION FLUX

S PLASMA ~JET

INSULATOR

CONSUMABLE

PLASTIC OR UQUID

FIG. A. HIGH PRESSURE ELECTROTHERMAL CAPILLARY

i00000

M = 20 8 degree forebody21% 02 by volume

I00 p000

I~10

10 10 20-

'U

- -16- 12

Equilibrium Chemistry 10-........ Perfect Gas

) 95 50 75 100

Pt-rccnL Argon in Test Gaun3 (by mass)

FIG. 2. EFFECT OF TEST GAS COMPOSITION ON LOCALSTAGNATION PRESSURE.FREE-JET MODE

Best Available COPY217

30000 8 degree forebody -- '20

21% 02 by voluie - -25000 q 00 -- "- flight Mach nunber

~20000 16- - 20

S1.6_ - .- ...... .. - --- O

- ----------------------- t2

equilibrium chemistry- -- ------- perfect gas

0 25 50 75 100

Percent Argon in Test Gas tby mass)

FIG. 3. EFFECT OF TEST GAS CO4POSITION ON LOCALSTAGNATION TEMPERATURE

F0.tgh Mach NO.

air 10

12

16

02

0.25 •-.•75Y 7rg 5 1.0 L2.5

u.ci

VS. NOZZLE HALF-ANGLE AND FLIGHT MACH NUMBER.

FREE-JET MODE

Best Available Copee

12.5

10

7.5

0 5Z

Z5Flight Mach No. =10 -20

0 2-5 5 7L5 1.0 125

Cone Half-Angle (degrees)

FIG. 5. FACILITY NOZZLE LENGTH VS. EXPANSION RATE.FREE-JET MODE

0.48 . , ,

0.179

0J 0.178

CTheta =10 2.5 degrees

U..

x 0.176

0.175

Theta = nozzle half-angle

0.174LOOOe-05 LOOOe-04 0.001 0.01 0.1 1 10

Distance from Throat (ft)

FIG. 6. FACILITY NOZZLE OXYGEN PROFILES - FLIGHT MACH 10FREE-JET CONDITION

Best Available C( 21

219

0.18

\1 0.16

C04-

0.15 Tht 0/ .degreesLL

S0 .I

0.44

Theta = nozzle half-angle0 .1 2 . . .. ., . . .. .= , . . .. . ' . . . ... ý . . . ..... . . . .....LOOOe-05 LOOOe-04 0.001 0.01 0.1 1 1O



0.15 17.5% argon by mass

0.125

C'j0C0 0.14o~

U_LLTheta = 10 2-5 degrees

0.075

0.05

Theta nozzle half-angle0.025

LOOOe-05 LOOOe-04 0.001 0.01 0.1 1 10




0.1 64% argon by massTheta = nozzle half-angle

N 0.0750C

0

LOOOe-05 1OO0e-04 0.001 O.01 0.1 1 1.0



1.3

Flioht Mach 16 Free-Jet Condition1.2 17.5% argon

:) 1.1 Mach number

<i. 0.07

a)

0.-0

Tht t)0 25 ere

0 2 . 0 0

Dozzstane Hrom Turatge

FIG. 10. FACILITY NOZZLE OXYGE PRFLES CODTIN FLIGHT MACH 16FREE-JET CONDITION

1.3

Fligh1~s Mach 16 Fre-e CCndtPo

i_21

1stMach numberatr0"

FGI0.8 FACILITYNOZZLE EIT FLOW ONDITION_- FLIGH _ACH 1

staticAvtelperature

0.26

[lih$n Micn 1.0 Free-Jet Conaitions640o argn

"•; 3cn nunuer

L

o,

k statC pres__re _

,Latlc temperature

I. .. .. . _ _ _ _ _ . . ._ _ _ _ _

0 3 4 5

No/.le Half-Angle. deg

FIG. 11. FACILITY NOZZLE EXIT FLOW CONDITIONS - FLIGHT MACH 20FREE-JET CONDITION

015

c\)0

X (n)

FIG. 12. EFFECT OF ARGON ON OXYGEN RECOM4BINAIIONFLIGHT MACH 20 FREE-JET C;ONDITION

222

02

0.3. .........

l/L, si AtA30l' 2.84.5 k N2. I

015

0j

"" C , 01

I- .

tf o84Ofl0 ot o 1 1o

X (iM)

FIG. 13. EFFECT OF ARGON ON ATOMIC OXYGEN -ECW1LIMATIONFLIGHlT MACH 20 FREE-JET CONDITION

?.5 d~oC~A1 nl1.A M, 30% 02. !)2.!)% N2 DY MUSs

007

b 02

NO y

ll~o8u0 X 2(A4 0001 001 01 I t0

x It)

FIG, 14. OXYGEN *EPLENISIHENT CALCULATIONS -FL1GHT MACH 16

IREE-JET NOZZLE

221

2.5 degree expansion angle64% Ar, 33% 02. 3% N2 by mass

0 0

4-)

L 0.1.

002

0.01LOOOe-05 L(00e-04 0.001 0.01 0.1 1 10

X (ft)

FIG. 15. OXYGEN REPLENISHMENT CALCULATIONS - FLIGHT MACH 20FREE-JET NOZZLE

02 (0.18)

0.1NO (0.0575)

C0

>-,

w 0LL

2.5 degree exp:an-sionage

0.0011.000e-05 1.000e-04 0.001 0.01 0.1

X (M)

FIG. 16. FACILITY NOZZLE SPECIES PROFILES - FLIGHT MACH 10DIRECT CONNECT CONDITION


o (0.176)

01 NO

o (0.0311)IL

• 0o0 02

0 J(,-"f ee eXpandfOn ang,(eIj./j --.. I •.o ,____________. ______________

I 'If)'*-O, 10C')e C4 0001 001 0.1

X (M)

FIG. 17. FACILITY NOZZLE SPECIES PROFILES - FLIGHT MACH 20DIRECT CONNECT CONDITION

FIG. 18. COMPONENTS OF FLECTROTH[IEAL HIGH PRESSURE WIND IUNNEL

22.5

2L -3.0

2.0

to-4

1.0

1000 2000 3000

TIME (Asec)

FIG. 19. CUVYA-& AND VOLTAGE TRACES FOR 10 MW PULSE

PSI

15000 INTENSIFIER PUMP PRESSURE

10000

5000 111.

-50 0 50 t00

TIME (msec)

FIG. 20. LIQUID PUMP PRESSURE. SHOWING THE INITIALSTARTING TRANSIENT AND THE PRESSURE BEHAVIORDURING THE POWER PULSE.

226

AN 3XJ'RIMZNTAL EVALUATION OF COMBUSTOR LINERMATERIALS FOR SOLID FUEL RAMJET TESTING

J. B. OppeltWright Research and Development Center

Wright-Patterson AFS, OH

ABSTRACT

The investigation determined the survivability and effect on combustionefficiency of several candidate liner materials for ramjet solid fuel screeningtests. A liner is necessary both to protect the combustor walls and to reduceheat loss from the combustion chamber. This heat loss can result in incompletecombustion and poor performance. Both a hydrocarbon and a boron fuel were usedto compare the liners' effect on the combustion of these fuel types. A finiteelement heat conduction model constructed prior to testing indicated that theboron nitride (BN) liner would allow nearly the heat loss of stainless steel, andthus result in poor measured performance compared to a better insulated combustor.The results of the testing show this to be true for the hydrocarbon fuel. Slagdeposits on the liner walls from the boron fuel help insulate the liners andreduced their effect on measured performance. A ceramic liner of a boronnitride/aluminum nitride composition (BN/AlN) withstood the thermal environmentand insulated well. Measured combustion efficiency was higher for the BN/AlNliner than for DC 93-104, the accepted liner of choice for ramjet testing. Thisshowed that the choice of liner material can affect performance results, and inthe case of DC 93-104, indicated care must be taken to use an adequate thicknessof insulator.

INTRODUCTION

In order to retard heat loss and protect the combustor case, the ramjetcommunity has been using an ablative silicone elastomer liner material, (DC 93-104) in most ramjet testing. Several problems have been experienced when usingthis material in a testing program. First, the material is time consuming andsomewhat difficult to cast onto the combustor walls. During a combustion test,outgasing of the liner increases the mass flow through the nozzle and affectsmeasured performance. In small scale testing, this effect is more dominant dueto the large liner surface area to mass flow ratio. Ablation of the liner causeschanges in chamber cross sectional area that affect performance calculations ifnot accounted for properly. This insulator can sometimes be reused, but hasgenerally been sufficient for only a few test runs. A reusable liner materialthat avoids these problems would improve the accuracy and ease of testing.

Several candidate liner materials were evaluated for survivability,reusability and insulating capacity. The materials selected were boron nitride(BN), boron nitride/aluminum nitride (BN/AIN), boron nitride/aluminum oxide(SN/A1 203 ) and a Dow Corning ablative silicone polymer (DC 93-104). A stainlesssteel (347) liner was used for a baseline comparison. Two fuel types were usedin the evaluation to see the effect on the measurement of combustion efficiencyfor both hydrocarbon and boron loaded fuels. Condensation of combustion productsmay affect performance differently depending on how they deposit on the liners.Soot will have different insulating properties than boron slag, and neither willdeposit the same on liners with cold or hot surfaces.

HEAT TRANSFER ANALYSIS

A preliminary look at the effects of different liner materials on fuelcombustion measurements was accomplished by computational analysis of their heatconduction properties. Conditions representative of the test rig environment werechosen: 0.567 kg/s (1.25 lb/s) of 2200 K (3960 R) air in a 60.3 mm (2.375 in)diameter chamber 45.7 cm (18 in) long. To simplify calculations, the flow in thechamber was assumed to be air and the combustion products were neglected in thecalculations with the exception of radiation heat transfer. In this case the flowwao assumed to consist of CO, and air.

Since the test duration for these small solid fuel grains was only 10seconds, which was not enough time to reach steady state, a transient heatconduction model was employed. A finite element computer model was chosen over

Approved for public release; distribution is unlimited

227

an analytical solution in order to account for the variation of thermodynamicproperties with temperature. Within the heating liner, a large temperaturegradient causes corresponding variations in thermal conductivity, specific heatand density which affect the insulating properties of the material.

Both radiative and convective heat transfer to the liner were included.Since the emissivity of a gas is increased dramatically with the addition ofcarbon compounds, the radiative heat transfer coofficient was calculatedaccounting for the luminosity of combustion products (CO.) in the air. Theconvective heat transfer coefficient was calculated based on Seider and Tate'sequation for the Nusselt number (Nu.) for turbulent forced convection heattransfer with large temperature variations'.

NU 0. )0,14

and since,

NuDO= , ReD pkIA

k FY , Ih pVAPr

then,

0 2 I•h 0(2

hg = .027 C4 p£0) (7 ) -Ps*J7 (2)

P.r067

The values of c,, 1, and Pr were computed at the mean temperature of the flow(T.), and ýt. at the temperature of the flow at the surface of the liner (T.).

The total heat transferred to the surface was computed over the inside areaof the liner by numerical integration of the following equation.

dQ = dt(h.+h,) (T 3 -T.)A (3)

Conduction through the liner was computed using the cylindrical coordinate

heat conduction equation below, and the following assumptions.

1 a aT÷ 1 a 3 aT) a (k a4) = aT

-i-T o-- - O V- rý g -M k I -re

(ý- a = a=

Finite-difference approximations were used to numerically solve the problem.The counter, nr, denotes the nth radial finite element section of the liner fromthe inner surface; the next one is represented by nr+l. Time is diseretized bydefining t - pAt where p represents the current time step where all cquantities areknown, and At is the size of the time increment. The next future time step isdenoted by p+l.

31 -nr -7nr &r~T~ T

Att ,A rIrl.r Ar

The approximations (Eqni. 5) were substituted in Eqn. 4 and rearranged. Thus,the temperature of the nth element at the next time (p+l) is given by:

22S

Calculations were performed forconduction through an 8 mm wall

AdiabaticOulerBoundary divided into eleven thin shells.0 wH10K The inner and outer shel).s were

made half the thickness of the11N, others. This arrangement has

I- been shown to yield good results00 0t for similar finite element

I -.-. models.'70D

700frTwo conditions were examinedS------------- for the outer boundary of the

liner: adiabatic and isothermal.I In the adiabatic case, no heat

0 .. .. was conducted from the linerda" "m &W WM yam {mm) during the run. During the

isothermal case, the temperature- STEL2S -.- •"TE•Io of the liner outer surface was

--- 2 --- NI held to its initial temperature.Because the liner was placedinside a steel pipe with a largo

Dig. 1 thermal mama, the actual. boundarycondition in the test combustorwas between these limits.

IsothermalOuter Boundary This analysis will focus onk kweuym300K comparing 347 stainless steel sand

ONi. A sublimation tompecatiirenear 3300 K (5900 R), &no a

700 thermal conductivity that varies700;. -inversely with temperature seem

! -- - to make ON an attractive choice.,0 347 holds up well for short runs

ooo-.'K --. -- where it doesn't get too hot, •the large tempetrature difference

400! between the liner and the flowcauses high heat losses.

02 4 ado&W fom hrW b,. WIP IMM) Based on this, ot\e would

conclude that ON is a much better-- SEELi. - OTFEL2 -- STEELIO0 choice than 347, but the

N2@ * following analysis ind-ic;atedotherwise.

Fig. 2 Figure 1 shows the thermalprofile through the liners at 2an'! 10 seconds into the run for

the adiabatic outer boundary conditicn. Surface temperatures are identical at 2a, and since heat transfer k-at:ei vary with the temperature differerca betwaer, thecomb.stor gases and the linei surface, this trarelatee to identical heat long t.iboth liners at the beginning of the run. By the end of the run, however, the 150is significantly hotter and the 347 heat transfer rate. 3 higher. Integrattun ofthe heat transfer showed the 347 liner lost 6% idore heat over the run than the bW.In the isothermal boundary case (Fig. 2) the effect was rovtirsed, with the ,:ovj4-BN liner conducting 3% more heat by the end of the run.

'hese results ;an '-e attributed to the high mpecific heat of the 347 andc-nductivity of BN. In the adiabatic case, tho steel's higher specific heatcaused its tetrperature to remain lower than ON and resulted in more hest loga.Tt# the iaothezmal boundary caue, the higher conductivity of the BN kept. Jt innerwall cooler by conducting away more hoat; the BN surfaco only reached 700 K (12(0R). Only above 1100 K (1980 P), does the conductivity of BN drop below that of347. Up to that temperature, the 347 is the better i-sulator.

From this analysis, it's clear that there are several factors that sffe':t aliner'e insulating ability. The outer boundary cordition alon& car, rhari'i, th-choice of liner material. For this reaion, tn exp-trimer.tal rvaiusrvur, innecessary to determine which liner is most suited to a given corifiUtiiatioti.

EXPERINENTAL PROCEDURESOLID FUELED RAMJET RARDWARE DIAGRAM

Tests were run in a solidfuel ramjet test rig sized for635 mm (2.5 in) fuel grains.Constructed of stainless steelhardware, the combustor was a Tu T,24OCOPas BYiuSSconventional solid fuel ramjet VALVvcoaxial dump combustor with a LIN=ma VI N. PUP=27.9 mm (1.1 in) inlet portdiameter and 44 mm (1.73 in) IDfuel port. The 45.7 cm (18 in)mixing section downstream of the m o-wfuel grain was insulated with the zKiim vALVZ MOO=J I?

liner materials under test. The

54.6 mm (2.15 in) ID mixerterminated in a .35 area ratio

convergent nozzle. Fig. 3

Figure 3 shows a diagram ofthe flow path. Compressor air was heated via combustion with JP-4 in the vitiatorand diverted through the bypass valve prior to the run. Makeup oxygen Was addedupstream of the vitiator to keep the mole percent of oxygen at 21%. Aprogrammable controller sequenced the test 'in a repeatable manner once the startswitch was thrown. The through valve opened as the bypass closed, and an ethyleneigniter initiated combustion of the grain with the help of an ethylene push. Thefuel burned until the through valve closed and a nitrogen purge quenched thegrain. The hardware was allowed to cool, the liner was removed for inspection andthe fuel grain was weighed. The weight loss of the grain was used to determinethe average fuel flow rate for the run.

DATA ACQUISITION

Data was sampled by a CAMAC A/D data system to be stored and processed by aVAXSTATION 3600. To remove 60 cycle noise, each data channel was averaged over64 samples taken in a 60th of a second. The data included all pressure transducerand thermocouple outputs as well as flowmeter and time data. All data channelswere averaged and scanned approximately once each second. A digital oscilloscopemonitored the output of a pressure transducer on the combustor mixer. The runtime was taken to be the time difference between the 90% points on the rise andfall of the chamber pressure trace.

TEST CONDITIONS

A harsh ramjet environment was created in order to assess the linermaterials' usefulness for ramjet testing. Two different fuels were used: ahydrocarbon (18,818), and a boron carbide loaded (B4C) fuel. The 18,818 grainswere cut to 32.7 cm (12.875 in), and the B4C grains to 43.5 cm (17.125 in). Thisresulted in an equivalence ratio of 0.9 at the air mass flow rate of 0.454 kg/s(1 lb/s). The preheated air temperature at the combustor dump was kept to 800 K(1200 R), and the run time was 10 s.

TEST ARTICLES

The DC 93-104 liners and the fuel grains were produced at the Naval WeaponsCenter, China Lake, CA, and shipped to Wright-Patterson AFB. The ceramic linerswere commercially available pressed powder products manufactured by Union Carbide,Specialty Products Group, Cleveland, OH. The steel liner was made in an in-housemachine shop and has been used many times in the past as a dummy liner forshakedown runs in the solid fuel test rig.

The liners were sized to fit in the mixer section of the ramjet combustor.The steel (347 stainless steel) and ceramic (BN, BN/AlN, and BN/A 20 3) liners weremachined to an 11.4 mm (0.45 in) thicknesn and a 53.5 mm (2.11 in) ID. The steelliner was a single piece, but the ceramic liners were constructed of two equallength sections butted together. The DC 93-104 liner was made of two unequalsections, one 12.7 cm (5 in), and one 33 cm (13 in) long. This liner was muchthinner than the others: 2.8 mm (0.11 in) thick, and wap cast into an aluminumtube that slipped into a sleeve to fit the mixer. Its inside diameter was 54.4mm (2.14 in); very close to the other liners.

230

.ielSft Av ai lable Copy

2WW I ~mb the condition of the Unarm After two rus With the given fuel.In gaeua1, the J1iins ma i t-,= the thewrm1 en!rmetws newt severe at theWwrIM ze mu Of the user. Soct dkosits max Jliuer and the rea1ts of heatin

bmc In ]Ayem xpm frem mbfte muteranla for the 1 ast 5 cmm (2 In) of the"md4 of Ithe e masna toi in 16,1mesresknt odLIAMa h

mmm

RmMM M3-M TMIMw- v& 3M Q:

___ __ __ ___ __ __ ___ _ __ __ ___ __ __ ___ __ __ ___ __ __ a m

~1 Icitmditio of Lamer Miter 70mtim

AtfTh~Mtem 7* ma rm Witmide izer After Tho tm

W19-. 41 Fig-.5

Materialse survived the testts in gund cnmit ion, !LMe inP003c~azitizm~ an ~ i~edto xemzaiM i-tact - MIhe Steel -and PIAIN l-iners

weze expusedto two zmm wfth 16,316 and two =inw with tbe Ax fxmel - mWMrW-U'W aU ft=r =M 3M VeX g=od Conditici_ ¶a Steel X0 -ld wa3~ 1ith

a gin dI.U alaw steel VMCL toe reMove a J'ape Of hR=X a~epsat w as inqVge to he ome- me ]//A= onr etily e a ffimgan laaer 'of sutzri

withb the box~n 1m~,. acd a %blaa d s~oolr;ltio= wiath the 16ff]g ][ t wseslc~m~ byr %myf wit& steel wool- 7be = 3-2-04 11ser al"o =XnvIved in goiod

OMM'L~q- W-Mtn the M tMMaC Of SMM Smwty i%3t-s CM the firSt 7..-of thellim0, t r"=0 em in all tmat xmm 12 Thebom foela Seemei Ito *0it a

ttlbiui q~lLw-7A (= the. ..T~re 7%e MaII lzmier snanrvilra thbe best rmmms, Rhmtt Sahwd dwmage'a" fb( tbkn=RfL Tt~q-&9 'Ifle Pi~eoeS == Withl the 1b[MmOM ft1m!n1 j&Vwejf j~mVD faCtmrez'aiftlt~ tkxe secMMIRf MIM RTUCt Weve urn (god1 oxmiui alhape t'O rr .agaim. 'Thl. Xedt~iims

Best Availa'b10- Q10

used for the 18,818 run were moreseverely fractured and the insidesurface began eroding away in COMBUSTION EFFICIENCIES WITH 18.818

layers during the second run.The BN/A1O, liner fractured badlyon the first run with boron fuel,and was unable to be reused.

AHALYSIS OF RESULTS

Since the goal of thisinvestigation was to find auseable liner material, theprimary emphasis was ondetermining the liners' ability Ito survive the test conditions.However, the effect a liner 0may 2 4 6 S 10

have on measured performance Thm (s)parameters is also of interest.The effect of liners oncombustion efficiency is examinedin the following section. Fig. 6

PERFOPMANCE CALCULATIONS

Performance calculationswere made based on the COMBUSTION EFFICIENCIES WITH B4Crecommendations of the JANNAFCombustion Subcommittee Workshop -

Report: Standardization of -

Experimental & Analytical Methods --

for Determination of Connected-Pipe Pamjet Performance 2 . -

Combustion efficiency (1T,) wasdefined as the temperature-riseefficiency based on equilibriumthermochemistry;

Ti c T ,- 2 (7 ) /

0 2 4 a - 10 12Tm (s)a

T12 is the combustor inlet rig. 7temperature as measured at apoint just upstream of thecombustor inlst dump. The experimental value for the total temperature at thenozzle throat (Tt.) was based on combustor static pressure, fuel and air flowrates, and chamber geometry. The ideal total temperature (Ts,) was based on fueland air flow rates into the combustor and on predictions for thermochemicalequilibrium. The June 1989 version, (and associated data files), of the NavalWeapons Center Propellant Evaluation Program (NEWPEP)3 was incorporated into aramjet performance code to do the equilibrium calculations.

LINER EFFECT ON PERFOPRMANCE

The results of the performance calculations on the test runs are shown inFigs. 6 and 7. The efficiencies were normalized for presentation, and the curveswere aligned so combustor ignition occurs at the origin, and the ignitiontransients were removed for clarity.

Figures 6 and 7 show that combustion efficiencies increased throughout therun. This was due to two main causes. First, the fuel was a progressive grainthat delivered a fuel flow rate that increased with time. Efficiencies werecalculated based on the average fuel flow rate which was higher than the actualrate at the beginning of the burn. The second cause for the shape of the curvesis the cold chamber walls. They removed heat from the combustor as the mixerheated up. This is the process through which the thermal properties of thedifferent liner materials could affect the measured combustion efficiency.

232

18,818 Fuel. As Fig. 6 shows, the choice of liner material did affect performancemeasurement. The differences varied combustion efficiency up to 10%. The BN/AlNliner insulated better than DC 93-104. Apparently, the thickness of DC 93-104used in this test was too thin to take advantage of all of its insulatingproperties. In the case of the BN liner, the data showed that as predicted in theheat transfer section above, BN is not much better than stainless steel forinsulating the mixer.

BC Fuel. Figure 7 shows that with the exception of steel, the choice of linermaterial didn't affect the measured performance of the boron fuel. Boron slagdeposits on the liner walls may help explain this. The insulating effect of thethick layer of deposits may have reduced the heat lost by the BN liner. Thedeposits on the BN/AlN were very thin, so probably made little difference. Thethin layer of deposits on the DC 93-104 may have helped increase its insulatingability by sealing the porous charred surface from penetration by hot gases. Thesteel liner had a thin layer of deposits, but not nearly as much as the BN liner.This layer wasn't enough to decrease heat loss to the level of the other liners.

CONCLUSIONS

This evaluation determined that the BN/AlN, DC 93-104, and steel liners allcould withstand the ramjet combustor environment for short duration fuel screeningtests. It also demonstrated that the liner material can affect the measurementof combustion efficiency up to 10%. While a steel liner may withstand the heatload for a short time, it conducts heat away from the combustion process.Although there are good arguments for the insulating properties and hightemperature capabilities of BN, at the cold initial temperatures and high heatingrates of the combustor, BN will neither insulate well, nor survive intact. TheBN/AlO liner couldn't stand up to the thermal stress either. Of the materialstested, the only two suitable for use as liners for these tests were DC 93-104 andBN/AlN. The performance of the DC 93-104 liner would have been greatly increasedif it had been thicker. 3ince the heating loads applied to the combustor wall aresimilar for these small scale and the full scale combustor, the DC 93-104thickness should not be scaled down. BN/AIN shows promise as a reusable linermaterial for ramjet solid fuel screening tests.

NOMENCLATURE

A areacP specific heat ratio of the gasesD inside diameter of the mixerhq convective heat transfer coefficienth, radiative heat transfer coefficientk thermal conductivityA mass flow rate of gasesNu Nusselt numberPr Prandtl numberQ energy transfer4 rate of heat generation per unit volumeRe Reynolds numberr radial distanceT temperaturet timeV gas velocityz axial distanceAt time intervalAr radial incrementm, combustion efficiency0 angular displacementA gas viscosityp density

subscripts

2 comrrbustor inlet station5 nozzle stationt stagnation (total, flow property

211

fte atbr woid like to reognaize the taicbea1 guidance of oft ParkerSac"" in the pezfozm of this work, and to thank Mrt Kenneth Scbwatztkopf forbis asiwsta Lim providing data acquisition and comuter amport.

NWERMU

1. Imcrapra, I.? *, and De~ltt, D.P.., FMMdMEM sls of Root aW "U 3~aTraf~ed., jobs wiley & so", New Tork, 1965.

2. Saer D .W., -uorksahp Report: 2tb Standardization of Uzparluenta1 and8m17lctiml Ilohods for Determnination of Connected-Pipe R~mjet Performance,

Pro dM_ &Mth 27th f 1Cacmatlge &uboamittee Meetin Ca aeyenne, WY,uab- 1990.

3. cruisem, D.A., ftqq~tical Ctatioau of Daullibrium C~omwitigns.ftevmodnmsic Prmertlee.- Ald PerfOrMa a~racteriutics of Propellant

WCT 6037, WLIl 1979.

2M4

Mock Usyu/3samsry Layer laterectiem Conto in a Gomric Uyprmnu Isl~et

3. a. Vbite An . 3. 1"a___ Jobn 1ophins ftimaraity

Applied Physics laboratoryLaurl. M

3. D. Coac An . Ge3abeGeneral Dynmaics

Fort worth DiwisiouFort Worth. IR

A serises of testa ha,. been parfo.un to investigate Adcharacterize sback __w/bomizylayer interaction typical of hyparsoulc inlets mad to develop =the& of cotrelling isgimnof &hock Indued spepratism by energizing doLsoming bounday layrm dwn h m oftangential nme; addition. A proof -of-concept test program has demntrated shock wov/baoerzylayer interaction control in a acranjet inlet using tangential smss adiioni at simulated MAC~Mach numers f rom 3.5 to 5.0. Discussed herein are results from a generic: hypermsiuc inlet ~program in which the effectiveness of tangential me"s addition was Invest igated At M1acb nubrsof 6. 0. 8. 0 and 10. 0

A generic hypersonic inlet model hoas been tasted at Mach umers of 6. 5 ad10 Utilizingtangential mass addition for control of the I- ,layer interaction resulting fromimpingement of the cowl shock on the forahedy bounadary layer. Tests were conduted In the ArnoldEngineering And Developmen Center (ASKC) Von Karme Facilities (WKV) Tunnls 3 and C. in Iletmodal Incorporated a tangential smes adition slot 0.1* in beight pith & onsi~nel emit Kicb moubrof 3.

General observations fro the teast program are &As follows:

"* At Mach 6. tangential mesa addition was required to control cowl sahck in~med separation

of the forehody bounda~ry layer and allow the inlet to start.

"* at Poch 6.* when the shock was very near the shoulder. once started. the inlet wouU remainstarted if the moo ad4ition was turned off. Hoover. if the abak viqgui atwasignif icantly forward of the shoulder. turning the injector off resultod in an Inlet inatart.

"* Inlet performance. in term. of maxiumm contraction ratio and mimin total pressur recoftzy.vas signif icantly improved for all Mach 6 cases with mass addition shu comared toceswhich remained stable when mass addition was turned off.

"* Mach 8 and Mach 10 teast conditions did not appear to result In cowl abock Indued ferabodyboundary layer separation an; therefore * there was nosignificaut chang io iasltperforueao due to maes adition.

one of the mest dominant sarodysmaic mochadamn controlling the performane adoperabilitycharacteristics of hypersonic air-breathing engines is the intaraction of shok wave withthlow eumrgy boundary layer flow. Efficient operation of a hypersonic vehicle utilizing air-breathing propulsion requires that,* to the extent possible. the air comressed by th Vehiclepasses through the angina and ise utilized to generate thrust. This results in a criticalreliance on propulsion/airfrmn integration and, as a result. the inlet on a typical Single-Stage -to-Orbit (SST0) vehicle utilizes the lon vehicle forehody to provide einternal comression.The inlet cowl subsequently captures the comressed air Ad provides additival cempreseivethrough genisration of Internally reflected shock wames. If tot properly accouted far. thehighly viscous layer developed an the forebody An ingested into e inlet will be proem toseparate when subjected to the shock gowaratod Adverse press=*e gradient&.

This min regions of interest for control of shock wavo/houndary layer interaction In planr.mixed compression. high speed Wneta are shown in Fig. 1. Inlet operability characteristics aeaffected by the Interaction of the strong cowl ahock with the forebody flowfield. Thisinteraction, if strong enough, can cause a large scale separation which results in exrmmavshock waves and reduced Zlelt efficiency. If an inlet is operating near th mulau gomtnicinternal contraction ratio, the cowl shok-k induaced separations can cause an aeredynmmaic over-contraction and reault in inlet unstart- Therefore. cowl- Aback-induced boundry layerinteractions are critical in defining the achievable inet contraction ratio. For maims

*This work was performed under Task APB of "as Contract OD039-89-C-0001 with The JohnHopkins University Applied Physics Laboratory.25Approved for Public release; distribution is unlimited.

Cowl Shock Terminal Shock

System

Figure 1 Schematic of most significant regions of shock wave boundary layer interactions in avertical compression hypersonic inlet.

operating conditions with high relative heat release in the combustor, the shock wave/boundarylayer interactions that occur in the region of the inlet terminal shock system are critical indefining the inlet maximum total pressure recovery, and supercritical stability margin. Theobjective of using of tangential mass addition is to replace the low energy flow near the wallwith a flow of sufficient energy to overcome the impending pressure gradients associated withthe inlet terminal shock system.

A significant amount of work has been done on the investigation and control of shockwave/boundary layer interactions over a wide range of conditions. A comprehensive review of suchmaterial is beyond the scope of this paper but the reader is referred to two excellent reviewarticles; one by Delery and Marvin' discussing work on shock-wave/boundary layer interactionsin general and the other by Delery2 reviewing the state-of-the-art in control of shock-wave/boundary layer interactions. The most common technique for control of shock/boundary layerinteractions has been the use of boundary layer bleed. However, as a consequence of the extremetemperatures typical of hypersonic boundary layers and the large system volume requirements,boundary layer energization with tangential mass addition has been selected as the principlecontrolling mechanism being investigated in the present effort. The source of such an injectantin an actual vehicle might be a hydrogen/air or hydrogen/oxygen gas generator which feeds asupersonic injector nozzle. A brief summary of work performed using of tangential mass additionto control shock wave/boundary layer interactions is provided in Ref. 3.

Discussion

A test program has been conducted in which tangential mass addition was utilized in a generichypersonic inlet model tested in the Arnold Engineering and Development Center (AEDC) Von KarmanFacilities (VKF) Tunnels B & C at Mach numbers of 6, 8, and 10. Tests were conducted for a rangeof Reynolds numbers between 0.5 x 10 and 2.0 x 106. The inlet model was equipped with avariable position throttle providing inlet performance information for a range of back pressuresand for back pressure induced unstarts. The inlet was also equipped with a variable angle finalcompression ramp for inlet starting and for providing a range of inlet contraction ratios. Thisramp was also used as a mechanism for determining inlet maximum contraction ratio.

The model schematic and instrumentation layout for the generic hypersonic inlet model isshown in Fig. 2 indicating the position of surface static pressure taps, heat transfer gages,skin friction gages, and high response dynamic pressure transducers. The inlet model utilizesan 8" span, planar geometry consisting of a short, flow-aligned plate followed by a fixed 4.5degree compression ramp (Ramp 1). Ramp 2/3, which housed the mass addition hardware, is avariable angle, straight compression surface with angles ranging from 0 degrees, for inletstarting, to a maximum value sufficient to cause inlet unstart. The baseline inlet performancewas taken with Ramp 2/3 aligned with Ramp 1 at 4.5 degrees. The inlet cowl was fixed during atunnel entry but could be moved between entries. The cowl was a straight ramp compressionsurface aligned at -6.5 degrees relative to horizontal (i.e. an ii degree compression of theIngested forebody flow for the baseline geometry) . The cowl could be positioned at two differentheights in positions consistent with cowl shock impingement on the innerbody either upstream,"at or downstream of the inlet shoulder. Possible cowl lip positions are indicated by dots on

236

Best Available Copy

C3 a X0 XX X

a o 9 ( x 0 X

RAMP I RAMP 2/3 RAMP 4

o000

-x x x X' X- -- o X 0 HEAT TRANSFER GAGE * HIGH RESPONSE PRESSUREx x x xTRANSDUCER

X STATIC PRESSURE TAP 0 SKIN FRICTION GAGECOWL

Figure 2 Instrumentation schematic of the generic hypersonic inlet model.

the schematic. The inlet flow was dumped into a large plenum which contained a variable positionthrottle.

In addition to the surface instrumentation shown in Fig. 2, the inlet model was equipped witha traversing mechanism to house the AEDC developed Mach-Flow Angularity (MFA) probe and a totaltemperature probe. These probes were used to measure centerline flow field properties at theend of the constant area throat for all configurations tested. In addition, for each tunnelcondition the probing mechanism was located such that centerline flow field properties weremeasured at the cowl lip plane providing detailed boundary layer information upstream of theshock interaction/mass addition region.

During a back-pressure performance run data would be taken at various supercritical throttlepositions, during an unstart, and, when appropriate, for a sub-critical operating condition.During a contraction ratio run data would be taken for the baseline ramp position of 4.5 degreesand for incremental increases in ramp position up to unstart.

The mass addition hardware used during the test pro ram was the same basic design as thatdemonstrated in an earlier proof-of-concept test programT. High pressure, room temperature airwas supplied to a plenum and accelerated through a supersonic nozzle with a design Mach numberof 3. Mass flow rates up to 0.16 lbh/sec could be supplied to the nozzle. The nozzle designutilized a centered Prandtl-Meyer expansion contour with the expansion focused inside the lowersurface so as to increase the radius of curvature of the throat. The I" high nozzle exIt spannedthe entire 8" inlet model and was located 4.42" upstream of the inlet throat for the baselinecontraction ratio. The exhaust flow was aligned with Ramp 2/3.

Earlier tests and analysis have indicated that the location of the mass addition must beupstream of the interaction region to be effective in energizing the boundary layer prior toexposure to the shock induced adverse pressure gradient. The location of the mass addition forthis test program was chosen far enough upstream so that the interaction could be controlled fora wide range of shock positions. For contraction ratio runs Ramp 2/3 was increased in angle,thereby strengthening the cowl shock and the resulting interaction, until the inlet wouldunstart. The Inlet maximum contrnction ratio for various mass addition flow rates wasdetermined.

Best Available Copy

Mw sbck , , wFI -y lajuwr ImmwraurzJIleu w&iizh orur Jim dme xvefm off Ulka ff~~ tmz~mm~sbck systum frw .sratigmdtius omtM balimut1ntz bI*i Iztl~am Thaair rVEMfmatchicriticoll In decommuimb the inletoh mzimm tmetal pmý remy anid d~irfi~dmg . 7w=MziraI1f-Dstability ainzmzu - DIA thiA~ pz~azn,, hijkt bwat Xvamem ozudfttiMu WAM xfsnwv tzwdithe vwiable pooeliii theuttlia omd., a]Ltbmx* the mass adaizitm mmme ~imlr ~nftwvUcczutl of tais xgism. dmt wota l o ami the effeet of mrns A&Hfflji mu[M lu~affr~zITjlrzJ mflfiaperfoin and stability chwamteristiris..

D~r% MX W~t" Iof Ist~e=I4 Strat dbatr WMW MPM at WM BM ES J ffI3Ct lhlii.. A sizwmiiyof the soaor obevmoswi~th trespw mo ifffert s Of to~mtfanmoss ar fl~f-mm EsprOwldd below

Usulegmm at Rmci S the 11ml t: omMil met he storart N crr~ Momp 72133 wftdkaw=mma~itim: bmwor. W&C mass abatsIMM Wn &dtltaumdr vaor meMErwseutofnnt mrlwa acb~md. 7*sms uee cmnrted vithi n~b 7mm~tam qsar% thont the sffiuxk wws I=

dowstOMof deinlt-m F1 (Cow Nosfitla 9ý,, at th Jimlem fimne (Qr~w]1 Rwaittd=n $)),, aidupstom f te sumamr witdowstramof the umm s&Wtff= im==1 ((C=Ifl tisiittis TD))..

shock iMyIWE ýsu thM Ue s*Mw11r cowl ftaitfia 3 and T"mae,~ smof o En~toe fliw WEesablished the zm= aitimimm com1d be tmmdoff amid the ffnaw smlet mmn4 HIIti

adhfftlomwl flowtla ra,,a the~o gne gu~ ffiutrmff 2-5 ofsi Smctc

111g. 3 shmthe ram stotff nmme -=f= for thfe inEt1C F2~ff aaWZ TD~ ~Wdti~time cxwl in positiim 9 Moe-35". da wtl~e ms widab apmm 3?-1. fmr aUl I'Jzs p~mox

SC, Salid ciztcle) the JinMlat Mest starme.. E asmedLtfmSc ~mvz nlgttm~

iiat ii 1w urnspmm

M-h6 Ne/fft - 2-Dx BfI

Ar

- 0

0I

s-]r ! 4 fo'fxofMs AMu s i ~eu r

Mom mmMmmu -AMm thmt~lagtu m aftU= fggmfnmryT5 ofw cam~ iffm MW crlaxsiEfam w rD dm mam mzmm, muilatrin im mr mri ulifamd &ah w zvs~

Hi If vba -ý NMaMfl±mm MUM offu & ttfir fln~at wamwd! xwnudf Smam s mf trikintt w

*mk3gam m mrx~ uq dura~w1ing t am slaxtn mm w mraI1 tAk sftwx %nundym gIntILr mezmiiio

Mcdmwjjmm mw~imb m 7,o)). a tm 1bunjmrund mma mtmw~iiam. M"ani ttffi mwss~ -mifi fnefl-n~a

am a n ~wimf tt anldrm tttLs.-D." mq ms fa tfris m hJaw tdkl sfimý ffn ý ; tdidmk-rjjMrm

thewi~ aamm SpmiU1r =Qmaffvm ((mm. aiii ffrII sanw rf~ctfm

ME imop mm (O=&, ff amnf 50 ME. MMWixr (lhi m crIM5il anitriMz-M c Imbm ft omaxmfurm!m M.m/ inommmlm mff Um EilinrfaJm tvntf a imflt ,91 ac m~o g flmmaD fi nm w-I a

Mi gnsum mW Uimm MES.-4.9a iff.%m is= &w~limu !irn UR fanMS f~OM MJW-RA-

&Waff (IC Mom ttlku ctam m~ mem ~ =waiil irmi mm and=mid! g mi ftrw r~ nmff tt~hmy~u

ugmmvm o iaUlmrs mfinm cfamr t tfimmmny 2R naurem if~mm of ~ m~ mrmtI

umaum f4 fkmi.. ttih murt ffDAre i rd~n fm mmmcww IA i yM

Ma m mm r~tfm hmfaie fs wmm rwd mutfin ffm- mml D17vxno/t Sdbmc 5sYMMKafikf

mike ]law ww2mw nw ritw zam. D m amffrb .. ift isfa Mmiti ftt~zml adlet affi Mar

24. =WhamqsrwdmAMMxwmtWmgm~rwm lmmm ~o N f~alUM

11 f3tnD EcWn. at i ~mmdtm rs d~f m 1TwimlXUfl

%N taw fflrir 1ldimTi mum snin' atnutma mE maw am ffilflam 5mtmwm wwuni a3~xR lgad

BesnAvilale op

Reducing the injectant flow rate to approximately 2.50 of capture, ikl-0.03 Ibm/sec. alsoresults in a started and well behaved inlet flowfield as indicated by the corresponding pressuredistribution. The nozzle exit flow, although still under expanded relative to the ramp pressureis wich more in line with the inlet flow structure. The pressure in the throat region is lowerfor the lover injection flow rate which is consistent with the reduce aerodyrnuaic contraction.The pressure distribution downstream of the throat is very similar for the different injectantflow rates.

The above discussion illustrates the effectiveness of tangential ms•s addition in controllingcowl-shock-induced separation to improve the starting characteristics of a the Inlet. Inaddition, for all Mach 6 conditions tested, the use of mass addition also improved the inletperformance and operating envelope as measured by maximum contraction ratio and maximum totalpressure recovery.

Once started at the baseline contraction ratio, accomplished by rotation of Ramp 2/3 aftersetting an appropriate mass addition flow rate, the inlet maximum contraction ratio wasdetermined for various mass addition flow rates by continuing to increase the angle of Ramp 2/3.Observations aade during the test included tho ramp angle for which the inlet unstarted for thevarious i•ass addition flow rates. Percent chinge in maximum contraction ratio as a function ofinjectent flow rate is presented in Flgn. 5 7 for cowl positions 9, 8, and 6. respectively.These cowl positions correspond to cowl thock impingement just downstream (position 9), rightat (Position 8). and well upstream (Position 9) of the inlet shoulder. Keep in mind that whatis plotted an-' observed ramp angles noted on the fly during the test so that there is someuncertainty in the absolute values; however, the trends clearly demonstrate a significantincrease in maximum internal contraction ratio achievable with the use of mass addition.

The inlet model was tested with a plenum and throttle mechanism to provide data on masscapture and inlet recovery as a function tf back pressure. For performance runs the inlet wasstarted using Ramp 2/3 and then this ramp was positioned for the desired contraction ratio. Athrottle sweep WAs then initiated for several mass addition flow rates. The resultingperformance parameter of interest is total pressure recovery and the maximum recovery for whichthe inlet operates started.

3 0 -

o 10u20

* Mach 6 Re/ft - 2.0 x 10610 Cowl Position 9

0

1 -10C

a. I p I

O.Mo 0.2 eO4 GASM 0.10 0.12

Injection Ratio (i0.)/kv..)

Figure 5 Rfiect of mass addition on maximm internal contraction ratio for cowl shockimpingeamnt just downstream of the shoulder.

240

30

0

o 20

C0

10 M(ach 6 Re/ft - 2.0 x 106

Cowl Position 8

C

0.,

-10 1

0.00 0.02 0.o4 O.AS 0.06 0.1o

Injection Ratio m/•,.r)

Figure 6 Effect of mass addition on maxium, internal contraction ratio for cowl shock

impingement at the shoulder.

i 30

A

Mach 6 Re/ft - 2.0 x 106

S~Cowl Position 6

C

S 20

0

v

(.1

S0 -

-10IIIII

0.00 0.02 0004 0.001 0.10

Injection Rati o (ilJ/,¢.,)

Figure 7 Effect of mass addition on maxiamum internal contraction ratio for cowl shockimpingement aetl upstream of the shoulder.

241

Plotted in Figs. 8-10 is the minimum inlet thrnttle opening for stable operation for cowlpositions 9, 8, and 6; respectively. For a given mass flow and total temperature, the throttlesonic area is related to the inlet total pressure recovery through mass continuity. Minim•throttle sonic area provides and indication of the level of the maimn total pressure recovery.In the plots, a reduction in D.,. indicates a corresponding increase in sxiwum total pressurerecovery.

These figures show the effect of increasing mass addition flow rate on the minimm throttleflap setting for stable inlet operation. The reduced stable throttle openings, (increasedaximim total pressure recovery) for increasing injector mass flow demonstrates the effectiveness

of the mass addition at controlling the shock wave/boundary layer interactions associated withmovement of the inlet terminal shock system into the throat (recall Fig. 1). In mixedcompression inlets with inasdquate isolator length back pressure induced unstarta can occur whenthe separation associated with the terminal shock system moves into the inlet throat therebycausing an aerodynamic over contraction and unstart. Use of mass addition to eliminate theseparation region associated with the terminal shock system increases the stable operating rangeof the inlet by allowing the inlet terminal shock system to continue increasing in strength withincreasing beck pressure. This increased stability derives directly from the reduced lengthrequired to contain the terminal shock pressure rise and, therefore, can result in a siprifitcatreduction in the required combustor-inlet isolator length. This can have considembla impacton dual-mode engines because of reduced internal friction losses at high speed mad overall engineweight savings achieved by elimination of structure exponed to the high internal pressures.

4. - '4

\

Mach 6

%% Cowl Position 9

"4 %A R/ft

%%

e' •e4 in---• •

Injection Ratilo (i1U'/i1CO.M.)

figure 8 g~ffct of mast addition on minimum throttle sonic or"4 for cowl shock impingement justdownstram of the shoulder.

242

4.'

Mach 6 I/ft - 2.0 x 104

COWL vosition S

3.7sS

U I I I JUS• o4t 0.84 SMO 0.5 @10

Injection Ratio

Figure 9 Effect of wass addition on minLuem throttle sonic area fo' covl shock impioement attha shoulder.

1A&

Mach 6 Ia/ft - 2,0 a 10

Cowl Position 6

I1 - I 1US Omt SM O0* ,Ue.

Injection Ratio (

Figure 10 Effect of -sgs addition on minimum throttle sonic area for cowl shock impingesent wellupstream of the shoulder.

241

The ability for the mass addition to provide control of tne separations associated with tý e

inlet terminal shock system and thereby allow an increase in the a ximum terminal shock pressurerise is further demonstrated by comparing the minimum throttle setting, D', for various ratesof injection. Figs. 11 1? -;how the inlet innerbody static r.ressure distribution for a wide rangeof throttle settings. The Oita in Fig. 11 is for Macih 6 opration vtth the cowl it, Posirion 6and for a mass aedition flow rate of approximately 2.5% of capture, irnJ-0.03 ibm/se., just abovethe minimum amount cequirev to maintain stable a 1ercritical inlet operation. Recall, areductior in D" corresponds to the throttle being closed and an increase in back pressure andtotal pressure recovery.

The stetic pressure traces for the different throttle settings clearly show the progressioncf the inlet terminal shock system tkpstream as the throttle is closed and the back pressurpincreased. At D'-2.•' the forward edge of the •hock system has reached the throat, the stationfor which is deiotoe by "T* on the plots, as indicated by the significant increase in thecorresponding locI static pressure. The inlet Is still operating supercritically as evidencedby the invsriant pr,-su "e upstream of the throat. The maximum static pressure ratio, diffuserexit prassure/freestrea., st.atic pressure, achieved with this mass addition flow rate was 59.Further ncrease in back pressure induced by closing the throt:le to D-2.0". resulted in inle..uotwart with the unstarted shock system affecting the static pressures well upstream on the ramp

for the same Inlet test conditions and model geoNPery, ruach 6 at baseline cov,nraction ratio,thv mass addition flow rate was increased from 2.5* of capture to appr3.imately 6% of captureby increasirg tne injector pressure. Again the t,,rottle was progressiv-ly closed to increasethe ituleat sack pressure and define the maximums pressure reovery. Fig. 12 shows the ramp staticpresauYe diatributir- fotr this case for a range of throttle settings. Recall that for anin ,i.ction rare of -. 5% of captiare the Inle. unstarted for D*<2.l" at a diffuser maximum staticpiessure ratio of 59.

so

Mach 6 P.e/ft . 2.0 x 106Injection Ratio - 0 025

Cowl Position 6

60

{1'(IA.) (RUN) /

- 7.3 (727)

40 - - -- 4.2 (230)

-..-.... 3.0 (231)

........... 2.0 (P.34)

307.0 (236,

20

10

00 2 I I

20 30 40 s0 60

Mode, Sta ion (Inches)

ligurz It Ramp static pressure digtribution for cowl shock impingemenr well upstream oi the

shoulder with near minimum vtabl.: masr addition flow rate.

'4-1

The static pressure distributions plotted in Fig. 12 indicate that at a mass addition flowrate of 61 the inlet did not unstart until the throttle setting was between 1.86 and 1.69,considerably below the minimum stable setting for an Injection rate of 2.5%. The ability forthe inlet to achieve significantly higher pressure recovery is indicated by the maximum diffuserexit static pressure ratio of 74. This is a 25 percent increase over the maximum static pressurerecovery achieved for the lover injection rate.

It is interesting to note that the primary focus of this effort was the control of theinteraction of the cowl shock wizh the forebody boundary layer, a determining factor in theplacement of the mass addition nozzle. However, this data highlights the potential forconsiderable improvement in stability margin and performance using mass addition to control theshock bounidary layer interactions associated with the inlet ter.inal shock system.

Tn. Mazh 6 test results for the generic hypersonic inlet discussed above can be summarizedas follows. Tangential mass addition has been demonstrated as an effective mechanism forcontrolling adverse effects associated with the shock boundary layer interactions. Inletstarcing characteristics were such that mass addition was always required to allow the inlet to

start. When the cowl was positioned such that the initial shock impingement on the forebody wasat or near the throat, with the resultant pressure rise was partially cancelled by the shoulderexpansion, the mass addition could be turned off and the inlet would remain started. However,when the initial shock impingement was considerably upstream of the shoulder, but stilldownstream of the injector, a reduction in injectant mass flow below approximately 2.0% resultedin an inlet unEtart. In this case the boundary layer was exposed to the full reflected shockpressure rise. Once started, the inlet maximum internal contraction was improved significantlywith the use of mass addition for all cowl positions.

U

Mach 6 Re/ft - 2.0 x 10sInjection Ratio - 0.07Cowl position 6

* / .' ...d.o

50 Wp) (R)/...-

- 7.2 (24)

- 4.1 (249) .S40 ."

S- - .4 - / ..- LaS........... 2- 425)I: " ., I. .. (252) ; ,

1.69 1 :)

I Q10

10 2D 30 4 os

Model Station (Inches)

Figure 12 Ramp static pressure distribution for cowl shock impingement well upsat.cau ot rheshoulder with increased mass addition flow rate.

24.!

TA ft2l 'A

41uION aWRp K~ml zugiflmwu EMz iuinelt WtIar low~ vmmn. mlitt ft ftmestt mm zt~

ffit ffir"iz ftgw- g F wtB aurwr.1 fft amrg~ ffLlmxm pJico S uiwmjj m afftfim ml

xffiviM~~~~~~~~~~~w' sm1toeffw 1wyom mLmwaa fmj

ADVANCED COMPUTATIONAL MODELS FOR ANALYZINGHIGH SPEED PROPULSIVE FLOWFIELDS

Sanford M. DashScience Applications International Corporation

Fluid Sciences Division501 Office Center Drive, Suite 420Fort Washington, PA 19034-3211

ABSTRACT

Recently developed 3D computational models (SCRINT, PARCH, SCHAFT, CRAFT) whichanalyze high speed propulsive flowfield problcms are des-cribed. All the models contain generalizedthermochemical capabilities and advanced turbulence models, and all integrate the governingequations using implicit/strongly-conservative upwind and/or central difference numerics. Themodels differ with respect to discretization (finite-difference vs finite-volume), solution formulation(spatial marching PNS, iterative PNS, time-asymptotic FNS, and time-accurate FNS), chemistrycoupling (matrix-split/loosely-coupled vs large-matrix/strongly coupled), and geometric/boundarycondition flexibility. The hierarchy of propulsive related problems analyzable by these models isdescribed, with specific model formulations found to be most appropriate for specific classes ofpropulsive flows. Applications to both air-breathing (scramjet/gas-turbine) and rocket propulsiveflowfields are described along with thoughts and experiences related to code validation, turbulencemodeling, and the development of "problem specific", user-friendly codes.

INTRODUCTION

The author and coworkers have had significant involvement in both the development andapplication of propulsive-oriented CFD codes (which solve the Reynolds-averaged Full Navier-Stokes(FNS) or Parabolized Navier-Stokes (PNS) equations), out of which a very practical "philosophy"has evolved. Some of thi- philosophy for scramjet related flow problems was discussed in earlierpublications"3 but the specifics aie now somewhat outdated. The philosophy evolved requiresdistinguishing between basig code attributes (Table 1) and 39m" code attributes (Table 2).

TABLE I - CFD CODESIBASIC ATTRIBUTES

* MODULAR CODING STRUCTURE TO FACILITATEMODIFICATIONS/ UPGRADES

• BOUNDARY CONDITIONS/FLEXIBILITY vs ROBUSiWESSa Explicit vs Implicit

* ALGORITHM/DISCRTIrZATION"* Implicit - Central/Upwind Differences"* Finite Difference vs Finite Volume

* GEOMETRIC FLEXIBILITY* Grid Patching/Blanking* Blocked Grids* Dynamic Grids

- Basic/ChimeralUnstructured

Approved for public release, distribution is unlimited. 247

TABLE 2 -.CFD CODESISPECIAL ATTRIBUIES

* CHEMICAL NONEQUILIBRIUM"* Matrix SplitiVeakly-Coupled"* Large Matrix/Fully-Coupled"* Thermo of Upwind Numerics

* THERMAL NONEQULIBRJUMa Multi-Temperature Model

* TURBULENCE MODELING* Two-Equation Model Emphasis

- EfficientiAccurate Inclusion into Code- Zonal Corrections for Compressibility,

Streamwise Curvature - Vorticity, etc.

* MULTI-PHASE FLOW"* Equilibrated Mixture"* Gas/Particle Nonequilibrium

- Robust Particle Solvers

The first basic attribute, code structure, relates to how the code itself isconstructed. Code structure impacts: run time - via an ability to vectorize;extendibility - via modularity features which readily permit one to upgrade orreplace the basic algorithm and add modules for enhancing the thermo/physics,etc.; and, overall utility - via an ability to calculate 1D, 2D (planar andaxisymmetric), and 3D problems in a single code. The second and third attributes- boundary conditions/algorithm-discretization, form the "guts" of the code, butshould not serve as the dominant selection parameters if the code is wellstructured and the algorithm can be readily upgraded. For example, it isrelatively straightforward to extend a central difference algorithm to incorporatean upwind run option for enhanced accuracy and robustness - for perfect gassimulation. It can be extremely complex to do this (with implicit/conservativeformulations) for real gas, multi-component mixtures. The last attribute,geometric flexibility, can be an ovsrriding one. Most research codes, asdeveloped at universities or government facilities (e.g. NASA centers), containthe most currt.nt algorithms and boundary condition formulations, but aregeometrically-restrictel to very simplistic flows. These are unit problem codes.The real work then becomes that of extending such codes to deal with complexgeometries which include struts, staps, cavities, etc., without requiring a gridgeneration expert in the loop for each new problem. Some recent codes havespecial blanking/patching formulations to facilitate dealing with such propulsive-related geometric complexities.

Tre next set of attributes (Table 2) relates to special features containedin the codes for dealing with real gas behavior, nonequilibrium processes(chemical, thermal, multi-phase), and wiLh the most problematic area, turbulence.The inclusion of real gas behavior and finite-rate chemistry into modernimplicit/conservative solvers is nonsimplistic, particularly in dealing withupwind algorithms which can treat strongly combusting flow. Loosely-coupled/matrix-split approaches can work well for less demanding steady-stateflows using time-a.,.nptotic solution procedures, and can provide significantsavings in cpu expense it "tricks" are employed in treating the nonequilibriumsource terms. Space marching or time-accurate problems require a strongly-coupledapproach, or an iterative procpd,:re built into the spatial or temporal step, if

matrix-split concepts are util;7cd. Inclusiort ol multi-phase and/or thermalnonequilibrium effects arc quite spe,-zialized, bot can be incorporated into the

2.v•

framework of existing codes by skilled personnel. This is often more expeditiousthan going back to earlier codes which contain such very special features, but

within the framework of "antiquated" fluid dynamic algorithms. Turbulence models,

at the two-equation level, are also readily incorporated into computer codes -

but, making them operate properly for different environments is complex, both

computationally and from the view point of selecting models and/or "tuned"

coefficients appropriate to the problem at hand, or to localized regions of flow,if a zonal turbulence approach is taken.

With the comments summarized in Tables 1 and 2 in mind, the philosophy thathas evolved towards the selection, development/upgrade and validation of CFD codesfor high speed propulsive applications is summarized in Table 3. The "baseline"codes we select are always "popular"

TABLE 3 - CFD CODE PHILOSOPHY

"0 OBTAIN "BEST" STATE-OF-THE-ART GOVERNMENT CODES

"* Upgrade Basic/Special Attributes"* Move to "New" Code (Every Several Years) if

Upgrades to Stay Slate-of-the-Art Become Substantial- e.g. Finite Difference -> Finite Volume

Matrix Split Chemistry/ -> Large Matrix Chemistry/Central Differences Upwind Differences

* SPECIALIZED VERSIONS OF CODES FOR SPECIFIC PROBLEMS

Different Versions with Different Algorithms/BC/Chemistry etc.- e.g. PARCH/RN, PARCHIGT, PARCH/VSL,

"* HIERARCHY OF VALIDATION PROBLEMS

n Level 1 Unit Problems to Level 4 Propulsive Component Analyses

"* USER-FRIENDLY CODE EXTENSIONS

"* Problem Specific Grid Generation"* Flowfield Initialization to Enhance Convergence

government codes which have the appropriate "basic attributes" [Table 1] and someof the "special attributes" [Table 2] required to solve the problem. The choiceof popular government codes is a practical one which recognizes that: (1) manypeople will work with these codes and thus there will be significant informationexchange amongst developers and users; and, (2) the government will providecontinued support to the "expert" developers for upgrades in the methodology, andfor remedying observed deficiencies identifi-ed -by the user community. Thedevelopment of "proprietary" codes by industry has, in this author's opinion, beena wasteful exercise of resources.

The next point made in Table 3 relates to the ability of an off-the-shelf FNSor PNS code to handle varied problems. While, "in theory" such codes are capableof handling a very broad array of problems, in practice they have to bespecialized for specific problems. The specialization entails the selection ofthe solution algorithm itself (a single algorithm is not best for all problems),the choice of boundary conditions, the thermochemistry to be incorporated (whattype, how to incorporate it, etc., etc.), and, the turbulence models incorporated(which ones, what specialized corrections are required, etc.). This viewpoint maybe counter to the philosophy of many organizations, but it has evolved from ourexperiences in working varied problems, starting with the same initial baselineFNS or PNS code, but, ultimately incorporating so many modifications or upgradesto address the problem at hand, that a special version of that code evolves, quitedifferent than the versions which have evolved for other applications.

Best Available Copy

The code validation process, with an emphasis on scrasjet propulsionflofai•lds, was discussed by the author in reference 4 vhere the 4 lvoyl. ofvalidation summarized in Table 4 were defined. The following comments regardingthese validation levels apply:

TABLE 4 - SCRAMJET CFD CODE VALIDATION PRQlOM

S' I000C FD 3O68 amC"COS DON sas CUS 0s SOOFU 8W10-

- %a ..l Akcwru. AM&-' -) xjwtv A"%&- -- N IOP 09 ="a~FVIXFOLT I JMU&WW Tza MCMIUMM _ *1u, M-

ONT PNLOSUMT o I1DC S VWPiAI

COS/OD COMINONS WX COMpJ.[lMUSM n DOU win €U IXIB•yOM1AMI O35 031? P3ies. DATA CI DATA COPWLAm PrQao•L--flUris -Dittgemta Data -Plow Details SIVB DATA COM--P*im/ taO Diff *Eeget 0a I MAMINS ISZ.0--rnI/IF cgoneant -ParamitLe W K Of I cow--Mler/PtaabelLo TIgMi W 0 00 AN68

.-so8 MWO .- 1So rI n .AD CM CAM .-Cws4m 919M.WE PAMUMUM tur IAAPAMiMI 813 A" MIM OF CF- 0066(01.Z. ~ZOAWBA5L WQ. XPC*4MT1"W~ A96- aSMAa1SAa AS

RK).1aAz~MhAMM MUNw mose- MIos Pi OSZMAWBL UNSIMMNI low MAL To INAMALOFSO15-

.OPAIOAL11 of 10 To8 a"O POSIA

LUYK I - Petrormance of the "Slainp rod,. This entails oode-to-code

comparisons, serves to remove "bugs", eatablishes run parameters (eqg.,artificial damping coefficients), and defines the operational range of thecode using the existing algorithms and boundary conditions. Level 1validation is generic and can be performed before the code is speolalised toan application.

LEY"ZJ - Analysi- of "rundamental" Unit-Proble This entails analysingunit problems such *a boundary layers, simple @hear layers, iaminar/turnulentdiffusion flames, etc., to ensure that the turbulence models andthermochomietry were incorporated into the code correctly, and, that thecoefficients ere appropriate (e.q., turbulence model coeffiolents,theraodynamic curve fits, chemical rates, etc.) Level 2 Is performed onlyafter "preliminary" specialization of the code takes place uincs theturbulence models and thermochemistry "presumed" appropriate to theapplication of interest must first be defined and then incorporated Into thecode. The unit problems of Level 2 are not generic and should be rapecilimedto be of direct relevance to the propulsive component of interest.

LZ!.ZL)_- Analysi, of "L *J.ft b.•oJm..L This entails the analysis or"advanced" unit problems which are submete of the overall flowfteld problems,and for which detailed date are generally available. Por a proplsAivecomponent, typical component problems would involve the simulation oflaboratory data obtained for model Inlets, coubistors, notrles, etv. wherethe initial/boundary conditions are well defined.

L3VEL 4 - 1 _" Tis entailthe analysis of a problem which could Involve additional complexities suchas multi-tone blocked grids, zonal turbulence modeling, etc. Validation onthis level should initially proceed to establish a basoline operational

250

capability (e.g., to get all the pieces in place), recognizing that the Level2 and Level 3 unit problems may not be easily resolved to support Level 4analyses (e.g., turbulence modeling issues may preclude obtaining goodagreement with Level 2 and/or 3 unit problem data).

Comparable comments on the validation process for exhaust plume interactionflowfields are given in reference 5. It should be emphasized that Level 2 and

Level 3 validation studies require very detailed data so that "canceling errors"do not occur (see, e.g., the discussion of Caruso and Childs 6). All to often,adequate agreement with "bottom line" data is achieved via the use ofinappropriate turbulence models and grids which are not sufficiently resolvedand/or aligned with the flow features. The areas of turbulence modeling and gridgeneration must be carefully scrutinized in propulsive flowfield calculationssince, more often than one would like to recognize, seemingly correct solutionsare obtained for the wrong reasons--the results must "make sense", not justcompare with data.

The bottom item in Table 3 is also of great significance with regard to"turning over" CFD codes to non-experts in CFD and/or non-experts in thephenomenology involved. Starting with a well-founded CFD FNS or PNS code, ittakes at least one man year of effort by an "expert" in the field to validate thecode for a very specific propulsive application. We have found out that it takesat least this amount of effort to subsequently take this validated '7ce and makeit sufficiently "user friendly" so that a "non-expert" can obtain reaLdble resultsin a consistent manner. The non-expert cannot apply the "bag-of-tricks" theexpert has learned with regard to grid generation, flow initialization, creativeuse of artificial damping parameters, etc., etc. Building this "expertise" intothe code is non-trivial as we will exhibit by examples later in this paper.

This paper will illustrate how the author and coworkers have dealt with theissues discusssd above. It will discuss our basis for the selection of "baseline"government codes, how we have specialized these codes for varied propulsiverelated applications via the inclusion of advanced turbulence models andthermochemistry--as well as specialized algorithms, and how we have validatedthese codes and developed "user-friendly" versions. At this point in time, ourpropulsive-related FNS and PNS work has involved the utilization of 4 "baseline"codes--the PARCH and CRAFT FNS codes and the SCRINT and SCHAFT PNS codes. PARCHand SCRINT have evolved from the most popular "current-generation" FNS and PNScodes ever developed in the United States, namely the ARC/PARC NS codes [ARC isthe aerodynamic Ames Research Code developed by Pulliam and coworkers7 ; PARC is thepropulsive extension developed at AEDC by Cooper 8 with advanced gridblanking/patching capabilities for treating struts, steps, cavities and otherpropulsive oriented geometric complexities], and, the Ames/AFWAL PNS code 9 . Thesecodes have many years of history behind them. They have had continuous supportby the government for varied upgrades (e.g., the upgrade of ARC3D from a finite-difference/central difference code to a finite-volume/upwind [TVD] code10 ), and forvalidation studies. There have been regular ARC, PARC FNS and AFWAL PNS workshopsfor the developers, "extenders" and users of these codes to share thoughts,discuss problems, etc. Much of our recent work has involved systematic upgradesto these codes in the areas of turbulence, thermochemistry, etc., to provide forthe analysis of specialized propulsive problems.

The above codes originated about ten years ago as central difference, perfectgas codes. The inclusion of generalized chemistry into these codes has beenperformed in a loosely coupled/matrix-split manner as described in references 11and 12 and has proven effective in many situations. However, for severeenvironments with strong-shocks and combustion chemistry, the inclusion of upwindnumerics and strongly-coupled/large matrix chemistry becomes requisite. As theauthor and coworkers were to embark on such extensions, the TUFF (FNS)/STUFF(PNS)aerodynamic codes developed by Molvik and Merkle13 (under NASA/AMES support) becameavailable. At that point in time (a little over one year ago), no comparablegovernment-owned codes were available which had properly dealt with the Roe/TVDreal gas mixture problem (see, e.g., references 14-16), which were efficiently andeffectively coded (these codes handle 1D, 2D (planar/axi] and 3D problems andreadily vectorize on CRAY machines) and, which incorporated chemistry in a robust,implicit strongly-coupled manner. Our screening of these codes indicated them tobe unique and truly reflective of the state-of-the-art. They have served as the"basis" for our codes of the future. Our propulsive versions, CRAFT(NS)/SCHAFT(PNS), have been extended to include upgrades in turbulence and ageneralization of the chemistry from clean air to H/C/N/O. Dynamic grid

3es t Availab!e Copy 251

capabilities have recently been incorporated into CRAFT for the time-accurate

simulation of unsteady flows with moving boundaries and/or fitted discontinuities(e.g., liquid propellant and electro-thermal gun problems, etc.). Currentfeatures of these codes are summarized in Table 5. An overview of thedevelopmental history of all the codes in Table 5 and applications to a varietyof propulsive-related flovfield problems will be discussed below.

TABLE 5 - ENS AND PNS COMPUTER CODES WITHADVANCED THERMOCHEMISTRY - OVERVIEW

FNS CODES PNS CODES

PARCH 2D/30 SCRINTX/30T

"* FlNrTE-DIFFERENCE DISCRETIZATION 9 FNITE-DIFFERENCE DISCRETIZATION"* BEAM WARMING ALGORITHM e BEAM-WARMING ALGORITHM"* MATRIX SPLIT CHEMISTRY * MATRIX SPLIT CHEMISTRY"* MULTI-PHASE FLOW (EQINON-EO) T TIME-ITERATIVE FRAMEWORK (3DT)"* ROE/TVD ALGORITHM - AIR ONLY * ROE/TVD ALGORITHM - AIR ONLY"* SEVERAL TWO EQ. TURBULENCE MODELS 9 K-c TURBULENCE MODEL."a GRID BLANKING - GENERALIZED * GRID BLANKING - SPECIALIZED"* GENERALIZED EXPLICIT BC e IMPUCIT OR EXPLICIT BC

CRAFT SCHAFT

* EXTENSION OF TUFF THIN-LAYER NS CODE e EXTENSION OF STUFF PNS CODE"* FINITE-VOLUME DISCRETIZATION (2D or 3D) FINITE-VOLLIME DISCRETrZATION (2D or 3D)"* ROE/TVD ALGORITHM * ROE/TVD ALGORITHM-TIME ITERATIVE"* STRONGLY-COUPLED•LARGE MATRIX CHEMISTRY * STRONGLY COUPLED/LARGE MATRIX CHEMISTRY"* FULLY IMPLICIT BC * FULLY IMPUCITY BC"* K-c TURBULENCE MODEL * K-c TURBULENCE MODELe RESEARCH STATUS - LIMITED BC/BLAING * RESEARCH STATUS - LIMITED BC/BLAMANG

FNS HIGH SPEED PROPULSIVE CODES-DEVELOPMENTAL HISTORY

Our development of FNS codes for application to high-speed propulsion (andthe related area of exhaust plume interactions) initiated in 1987. Just prior to1987, we had been involved in "early-NASP" activities to assess variousgovernment-owned FNS codes for applicability to high-speed propulsive flowfieldproblems. Our assessment" 3 indicated that implicit/conservative codes utilizingBeam-Warning type numerics (as typified by the ARC aerodynamic code 7 , and improvedvariants with finite-volume/upwind extensions"0 ) represented the current state-of-the-art, and were clearly preferable to earlier explicit/conservative codesutilizing the MacCormack algorithm and implicit/nonconservative codes utilizing"TEACH" based algorithms. The earlier codes already had many propulsive featuresavailable, such as finite-rate chemistry and/or generalized boundary conditionsto facilitate analyzing struts/steps/cavities etc., but were not computationallyefficient and/or robust. Prior to 1987, such propulsive features were notavailable in government-owned implicit/conservative codes.

During 1987, two-major breakthroughs occurred with regard to propulsive-related Beam-Warming based FNS numerics, namely:

(1) the Air Force AEDC PARC code 7 was completed and made available,extending the ARC code into the propulsive arena via a generalizedboundary treatment, obtained implementing "patching" methodology; and

(2) generalized finite-rate chemistry was made operational by the author andcoworkers in the ARC/PARC code framework employing matrix-splitmethodology 11' 12.

The PARC code has become the most widely distributed FNS propulsive-solver in theUnited States, for non-reacting flow problems (it is currently operational at over50 governmental and industrial entities) because of its user-oriented boundarytreatment, and other user-oriented features (see the discussion in reference 17

SBest Available Copy

which also provides references describing PARC applications to test facilitysupport, turbine engine tailpipes, high-speed inlets and nozzles, etc.).

We obtained PARC from AEDC in 1987 and developed an extended version entitledPARCH to which we incorporated advanced thermochemistry (e.g., matrix splitfinite-rate chemistry) two-equation turbulence-models, multi-phase nonequilibriumand thermal nonequilibrium under the varied programs summarized in Table 6. Aswork progressed on these programs, it became quite evident that the features

TABLE 6MAJOR PROGRAMS INVOLVING PARCH

"* ROCKET NOZZLE VERSION, PARCH/RN, FOR SDIO"* User Friendly for Non-CFD Community" Chemical and Multi-Phase Nonequilibrium"* "Tricks" for Nonequilibrium -> Fast Run Times

"• PLUME&MISSILE AIRFRAME INTERACTION VERSION,PARCH/TM, FOR MICOM

"* 3D Blocked Version with Above Capabilities"* Time-Accurate Version Under Development

"* GAS TURBINE VERSION, PARCHIGT, FOR AFEWC"* User Friendly for Non-CFD Community"* Swirl/Advanced Turbulence Models

"* LASER VERSION, PARCHILF, FOR MICOM, KIRTLAND AFBa Adaptive Gridding, Multi-Component Diffusion

"* VISCOUS SHOCK LAYER VERSION, PARCHIVSL, FOR EGLIN AFBW Thermal Nonequilibrium

required in a version for one application, could be quite different than that foran alternate application. As such, we developed very specialized versions ofPARCH that contained the best "basic" and "specialized" features for the problemat hand and performed Level 2, 3 and 4 validation studies with these specializedversions. However, "master versions" are maintained containing all the namedcapabilities under one roof as exhibited in Table 7, which also shows several ofthe specialized versions. Note that thermochemical and transport data is nothardwired into PARCH - rather, it is contained in data banks which PARCH accessesthrough preprocessors (specialized for each application, as will be discussedbelow). In addition to grid generators, many PARCH computations are performedusing a solution adaptive procedure to concentrate grid points in regions of rapidchange. The SAGE Code 18 developed at NASA Ames is presently employed and its usagehas to be specialized for each problem class.

The sequential development of PARCH and related validation studies aredescribed in references 19-27. The matrix-split chemistry approach has workedquite well with the central difference numerics, and the use of a point-implicitsource term treatment has permitted very significant CPU cost savings. For manysteady problems in hypersonic propulsion, the central difference numerics andmatrix-split chemistry in PARCH are not sufficiently robust to deal with the verystrong shocks and highly exothermic combustion environment. For many nonsteadycombusting flows, where a time-accurate treatment of the chemistry is required,the matrix-split approach is not well suited unless it is iterated uponsignificantly within the time step. The upgrades of PARCH to deal with suchproblems would be substantive, and would require moving: from finite-differenceto finite-volume discretization; from central difference to upwind numerics; and,from loosely-coupled/matrix split to strongly-coupled/large matrix chemistry. Asper the philosophy summarized in Table 3, this warranted a move for us to a newbaseline code. With the significant emphasis on CFD for hypersonic/chemicallyreacting flows in national programs such as NASP and SDIO, one would have expectednew government codes with the appropriate attributes to be available. At the endof 1988, this was not the case-work towards this end was still in progress, butno government-owned codes were available with the required attributes.

Best Available Conp5

TABLE 7PABCH-MASTER CODE

_ _ _ _._•• -- Fh* SMpl Alebrlac Thermdnm* Canie ID/3D (Soni) Chmica Kinetics* ragle (Thoopeon) PAIWDJ) MN_ COW Particulate Proper-

2Z_ tie3, Algorithn .laminar Transport. Thermoch ry options Data

M .ulti-Phase Options- Turbulence Model Options-Grid Blanktrq/multi-zone

X POC5D (Ame)* Zn-Howse CSiC)

SM? mmr' VJT( M LTUM TnCAL MIS VLEUr

, De &M-Wa .te/ ReND. All Algoritlvr/Zone

. 7nitt-ma te (Sol,1l • Air rq/ Non eq . C e ..

Liqui) T. Finite-Rate Chm.

, IP Iones~ilibrium .~iarOl GIP Equilibrium•Laminar Only qiirw

A9. ry VLsity SnlelUlankedT-Eq. Trbule

Xi. Brgle/UnblaMed Grid Multiple Zones/GIid .dvared Blankinr

An FNT government code having the proper "baseline" attributes at this timewas the TUFF code developed by Molvik and Merkle' 3 at Penn State under a research

1ant from .NAA/Ames. A comparison of the PARCH and TUFF FNS codes is describedn Table 0.

A comparable PN8 code entitled STUFF was also developed by Molvik and Merkle. TheTUFF/STUPF codee were developed for hypersonic external aerodynamics with finite-rate clean air chemistry. We have started ur- -I ng these codes to beplume/propulsive flowfield solvers entitled CRAFT (t and SCICAFT (PNS), whosepreliminary capabilities were summarized in Table S. The work to date with bothcodes has involved the inclusion of: the full stress terms (not Just thin layer);the ke turbulenue *odel; and, H/C/N/O chemical systems. Our work with SCHAFT wasinitiated a year ago and we now have significant experience and a validation basis(Levels 1-3) for propulsive flows as will be described. Our work with CRAFT isvery preliminary and has additionally involved the inclusion of dynamic griddingfor problems with moving boundaries. Preliminary exploratory studies are now inprogress with validation at Level 1 near completion, and Level 2 studies in

progress for various jet/plume studies and for gun-related problems (liquidpropellant/electrotherasl guns).

214

IARLE8PARCH YS TU•EE

FLMd TYFM - 2D/AXI IN PARCH2D - 2D/AXI/3D IN SINQ..E CUE- 3D IN PAU=t3D - 20D/I WITH SINGLE KANE

EOUATI2IS. - FILL OR THIN-LA.YER NS - THIN-LAYER NS- STROINGLY CM14SUVATIVE - STRONGLY CONUMtVATIVE

DISCRETIZATION: - FIIITE-DIFFEJEICE - FINITE-VOLU.I- MAPPED GENERALIZED COOR- - MAPPED GEKED COM-

DINATES DINATES

ALGORITIM: - DIAGNAI7E BEAR-WARMIN6 - BLOCK TRIDIA6ONAL RouTViD- SL0 TRIDIA6OKAL BW- DIAGONALIZED Ro./TVD

ADDITIONAL SCAL R ELUAONS: - WAIM.Y COUIP, CENTRAL- - STRONGLY COUPLED. Rn/TYD(SPECIES/TUDILLNCE) DIFFERENCE SOLUTION SOLUTION VIA MATRIX EXT.

0INISThY: - POINT IMPLICIT/LOOSEL Y - FILLY COUPLED/ AILEDCOUPLED LINEARIZATION OF CWIAICAIL

- SPF THERY&JOISTRY SOURCE TERM- AIR THERIMOCC)ISTRY

TURBULENCE: - Ke -NE

BOUNAARY CONDITIONS: - GEMERALIZED EXPlICIT - LIMITED.IMPLICIT, UO# BLANKING BLANKING

PNS HIGH-SPEED PROPULSIVE CODES - DEVELOPMENTAL HISTORY

As per the FNS discussion above, for developments in the PNS arena, we againstarted with "modern" codea which utilized finite-difference Beam-Warming numericsas typified by the popular AFWAL 3D PNS code 9 - the most widely used PNS code inthe United States in the early to mid 1980's for high speed aerodynamicsimulation. (Our earlier NASP SCRAMJET propulsive work at the 2D level did employnonconservative/implicit numerics, viz-SCORCH combustor code 27,2, and,conservative/explicit numerics, viz-SCHNOZ nozzle code29 -3, for expediency sincethese codes had well-established finite-rate chemistry capabilities from work inthe rocket-plume community.] Significant improvements to the AFWAL spatialmarching PNS methodology were provided by the recent 2D and 3D UPS PNS codes ofLawrence et al 31 ,3? which utilized Roe/TVD upwind numerics with a finite-vo]umediscretization of the equations. Air chemistry was incorporated into such codesby Tannehill and coworkers as described in references 33 and 34 usinq loosely-coupled methodology. We had developed extended versions of the AFWAL and UPS FHScodes for NASP aerodynamic and propulsive applications. Our internal flow code,SCRINT, was an AFWAL extension and our external flow code, SCRAimP, Was .1 UPSextension (see references 35 and 36).

The original versions of SCRAMP and SCRINT cintained only equilibrium airthermochemistry. A research version, SCRINTY, yas upgraded to include combustion-oriented finite-rate chemistry into the Bet.m-Warming numerical framework using amatrix-split approach".. 2 . SCRINTX was applied to a variety of combustor andnozzle related flowfield problems as well as to inlets. Its performance w-scompared with the earlier SCORCH and SCHNOZ PNS cudes as described in refer-ence37. SCRINT did quite well for problems where the central difference numerics wereadequate to capture shock waves.

In our 2D research with SCRTNTX, we had inco•rporated tirst and sec3und order(TVD) Roe upwind numerics into the code after the matrix-zplittinq was performed.This "ad hoc" approach worked quite well for flows with air chemistry, but workedpoorly for strongly combusting flow problems where the approximations entailed"broke down". Similar comments apply to its inclusion in PARCH. Pence, thesuccess obtained by us (and others, viz, the work describod in references 33 and34) for air chemistry, did not transfer to strongly t-ombusting flowfields where

255

the chemistry influences the wave field. For such flows, the Roe formulationrequires use of strongly coupled fluid and chemical species equations. For spacemarching, such a Roe formulation depends upon an eigen-decomposition whosederivation is quite difficult to performi. The Roe (PNS) formulation withchemistry is much easier to develop if the approximate Riemann problem is posedin time rather than space. For time marching, the eigen-functions for a muchsimpler matrix are required. At the 3D level, our recent PNS work has

concentrated on the inclusion of time-iterative methodology into existing finite-volume (SCRAMP3D) and finite-difference (SCRINT3D) spatial marching codes".Significant improvements in robustness and accuracy, above that provided by theinclusion of upwind numerics, have been obtained by the use of time-iterativerelaxation methodology as discussed in the recent PNS survey paper of Krawczyk at

al 40. Also, significant work was performed involving the use of the advanced gridgeneration techniques in the cross-flow plane, and grid blanking for the treatment

of cavities and swept surfaces, following procedures similar to that introducedin PARC.

Our preliminary 3D PNS work with chemistry had concentrated on the inclusionof chemistry into the SCRINT3DT code which was a 3D extension of the SCRINTX codewith time-iterative PNS numerics. SCRINT3DT was made operational with matrix-split chemistry and operated with Beam-Warming central difference numerics or "adhoc" real gas Roe upwind numerics (splitting done before the Roe decomposition).As per the spatial marching experience with SCRINTX, its ability to analyzestrongly cosbusting flow problems with "ad hoc" Roe real gas upwind numerics wasfound to be problematic - the obvious remedy was a reformulation of the Roe realgas methodology using a large matrix/fully-coupled framework. As this ratherambitious effort was to initiate, the STUFF 3D iterative PNS code of Molvik andMerkle1 3 became available which had the large-matrix/fully coupled Roe formulationdone properly for clean air chemistry. It was deemed more expeditious to use theSTh2F code as our new baseline code, than to upgrade the SCRINT3D code, and, theadded benefits of the finite volume formulation of STUFF became available withthis Jecision. Our upgraded version of STUFF is called the SCHAFT code, and workto date has involved the generalization of the chemistry and the inclusion of two-equation turbulance models ,41-4 3 , very much akin to the CRAFT code upgradesdescribed above. The nearly identical structure of TUFF/STUFF (and thusCRAFT/SCHAFT), facilitates concurrent FNS/PNS upgrades.

In referring back to Table 5, we thus have 2 classes of codes for FNS and PNSsolutions: finite-difference codes with central difference/Beam-Warming numericsand matrix-split/loosely-coupled chemistry (PARCH/SCRINT); and, finite-volumecodes with Roe/TVD upwind numerics and large-matrix/strongly-coupled chemistry,CRAFT/SCHAFT). PARCH is currently the FNS code of choice for steady-statepropulsive flowfi-ý1ds with elliptic features and there are distinct advantages tothe matrix-split approach utilized for chemistry for dealing with large numbersof chemical species in a time-asymptotic simulation. CRAFT is currently usedwhere shocks are very atrong and more robust upwind/finite-volume numerics isrequired, and, for time-accurate stvdies. PNS requirements are generally morestringent thin FNS requirements and SCRINT has not proven capable of routinelydealing with the hypersonic propulsive flowfield environment. The current codeof choice is SCHAFT due to its enhanced robustness for such flows. This paperwill thus focus on the PARCH FNS and SCHAFT PNS codes, with a very briefdiscussion of the new CRAFT code under development and some preliminaryapplications.

OVERVIEW OF PARCH CODE AND SPECIALIZED VERSIONS

PARCH CODE OAS-PJASS EQUATIONB - The gas-phase equations in PARCH are cast ingeneralized mapped coordinates as listed in Table 9. In addition to the 5 fluiddynamic equations, they include NS-1 chemical species equations and 2 turbulence

model equations. The RHS source terms in these equations (representing particle/..chemical/turbulence nonequilibrium) can be quite stiff and are generaliy treat*:implicitly. If these equations (concurrently, with the particle equations fL•multi-phase flow) are integrated, in a strongly-coupled manner, the resultin.1block tridiagonal solution matrix has a block size of 5 + fNS-l] + 2 which can beunwieldy for propulsive syrtems where the chemistry is complex and NS is large.In addition, the strongly-coupled integration procedure requires that the sametime-step be taken for integration of fluid dynamic/chemical/particle/turbulenceprocesses. This can be problematic since the fluid dynamic equations are

256

TABLE 9PARCH CODE GAS-PHASE EQUATIONS

8Q + E !7 +8!at a a, 1 J6

P0 1~u~ 1A 1 .VX N*ZZVXPU pu~iiP *iq*x t "'

P, PWU4t P 4 X y211 C HP W

S (EPU EV y HPC

P ru I !(t x ZaLx Y *tY #Z I H -i

ok pUk p(t k - yky + ik Z p . vs.

Pus P~tX41 y6Y Yz

-'accelerated" to steady-state using independent time-steps (e.q., each gqid poirnLis advanced utilizing a time-step which is a multiple of the cell Courant number)whereas the nonequilibrium source terms contain "real" time scales which can beinconsistent with the "fictitious" numerical time steps. Via a matrix-split/loosely coupled approach, the fluid dynamic, particle and chemicenonequilibrium equations can be separately integrated using appropriate timscales.

HATRIZ-SPLIT FORMOLATION - Applying the conventional Beam-Warming algorithm to 0Jequations of Table 9, utilizing Euler implicit time-differencing, yields L.-1,"delta" focm finite-difference expression written below (for simplicity) for the2D system of equations:

11 * Atit AN * a 5N - MI)]AQ()

a -At(&.E C E PN - RN) I N

In equation (1) 6 is a central difference operator, AQN - 0" - Q0 where N denotesthe time step level, and, the flux vectors E and F, (which are nonlinear functionsof Q) have been linearized about QN, viz., E*1 = - + A"&QU where A - fE/aQEquation (1) is the unfactored form of the block algorithm and represents a systemof 4 fluid (5 for 3D flows), n (= NS-I) species equations, and 2 turbulence modeiequations (w;hich we will now dismiss to simplify our discussion of matrixsplitting). We thus seek to decompose an n+4 system of coupled equations, to t;ystem of 4 coupled fluid dynamic equations, ard n scalar chemical specie!,equations. Using the nomenclature

Qj(p. PU, PV, E.,)T(2 (2)

Q,= (oh , 02, .

257

the n+4 systt.m is decomposed am:

p A Go L j + aL (3)

X AC I A~ AG c I C

The above syrtem of matrices is split about the horizontal (indicated by thedashed lines) leading to the fluid dynamic system written as:

1! + At(& AN +, • &4•)IA

* - At(& A€ ,1r )AQ (4)

(where the source term Mf, having particuiate contributions is treated explicitly,and has thus been incorporated into P,.)

This exercise in matr:x partitioning recovers the original 4x4 blockstructure of the perfect-gas Luid dynamic formulation, with addition of a forcingfunction on the explicit right hand side and revised elements of the Aff and Bffmatrices on the implicit left hand side to account for generalized multi-componentspecies and calorically imperfect behavior (see references 11 and 12 for details).The elements of the forcing function term, AQC, contain the net influence of thechemical species change (due to convection/dif fusion/) inetics) on the fluiddynamic solution through pressure/species derivatives. The inclusion of this termon the right hand side is not essential to obtain a converged steady-statesolution and it can be discarded. However, it cannot be discarded for timeaccurate computations. For the same reason, it cannot be Miscarded for a spacemarching PNS computation unless time-iterative methodology is incorporated.

NUNENICAL NITNODOLOGY IN PARCH - All work with PARCH has focused on the analysisof steady flow problems using non-time accurate procedures to expedite convergenceto steady-state. The block unfactored, matrix-split, fluid dynamic equatioas(equation 4), with the explicit species term, aQ¢, removed, are approximatelyfactored as follows:

II * At& fA) I11 * At& S~ttGN - ( N4 it 9 f1 AQ

Equation (5) can be solved by block tridiagonai inversion, which is computa-tionally expensive per time step, but generally permits taking large time steps(e.g., Courant numbers of 5 - 10) and obtaining convarged solutions in a veryreasonable number of iterations (e.g., 300-3,000, depending on the problem athand, the grid, the initial and boundary conditions, etc.). Equa* ion (5) can moreefficiently be solved by using a diagonalized scheme which uncouples the blocksystem and reduces the work to the inversion of a scalar tridiagonal system.However, the implicit diagonalized solution is restricted to the Euler equations(the viscous terms do not diagonalize and must thus be treated explicicly), and,the path to convergence for compley, viscous dominated flows can be slow andsometimes problematic.

258

The choice of the blocked tridiagonal inversion or diagonalized solutionprocedure for the fluid dynamic equations is available as a user option in PARCH

and is problem dependent. With multi-zone versions, ta:' .- 5%gonalized option can

be used in some zones and the blocked option in others (e.g., for a missile/plumeinteractive problem, the zone of strongly interactive flow can be handled by theblock procedure with all other flow zones analyzable by the generally moreefficient diagonalized method). With central difference numerics employed,artificial dissipation is required in nonviscous regions to ensure stability anddiagonal dominance. The implicit/explicit second and fourth order dissipationmodel of Jameson, et al", is employed in PARCH. Independent time steps are usedto advance the equations based on a user-specified Courant number (locally appliedat each grid point) with a flux change limiter (e.g., 20 percent change in AQ/Q),which cuts back on the local time step in regions of severe change.

Chemical Species/Rate Kinetics Algorithm - Upon matrix splitting and decompositionof equation (3), the lower half contains the chemical species transport equations,written as:

N N[_ tN]SS NAt($ Acc * &•Bcc KAQc Rc (6)

- Acf * a tc.)AGf

The second term on the right hand side contains the forcing functions fromthe fluids upon the species, which is discarded for time-asymptotic steady-statesolutions. Subsequent factorization leads to:

[(I-AtMN) + At&_AN I [(I-AtMN)

+ At&BN AQ - N (7)i tB ccAQ - Rc

Equation (7) represents the numerically intensive task of inverting block NxNtridiagonal matrices. CPU costs and memory requirement can become prohibitive asthe number of chemical species get large. An efficient alternative solutionstrategy has been devised, which breaks up the solution sequence into two steps:

(1) a point implicit solution of the chemical kinetic rate equations toyield &,; and,

(2) a globally implicit time integration of the species equationssequentially with the chemical source term specified from step (1).

Step (1) requires the inversion of an NxN matrix at each grid point. Recentnumerical studies have indicated that this computation need not be performed ateach time-integration. For rocket nozzle flows, performing step (1) once forevery 50 fluid dynamic time steps, and holding the source term constant at theintermediate time steps has yielded the same converged solution obtained byperforming step (1) for every integration step, with no convergence penalties.Hence, this splitting can be extremely cost effective for chemically reactingflows with a significant number of chemical species.

MULTI-PHASE FLOW CAPABILITIES - The PARCH code contains multi-phase flowcapabilities for the simulation of solid propellant rocket nozzle/exhaust plumeflowfields where the dilute particle assumption applies and eliminates particlevolumetric effects45 . Gas/particle interactions can be treated in both theequilibrium limit (where particle velocities and temperatures are taken to be thesame as that of the gas-phase) and the nonequilibrium limit (where particlevelocities and temperatures differ from those of the gas-phase). In thenonequilibrium limit, the analysis is presently restricted to flows with a primarystream wise direction where the particulate equations can be spatially integrated.The particulate equations, cast in strong conservation form in generalizedcurvilinear coordinates, are listed in Table 10 for two dimensions.

3est Available Copy 259

TABLE 10PARCH CODE PARTICULATE EQUATIONS

a÷ 7. -H l[p.= I1 1PP u p up Pr up .y HP WPv ptp- c

P hp up Lpp hp V hh - Pp (hp h)lvh

Particulate equations are solved for different particulate types (e.g., Al03)and for several representative sizes (e.g., 1 gm, 3 um, 5 lug,...). Thenomenclature to designate types and sizes has been eliminated for simplicity. Inthe above equations, u. and v. are the Cartesian particulate velocity componentsin the x and y directions and U. and V. are the contravariant velocity components.Tu, Y , and Y. represent characteristic particle times for velocity and thermalequilibration. Details of the predictor-corrector based particle space marchingalgorithm utilized is provided in reference 46. Coupling between the gas andsolid phases is provided through the gas/particle interaction source terms in boththe gas-phase and particulate equations. The particle equation integrationprocedure has recently been improved by a fully-implicit treatment of the sourceterms (which can be stiff for small radius particulates which are nearPquilibrium), including special methodology for dealing with phase-change effects(viz, solidification in the nozzle/plume expansion process). In nozzle/plumeapplications, the particle solution is typically updated (by marching down thelength of the nozzle or the exhaust plume) every 50 iterations of the gas-phasesolution. The highly efficient particle spatial marching technique is, of course,limited to flows where there are no recirculating features. A new, time-accurate, 3D particle solver is under development for use in CRAFT, which willbe "efficiently" implemented in PARCH for time-asymptotic, recirculating flowregions, via straightforward modifications.

GENERALIZID BOUNDARY TREATMENT USING PATC•ING - The discussion here followsdirectly from reference 17. The AEDC PARC code was designed so that specificationof boundaries and boundary conditions is done entirely through inputs to theprogram. Any portion of any grid line may be designated as a boundary. (For 3-Dthis would be any portion of any surface - for economy of exposition, the 2-D casewill be discussed). Thus flow problems of arbitrary geometric complexity can besimulated as long as satisfactory grids can be generated for them. In addition,the PARC code is so structured that no noticeable performance penalty is paid forthis capability. These features make the use of the PARC code very flexible andeconomical and they have been retained "intact" in PARCH, with upgrades to the BCfor chemistry, particulates and advanced turbulence models.

Figure 1 (from reference 17) illustrates the boundary condition treatment inPARC for an idealized ramjet engine conafiguration. Note that the entireconfiguration is contained within a single grid. Also notice that the cowlsurfaces are entirely positioned on internal grid lines and that the grid cellswithin the forebody are not part of the flow field. The particular geometry ofthis flow problem must be communicated to the PARC code. This is done byconstructing a list of the grid surfaces which are also boundary surfaces asschematized in the table included in Figure 1. Notice that, within each indexclass, the order of the boundary segments is not important. Ales, note that theccwl tips do not need to be specified (boundary edcges are automatically recognizejas such). Boundary specification for more complex configurations is as simple asthis example, requiring only the delineation of additional boundary segments.Three-d,.mensional specifications can be difficult, but only; because there does not(yet) exist an1 easy way to visualize the interrelationships ot 3-D grid surface!;.

The PARC code contains a special "patchinq" algorithm which logicallydecomposes a grid containing internal boundaries into a family of patches. Thistechnique allows the flow simulation to be done nearly as efficiently as for aflow problem without internal boundaries. This patching technique is especiallyeffective for ADI-based schemes, such as the Beam and warming algorithm used inthe PARC code. The basic ccncepts involved are illustrated in Figure 2 whichcontinues the sample ramjet problem discussed above. Notice that there areactually two families of patches, J-patches and K-patches. The FOR1hAY DO loopsin the PARC code are structured so that the appropriate family of patches are usedto set the loop limits depending on the calculation to be performed. For example,the pentadiagonal algorithm requires matrix inversions for each grid linecontained within the flow; the PARC code accommodates this requirement by usingthe appropri.aite family of patches depending on the sweep direction of the matrixinversion algorithm. Use of the patching concept is practically transparent tothe use of the PARC code as the patches are generated automatically based solelyupon the boundary specifications.

PAJCE/RN ROCKET NOZZLE VERSION OF PARCH - The details of this version aredescribed in references 21 and 24. PARCH/RN was configured as a "user-friendly"version of PARCH and its structure is schematized in Figure 3. A preprocessoraccepts user input defining the chemical system and nozzle wall geometry. Allwork in defining the thermochemistry (e.g., selecting species, reactions,thermodata and kinetic data from generalized data bank), the grid, and the initialflow solution, is performed by the preprocessor, permitting this code to be runtotally in a "black box" manner.

The diagonalized version of the Beam-Warming algorithm is used to integratethe fluid dynamic equations; the finite-rate chemistry is solved in a looselycoupled/matrix-split manner; and an algebraic turbulence model is used for thewall boundary layer. This version contains detailed methodology for dealing withnonequilibrium multi-phase flow and can concurrently treat different types ofparticulates, each type distributed into a range of size bins. The code can dealwith an arbitrary number of chemical species and reactions (via use of variabledimension statements), and is presently configured to treat up to 5 particletypes, with 5 size group allocated for each type.

The "brute-force" solution of 4 fluid dynamic, NS-l chemical species, and upto 25 particulate equations can involve very substantive CPU time. A practicalapproach for dealing with this problem has involved using a local "steady-state"method for dealing with the nonequilibrium processes. Via this approach, theparticulate equations need be integrated "spatially" only once, for each 50 time-iterations of the fluid dynamic equations. With the "loosely-coupled" pointimplicit solution of the chemistry, the species equations are solved concurrentlywith the fluid dynamic equations at every time-step (since theconvection/diffusion of chemical species has a strong influence o.. the local wavespeed), but the chemical source term, 6,, is held fixed and updated every 50 time-steps. The determination of the chemical source term at each grid point must beperformed implicitly (for stability) and involves the very costly inversion of an(NS-l) by (NS-l) vatrix at each point. Being able to perform this matrixinversion (and the integration of the particle equations) every 50 fluid dynamicsteps, rather than every time-step, has provided very significant time saving,viz., if the solution converges in 500 steps, only 10 chemistry matrix inversionsand 10 particulate sweeps are required.

In performing the chemistry and partic-ilate solutions, the local convectivetime (1/at - u/ax + v/Ay) is employed, not. the numerical "unsteady" time-step,which is nonphysical. Then, the nonequilibrium processes see a sequence oflocally "steady" flows and their accurate solution thus depends on having the gridsize fine-enough to resolve details of the nonequilibrium scales. Via the fullyimplicit treatment of both chemical and particulate source terms, grids which areadequate for resolving fluid dynamic processes have been adequate for providingstable and resolved solutions of the chemical/multi-phase nonequilibriumprocesses. Details of the new chemical source term "decoupling" approach and thenew implicit treatment of the particulate source term is given in reference 47.To illustrate PARCH/M capabilities, several representative test calculations areexhibited. Figure 4 exhibits the prediction of a solid propellant nozzle flowwith and without particulates to show the influence of particles on the qas-Thasestructure. The calculation with particles was performed jith a 30% mass loadingof 3pm, A12 03 particulates. The particles separate from the wa'l in the throatregion and are bounded by a limiting particle streamline which is "captured" using

261

the specialized filtering approach discussed in reference 46. The flow withoutparticulates exhibits a significant shock in the supersonic portion while the flowwith particulates suppresses this shock. Figure 5 exhibits a comparableprediction made with 3 particle size groups.

PARCR/IT G ?UR ZMU TAILPIP3 VMIZON O PARCS - This version is being developedas a "user-friendly" component to upgrade the SPIRITS aircraft signature code".Preliminary work performed to date is described in reference 26. PARCH/GT iscoupled with a thermal balance code to yield tailpipefnozzle surface temperatures,and serves to provide startline conditions for aircraft plume codes. The numericsin PARCH/GT are quite different than those in PARCH/Ri. In PARCH/GT, the dominantprocess is the internal mixing of the core and fan (bypass) flows. This analysisrequires the use of a ke turbulence model, and since free shear layers can hequite unstable, more robust numerics. PARCH/GT solves 7 strongly-coupled fluiddynamic/turbulence model equations [the flow is presently treated with anaxisymmetric swirl equation), which are integrated using the more robust blocktridiagonal matrix inversion procedure. To make PARCH/GT "user-friendly",significant pre-processor work was performed as summarized in Table 11.

Table '11

PARCH/GT PREPROCESSOR

e GEOMETRY 9NPUT/GRID GENERATION

- SMOOTH GRID CONSTRUCTED USiNG ALGEBRAIC SCHEMES

- GRID POINTS BLANED OUT IN REGIONS OF PWUG/SPTflTER

9 NTIAL FLOWFIED SOLUTION CONSTRUCTED

- USES CYCLE DECK INFORMATION (TOTAL CONDITIONS)

- ASSUMES SONIC CONDITIONS AT GEOMETRIC THROAT

- GENERATES 1D ISENTROPIC SOLUTION wrrH MASS AVERAGED PRESSURE AT

EACH STATION

- SHEAR LAYER BETWEEN CORE/BYPASS STREAMS GENERATED USING MIXING LENGTH

TURBULENCE MODEL AND INCOMPRESSIBLE SHEAR LAYER SPREADING MODEL

Figure 6 schematizes a typical turbofan tailpipe flow field domain and the mappedcomputational domain used to integrate the equations. Figure 7 shows the flowfield initialization and the PARCH/GT converged solution. Figure 8 shows thenumerical instabilities in the shear layer that occurred in earlier calculationsusing the diagonalized algorithm in the original version of PARC. Figure 9 showsthe baseline grid and and improved shear layer adaptive grids obtained using theSAGE code's. Detailed comparativo studies with AEDC data for an F-1O0 engine havebeen performed which show a strong dependence of the predictions on inflowparameters that were not initially made available (e.g., profile detail a/splitterplate boundary layers, core/fan turbulence levels, etc.).

PARCH/TWE TACTICAL MIBRILN/PLUXE INTERACTION VERSION OF PARCH - This 3D version,developed for MICOM, has been applied to the analysis of varied tactical missileconfigurations with conventional and unconventional propulsive systems (e.g.,scarfed nozzles). Complete details are given in references 5 and 23. A typicalscarfed nozzle missile/plume application is schematized in Table 12.

262

TABLE 12PARCHfTMP ANALYSIS OF TACTICAL MISSILE AIRFRAME/

PLUME INTERACTIONS WITH BIFURCATED/SCARFED NOZZLES

"* COMPLETE FLMOFWD A3? bWrrH SIabE GMO/BLMNFOR BASE RB JO N m mOmn M 1 g

"* LIUO PROPEXLX4T (AMWE)CHEMISTRY

"* Ke TUFRLJUENM MOOEL

THPlOUGHOUT

"* RAJM AN PROTWXWGNOZZLES

Y

Fy •$SS,1e ax;s

This code has various algorithm options, but has primarily been utilized with theoriginal diagonalized Bean-Warninq algoriths. The code has provisions for finite-rate chemistry and equilibrated particulates, and uses the ke turbulence modelwith a Chien low Re extension for the near wall analysis. The species andturbulence model equations are integrated in a loosely-coupled Banner. PARCH//RPprovides initial conditions at the nozzle exit planes which serve as inflowboundary conditions on the missile surface (for flush scarfed nozzles), -r onangled surfaces protruding from the missile body (for non-flush scarfed nozzles).Figure 10 depicts temperature, turbulent kinetic energy and velocity contours fora generic missile (sans wings/fins) with a flush scarfed nozzle, flying at Mach.6 at sea level. The plume was particle free and had amine chemistry. Yigures 11and 12 exhibit Mach number and turbulent kinetic energy contours for this sameconfiguration at an angle-of-attack of 108. Figure 13 exhibits cross-flowcontours at the base and just downstream showing the complex interaction of thetwo plumes, the misaile boundary layer and its downstream wake.

Validation studies are in progress using Aero Spatiale dat4e" as reported inreference 50. Figure 14 shows the experimental set-up for Mach 2 flow over ageneric missile interacting with a Mach 2.5 isoenerg•tic air jet inclined at 70.Figure 15 shows the grid utilized in a vertical plane passing through the centerof the jet and the missile axis, along with predicted Mach number contoursobtained using the Roe/TVD run option. Figure 16 compares predictions with cross-flow pressure data at an axial location 6.5 missile diameters downstream of thejet--the predictions are seen to be in oxcellent agreement with this data.Further details, turbulent predictions, and other cases are described in reference50.

OT•tU VSzIOlS Srp PARCe - PARCH has been applied to many other propulsive-relatedflow field problems. Scramjet studies were performed for inlet/combustor/norztzeflows as described in references 19, 20, 22 and 24. [Specialized scranjetpropulsive versions of PARCH were never developed.] A laser version, PARCUI/LY,has been applied to low Re laser nozzle/cavity flow fields as described inreference 51. A jet research version, PARCH/JR, is under development for cotoplax2D and 3D jet flow field studies (under NASA/LPC and NSWC support), as will bedescribed in reference 52.

2',

OVERVIEW OF SCHAFT (PNS) AND CRAFT (FNS) CODES

9M To D&T - As per earlier discussions, BCHArT and CRAFT are extendedpropulsive-oriontId versions of the STUFF and TUFF aerodynamic codes develop" byHolvik and Nerkle11. STUFF in a thin-layer, iterative Poo code whose equations arediscretised using finite-volu"e methodoloqy, and integrated using upwind *toe/1"JDntmarios. TUTF In a comparable thin-layer FNS code. The upwind Inviocid fluxmeare constructed employing a nov temporal Plieann solver that properly acoounts forthe real gas behavior of a uulti-oomipnent gas mixture. The fluid and speeles"equations are strongly coupled in STUFF and TUFF, and thu solution Is Rads fullyimplicit by linearising all oonveotive/diffusive flux terme as wall as thechemical mour• s term 4,. A modified Newton iteration is employed to eliminatelinearization and approximate factorization errors in the Iterative process, The:ublayar procedure of Vigneron is employed in sTurF to eliminate depertureSolution in marching through thin subsonic regions. The STUFF and lurp codes

contain clean air finite-rate and equilibriu& chemistry, and laminar tshasportdata.

With the well founded basis in SlTrU and TUrF, the extensions performed bythe authors and coworkers Jn developing SCHAIT and CRAMT have includedi

(1) inclusion of the Cull cross-flow stress terme, as sequired for anly|iny3D propulsive problems and plusesi

(2) generalitation of th. boundary oonditions for plume/propulsive prublems(e.g., inclusion of a subsonic *ntrainment boundary to permit theanalysis of the jet into s.till air Irbloe)l

(3) generalitation or thi ohemtsLry, theradyemamlu and trnuiopurl date #,?provi-.e for iha analysis of HI/C//O systemsI and,

(4) inoll,,ion of two-equation turbulence &odels (Ie, kh-oo, kW) end low tovrliats (e.,q,, the ChIan eytonulnn of Iq) in * atr-nrly-clamIe soifnnetimpltmentirvi full lineerilstion of conveotivn/dirfuelv%/svule:a term*,

The finite volume equations in the SCHAIID code are ilvan in Table 1) blow,SCHAF? and Cs.I'T Integrate a system of & (fluid) 4 HS - I (OPOeUis) * )(turbulence) equations in a fully-coupled mainer with eli auurum talme ri1eLedfully-implicitly.

TABLE 13SCHAFT3D EQUATIONS

1f1SOdor ' JJA(Ei14Jir * JAd 0¢Ifdi

d4i

where S * il (0, On) 1 $t (0, Or) I FULL PAJiA5I IfIoSTnti*l/OII'UIVE 41& ;T, Tt7 (Q. 0) ,T2 (0, On)

"LOIIIINAL STUFF COO( HAO ONLY $1 U01s4]

Q * Co. iU. Y, pw , *t, Po i. dt . .... . Dp$ .I , Pb, Pij1

I FLUIO Ni • I C)TMICAL 7 luIsUeviI .I

S.l ViIAIASLIS 4P01011 VAbIAJIA

0 a CoA.•,,0,o, i. ..... , W .1. Ob. oil

ci(Mlfml Y IQU•CCf IUgi'JL|N( I

0 k. 0d, LOW toe C01ll4 lOm Of kr f(Ip0 A! IlOi

ALl. SOUICC INI - LIhIAiHMtO/?ULLY IN1PI.II INIAIMNIN

7M.

A major upgrade to CRAFT has involved the inclusion of dynamic gridding. Theobjective has been to develop a capability for handling problems with movingboundaries and/or to have the grid adapt to time-evolving features in a time-accurate manner. The approach taken employs an integral formulation which isconsistent with the finite-volume numerics in CRAFT. The finite-volumemethodology for moving grids entails:

(1) dividing the flow into cells whose volumes vary with time; and,

(2) formulating general conservation laws (equation 8) for each cell

t2 -. . t2

j QdV-f QdV + f f n FdSdt - JJPdVdt (8)t' t' t' S(t) t' V(t)

TERM I TERM II TERM III

where:

TERM 1: PRIMARY CONSERVED VARIABLESTERM II: CONVECTIVE FLUXTERM III: SOURCE TERMSV(t: TIME VARYING VOLUME OF CELL

In the calculation methodology:(1) we treat the grid motion as a purely geometric quantity, i.e. definition

of contravariant velocity remains unchanged

(2) we evaluate Term I as:

(Q.+l _ Q") V"W' + Qn (VW÷' - Vr)

(3) we evaluate the convective flux (Term II) by assuming geometric (grid)quantities as being held constant at time averaged value; and

-- Fn"+' SVtWHEREFn = n FandS IS THE TIME AVERAGED METRIC.

(4) the source terms (Term III) are evaluated at the new level

- p" '+1 V .+'

Then, the numerical algorithm takes the form of equation (9)

.(Qan+l Qn) vn+1 + +7(vn+ _ Vn) Fnn+ 1 SAtAt At (9)

= pnf+l Vn+l

where we: evaluate Fn as third order upwind-biased flux; linearize in time; solveimplicit operator with an ADI procedure; and use Newton iteration for higher orderaccuracy in time. Complete details of the features and numerics in SCHAFT andCRAFT are given in reference 53. est Available Cop

S~265

8CAFT NUMINCAL AND VALIDATION STUDIES - The studies performed to date have beenquite extensive and have included Level 1 studies comparing SCHAFT with theearlier SCRINT and PARCH codes for a variety of propulsive and plume flowfields.Many of these earlier studies were described in references 38, 39, 41 and 42. Atypical comparative study is exhibited in Figure 17 for a square converging-diverging nozzle where the SCHAFT and SCRINT3DT comparisons are seen to be quitecomparable. A more severe case is schematized in Figure 18 for the combustion ofmultiple H2 jet in a mildly diverging duct. The SCRINT3DT code with either Beam-Warming or Roe/TVD upwind numerics would not run for these severe conditions.PARCH code calculations for this case were described earlier. SCHAFT had noproblem starting from top-hat initial profiles. Figure 19 exhibits contours oftemperature and HO in a vertical plane passing through the middle of the inner H2jet. Figure 20 exhibits contours of these variables in a horizontal plane passingthrough the middle of the H2 jet. Figure 21 shows cross-flow contours oftemperature and H20 in a plane 5 duct heights downstream of the inflow plane.These calculations were performed utilizing the ke turbulence model and 7species/8 reaction formulation of the H/0 finite-rate chemistry (with N2 inert).

At Level 2, simple validation studies to check out the turbulence m '..ls andchemistry were performed for boundary layers, shear layers, etc. These studiesare described in reference 54. A typical Level 3 study using the data of Stalker(Figure 22) was performed and compared with the SPARK code prediction (Figure 23)described in reference 56. Figure 24 shows SCHAFT predIctions with (solid line)and without (dashed line) finite-rate chemistry. Figures 25-27 exhibit pressure,temperature and H 0 contours - ignition occurs when the underexpansion shockreflected off the lower surface impinges on the shear layer. Further details andfurther comparative studies are given in reference 54.

CRAFT NUMERICAL AND VALIDATION 8TUDIES - The Level 1 studies performed to datehave been quite limited primarily involving ID shock tube studies with fixed andmoving pistons to check out the dynamic gridding, and the related 2D shock-tube-like problems of under expanded jets impulsively discharging into a duct withfixed and moving end-walls. In addition, fundamental studies of steady-stateunderexpanded jets are being performed and compared with earlier SCIPVIS57 PNSpredictions (see reference 52), and, time-accurate studies of short durationtransverse jet interactions are in progress. Several of these studies will beexhibited below. Figure 28 exhibits pressure variations with time for the 1Dshock tube problem started impulsively with a 1000/1 pressure ratio. Figure 29exhibits pressure contours at three distinct times for an air jet discharged intoa duct with a moveable end-wall. The calculation was an air/a4r case with a 5/1pressure ratio at t - o. Figure 30 exhibits pressure and H20 contours for thissame problem performed for H, into air with a fixed end-wall. This laminar casewas performed with 7 species/8 reaction finite-rate H/0 chemistry (N2 in duet wasinert). The final calculation is that of an underexpanded Mach 2 Jet exhaustinginto a Mach .25 uniform external stream. Predicted density contours are exhibitedin Figure 31. A comparison of CRUFT centerline pressures with those of SCIPVISare shown in Figure 32. These calculations were laminar--comparative studies withvarious turbulence models are in progress. Details of numerical studies withCRAFT are described in reference 55.

CONCLUDING REMARKS

Capabilities of the PARCH FNS code and the newer SCHAFT (PNS)/CRAFT (FNS)codes were described. PARCH is an extension of the PARC propulsive solverdeveloped at AEDC which contains "unique" capabilities with regard to boundarycondition flexibility via "patched" methodology. On extended versions containfinite-rate chemistry, advanced turbulence models, multi-phase flow capabilitiesand problem-specific/user-friendly specialization. These extensions havepermitted the analysis of extremely complex propulsive flow problems by the non-expert user. The areas of concern are currently gridding and turbulence. Withmodern grid generation packages and new solution adaptive methodology--the abilityto generate adequate grids is under control. The turbulence arena is stillproblematic with some thoughts on current abilities and approaches for validationgiven in references 58 and 59.

SCHAFT and CRAFT are new tools and will require Level 1 validation,specialization to specific problem areas, and further validation at Levels 2-4before being used in an engineering environment. The technology in these codesis truly state-of-the-art and very significant progress has been made in arelatively short period of time to evolve these codes from their original researchstatus, to predictive tools with demonstrated reliability for steady and unsteady

266 propulsive flows.

Best Available Copy

REFERENCES

IDash, S.M., "Scramjet Technology Assessment, Volume II - CFD Assessment forScramJet Propulsion Flowfield Analysis", AFWAL-TR-88-2015, Vol. II, Aero Prop.Lab, WRDC, WPAFB, Ohio, Kay 1989.2Edelman, R.B., Kollrack, R., Dash, S.M. and Loomis, W., "Scramjet TechnologyAssessment," Third NASP TechnolocM Symposium, NASP CP-3020, June 1987, pp. 185-198.3Edelman, R.B., Kollrack, R., Dash, S.M. and Loomis, W., "Scramjet TechnologyAssessment," 1987 JANNAF Propulsion Meeting, CPIA Pub. 480, Vol. IV, December1987, pp. 221-239.4Dash, S.M., "Turbulence Modeling, Chemical Kinetics and Algorithm Related Issuesin CFD Analysis of ScramJet Components," Second NASP Technologv Symnosium. HASPCP-2012. Nov. 1986. p.i 39--80.5Dash, S.M., Sinha, N., and York, B.J., "Computational Models with AdvancedThermochemistry for the Analysis of Missile/Plume Flowfield Interactions," AGADSvmoosium on Missile Aerodynamics. Friedrichshafen, Germany, April 23-26, 1990.OChilds, R.E. and Caruso, S.C., "On the Accuracy of Turbulent Base FlowPredictions," AIAA Paper 87-1439, June 1987.7Pulliam, T.H., "Euler and Thin-Layer Navier-Stokes Codes: ARC2D, ARC3D," Notes forComputational Fluid Dynamics User's Workshop, The University of Tennessee SpaceInstitute, Tullahoma, TN, UTSI Pub. E02-4005-023-84, March 1984, pp. 14.1 -15.85.sCooper, G.K., "The PARC Code: Theory and Usage," AEDC-TR-87-24, October 1987.9Kaul, U.K. and Chaussee, D.S., "AFWAL PNS Code: 1983 AFWAL/NASA Merged BaselineVersion," AFWAL-TR-83-3118, AF Wright Aero. Labs., WPAFB, OH, May 1984."1°Pulliam, T.H. and Steger, J.L., "Recent Improvements in Efficiency, Accuracy, andConvergence for Implicit, Approximate Factorization Algorithms," AIAA Paper 85-0360, Reno, NV, January 1985."1Dash, S.M., Sinha, N. and York, B.J., "Matrix Split Approach for Inclusion of SPFThermochemical Capabilities Into Beam-Warming Based PNS and NS Models, "JANNAF17th Plume Technoloay Meeting, CPIA Pub. 487, Vol. I, April 1988, pp. 91-110."Sinha, N., Krawczyk, W. J. and Dash, S.M., "Inclusion of Chemical Kinetics intoRoe Upwind/Beam-Warming PNS Models for Hypersonic Propulsion Applications," AIAAPaper 87-1898, San Diego, CA, June-July 1987.'13 olvik, G.A. & Merkle, C.L., " A Set of Strongly Coupled Upwind Algorithms forComputing Flows in Chemical Nonequilibrium," AIAA Paper 89-0199, Reno, NV, January1989.14Liu, Y. and Vinokur, M., "Nonequilibrium Flow Computations: I - An Analysis ofNumerical Formulations of Conservation Laws," NASA CR 177489, June 1988."Montagne, J.L., Yee, H.C. and Vinokur, M., "Comparative Study of High ResolutionShock Capturing Schemes for Real Gas," NASA TM 100004, 1987.' 6Grossman, B., Cinnella, P. and Garrett, J., "A Survey of Upwind Methods for Flowswith Equilibrium and Nonequilibrium Chemistry and Thermodynamics," AIAA Paper 89-1653, June 1989."Cooper, G. and Sirbaugh, J., "The PARC Distinction: A Practical Flow Simulator",AIAA Paper 90-2002, June 1990.laDavies, C. and Venkatapathy, E., "A Simplified Self-Adaptive Grid Method, SAGE,"NASA TM 102198, October 1989."19York, B.J., Sinha, N. and Dash, S.M., "Computational Models for Chemically-Reacting Hypersonic Flows," AIAA Paper 88-0509, Reno, NV, January 1988."2 0Sinha, N., York, B.J. and Dash, S.M., "Applications of a Generalized ImplicitNavier-Stokes Code, PARCH, to Supersonic and Hypersonic Propulsive Flowfields,"AIAA Paper 88-3278, Boston, MA, July 1988.2iYork, B.J., Sinha, N., Ong, C.C. and Dash, S.M., "PARCH Navier-StokesReacting/Multi-Phase Analysis of Generalized Nozzle Flowfields," AIAA Paper 89-1765, Buffalo, NY, June 1989.22Sinha, N., York, B.J., Ong, C.C., Stowell, G.M. and Dash, S.M., "3D Navier-StokesAnalysis of High-Speed Propulsive Flowfields Using the PARCH Code," AIAA Paper 89-2796, Monterey, CA, July 1989.•Dash, S.M., Sinha, N., York, B.J. and Ong, C.C., "3D Navier-Stokes Analysis ofTactical Missile External/Plume Interaction Flowfields," JANNAF 18th PlumeTechnologv Meeting, Monterey, CA, November 1989.14York, B.J., Sinha, N. and Dash, S.M., "PARCH Navier-Stokes Code Analysis ofRocket and Airbreathing Nozzle/Propulsive Flowfields," JANNAF 18th PlumeTechnologv Meeting, Monterey, CA, November 1989."Dash, S.M., Sinha, N. and York, B.J., "Computational Models with AdvancedThermochemistry for the Analysis of Missile/Plume Flowfield Interactions," AGARDSvmDosium on Missile Aerodynamics, Friedrichshafen, Germany, April 23-26, 1990.' 6Dash, S.M., Sinha, N., York, B.J. and Kenzakowski, D.C., "Turbine EngineFlowfield and Hot Parts Module Upgrade", SPIRITS Users Group Meetinci, HanscombAFB, MA, June 1990 (article Aerodyne Research, Inc. RR-788, Billerica, MA)

Best Av&-ibki Gý, 267

"2 7Sinha, N. & Dash, S.M., "Parabolized Navier-Stokes analysis of Ducted SupersonicCombustion Problems," J. Propulsion, Vol. 3, No. 5, Sept.-Oct. 1987, pp.455-464.28Sinha, N. & Dash, S.M., "Analysis of Aerospace Vehicle Scramjet PropulsiveFlowfields: 2D Combustor code Development - Phase I," NASP CR-1013, April 1988.29Wolf, D.E., Lee, R.A., & Dash, S. M., "Analysis of Aerospace Vehicle ScramjetPropulsive Flowfields: 2D-Nozzle Code Development-Phase I," NASP CR-1005, Aug.1987.3Wolf, D.E. Lee, R.A., Sinha, N. & Dash, S.M., "2D & 3D Euler/PNS Analysis ofScramjet and Rocket Nozzle/Exhaust Plume Flowfields Using the SCHNOZ Code," JANNAF17th Plume Technoloav Meetina, CPIA Pub. 487, Vol. I, April 1988, pp. 9-52.31Lawrence, S.L., Tannehill, J.C. and Chaussee, D.S., "An Upwind Algorithm for theParabolized Navier-Stokes Equations," AIAA Paper 86-1117, May 1986.32Lawrence, S.L., Tannehill, J.C. and Chaussee, D.S., "Application of an UpwindAlgorithm to the Three-Dimensional Parabolized Navier-Stokes Equations," AIAAPaper 87-1112, June 1987.33Prabhu, D.K., Tannehill, J.C. and Marvin, J.G., "A New PNS Code for ChemicalNonequilibrium Flows," AIAA Journal, July 1988, pp. 808-815.34Buelow, P., Tannehill, J.C., Levalt, J. and Lawrence, S.L., "A Three-DimensionalUpwind Parabolized Navier-Stokes Code for Chemically Reacting Flows," AIAA Paper90-0394, January 1990.35Krawczyk, W.J., Rajendran, N., Harris, T.B., York, B.J. and Dash, S.M.,"Computational Models for the Analysis/Design of Hypersonic Scramjet Components,Part II: Inlet and Ramp/Forebody Models," AIAA Paper 86-1596, Huntsville, AL, June1986.36Sinha, N., Krawczyk, W.J. and Dash, S.M., "Inclusion of Chemical Kinetics intoBeam-Warming PNS Models for Hypersonic Propulsion Applications," AIAA Paper 87-1898, San Diego, CA, June-July 1987.3'Lee, R.A., Sinha, N. and Dash, S.M., "PNS Code Assessment Studies for ScramjetCombustor and Nozzle Flowfields," AIAA Paper 89-1827, Buffalo, NY, June 1989.3Dash, S.M., Sinha, N. and Lee, R.A., "Time-Iterative Upwind/Implicit 3D PNS Codesfor the Analysis of Chemically-Reacting Plume/Propulsive Flowfields," JANNAF 18thPlume Technoloav Meeting, November 1989.1"Dash, S.M., Harris, T.B., Krawczyk, W.J., Rajendran, N., Sinha, N., York, B.J.and Carlson, D., "Three-Dimensional Upwind/Implicit PNS Computer Codes forAnalysis of Scramjet Propulsive Flowfields," Sixth NASP Technology Svmposium, NASPCP 6035, April 1989, pp. 129-172.40 Krawczyk, W.J., Harris, T.B., Rajendran, N. and Carlson, D., "Progress in theDevelopment of Parabolized Navier-Stokes Methodology for External and InternalSupersonic Flows," AIAA Paper 89-1828, Buffalo, NY, June 1989.41Sinha, N. and Dash, S.M., "Implicit/Upwind 3D PNS Scramjet Propulsive FlowfieldCode: SCHAFT3D," Seventh NASP TechnoloaV SvmDosium, NASP CP 7041, October 1989,pp" 151-180.

Sinha, N., Dash, S.M. & Lee, R.A., "3D PNS Analysis of Scramjet Combustor/Nozzle& Exhaust Plume Flowfields," AIAA Paper 90-0094, Reno, NV, Jan. 1990.' 3Dash, S.M., "Advanced Computational Models for Analyzing High Speed PropulsiveFlowfields," 1990 JANNAF Propulsion Meeting, Anaheim, CA, October 1990."Jameson, A., Schmidt, W. & Turkel, E., "Numerical Solutions of Euler Equationsby Finite-Volume Methods Using Runge-Kutta Time Stepping Schemes," AIAA Paper 81-1259, June 1981."5Dash, S.M., Wolf, D.E., Beddini, R.A. & Pergament, H.S., "Analysis of Two-PhaseFlow Processes in Rocket Exhaust Plumes," Journal of Spacecraft and Rockets, Vol.22, May-June 1985, pp. 367-380."4Dash, S.M. & Thorpe, R.D., "Shock Capturing Model for One & Two-Phase SupersonicExhaust Flow," AIAA Journal, Vol. 19, July 1981, pp. 842-851.47York, B.J., Sinha, N., & Dash, S.M., "Upgrades to the PARCH/RN Rocket Nozzle Codefor Treating Chemical & Multi-Phase Nonequilibrium Processes", report inpereparation for MICOM, Sys. Simulation & Development Dir., Redstone Arsenal, AL.

Anon, Description of the SPIRITS IR Model, Aerodyne Research, Inc., RR-480,Billerica, MA, May 1987.49Dormieux, M. and Mahe, C., "Calculs Tridimensionnels de L'Interaction d'un JetLateral Avec in Ecoulement Supersonic Extenne", Validation of Computational FluidDynamics, AGARD Conf. Proceedings 437, Lisbon, Portugal, May 1988.nYork, B.J. Sinha, N. & Dash, S.M., "PARCH/3D Computations of Inclined JetInteractions with an Ogive/Cylinder Airframe Flowfield", SAIC/FW TM-61, preparedfor Aerospatiale-ESC/AT, Sept. 1990.5ISinha, N. "Analysis of Chemical Laser/Cavity Flowfields using PARCH'LF Code, inpreparation for MICOM Directed Energy Directorate (Overtone Chemical LaserProgram), Redstone Arsenal, AL.52Dash, S.M., Sinha, N., York, B.J., Lee, R.A., & Hosangadi, A., "Progress in theDevelopment of Advanced Computational Models for the Analysis of GeneralizedSupersonic Jet Flowfields," AIAA Paper 993915 C

26•

"•Sina, N. & Dach, S.M. "Advanced Computational Models for Hypersonic PropulsionApplicationg: Vol. I-SCrAFT (PNS) & CRAFT (FNS) Computer Codes", HASP CR in9reparation.

Lee, i.h., Sinha, N., & Do.5h, S.H., "Advanced Computational Models for HypersonicPropulsion Applications: Vol. 1I-SCHAFT PNS Code Numerical & Validation Studies",HASP CR in preparation."SSHosangadi, A., Sinha, N. & Dash, S.M., "Advanced Computational Models forHypersouiic Propulsion Applications: Vol. III-CRAFT FNS Code Numericai andValidation Studies", HASP Ck in preparation."Dfrumn-.nd, J. P. and Rogers. R.C., "Summary of Workshop on Superaonic Combustion,HASP Workshop Publication 1003, Juno 1988, pp. 34-37.57Dash, S.M., Welf, D.E. & Seiner, J.M., "Analysis of TurbiO.ent Under-ExpandedJets-Part I: Parabolized Wavier-Stokes Model, SCIPVIS," IAA Journal, Vol. 23,A ril, 1985, pp. 505-514.fash, S.M., "Turbulence Model Validation for High Speed Propulsive Applications,"

r4fh 6--•Hn, HASP C? 5029, Oct. 1988, pp. 467-504.aDash, S.M., n1-rbulence Modeling for Computations of High Speed; Flows State-of-

the-Art," FIlTMH N cn SvmDOHASP, NASP CP 5029, Oct. 1988, pp. 287-320.

269

J-Constant Index Clas K-Constent Index Class

- J Sign of Segment K Sign of

Number Index Nofmal Number Index NormalMin Max. Min Max.

1 1 75 1 1 1 1 20 1

2 21 1 31 .2 1 42 6 1

3 41 1 30 3 75 2 61 -14 1 61 22 74 4 31 22 41 1

I5 51 1 41 -16 1 52 1 21 1 411

I----------' . -',- . .-. .-. t -- T '-:_.I_

I 1 I I

M PYSCAL (OWL 4 PAT S -1-

~~~~~~Figure 2. Ramjetegeoieurytinn gridadbu dar patexs.(r•re.1)

2 70 T -7 -- -T

31 J

-L- L__J J 4Q I --J--L-j I i lr- -}

.PHSLDOAIN E C K-PUATIONL OS

Figure 2. Ramjet simulation grid patches.

270

UPAICOHIN STM)MCrIn SOUD PROPELLANT NOZZLE FLOW- NONEQUIUBRIUM CHEMISTRY- 30% LOADING At203

CHEM ____________ - EULER ANALYSIS•PCxE 1 sW DATA L GAS-PHASE ONL'L CALCULATION

MACH CONTOURS SHOWRECOMPRESSION SHOCK

0201HPUT MACH CONTOURS

F.OisE'TRY ( /RI IiNITIALIZATIO.

0.0 C-3.0 -1.0 1.0 3.0 S. 0 7.0 9.0

AUN PARCH/IRFILEU:." GAS/PARRTnCE NONEQUILIR'UM CALCULATION

PARCM

SPECIES MACH CONTOURS

PARTICULATE SOLVE 2.0

CIOPLING ROUTINES. C

-3.0 -1.0 1.0 3.0 S.0 7.0 g.o

fI o Oft S.0 PARTICLE DENSITY CONTOURS

Figur'e 3. PARCH/RN Structure. 2.0

-3.0 -1.0 I,0 !.0 5.0 7.0 9.0

Figure 4, PARCH/RN Nozzle calculation with/without

particulates.

271

nl I mffl" III - I0 I MINUTEMAN III - "TAW I91 x 51 •IO PFR5&RE CONTOUr5

.0 2.0 00 .2-0 4." 6.0 6.0 1C.0 4 ,0 60 1 0.

M lNUiVrlI:t I I I - 5TMEC I M I NVIANt III I STAGEC ITEHPTUR- • WMTOU5 MACH NLrMJER LTOUR5

0. -J.0 0 2.0 I U 6.0 i.0.0 -2.0 0.0 O . 4.0 60 0Z0 0ý,

Figure 5. PARCH/RN Nozzle calculation with 3 particle size groups.

"10.

-10. 0. 10. 20. 30. 40. 50. 60. 70. 80. 90. 100.110.

x (IN)

SLIP WALL

FAN

0 -713-J

Z1. 0

0

AXIS

Figure 6. PARCH/GT Physical and computitional domains foi tu-bofan tailpipr/nozzlc.

272

INIIAL GUESS PARCH/ GT SOLUllON

TCnPCRFITUrlC TEMPERATURE'INITIAL. GL&VS 'TUROULCNCC Cn5C*

MIMI rnisr , SIS& usimc rAi~cw Pwrc

U VELOCITY U VELOCITY*INITIAL WS, "TURLLE CNL ISC'

FUG rAIRCO PUA CSC' 'USING FAIRCO PWG

Figure 7. PARCH/GT Initialization vs. converged solutions.

PARCH/GT PARCTEMPCRATURC

'USIN~o h lA PUS' IMMPEIITURE

U VCLOC17YInteUASC(Z UISC

Figure b. Comparison of solutlion with PARCH block tridiagonal solutiontid using original PARC diagonaiizfld solution procedae.

273

BASELINE GRID

20.

o0.

0. 10. 20. 30. 40. SO. SO. 70. 80. 90. gtO. 110.

x (IN)

SOLUTION ADAPTED GRID

?C.

z"-,- 10.

o.

0. 10. 20. 30. 40. SO. 60. 70. 90. 90. tOo. I10.

x (IN)

Figure 9. Baseline grid and solution adaptiv, grid obtained using SAGE code.

T CO97oMURS

T CONTOURS 1-,,

___Q

TURB LEVEL CONTOURS

U CONTOURS X s 3

5.0 15.0 25.0 35.0 45.0

Figure 10. PARCH/TMP analysis of Mach .6 tactical missile plume/airframeinteractive flowfield.

274

0.0

-5. 0. S. 10. IS. 20. 25. 30. 35. 40.

-5.0

-0. 0

-5.0 -E

-50 4.0 13.0 22.0 31.0 40.xFigure 11. Marchn knti neg contouJrs for tactic.al missile problem a 0 nl-fa~ak

Rt0. j 0 rg~O-t~k

-5. . S. 10. IS, 0. 2. 3. 3S 40

LU coU ou.s * . .CONTO w W C"URS .22. U CONTOURSX a 23.3

.N4.0 N GA

,Do / 1T CONfTOURtS

4.6 - -*,. LS

A TURB LEV CONTOURS. L CONTOURS LS 2U E

Li-4

Li* * ,.4

0 .M 6

-8. 2.4 6.8 - ). CS0.4.6 -y.6 6.O .0 ".. 4.. *. 6-6 1.a t6 -. O ,'II ' 4.,, -'. Lb 1.0 6.6

yFigure 13. Cross flow contours for tactical misole prohlt.-, t bas- and

juat dowrstream.

probe planes8 50 100 120 150

Mo I

~~IENT]

- Probe in 3 plones(x = consinit)

Figure 14. AuriSptiti-le experlmental -- tu;:. , 2. Mp' ¶. P . Pl .. ',i)

276,

o. - PARC1H/TMP PREDICTIONAERSPAD-L USING ROK/TMP

- .10. .09

.*06- .03

- -. 03- -. 02

. .01- 00

- 01

FUSELAGE (Mo 2 -Mj=2 Z-0-1~uTPRESSURE COEFFICIENT CONTOURS IN THE TRANSVERSE PLANLOAE6.1

BEHIND THE NOZULFigruro 15. PARCIH/TMP grid and predicted Mach iiumbor contlours (or

AtiroapatiaI.p Maevh. 2 P,-st c'ame ohtolned IOI~l!TVDIFutl uptilill.

VIpturt 16.omnporlann ipr PAN(C1I/TMP

pre'dlelinon with Ar~lo otvrome flow Ipromablus,

0.0

-5.L 0. t- (jj~r 1 0JU Id ?0) 0 2 IX

SCRmNT30TC-D NOZZLEY=O PLANE

2.0

NI.s

O.S_

0.00.0 5.0 10.0 15.0 20.0

SCIIAFTC-O NOZZLJEyz0 PLANE

2.0

Ol.S

0.00.0 5.0 10.0 15.0 20.0

Figure 17. SCHAFT and SCRINT3D pridictions for square converging-diverging nozzle.

INITIAL CONDITIONS:

"" 0 - 13000 ft/a Tor view

N - 3.0 H•s-"T - 300 KS-, 0.45 atI m::: If

all- - I .= - - - .eitAZI 0D- 5000 tt/s S x IC '5

H - 7.0 I11C viewT - 1100 1P - 0.15 atm lie

AI

Figure 18. Schematic of H 2multi-jet combination in duct.

2i

Best Available C0r.278

T CONTOUR

8.0

7.0

6.0

5.0

0.0 5.0 10.0 15.0 20.0H20 CONTOUR

7.0 J!!

6.0

5.0

0.0 5.0 10.0 15.0 20.0xFigure 19. SCHAFT Teaperaturel H12 0 Contours in vertical plane passing

through inner H 2 jet.

T CONTOUR M420 CONTOUR

4.0

3.0

2.0

000.0 5.0 10.0 15.0 20.0 0.0 5.0 10.0 15.0 20.0

x xFigure 20. SCHAFT temperature/H-2 0 Contours in horizontal plane passing through

center of H2 jet at injection plane.

Best AvailableCo~279

T CONTOUR M20 CONTOUR

8.0

7.5

7.0

>q6.s

8.0 . . . . . . . . .. . . . . . . . . .0.0 0.5 1.0 L.5 Z.0 a 0.5 1.0 1.5 2.0

z zFigure 21. SCHAFT Temperature/H 2 0 Contours at axia station 5 duct heights

downstream of injection plane.

A Iir

200

~~Figure 23. SPARK aemtyfo hcnalysisbstonstd of Stalker.eprmn.pv~~l o1

6.0

5.0

0043.0

.0

0.0310.0 0.30000

042.0

1.0.

0000101.2 0 .0 3 0 .0 T0 30000.

)(x (cM)3.0 ur Figur 25.F SCHAFTi ofer Stalkers Stalkermcase

2.00

-2..0

20.0 3. l.

0.0 10.0 xn (cM)0 0.

Figurigu26.2S.HAFT temperature contours/ Stalker case.

Bet2vilbe08

3.0

2.0

~1.0

-2.0

-3.00.0 10.0 20.0 ( M 30.0 '10.0

Figure 27. SCHAFT H20 contours/ Stalker case.

0. 4 . . .0 5.0 6.0 7.0 0.0 9.0 10.0

Ii Figure 28. CRAFT 1D shocktube calculation,.~ II II'~\1000/1 pressure ratio.

Figure 29. CRAFT analysis of 5/1 air jet into duct with moving end-wall at3 time.

282

PRESSURE CONTOURS H2O CONTOURS3.

3

2. 2.

2. 0

0. 1. 2. 3. 0. 1.2. 3. 4.

X/H X/H

Figure 30. CRAFT ,~..alysis of 5f 1 combusting 1:2 jet into air duct

3.0 vith fixed wvall.

2.0

1.-\

-2.0

0.0 10.0 20.0 30.0 40.0

-X/RJFigure 31. CRAFT analysis of Mach 2 underexpanded jet into Mach .25

stream- density contours/Ilaminar.2.0

>1. 0

0.00.0 10.0 20.0 30.0 ql0.0

X/RJBe t vala~eCcFigure 32. Comparison of CRAFT SCIPUrS centerline pressureB s v i beCpredictions for Mach 2 jet. 283

MODEL 320-2: A COMPACT ADVANCED UAV TURBOJET

Eli H. Benstein, Chief ScientistBrian Cassem, Principal EngineerKathy Elliott, Senior EngineerTeledyne CAE, Toledo, Ohio

ABSTRACT

The Model 320-2 is a 355 lb. thrust, outgrowth/up-rating of the family of small turbojets developedon TCAE funding since 1985. It follows the thrust growth pattern of its predecessor J69 engines: a40% increase is achieved by supercharging the simple centrifugal stage of the basic 320 turbojet witha high pressure ratio transonic axial stage, without change of engine diameter or core flowpath. Theturbine was predesigned to allow for the increased work level without excessive efficiency loss.

The engine retains the simple/low parts count design of the new family, which leads to very lowproduction cost targets. Thus, for the intended expendable or limited life/reusable applications, acompact, high-performance, high-technology missile or UAV powerplant results.

-iie ppner f cu~-s -. r1,- aerodyn-'_ Ps'. It -._li-, t'he , . ;,,s - ," th- O.ndel ,ýUbsyt,-.:.,-. the soL.jc zu oL .tow-n-sacalea upercharging axial, and the approach to deletion oi aninlet bleed valve at the 7.85:1 pressure ratio of the compressor. Engine test data are presented incomparison to model predictions, and projections are made for future directions for development.'

INTT~ YrR, :: FAMILY ORIGINS

At 355 lb. thrust, the Model 320-2 is the highest thrust version of a family of six Unmanned AirVehicle (UAV) turbojets prototyped by Teledyne CAE in the last five years under company-sponsoredR&D. Figure 1 summarizes the family and its test history to date. References 1-4 describe membersof Lne family and its approach to modern, high performance, low cost design.

FMODELI [307 3127 [320THRUST* gO 135 200THRUST GROWTH 168 240 355SFC - LBIHRtLB 1.23 1.06 1.06EGT - F' 1630 1138 1140EQUIPPED WEIGHT - LB 18 33 56DIAMETER - IN. s.5 5.61 9.9

"SEA LEVEL STATIC VALUES 60+ TEST HOURS473+ STARTS5 HOUR SEA LEVEL ENDURANCE8 HOUR ALTITUDE ENDURANCE

HARPOON STARTING RELIABILITYAWuo iPERFORMANCE DEMONSTRATED

Figure 1. TELEDYNE CAE SMALL TACTICAL ENGINE FAMILY: Models 305, 312, 320

The engine series is desn-' r-r 7 '-ited life application, which can range from a one-shot,three minute! missile mission tu 'ý -. lours of repetitive reconnaissance flying. Ratings and theaccessory variants illustrated in Figure 2 are tailored to specific UAV installation, operability andmission needs. The generic design drivers for each member of the engine family are acquisition cost,engine diameter and/or volume and fuel consumption. Acquisition cost is critical because it involvesover 90% of the life cycle cost of a UAV engine, as shown typically in Figure 3. Diameter and volumeare important because the system installations are limited by launchers, fuselage diameter or thecombined fuel tank plus engine length, as illustrated in Figure 4 . Fuel consumption is significantin providing either minimum fuel tank volume or sufficient loiter time for some missions.Approved for public release; distribution is unlimited.

P R E C E D I N G P A G E B L A N K.• # > ,285PRECEDING PAGE BLANK •• tA,,-,••,

Uin30APICATWhlm FAILYv FIAUM A1E35

DUTY cY=. - ENDURANCE LUSICAI1O - 909AN 00s

-WA911 OIL

PAM5 UUWEiPI PAP5 STUIIIN - AM & MSPA V

FUE SYUEMINU..*AA PiM SafftT LOW PO141* fAaMS-LOW &D aPumaKEP WWO-0O

CONW~ - 3TMOALONS ElOCTFWMOS A &OtrKNPUW VALVE 06

-OLM RiPAi LCTROSIc5IwTIUM PUW

IECT7480MU A VALVE

1LXCTmIVI.A FW S.IA4TT5OWAL A.TMSAIATOR & OCaI______ -174.16414AF Fagure 3. TYPICAL LIFE CYCLE COST DIMSRIBUMON:

FigwZ SALLTAC7CA UA ENINEAPPICA70N91% of cost it in prod uction for wooden rownd.

FLIYJBIITY - OPTIONS iae f nge.

LAUNCH CARRIERS WONT BE CHANGED. THEY hLM1"* OIAMgTCA PYLON A CSAL 11449 VLA. ..L CL.3"* WRIGHlT A46 F-5 14. OS& 1 . MLPM"* LEIWOT1. VOLUME. 5N roAPffOO lu9L WIAS"* ENGINE SHAPE FOOaM

PIRVAJIW THEME

Figure 4. UAV PROGRAM ILAUNCH IMISSION CONSTRAINTP1S

MODEL 320-2 DESIGN

The 320-2 was designed with these constraints in mind and as a simplest-possible growth versionof the 200/240 lb. thrust Model 320-1 (Figure 5). The path followed was the same as that used inraising the J69 engine series from an initial thrust level of 880 lbs. to 1.920 lbs. - i.e.. byadding a transonic axial zero stage and increasing turbine inlet temperature. The growth plan wasstipulated at the time of original design of the 320-1. in that the combustor and turbine wereaerodynamically and structurally (materials selection) designed to accommodate it. Table 1suma rizes the two cycles, with emphasis on the turbine and thermodynamic cycles.

-IS SAM 7.21 -

0 54.

THRUST .. 240 LO TURGINE IN TFMP - 1727 TFSC- IlIl L&MPASt 60T . 1A0D*F

AIRFLOW . 3AG ISS EJNAUMT PRESS RATIO . W0

figvmre . MO0DEL 320-I TURBOJET CROSS SEC77ON: Avkrrpe. back-to-Mncsturbojet.

TABLE 1 - COMPARISON OF 320-1 AND GROWTH 320-2

PARAMETEP M~QDEL 320-1MOE322

Thrust-lbs. 200 (240 uprate) 355SFC-Ib. per hour per )b. 1.056 1.090Corrected speed-RPM 60,000 60,000Airflow-lbs. pcr xecopd 3.71 4.82Coapressor pressure ratio 5.68 7.85Turbine inlet temperature-*F 1535 (1800) 1950Turbine corrected speed-RPM 31,000 28,270Turbine corrected work-BTU per lbs. 28.72 29.71Turbinr nozzle flow parameter 1.421 1.449

286

As shown, the turbine work level is almost unchanged in the -2 version; the turbine inlet nozzlearea is increased about 2% and the corrected speed is decreased by 8.82 to accommodate the 772 thrustincrease. The high reaction turbine design was optimized at the mid point, thus the turbine matchpoint efficiency fall-off was minimized - as has been proved in the engine testing described below.

The resulting engine configuration is shown in Figure 6; it is a 9.9 inch diameter simpleturbojet weighing only 60 lbs. fully equipped. As shown in the figure, the 320-1 core engineflowpath is retained unchanged in the uprated design. Its low SFC, small diameter and near 6:1thrust/weight ratio place it above the state-of-the-art, as summarized in Figure 7a, b and c(Reference 5). Engine materials are completely conventional and selected for low cost net shapeintegral bladed casting, as defined in Figure 8. The engine design uses only 30% of the total partscount of the J402-CA-400 (HARPOON) turbojet and has only 12 major parts. The core Model 320-1turbojet can be assembled or disassembled in approximately 5 minutes, with Sears Craftsman tools,exclusive of bearing insertion.

FIG 7a ENGINE SFC

kg,;N h LBHRLD

016.

I.-mooncm.rITUSoMMT --- i -15 .7 tREF 5)

V • -014

4 ,0_i 1 3 TPP320 8... ~SF ," : . . ,•13 - " 1 2, . 4.'!/' " " 013" ,12 E0 2. 1 * TPP i0o

t1? 9'320-I 320-2

o Io T 2W 2o 0 30o010 0 .5 If50 (LB)

AD --- 0 200 400 600 800 1000 1200 1400 16001N)THRUST

THRUST a 3,6 LB. FIG 7b. ENGINE MAXIMUM DIAMETERSFC m 1.0O6 L/HR/BAIR FLOW - 42 LB/SWEIGHT - 56.0 LB. 4mm iHuSTR AREAS•150 LB&F-'f

400 - S. 6e _TPP 100W 25Figure 6. MODEL 320-2 IVRBOJETCONFIGURA7ION 4'1 t4. - -9-

MAX 300 -12 TPP320 ,-S

DIAMETER 20 0" 320-?

~ *- 312100'

o 100 15 200 250 300 350 (LB)

0ý • I I ,l , I , I I I~_-drCO.PR!ES&534TA•t CO14PPESSOq COVER rEAR FAAME 200 400 600 SW9 10D 1200 1400 16B (NICAMI I74PH1 P-AOiAL I"AFUS(A ICAS INCO NWS6,

CAS.T J47 STSl)- . -CO MA' SNELS THRUSTS- _ JICO &25•O~ SHEET)

INtET NO115/ .klAB1 INLET NOME•-ic~ .~. L'JM 1 .-. ~:CAST. 11)~,

S,• MOE 321OOS CAf O CAST MASOM2TS FIG 7c ENGINL WEIGHT

THRUSTWT

- 7/ TPP 1000 320-2WTEIGHT -50 81

200 1FF 3?13 2. ~-43 2 ..

Figure 8 MO0DEL 320 LOW COST CAIT1 COMPONEN'TS -2 /.< __10025,- j> f$357E

1-, 010 150 200 25.0 30 350(1.81

20 0 400 60W 800 1000 1200 140D 1600 (N)THI)PU ST

Figure 7. fVMALL. UM[TED UJFE UAV ENGINE STATE -OF-UIIE-AR7- Model 320-2 is at thme lea~ink edge

The Prototype engine hardware is shown. in Figure 9. This version incorporates a 10* cantedexhaust system tor a special installation, but a conventional axisymmetric exhaust can readily beused.

287

10" CANTEDIPE,

Figure 9. MODEL 320-2 PROTOTYPE:43 total est hours, over 340 lb. thritr.

Engine starting in the test cell is rnormally accomplished by high pressure air impingement onthe centrifugal compressor rotor tip aid ignition via a spark igniter; future testing wil.incorporate pyrotechnic cartridges as used in the .J402-CA-400 Harpoon turbojet (Reference 6). Thecurrent prototype is configured with a low pressure rise positive displacement pump. The patentedliquid ring pump fabricated into the engine shaft (Reference 4) allows for near-atmospheric pressuredelivery into the front shaft fuel entry tube, thus a low pressure UAV fuel management system. Fuelcontrol electronics are vested in the air data computer of the UAV in current applications andemulated in the cell on an IB L-PC. A stand-alone, E141-qualified Full Authority Digital ElectronicControl, developed by Teledyne CAE, will be used in other applications. A 2.4 kilowatt integralsh.-it-muuited .lleinatoz .ith a ,umariu& tb.1lt rotor is lo.ated immediately aft of the thrustbeairig, with field windings mounted in the static housing. The alternator leads are drawn throughthe flowpath to a MIL-C-38999 connector outside the engine.

Engine lubrication is by waste oil on the front bearing and axial compressor dischargepressure-driven fuel lube/cooling on the rear, rcller bearing. The bearings are conventional; thethvrst bearing is an angular contact, 15MH split inner ring design, and the roller a 10 millimeterconvenzional dcible lipoed outer rear configuration.

Each of the components and accessories has been successfully tested, with over 43 hoursaccumulated on the rore and supercharged engines; the engine will soon be run at altitude in theleledyne CAF chamber to verify its cartridge start capability and demonstrate readiness for aprojected flight test.

C2PRESSOR STAGE MATCHING PHILOSOPHY

Design of the Model 320-2 compression system was governed by the following ground rules:

o 355 ibs. thrust at SLS

o No interstage bleed valve, for low cost, reduced cnmplexity and high reliability

o Axial compressor: Existing transonic stage design scaled to Model 320-2 size

o Centiifuz'al compressor: existing design

The requirements for 355 lbs. thrust and use of an existing centrifugal compressor defined theengine airflow and pressure ratio. Successful operation of the Model 382-10 (U.S. Navy MR-UAV) fanappeared to provide close to the re.,aired performance with some margin. However, in a turbojetconfiguration, matching of a transonic stage of this pressure ratio level with a 6:1 pressure ratio(unburied) centrifugal stage normally requires ar interstage bleed valve or variable geometry forsuitable starting and acceleration operation. Engine cycle analysis was used to determine the axial.nd cerFtrifug4 l bt,ig` match pointa. To pro-.ide for low specd surge margin without an intcrstagebleed valve, the Model 382-10 axial compressor was scaled at 84.5% of its design speed; performancecharacteristics of the reference axial compressor are shown in Figure 10. The compressor was scaledat this speed to take advantage of the broader flow characteristic, thus to insure adequate surgemargin at idle speed.

20%

ADIABATIC ..

EFFICIENCY &O 0"- - -

% % 9

7TO s 8 34% 100%

_.-W 60 70

26

24

PRESSURE RATIO2?

20 ENGINE 20.OPERATING LINE

SL90 g6 00 %CORRECTEDso. SPEED16 .• - -- 80 85\ 34%

SURGE MARGIN

70

2 so 6

10 1 1 3 14 15 15

CORRECTED AIRFLOW - LBMSEC

Figure 10. REFERENCE MODEL 382-10 TRANSONIC AXIAL STAGE MAP - baseline for scaling.

The axial compressor was scaled from a point in the choked flow region of the map. Axialcompressor match points with 20% and 34.1% surge margin at the 3S5 lbs. thrust point wereinvestigated. Turbine inlet temperature was the same for both of the simulations, only thecompressor scale factor was changed. As shown in Figure 10, the 202 surge margin match pointprovided a higher axial compressor efficiency but had no surge margin at idle speed (approximately70% of maximum). The 342 surge margin match point provided 12.4% surge margin at idle speed anihisher efficiencies in the cruise area of operation. Figure 11 shows that the 34% surge margin matcl,point also provides higher thrust and better specific fuel consumption at part speed conditions.

NET THRUST - LB. RELATIVE SFC -%

400 180' \

3W 170/7 \

300 / 16025034% S /M

250150 S---- 20% SM

200 140-

100 120

50 110

0, 100 .0 70 80 90 100 60 70 80 90 100PERCENT CORRECTED SPEED PERCENT CORRECT SPEED

Figure II. MODEL 320-2 AXIAL COMPRESSOR DESIGN STUDY: 34% surge margin improves part power performance.

From this study, the design match point for the model 320-2 axial compressor stage is:

TABLE 2 - COMPRESSOR DESIGN MATCH POINT

AXIAl. OVERALL

Correvted Airflow, lb/sec 4.82 4.82

Total Pressure Ratio 1.685 7.85

Corrected Speed*, RPM 60,000 60,000

Adiabatic Efficiency, 77 ad, % 78.3° 73.2

( 80.8% less 2.52 for reduced size effects)

2I)

COMPRESSOR ARO DESIGN

A geometric scale of the Model 382-10 compressor was used for the axial compressor rotor design.However, since the match point is near the choked region of operation, a direct scale of the statorblade row would yield an extremely high stator exit Mach Number. This would require a severetransition duct between the axial stage stator exit and the centrifugal stage impeller inlet.Unacceptable losses would occur in the transition duct if the flow were diffused from a Mach Numberof 0.'5 to 0.45.

To alleviate this problem, the stator blading was modified to retain some of the tangentialvelocity. It was designed to provided O" prewhirl at the hub streamline, with a linear variation to10" prewhirl in the direction of rotation at the tip streamline. This allowed the exit area of thestator to be opened up without overloading the blading. Figure 12 compares the Model 382-10 andModel 320-2 axial compressor flowpaths, with the former scaled to the 320-2 size. The number ofstator vanes was also reduced from 37 to 31 for the Model 320-2 compressor for ease of manufacturing.The effect of the prewhirl on the centrifugal inducer at the reduced corrected speed and Mach numberswas judged to be non-detrimental.

MODEL 38210 FAN

1 2 3 4 5 6AXIAL DISTANCE - IN

Figwre 12. FLOWPATH COMPARISION, BASELINE AND SCALED STAGES

As stated, the rotor was a direct scale of the Model 382-10, except for the leading and trailingedge radii. They were left at 0.010 inches at the hub and 0.006 inches at the tip, which is theminimum allowable for casting purposes.

The resulting hardware is shown in Figures 13 And 14.

.. .................. : * .,

Filur• 13. MODEL 320-2 TRANSONIC AXIAL ROTOR Figure 14. MODEl ?20-2 TRANSONIC AXIAL STATOR

29W

COMPRESSOR TEST RESULTS

Performance characteristics measured for the Model 320-2 axial compressor on the engine withlimited interstage instrumentation are shown in Figure 15. Maximum corrected engine speed obtainedduring this prototype test was 57,000 RPM (95% of design), due to a temporary mechanical limitation,which is now being fixed. As shown in Figure 15, the compressor flow and pressure ratiocharacteristics are in excellent agreement with the pretest simulation. The test efficiency at 95%speed was 84.22, which was higher than the simulated prediction. At lower speeds, the axialcompressor stage efficiency was somewhat lower than the simulation predicted, for reasons unknown atthis time. The engine has been started, accelerated and run surge-free with a 12.0 inch jet nozzlearea, and without a tailpipe.

STAGE EFFICIENCY - %

0.90

0.80

0.70

0.60

2.20

0. - TEST DATA

JNA = 12.0 IN'(BUILD

02)1.80

106.5% CORRECTED

1 00.6 SPEEDSTAGE

PRESSURE 1.60 MATCHRATIO POINT

140

71.0

120 59.2

1.00 1 1 A-0.00 100 2.00 3.00 4.00 500 6.00

CORRECTED AIRFLOW - LB/SEC

Figur 15. MODEL 320 2 kLIAL SIAf;E PERFORMANCE via enpine tesI

It was concluded that the scaled compressor met its design objectives without a.)dification.

CENTRIFUGAL STAGE DESIGN

The centrifugal compressor design point is summarized in Table 3. The rotor has 13 full blades,13 splitter blades and takes advantage of 26 degrees of backbweep. The _idial diffuser is comprisedof 24 cambered vanes with a pressure recovery coefficient of G.548 and _n exit to throat area ratioof 1.37. There are 100 constant thickness plates for the axial desw,rl row, with average diffusionfactor loadings of 0.48 to turn the flow to the axial direction and provide static pressure rise forthe 0.31 Mach number combustor entry conditions, The diffusion s)stem is compact, with a 1.37 ratioof the axial diffuser flowpath 0.D. to rotor O.D.

FABIE 3 - MODEL 320 £ENTRIFUr'AL DESIGN

Corrected Inlet Airflow, lb/s ,c 3.66Stage Total Pressure Ratio 5.73Stage Adiahati. Efficienc y, • 77.1Corrected Rotational Sp-J, ;pm 60,000

2')1I

ENGINE TUT PERFOR•"ACE

The Model 320-2 engine has accumulated over 15 hours of sea level testing. The engine has beencalibrated over a range of jet nozzle areas to determine the optimum match configuration. The steadystate performance measured for the design configuration is compared to the analytically predictedmodel in Figures 16 thru 18; the model represents the Full Scale Development goals for the engine.The measured performance does not account for the 10 degrees of cant in the exhaust duct; thrust andSFC should thus be adjusted by I.S for a conventional axisymetric exhaust. The Figures indicatedthat the enaine is operating within 5% of the goals set. At 60.000 rpm, the engine would provide anet thrust of 365 lb. (370 lb. for an axial exhaust), at a specific fuel consumption of 1.15 lb/hr/lb(1.133 axial) at a turbine rotor inlet temperature of 2050 degrees F. The results of this testingsuggest that the turbine nozzle area should be reduced by 2 percent to yield an exact match at 350lb. thrust at the design 1950"F temperature with a consequent reduction of SFC.

350 TEST - EXTRAP CRT 140-eo TEST- EXTRA P

SFC -- LOnMR/LB 1 30' ETDT

250,

CORRiECTED12jCACTDMODEL 110._' DE

NET THRUST 200- Fn/h2 - (L.OF) YTEST OATA 100 •.

150 1 50 100 150 200 250 300 350 400CORRECTED NET THRUST - FrILBF)

100

o Figs't 17. MODEL 320-2 ENGINE SFC PERFORMANCE @ SLS

NAW 3500 40000 45000 5000 556000 eo0 65000

CORRECTED SPEED - N1,?"(RPM)

Figure 16. MODEL 320-2 ENGINE THRUST PERFORMANCE @ SLS

215001

2500 T EST EXTTRAP.

2400

2300 TEST DATA

MODELCORRECTED 2100

FUEL AIR TEMP

T 1' (R 000laow

3000) IOM 4U0 00 45 00 /'00 55 60O 550.WM

IGIORI".3 2 (;01 ( t DSI"E.I l. N'I, R PiI RHPM)

Overall, the tested engine characteristics are steeper than the predicted model, such that thelow speed performance falls away from the model. This is influenced most strongly by the centrifugalcompressor operating characteristics. When tested alone, the centrifugal compressor airflow-pressureratio characteristic was also steeper than the characteristic used in the analytical model. Themodel will thus require modification.

Analysis of the test data indicates that the SFC increment above design can be attributed to a 3point deficiency in the centrifugal compressor and a 1-2% additional pressure loss due to bumps inthe as-fabricated exhaust duct (Figure Q).

Detail static pressure data shows that the centrifugal compres:or impeller is performing wellabove design performance levels (6.8 pressure ratio), and that the additional losses are in thediffusing system. To increase the stage efficiency, the radial diffuser is currently underdevelopment. It is expected that this development effort will attain the design objectives for thecentrifugal stage efficiency and thus provide the overall engine goals for both thrust and SFC, witha conventional exhaust duct.

The engine starts at a speed of 16% with impingement air on the centrifugal impeller, with theinstallation fuel control simulated on a PC emulator in the test cell. Accel times from start toidle ranged from 7-8 seconds.

Future testing will include altitude chamber evaluation with pyrotechnic cartridges, performancematching and the diffuser improvements. In general, the engine testing to date has achieved itsprimary objective: to substantially reduce the risk of Full Scale Development.

SMARY AND CONCLUSIONS

1. A 77% thrust increase has been successfully demonstrated in a competitive, low parts-countUAV-applicable turbojet with only a minor SFC change from the baseline.

2. The resulting engine is at the leading edge of the open literature state-of-the-art in terms ofSFC, thrust/frontal area and thrust/weight factors.

3. The engine design retains major commonality of parts with its lower thrust predecessor. Thisfeature, together with its inherent low parts count by use of high performance staging andmultifunction structural members, leads to low cost in production.

4. The key component in achieving the performance with minimum flowpath component change is thetransonic axial zero stage compressor. It was geometrically scaled over a 3.2:1 flow reductionfrom an existing 2.2:1 pressure ratio fan stage, with only the stator adapted to match the newflowpath. Stage matching with the centrifugal was carefully selected to eliminate the need foran interstage bleed valve or variable geometry, in spite of the near-8:1 cycle pressure ratio.

5. The prototype engine has performed well within its goal of 5% below FSD requirements withoutmajor redesign modification of the as-designed aerodynamic hardware. It has started cleanly andaccelerated well using a low pressure rise, positive displacement pump, with the controlauthority and algorithm emulated in a PC.

6. The centrifugal compressor radial diffuser has been isolated as the only flowpath componentrequiring optimization to achieve FSD performance.

7. The engine suits limited life applications in providing a viable combination of advanced missionand installation performance and low cost.

ACKNOWLEDGENENTS

The authors would like to thank the management of Teledyne CAE foi permission to publish theresults of their many colleagutes who contributed to the definition, design and successful testing ofthe engine.

REF EREr CES

1. Lill-'y, J. S., "Demonstration arid Evaluation of Low-Cust Turbojet Engine. for FOG-M". USA MICOM,November, 1987.

L. Papandreas, L.., "SCAT: A SmalI i.ow-trrst Turbojet for Missiles .tfid i'V's", AIAA- g--241,Pioulsion "Tonference 12 July 1988. Boston. MA.

3. Ren.t,:in, t.. 11., "A New Engine lani lv fur C'A\ Prouil.,.io'n", AI'S l~rvt:• Svmpusium, 1 Match 1989,bDavt 'jr, OH.

4. Benstein, E. H., "A New Family of Turbojets for Expendable Air Vehicles", JANNAF PropulsionMeeting, 25 May 1989, Cleveland, OH.

S. Barbosa, J. R., Takeda, A. S. et. al., "A Small Gas Turbine for Drone Aircraft - DesignPhilosophy", Institute of Research and Development, Aerospace Technical Center, Sao Jose dosCampos, S.P. - Brazil; 1990.

6. Barbeau, D. E., "A Family of Small, i.-!nost Turbojet Engines for Short Life Applications", ASME1981 - GT-205, Gas Turbine Conferen7-' Products Show, March 9-12, 1981, Houston, TX.

294

THE CRYOJET REVISITED

B. Singh, W. Roberts and S. HarperThe Marquardt CompanyVan Nuys, California

ABSTRACT

The viability of a cryojet engine is evaluated for high Mach number applications. Cryojetis an advanced thermodynamic cycle that offers high specific impulse (to 5000 seconds), highthrust-to-weight ratio (up to 25:1) and increased Mach number capability (to 5). This paperquantifies these advantages and addresses key design parameters.

The turbine engine operation is inherently limited by compressor inlet and dischargetemperatures. Therefore, for a given compressor pressure ratio there is a maximum allowablefreestream Mach number. However, cooling the compressor inlet air reduces these temperatures thusresulting in increased Mach number capability. The reduced compressor inlet temperature alsoresults in increased airflow and pressure ratio. These performance improvements are derived fromthe component rematching effects due to precooling. These benefits may be small at low Mach numberoperation, however, they increase significantly with increasing Mach number. This study is basedon cooling the compressor inlet air with high heat sink cryogenic hydrogen fuel.

The realization of these benefits has been limited in the past due to heat exchangertechnology. The prohibitive heat exchanger size, weight and pressure losses for practical designshave significantly reduced the realizable benefits. However, with the latest heat exchanger designand fabrication technology enhancements a significant fraction of the ideal improvements resultingfrom the thermodynamic rematching can be achieved.

Various engine cycles were evaluated during the study. The results of this study indicatethe feasibility of using cryojets for high Mach number applications. However, component designs,configurations and component integration must be carefully addressed for each design concept.

INTRODUCTION

System mission requirement; consisting of broad Mach number and altitude operability presentdesign challenges for the propulsion system designer. A single thermodynamic cycle is incapable ofmeeting these unique requirements. Therefore, it is desired to incorporate the best features ofthe appropriately selected propulsion elements based on the mission needs into an integratedpropulsion system. These propulsion concepts are knnwn as hybrid, composite, mixed or combinedcycles. The cryojet is a unique combined cycle consisting of heat exchanger, turbomachinery, gasgenerator and afterburner as its key propulsion elements.

Propulsion systems based on turbomachinery provide excellent performance at low speeds fromstatic to approximately Mach 3. The cryojet engine concept extends the upper Mach numberoperability of the turbomachinery without sacrificing the low speed performance. The key designlimiting gas turbine engine temperatures depend primarily upon the flight Mach number and thecompressor pressure ratio. in a cryojet cycle the engine inlet air is precooled in the cryogenicfuel (hydrogen)-to-air heat exchaniger(s) thus reducing Mach number effects.

Also, a gas generator driving the turbine which in turn powers the compressor is lessdependent on flight conditions. Therefore, the turbine inlet temperature can be held constant andthe turbine would operate at or near its design point.

NEED FOR A COMBINED PROPULSION SYSTEM

The need for a combined cycle originates from a desire to design a compact well integrated,high performance and lightweight propulsion system capable of operating from sea level static tohigh supersonic or hypersonic speed regimes. Specific impulse for individual propulsion conceptssuch as turbojet/turbofan, ramjet, scramj, 1. is limited to a relatively narrow range of flightspeeds, Figure 1. The combined cycle propu.sion system provides a means to operate efficiently inmore than one thermodynamic cycle configuration for a broad flight regime.


-95

I i i s

Figre . Eo~uionof dvacedComined

&L&CH NO

Figure 2. Evolution of Advanced CombinedEngine Cycles

Figure 1. Various Thermodynamic CyclesProvide High Specific Impulsein a Relatively Narrow MachNumber Range

The primary attributes of the combined cycle engines thus are:

Wide Mach number operability - A wide Mdch number operability is aftorded by propermodulation and control of the two or more elements of the propulsion system.

High thrust-to-weight - The combined cycle engines result in higher thrust-to-weightratio due to aerothermodynamnic and structural integration of the propulsion elements.

* Compact Installation - The combined engines tend to be more compact due to commonflowpaths and hardware use of the propulsion elements than the combined powerplant.

* Synergistic benefits - The synergism aftorded by the combined cycle engines improvesits overall operability, performiarce and installation through interactions between thepropulsion elements. Performance can be precisely tailored to specific :ýsionrequirements through cortrol logic.

COMBINED CYCLE ENGINES AND TkLHNULUUY

In the late 1960's, Marquardt investigated various combined cycle engines, their payoffs andbenefits for systems capable of taking-off from sea level static, climb/accelerate and cruise athypersonic speeds. These systems included rockets, scramjets, liquid air cycle engines, ramjets,ejectors and turbomachinery as primary propulsion elements. These concepts have been examined andevaluated by industry since their original inception. Typical propulsion elements for RAM-LACE andsupercharged ejector ramjet combined cycle engines are depicted in Figure 2.

Marquardt has built prototype hardware and demonstrated various component technologies.Supercharged ejector ramjet (SERJ) engine was designed and tested at Marquardt in 1968, Figures 3and 4. Liquid air cycle enqine (LAC-) feasibility demonstration and verification tests weresuccessfully conducted at the Marquardt facilities, Figure 5.

290

WKRtCKAMAGW LEXTOR AM*rT

Figure 3. Genesis of the SuperchargedEjector Ramjet (SERJ) Cycle

Figure 4. Ejector Ramjet and SuperchargedEjector Ramajet (SERJ) Testing

Key cryojet technologies include heat exchanger, cryogenic fuel distribution and handlingsystems, controls and integration of the propulsion elements, Figure 6. These technologies eitherhave been demonstrated, as discussed above, or are considered to be available for cryojet conceptfeasibility demonstration. Other items depicted in Figure 6 are matured technologies. Their usedoes not present any barriers, it only requires development programs to integrate them into acomplete combined cycle propulsion system.

Figure 5. Liquid Air Cycle Testing Figure 6. Cryojet Engine TechnologyR~equi rements

Best Available Copy

297

CRYOJET ENGINE CYCLE CONCEPT AND BENEFITS

The cryojet cycle concept is based on using the on-board cryogenic fuel's heat sink potentialto cool the compressor inlet air. Cryogenic hydrogen has emerged as the most promising fuel toachieve maximum potential of the advanced propulsion concepts such as the cryojet. Hydrogenstorage and handling technology is under grasp but may impose some technological and logisticschallenges.

The basic cryojet engine consists of an air inlet, an air-to-fuel heat exchanger, acompressor, a gas generator to power a turbine, the turbine to drive the compressor, a main burner(afterburner) and a propulsion exhaust nozzle, Figure 7. Numerous component packaging schemes andconfigurations are available within the basic cryojet concept. These include fan tip mountedturbine, annular or axisymmetric gas generator and annular or circular heat exchanger arrangements.

xG"1O MA - I

I /# ,-- 3

I L/ /

iOUPII SWR W" TI[a

Figure 7. Regenerative Cryojet ConceptSchematic (Flowpath)

Figure 8. Compressor Work RequirementsDepend on the Inlet Air Temperatureand Free Stream Mach Number

The compressor inlet and discharge temperatures (CDT) are primary limiting design parametersfor a cryojet cycle due to thermodynamic, structural and materials limitations. CDT depends uponthe inlet air temperature and the compressor pressure ratio. Inlet air temperature in turn dependsupon the flight Mach number, Figure 8. Thus cooled inlet air will reduce CDT resulting in enhancedMach number capability.

Thermodynamic benefits are also realized by the cryojet cycle due to rematching effectsincluding increased corrected speed which in turn causes increased airflow and pressure ratio.Additionally, the compressor work is significantly reduced at lower compressor inlet temperature,Figure 8. At 1080*R (corresponding to Mach 3 at 36,200 ft.) inlet air temperature the compressorwork is reduced by approximately 45% relative to the 1590*R (Mach 4 at 36,200 ft.) inlet condition.This also implies that a 5100 reduction in inlet air temperature would increase the Mach numbercapability from 3 to 4 for the same structural and materials design requirements.

In addition, a smaller turbine will be required to drive the compressor resulting in reducedengine weight and size or higher compressor pressure ratio can be achieved with the same turbinethus improving the cycle. It must be realized that the design parameters such as inlet airtemperature, flight Mach number, compressor pressure ratio and amount of cooling are inter-dependent. Therefore, proper trade studies must be conducted to define an engine cycle to satisfyspecific mission requirements.

Best Available Copy

298

Another feature of the cryojet cycle is the gas generator driven turbine. Gas generator isessentially a rocket engine powered by hydrogen, hydrogen and oxygen combustion products orhydrogen and air combustion products. Thus, the gas generator may be of any following type:

* Expander: High pressure heated hydrogen is discharged through the gas generator todrive the turbine. No combustion takes place in the gas generator.

* Hydrogen-Oxygen: Heated hydrogen combusts with oxygen in the gas generator. Thecombustion products drive the turbine.

* Hydrogen-Air: Heated hydrogen mixes and combusts with a small fraction of the inletair in the gas generator. The combustion products drive the turbine.

This feature fqas generator powered turbine) of the cryojet cycle allows the turbineoperation at or near design point thus decoupling the turbine rotor inlet temperature from theflight Mach number.

The cryojet cycle benefits thus are:

* Increased Mach number capability

* Increased thrust due to compressor rematching effects

* Improved specific impulse over a broad Mach number operation

* Turbine operation is independent of flight condition

0 Reduced compressor work results in a smaller and compact turbine

a Higher thrust-to-weight ratio engine due to smaller turbine size and rematching effects

* Cryojet performance can be improved further by a regenerative cycle. A fuel-air heatexchanger is introduced in the exhaust duct to further increase hydrogen temperature beforeentering the gas generator. This approach allows lower hydrogen flow rates to power the turbine.Therefore, overall engine equivalence ratio can be controlled for maximum specific impulse.

CRYOJET SCHEMATIC AND FLOWPATH

A regenerative cryojet concept is depicted in Figure 7. Air enters an inlet where it isdiffused to a subsonic Mach number. Diffused air passes through a hydrogen-air heat exchangerwhere it is cooled to a predetermined temperature. The cooled inlet air is compressed as it flowsthrough a compressor. The compressor discharge air is mixed with the turbine exhaust (hydrogen)and combusts in the burner. Additional hydrogen is added in the burner if required to achievemaximum thrust corresponding to equivalence ratio of unity. High temperature burner exhausts flowthrough a regenerative hydrogen-air heat exchanger before expanding through a nozzle.

Hydrogen is directed to an inlet heat exchanger from the fuel tank at a specifiedtemperature and pressure. The heated hydrogen then flows through regenerative cooling passes onthe combustor and nozzle where it is further heated. A fraction of this heated hydrogen thenenters the gas generator. Hydrogen flow rates, pressures and temperatures into the gas generatorare specified to have sufficient energy to drive the turbine and other auxiliary pumps for thecycle. Turbine exhaust (hydrogen) mixes with the compressor discharge air and combustion takesplace in the main burner. The remainder of the heated hydrogen from the regenerative heatexchanger is also directed to the burner to provide a maximum equivalence ratio of unity.

A liquid air based cryojet engine is a more complex cycle with a potential of improvedperformance. These cycles vary in flowpath and complexity depending upon the performance andmission requirements. A flowpath of a cycle requiring liquification of a fraction of inlet air isdepicted in Figure 9. A fraction of the cooled inlet air is directed to condensers where it isliquified. Liquified air then flows through a regenerative liquid air-gaseous air heat exchangerbefore it. enters the gas generator. The remaining cooled air is directed to the compressor beforeit is burned in the main combustor. Hydrogen flows through condensers, a cooler and a precoolerbefore it enters the gas generator where it mixes with cooled air and burns. The gas generatorpowers the turbine which in turn drives the compressor.

Various other advanced cryojet cycles can be established to operate over a wide speedrange, Figure 10 illustrates some of the LACE based advanced gas turbine cycles evaluated atMarquardt. Best Available Copy

299

-7~

Figure 9. Liquid Air Based CryojetCycle Flowpath

Figure 10. Liquid Air Cycle BasedCryojet Concepts

HEAT EXCHANGER CHARACTERISTICS

The flighit weight and flight type Lryuy•eflC htet eAchar,fier (1iX) has beer. analyzed, designedand tested by Marquardt during selected intervals over the past 30 years. The major objectives ofthis activity were to develop minimum weight and volume heat exchangers which efficiently deliveredcold air as required with an acceptable air pressure drop using cryogenic fuels as coolants - inparticular, hydrogen.

As a result of this activity, performed during the programs noted earlier, the technologynecessary to design heat exchangers having the above characteristics has been made available. Thistechnology was utilized in the subject study of the cryojet engine.

Numerous trade studies showed that an efficient, cost-effective heat excnanger design whichproduced effective heat transfer at an acceptable air pre3sure (thrust) loss was a closely spaced,in-line tube geometry where the warmer air flowed normally over the tube exterior and the coldhydrogen fuel was forced through the tube interior. Within the constraints of the particularapplication, tubes were manifolded to provide enough multiple passes to deliver the required airtemperature drop. Iterative thermal balances werp pertormed across a matrix 3f tube length, tubebank width and numbers of passes (HX length) to identify an acceptable, or minimum, tube assembly(HX core) volume which met the propulsion system requirements. These requirements included theengine air flow, fuel flow (preferably at an equivalence ratio of unity) along a system designtrajectory.

lhe HX design was then developed toward a minimum weight configuration. This involvedutilizing high strength, minirum gage material in the tubes themselves, a snugly wrapped HX shelland minimizing the number of tube (hydrogen) manifolds. In a cryojet application, the pressuredifferential across the HX shell is usually quite small (a few psi). Conversely, the hydrogen wallpressure differentials can be large (100's of psi), so minimization of hydrogen containing walls isessential tu a lightweight design. Using the above technioues on contemporary cryojet HX desionshas provided a HX weight to volume ratio of about. 35 lb/cu. ft. This was vith nickle-steelmaterials. The feasibility of using extremely thin, brazed, beryllium tube assemblies will furtherreduce HX weight.

300

A selected cryojet engine system, with a fixed HX configuration, was analyzed along a designtrajectory from SLS to Mach 4.5. The HX equivalence ratio was unity througho;it. This produced thecompressor inlet air temperatures both uncooled (for comparison) and cooled (for performance andmaximum Mach extension) as shown in Figure 11. An HX design weight curve was developed for thesubject cryojet engine. Using the design requirements at SLS along with the above desigi;philosophy, an HX weight was found for selected values of compressor inlet temperature and pressurefor state-of-the-art and advanced materials. This curve is shown in Figure 12 and was used in theengine design evolution. Turbomachinery and propulsion nozzle sizing waý done using the curve andthe cooled temperature values shown in Figure 11. The engine performance also reflected the cooledcompressor inlet temperatures and is presented in the following paragraphs. With the above HX, andthe shown reduction in compressor inlet temperature, it was estimated the maximum flight Mach numbercould be extended to above Mach 5.0.

SA LEVEL STATICMAJRLOW 100 MB'SECfouNALIEWct RATO 10

2000 4001 X --\I ... T• k ... .

1~~ 300-1

U)_2__00

0 -4 S43 310 3;0 400MACH NO. HEAT EXCHANGER EXFT TEMP., 'R

Figure 11. Inlet Air With and Without Cooling Figure 12. Heat Exchanger Weight Dependstor the ellected irajectory upon tnt( Compressor nim t Air

Temperature and Pressure

Requ i rements

ENGINE PERFORMANCE

Cryojet engine concepts with various gas generator configurations were examined includinghydrogen-oxygen combustion, hydrogen-air combustion and hydrogen expander. The regenerativehydrogen expander cycle (see Figure 7 for configuration) was selected for further investigationsdue to its high performance potential. Engine/airframe interaction effects have not been includedin the performance computations presented in the following p-rdgraphs.

Engine performance was computed for a specified flight trajectory, ligure 13. The selectedtrajectory has a dynamir pressure of 1000 lb/ft 2 at and above 30,000 feet altitude. Additionalassumptions used in performance computations are surmariz,:d in Figure 14.

SI-hFigure 3. Flight Trajectory Used forCryojet Analysis

21.

0 1 2 3 4 5

VACH NUMBER

•iIJI

wKV S" INMETI PRE5SSURE RECOVERY

rUtrY vARVLImE 1NAFT

P71FPESSU71i, RATIO 41 AT SEA LEVEL STATIC

EFFICIENCY 067

COUSUSTOR

!rXTi1~ ER TQ CORRTESPONDS TO ST .Ou-: TR,

UEL AIR RATIO

C ILMICAL. EOUftICe~iv#

TURBINE

*ROTOR INET TEMPERATURE 2000*A

I EFFICIENCY 09

I OZZL9*CONVEAGDAt OtWIFIRGETT

*FULLY VAITIARLE [XIT jAlLA

*TTITTST COErFICIENT. ogo

* 4AT EXCKANGEAf

MNEl BARE TVB1 MATR.X

*0125 IN 00150ET (".ME TEA TUBES

* MATE94AIA FAIr-ILE STEEL

Figure 14. Cryojet Engine CycleAnalysis Assumptions

Coi,ýiressor characteristics Indicate that this engine can operate up to flight Mach 4.5condition as a ti~rt~oraTmfet, Figure 15. At higher Mach numbers the engine would behave as an,ejector ramjet or a r~uiiet Cycle (compressor pressure ratio approaching unity). The Mach numbercapability without cooling inlet air would be less than 3.5. This provides an improvement of deltaM3c0. number of 1.0 for this cycln. H~owever, preliminary evaluation indicated that this improvementcould be greater than 1.5 depending on the selected cycle parameters, compressor characteristicsarid techno-logy levels.

The other Mach number limiting parameter is tile compressor discharge temperature (CDT). CDTat Mach 4.5 is relatively low indirating that the cycle can afford higher pressure ratio than theselected value. Therefore, cycle optimization requires that both of these Mach number limits mustbe satisfied.

'01/1

AD(0

A 7

figurE, 15. Compressor Oprerating P'-rfrjrminl.eLryojet Analysis

3012

As shown in Figure 16. specific impulse between 4500 and 5200 seconds is achieved for thespecified flight trajectory from sea level static to Mach 4.5 at 70,000 feet. This performance isbased on hydrogen with an overall equivalence ratic of unity as stated earlier. Specific thrust135-155 lb/lb/sec is obtained for the same flight trajectory, Figure 17. Exhaust nozzle tempera-tures range from 4518 to 4695°R.

Engine performance is sensitive to the inlet pressure recovery which in turn is influencedby the heat exchanger design. For one percent change in inlet recovery 0.3% change in thrust andspecific impulse (ISP) is noted at sea level static condition. Engine overall equivalence ratioalso impacts the specific impulse significantly. One percent increase in equivalence ratio resultsin approximately 0.8% reduction in ISP. However, thrust is not appreciably effected. Gasgenerator pressure did not seem to impact the engine performance. This is an important conclusionas it will have significant influence on gas generator and hydrogen pumping system designs andweights.

0 10 20 30 40 s,0 0 "12 20) 30: .0" so0 "

MACH NOIER 0MACH PVlM R

Fiure 1i6. Cryet ProvideS High Snpe ifir Fig,,rs- 17. Crynypt PrnvidPs HiatImpulse for a broad Maci Nunmber Specific Thrust for a Broad MachRange Number Range

The fundamental benefit of a cryojet cycle is illustrated in Figure 18. The cryoJet providesan optimum performance throughout the desired trajectory. Above Mach 0.5 the cryojet exceeds theturbine engine and the ramjet performance for equivalence ratio of unity. The cryojet cycle can beoptimized to satisfy thrust and fuel consumption requirements for any desired flight trajectory.It is anticipated that the upper Mach number operational limit (f a cryoJet is 5.0 or higherdepFending upon the technology and materials used. Since this concept doits nct operate on amulti-mode principal, it results in a smooth thrust variation as a function of Mach number. Also,the enginft control system is simplified.

,CR YOA ,

SPECiriC"iUPULSE(SEC, •Figure 18. Cryojet Specific Impulse.

"Lomparison with, tr,* funddn:entdlThermodynamic Lycles

0 2 4 4

UTHErk UESIGN CUNSIUERATIUN,

Cryojet OnqJifi'S retquiro ao, innovati, 1 engine/atrfrar,' integratton to r,.n y, th'. , rl .:performance potential ot thf, cychi. Inlet and nunle aria varlatl.n rfr,.. m-r,.h , p ý ,- i'velstat i c to high sJpersofi ic cruise" Machi riu.ciber c on it t i,,ns arm I arg(,. Ihtire(ior,', a u',(. .,K aI r I l IIIc'

external surfaces for inlet and exhaust nozzlli, may provide' a pctotntial sol'ittioi tu Lt. •,i(ijraton orprobil1m.

Sm iI

Low altitude cryojet engine operation may cause ice accumulation on heat exchangers andother engine surfaces. This problem can be rectified either by limiting air cooling to just above32"F at low altitude and low Mach number operation where stagnation temperatures are low orintegrating dehumidifiers or water separators with the heat exchanger. Additionally, the extent ofair cooling at low stagnation temperatures is limited due to compressor chocking effects. Furthercooling will not be beneficial to the cycle. Thus cryojet cycle benefits, while very significantat high Mach numbers, may be moderate at low Mach number operation. Therefore, careful enginematching and cooling modulation will be required to achieve maximum overall cycle improvements forspecified mission requirements.

ENGINE WEIGHT

Engine weights are estimated based on the current technology and materials including nickel-steel alloys for the heat exchangers. Preliminary estimate indicates an engine thrust-to-weightratio of 13.5. However, with advanced materials such as carbon-carbon and fiber reinforcedceramics for turbomachinery. beryllium for heat exchangers, innovative packaging (e.g., tip turbinedesign) and airframe integration, the thrust-to-weight up to 25:1 con ie achieved.

CONCLUSIONS

Cryojet cycle presents a viable propulsion system for 1itsijns iith wide Mach numberoperational requirements. This cycle provides an attractive ,'i2riate solution to two or moreindividual power plants or other potential combined cycles to cover the wide Mach number operatingrange. No new technologies are required to make the system operational. !he following conclusionscan be reached from this study:

1. Turbomachinery Mach number operating range is enhanced to Mach 4.5 for the selectedcycle. However, with proper matching and advanced technology and materials this limitcould be extended to Mach 5 or higher.

2. Increased corrected speed resulting from inlet air cooling causes the engine to operateat higher airflow and pressure ratio. These effects enhance the turbonarh!nreryperformance (thrust and specific impulse) at high supersonic Mach numbers. Specificimpulse and specific thrust levels up to 5200 seconds and 155 lb/lb/sec respectivelyare obtained.

3. Gas generator powered turbine rotor inlet temperature is independent of the flightMacth number. This feature of the cryojet improves engine flow characteristics andsimplifies tL rbine design.

4. The key required technologies (e.g.. heat exchanger and hydrogen handling systems)have already been demonstrated.

b. Modern heat exchanger design has significantly reduced its weight to the extent thatthe cryojet engine cycle is now mechanically viable.

b. improved performance and synergistic effects result in LumpaLt and higyhl thrust-to-weight ratio engines.

2. Large inlet and nozzle area variation requirements due to inlet air cooling willrequire innovative engine/airframe integration designs to realize full potential ofthe cryojet thermodynamic cycle.

PREMIXED, TURBULENT COMBUSTION OF AXISYMMETRICSUDDEN EXPANSION FLOWS

S. A. AHMED AND A. S. NEJADAdvanced Propulsion Division

Aero Propulsion and Power LaboratoryWright-Patterson AFB, Ohio 45433-6563

ABSTRACT

Velocity and low frequency combustor pressure oscillations have been measured in a ramjet dumpcombustor model. The mean and RMS values of the. turbulent velocity field were obtained using a two-component LDV system operating in the backscaxter mode. Reacting flow data were obtained for apremixed propane/air, while isothermal results were collected replacing the propane with nitrogen. Thevelocity data indicated substantial differences between the two cases. Coinbustor pressure oscillation datawere obtained utilizing high frequency dynamic pressure transducers. The intensity and frequency of theoscillations were measured as a function of inlet velocity, combustor length, and equivalence ratio. Resultsshowed that pressure oscillations were controlled by both vortex kinetics in the combustor and acousticresponse of the inlet section.

INTRODUCTION

Turbulent sudden expansion flows are of significant theoretical and practical importance. Suchflows have been the subject of extensive analytical and experimental studies for decades, but many issuesare still unresolved. Because of the complexity of the flowfield and the associated difficulties withweasurements, detailed information on reacting sudden expanmi,m flows is very limited. On the other

hand, development and evaluation of analytical models, especially for confined recirculating flowconfigurations, have been hampered by the lack of reliable and detailed experimental data. To furtherdevelop these codes to become general and applicable to more than a limited range of simpleconfigurations, reliable and well documented experimental data are a must. Therefore, the objective of thecurrent study reported herein was to obtain a credible and detailed experitental data base and to help inthe understanding of the behavior of such flows. There is some limited dam available in geometries whichparallel the current study. However, in one way or another, they lack details and accuracy, or do notrepresent a realistic geometry or combustor operating condition. Several researchers have reported data ingeometries which ma:y compliment the current study. The most recent and relevant to the present study arediscussed in the following paragraphs. Flow characteristics of interest which were examined included thevelocity field characteristics, the length of the comer recirculation zone "CRZ", and the pressuredistribution.

Pitz and Daily (1983) investigated reacting flow behind a step in a rectangular duct with a 2:1 arearatio. They reported that the presence of reaction (equivalencc ramio = 0.57) reduced the length of the CRZby 20 to 30 percent. Turbulence parameters were not of prime interest vid were not heavily discussed.The limited data that was plotted indicated combustion had insignificant effect on turbulent kinetic energy"TKE".

El Banhawy et al. (1983) studied reacting flows; with equivalence ratios ranging from 0.77 to 0.95for expansion area ratios of 2.0 and 4.0 in a coaxial dump combustor, and observed a CRZ extending aboutthree step heights. The reported value, less than half of the usually observed value in nonreacting flows, is30 to 40 percent lower than the documented results of Pitz and Daily (1983).

Stevenson et at. (1983) reviewed relevant results and made measurements in an axiiynmmelricsudden expansion with an equivalence ratio of 0.28, suggesting non-prenuxed flow conditrns They foundthat the reaction shortened the CRZ by roughly 15 percent, increased the marximum backflow velocity and

"Approved for public release, distribution is unlimited."

305

the vxial TKE on the centerline while reducing TKE elsewhere. In agreement with prior investigators,theyreported the significance of axial pressure gradient on CRZ length in nonreacting flows, indicating that"this pamneter needs to be monitored closely and riolated before considering the effects of combustion onthe behavior of the CRZ.

Smith and Giel (1980) evaluated a coaxisl hydrogen-fueled confined burner for an equivalence ratioof 0.14. The diameter ratio was approximately 2.5:1. In their experiment, the central air velocity was102.1 m/sec (335 ft/sec), and the hydrogen flow in the annulus was 0.91 m/sec (3 ft/sec) which was so lowthat it did not prevent the flow from exhibiting a recirculation pattern near the wall region. Reaction andheat release caused a significant change in the axial pressure distribution and the length of the wallrecirculation, i.e., decreasing axial pressure gradient and doubling the length of the recirculation zone.Similar studies by Chriss (1977) showed no change in wall recirculation length when the diameter ratiowas 10: 1. The maximum TKE showed little change with reaction, although the peak value occurred fartherdownstream for the reacting case. None of the above studies reported, in a systematic manner, the totalcombustiom tU.',ts on the combustor flowfield. Therefore, it is necesuary to try to establish a data base forrealistic coir.;- 'to 'qeometries and document the results in sufficient detail which can be useful to thecomputational comn..•.. ty.

EXPERIMENTAL SET-UP

WATER COOLED COMBUSTOR

Figure I shows the test section ,r detail. New design methodologies were implemented to allowstable and continuo&; combustor opermiiou over long periods of time. For example, an air-cooled quartzwindow used for LDV nmeasurements wow s.*.; ;.bled on a flexible mount (combination of springs & bolts)to allow for the differemal thermal expaiii' "f the quartz and the metal surfaces. The water cooledstainless steel combustor w,, conceptually wwi (4'ysically identical to the plexiglas rig used earlier byFavaloro et al., 1989. For t!. , -l'rent ex, -riment, 4 side-injector tubes (90 degrees apart, 1.52 mm (0.06in) diameter ports) were -,. for premixz-i, %tudies. Injector ports were located 12.7 nun (0.5 in) fromthe 101.6 man (4.0 in) ri:t pipe centerline jus% L,-wnstream of the 50.8 mm (2.0 in) 04ameter orifice plate.The pons wo.w. a.'jrad to point toward the 6ov.,stream direction. This configuration provided goodpremiixe,) m'. sid minimized flow oscillations * bile eliminating flashback for the range of operatingconditi.,, c- yied herm. To ignite the fuel/air niixtur,ý, •wo spark plugs were flush mounted 180 degreesapart b: . 1-t tY f,.x.e.

INSTRUMcir('f,~ 'i0

"Both n"t-,mt-uuive and conventional techniques were used in the :.sent study. For the velocity andturbulence measurements, a two color, four beam backwcatter LDV syncran was used. The details of thissystem were reported earlier by Favaloro et al. (1989). Seeding •,• acc.%mplished via reaction of TiCILand N1O producing TiO, particles of approximately one micron in Jin•me,

Temperature and pressure measurements were accomplished by using %onventional instrumentation.All temperature measurements were accomplished by K or R Type thermocouplee. Measurements in lowto moderately high temperature environments, such as the inlet air, water jacke', fuel, combustor outerwall, and exhauster temperatures, were accomplished by using K type thermocouples. R typethermocouples were used to measure combustor tunperature profiles. Pressure measurements were madeby using thin-film pressure transducers (CEC 55000). Two High frequency Kulite pressure transducerswere installed in the inlet pipe and the combustor chamber for dynamic pressure measurements. A NicoletScientific Corporation Model 660-A dual-channel FVT analyzer was used to obtain spectral densities ofpressure signals.

A Yokogawa vortex flowmeter Model YFI00 was used to mrasure inlet volume flow rate. A FlowTechnology Model Fr-08AER I -UEA- I turbine type flowmeter was used to measure the fuel flow rate. ANeff "System 470" data acquisition system was used to calculate, display, and monitor all pertinent flowparameters such as mass flow rates, temperaurs, and pressures.

306

EXPERIMENTAL RESULTS

The control parameters chosen in operating the water-cooled combustor were the upstream meanvelocity U., the length of the combustor L, and the equivalence ratio *. Table I shows the operatingconditions chosen for the instability studies. The results were accomplished through detailed fluctuatingpressure measurements. In addition, some veloc,.v measurements were taken for various combustorlengths to check velocity-pressure correlationr. For most of these operating conditions, periodic oscillationin the pressure was noted and recorded. "jhe details of the results will be discussed in the followingsections. The case of U,,, = 18.25 m/se=; s60 ft/sec) and * = 0.65 was selected to examine velocity andwall static pressure measurements in oj',ch.

# L&h U", ft/sec

0.6 28.5 31.21 44.8 61.37 75.47 88.8. 104.67 120.35

0.6 40.5 31.52 44.81 60.65 76.77 92.33 105.31 120.41

0.6 52.5 31.09 4522 61.52 .7 _.'2 8821 106.41 120.88

0.65 28.5 30 45 60 ,2 92.14 106.97 118.67

0.65 40.5 30 45 60 75 90.86 105.72 120.73

0.65 52.5 30 45 60 75 89.81 105.58 119.15

0.70 28.5 30.82 40.8 60.5 74.26 88.77 104.14 118.22

0.70 40.5 31 O 45.63 161.30 i4.90 90.93 107.17 119.46

0.70 52.5 ..30 It5.59 [l61.31 75.16 90.) 105.78 1l9.48

"�..'. 1. 0(-frerir Co.aditions for the Intability Vesults

MEAN FLOW RESULTS

Initial Conditions. Figure 2 sh3,.%s typical inlet velocity prof'le upstream of the dump plane. Theplots show variations of axial velocity and normalized RMS velocity profiles actm the inlet pipe whichare similar to fully developed pipe r, pdata. It is worthy to %Ae that tde iniet flow is smooth andsymmetric indicating a well-controlleci c€mbnstor inlet flow. The turbulence level a the center was about5%. This increased away from the cerwhroine and reached a maximum value of 9. S% in the proximity ofthe wall, which is in agreement with pW.vious results reported by Samimy et al. 0988) taken in a similarplexiglas rig.

Mean Velocity Results. As explait-.i -:airs, one case was selected to he examined in detail. Theselected case had a reference velocity of 18.25 aLrhx. (60 ft/sec), co•J,,puty',g to a Reynolds number of1.18 X 10( based on inlet velocity and pipe diamel:.:,_. '"1" cc•l'oo•tor t:,41h was increased from X/H =0.38 to X/H = 24 to allow for velocity measurements at dithicent axial locations (i.e. X/H = 0.38. 1, 2, 3,4, 5, 6, 8, 10, 12, 15, 18 and 24). However. only the results of four axial locations are reported here toillustrate the flowfield characteristics of reacting and coon-reacting cases. These representative axiallocations are X/H = 0.38, 3, 6. and 12. Figures 3-6 display the substantial differences between the twocases and illustrate the significant effects of combustion on the flowfield characteristics. Similar resultswere reported earlier by other investigators such as EI-Banhawy et al. (1983) and Stevenson et al. (1983).Figure 3 shows the evolution of axial velocity profiles which exhibited insignificant differences betweenthe cold and hot flow in the central region 0 < r/h < 2, especially in the near field. Differences were notedin the shear layer around the fl - 2 which grows in size with increasing axial distance. Thisdemonstrates that in the near field, combustion is mostly limited to the shear layer. Away from the dump

•;'7

plane (Le. X/H > 6), the shear layer increased in thickness tremendously to the limit of occupying theentire cavity of the combustor. In the far field (X/H 2 12) due to combustion and heat release, the velocityis substantially higher and the profile is more unifonm indicating flow recovery in a short distance ascompared to the non-reacting case. In this aspect, combustion effects are similar to inlet swirl effects.especially their effects on reducing the corner recirculation region length by more than 40%. Tangentialvelocity measurements showed insignificant values scattered around zero, and for brevity they are notshown.

Figures 4 and 5 show the corresponding radial distributions of the axial and tangential RMSvelocities. In general, the axial turbulence intensities are larger than their corresponding tangential valuesin both cases of hou and cold flows. Also. the maximum intensities of both components occurred at thesame radial location. It appears that the maxima in the hot flow case are located closer to the wallsuggesting shorter recirculation region length which is in agreement with the results of Figure 3. It isobvious that the turbulence level increased due to combustion in both directions except at X/li = 6. inspite of this, the local turbulence intensities, u'/U and w'/U, are smaller in the hot case than theircorresponding cold flow values. This is in agreement with some of the earlier results of EI-Banhawy et al.(1983).

Contour plots am usually presented to help with the interpretation of the meawurenments and toprovide a better picture of the flowfield. Valuable information such as regions of maximum or minimumand zero lines can be easily identified. For example, contours of the normalized axial velocities are shownin Figure 6. The reverse flow boundaries defined by U/U,, = 0.0 are shown. In the presence ofcombustion, the length of the comer recirculation zone decreased 44% from 6.7 to 3.75 step heights. Aclose examination of the plots show that, in spite of the reduction in the comer recirculation region length,the flow inside the bubble is more 9, ive as demonstrated by higher values of negative velocities. Thecontour lines show substantial increase in axial velocity across the combustor downstream of X/H = 10.This is an indication that combustion has taken place across the entire cross section of the combustor atthat axial location and also downstream of it.

INSTABILITY RESULTS

Several cases were investigated to examine the effects of combustor geometry and flow parameterson combustion instability. The different values of these parameters are listed in Table 1. Although the leanblow-off limit occurred for an equivalence ratio of 0,51. all experiments were conducted at higherequivalence ratios (# > 0.60) to avoid flashback and combustion oscillations associated with operating nearlean blow-off limit Here, the main objective was to characterize the pressure oscillations of the flowfieldfor different combustor lengths and different inlet flow velocities. Depending on the operating conditions,higher harmonics or subharmonics may appear and become quite intense. Of course, these differentspectral components contribute to significant cyclic variations in the amplitude of the oscillations, For allspectral analysis the sampling rate was 2.56 times the maximum frequency for dual channel analysis. Formost of the data, 150 averages and 500 Hz frequency range were selected. Each block of data contained400 samples which resulted in a resolution of 1.25 Hz.

Power spectral analysis of zero flow did not reveal any distinct frequencies in the range of interest <200 Hz. However, lower frequencies (17.5, 60, and 160 Hz) were recorded in the cold flow case of U., -32 m/sec (105 ft/sec). The 60 Hz frequency was slightly more dominant and narrower than the other twofrequencies. For free shear layers generated from an initial boundary layer with a uniform momentumthickness, such as the present configuration, linear stability theory (Michalke, 1965) predicts the strongestroll-up frequency to be approximately 220 Hz (i.e. F = 0.017 U,, / 0o where U,,, and 0. are the inletvelocity and momentum thickness which was determined to be 2.9 mm, 0.11 in). On the other hand, therange of the first acoustic mode of the combustor (i.e. L varied between 0.87 - 1.48 m, 34.25 - 58.25 in) is120-200 Hz. Similarly, the longitudinal acoustic mode of the inlet pipe is 58.5 Hz considering the totallcngth of the inlet pipe which is approximately 3m (118.1 in).

Apparently, the recorded frequency of 60 HIz is excited by the longitudinal acoustic mode of theinlet pipe. Generally speaking, the system acoustics play an inponant role in determining the nature of theinstability. However, for most of the operating conditions, the results indicate that the instability

mechanism cannot be purely acoustic but rather a combination of the system acoustics an-d the flowinstabilities which is mainly dominated by vortex shedding. One can see how complicated the flow hasbecome when combustion is introduced. Figures 7 and 8 reflect this fact. Figure 7 shows the differencesbetween the two signals recorded by the pressure transducers for # = 0.7, # = 0, and U,,, = 32 m/sec (105ft/sec). Note the presence of higher frequency harmonics recorded in the inlet (160 and 208 Hz) and howthey were amplitude modulated showing wider band. On the other hand, Figure 8 illustrates the effect ofU,,, on the coherent output power of both transducer signals showing the monotonic increase of theoscillation frequency with increasing U,, for = 0.7 and LH = 40.5. These frequencies were recordedand the results are summarized in Table H. Most of the time, these frequencies did not match inletacoustic resonance frequency. Therefore, it is worthy to examine the relationship between the inletvelocity and frequency as a function of combustor length while keeping a fixed value of +. When thevelocity is increased, the frequency increases approximately in a linear fashion as shown in Figure 9. Thissuggests that the frequency is closely related to the vortex kinetics in the reaction zone. On the other hand,the effect of the combustor length is also obvious since the frequency of oscillation increases as combustorlength decreases for the same values of # and U., This further suggests that there is a relation betweenthe pressure oscillation frequency and the combustor acoustic mode. Hence, using the data of Figure 9 andreplotting I/F versus L/U., (Figure 10) shows that most of the data points fall on a straight line. A linearregression through these points has a slope of 0.095 and an intercept of 9.2 msec. The intercept time of9.2 msec closely corresponds to the round trip travel time period of a pressure wave generated in thecombustor and reflected from upstream boundaries to return to the combustor. This corresponds to afrequency of 55.5 Hz, which is roughly equal to the calculated acoustic mode frequency of 58.5 Hz. SinceLJIf is approximately equal to the vortex lifetime (i.e. assuming the vortex breaks up when it impinges onthe exit nozzle, in reality for long combustots the vortex might disperse before reaching the nozzle), 'hetime period appears to be the sum of the vortex characteristic time LIJU.,, and the characteristic inletacoustic time. In other words, the vortex lifetime is associated with the characteristic convection time inthe combustor while the round trip time characterizes the acoustic response of the inlet. In sunmmry.combustor instability frequencies are determined by both voitex dynamics and the inlet section acousticresponse time. Also, a large range of other frequencies were excited which were not related to the naturalsound wave modes of the system.

- LA= F (Hz)0.7 28.5 53.75 6 0 76.25 80 90 97.5 100

0.7 40,5 51.75 60 68.75 75 82.5 88.75 93.75

0.7 52.5 47.5 53.75 1 61.25 67.5 76.25 86-25

Table If. Summary of the Instability Results for * = 0.7

The distribution of total RMS of the pressure signal with combustion is significantly higher than thecorresponding isothermal flow case. All data show an increased RMS value when either the upstreamreference velocity orf is increased or the combustor I ngth is decreased. For brevity, only the results of= 0.7 are piesented, see Figure II.

SUMMARY AND CONCLUSIONS

A detailed experimental investigation was carried out to determine the effects of combustion on theflowficld chamratrislics of a model ramjet engine. The study also described the low frequency vortex-driven pulsating i.ombustiont modes and the natmre of the pressure oscillations observed in the flowfield.The results showed the significant effects of combustion on the development of sudden expansion dumpc.mb•uior flowfield. For example, the mixing layer, inferred by velocity measurements, shifted towardsthe combustor wall. As a result of this, comer recirculation region length decreased by about 444%. Thiswas accompanied by a much faster flow recovery, i.e., flat velocity and turbulence profiles not far from thedump plane.

. _- = = = 1 )

Systematic manipulation of the combustion instability frequencies showed that the resonant period ofthe oscillation is determined by the sum of the vortex convection in the combustor and the acousticfeedback time of the inlet. For shorter combustors, the oscillations were more intense and their frequencieswere higher. Similar effects were noticed when the reference inlet velocity was increased.

Further detailed and refined data should become available in the near future. Future studies willinclude the effects of combustion on the flowfield characteristics of combustorq with inlet swirling flows.

ACKNOWLEIDGEMENT

The authors would like to thank Messrs John Hojnacki and Charlie Smith for their continuoussupport. Special thanks goes to the Air Force Office of Scientific Research and to Dr Julian Tishkoff forproviding sponsorship.

REFERENCES

1. Chris, D.E., "An Experimental Investigation of Ducted Reactive Turbulent Jet Mixing withRecirculation," AEDC-TR-77-56, AFOSR-TR-77-0749, ADA 044110, 1977.

2. El Banhawy, Y., Sivasegaram, S., and Whitelaw, J. H., "Premixed Turbulent Combustion of aSudden Expansion Flow," Combustion and Flame, Vol, 50. 1983, pp. 153-165.

3. Favaloro, S.C., Nejad, A.S., Ahmed, S. A., Miller, T., and Vanka, S. P., "An Experimental andComputational Investigation of Isothermal Swirling Flow in an Axisymmetric Dump Combustor," AIAAPaper No. 89-0620, 1989.

4. Michalke, A., "On Spatially Growing Distributions in an Inviscid Shear Layer," Journal of FluidMechanics, Vol. 23, 1965, pp. 521-544.

5. Pitz, R.W., Daily, J.W., "Combustion in a Turbulent Mixing Layer Formed at a Rearward-FacingStep," AIAA Journal, Vol. 21, No. 11, 1983, pp. 1565-1570.

6. Samimy, M., Nejad, A. S., Langenfeld, C. A., and Favaloro, S. C., "Oscillatory Behavior ofSwirling Flows in a Dump Combustor," ALAA Paper No. 88-189, 1988.

7. Smith, G. D., Giel, T. V., and Catalano, C. G., "Measurements of Reactive Recirculating Jet Mixingin a Combustor," AIAA Journal, Vol. 21, No. 2, 1983, pp. 270-276.

8. Stevenson, W. H., Thompson, H. D., and Gould, R. D., "Laser Velocity Measurements and Analysisin Turbulent Flows with Combustion," Part I and H, AFWAL-TR-82-2076, 1983.

Upstream Sliding Swirler BracketAccess Piston louginWindow Spacer -

Inlet ..... az %astr/ /////2ZAir & tS e e d ; " " ' r

+FuelOr N2•

Water Cooled Piston Graphite Measurement -Pipe g ins cao

Fig. 1. Schematic of the Dump Combustor Model

310

t o 2 .0 - - - - - T -. . . - - -. - "- - -- - I

1.4 o UUtef X/H 0.38•U/t~k'f le X/H •3.0•

z X/H " 1.0

1.2tt2

.6X/H- 10 .4

1.4

.2COLD NOT

0 .4

.2 -1 0 1 2-3 .2 -

RADIAL LOCATIO, ruH

Fig. 2. Upstream Inlet Conditions Fig. 3. Evolution of Axial Velocity Profiles(D 0.65

to 1.O0[

o XiH 0.38 o X/H - 0.38

.8 XH - 3.0 . X / -3.0

XH .0 /H - 6.0* -X/H - 12.0

.6

S: .44

COLO HOT COLD HOT

-3 -2 1 0 1 2 3-3 .2 -1 0 1 2 3

r H fH

Fig. 4. Evolution of Axial Turbulent Fig. 5. Evolution of Tangential TurbulentVelocity, 0 0.65 Velocity, 4) = 0.65

3.0

2.0LVEL V W.,e

109 1.40O 028 1.20

S"HOT ,• 7 1.O03.0 t 0.80S5 C to

3 a.,zO22 ..0

01 -0.20-1.00

44.'

-3.02.5 5.0 7.5 10.0

X/HFig. 6. Contours or Mean Axial Velocity, ) : 0.65

311

2,8"F 94j3

.76 MRc

S~ 0 - 0.7_-.,.W 120 ft 094;

COMBUCTOR 4161.76 MR

0.1f Is ft *9C

2-. *- thsf 46 11**

A- -- -- 8f 30 ft &.c

0 0 0 00 40 50 0 to0 200 300 400 500

FREQUENCY, Hz IEMC.H

Fig. 7. Power Spectra of Pressure Oscillations Fig. 8. Coherent Output PowerL/H =52.5, Urnf 32mlsec Spectra

120 L

*L M40.6 .

100- 40-

so 20 0

25

0 -. ..........- ,---.-.,--.-,.0

0 0

0 25 so 73 too 125 1!0 u so 10o ISO

t FRE UE ft Yc L klet,. m.s

Fig. 9. Pressure Oscillation Frequencies vs. Fig. 10. Oscillation Period vs. CharacteristicInlet Velocity, 4) 0.7 Flow Residence Time, r) 0.7

700 - ______-

.LH 405 0

So L H -6Z'.6

600

300 /

I. : .: H -40A• .

2601 -, •b 07

0 25 50 75 100 125 ISO

U1411 It sec

Fig. 9.. Turbulent Pressure RMS Values vs.Inlet VelocitV y I = = 0.7

,12

Tee ONKLOPI1Y or A LIQUID FUELXD OAS MM".TG1kr0lt OUCT3D PDCIWT RFU3AACU

P. L. Buckley and E. CorporiaoW4right It'asearch and flevelc.pment center

Wright-Pattersu-n M.?B. Ohio

A liquid fueled gas generator is described which was designed to simulate theeffluent produced by a solid propellant )Lydrouarbon fuel, typical of those usedfor ducted rocket propulnion, by burning at very fuel-rich mixture of liquidhydrocarbons with gaseous oxygen. The gas generator was built to function as atool for the dsvelopmrent of :,omponents of the xamburner portion of the ductedrocket system and to irovida a mkeans for scudyiag thc; tk~ermochemical processesinvolved ink the production of the fuel-rich effluent and _;t-, aecondazy coMLbustion.'th 4ir ir. the rairburner. Successful operation has been documented at total fuela~id oxid.lier flow rates throuqh the gas generator of .20 kg/sac (.44 lbs/sec)during direct--connect ducted rocket, combustion testn.

INTRODUCTION

Many facilitifes ifhich havj beign usid in the past for liquid or solid fueledramfjet research are restricted from doi.-c similar work with the ducted rocket, dueto th6 nature of the solid propeillant fuels used in the ducted rocket system.Because the facilities in the Aero Propu:.eion and Power Laboratory fall into thiscateyory, a program was undectaken to develop a suitable simulation of the solidpropellan,ý ivalm for ut-c in our cucted rocket: combustion research.

Previous remearch1 with a relatively low pressure gas generator, which burneda mixture of ethylene, toluenie, and gaseous oxygen, demonstrated the simulationconcept and provided1 valuabl" insight into determining the gas generator effluxcompoqiticn. baned on thim work and on liquid fueled gas generators used forliquid rocket &yutems and auxiliary power units, a gas generator has býeendeveloped which burns a mixture of J'P-7, toluene, and gaseous oxygen at resauvesand flow rates relmvant to current ducted rocket research requirements. In thispaper, the deisign azio operatiorn of the gas generator will be described, andperformance data will be presented for comparison to references used in thedesign, arid to thermochem4 cal predictions ba~sed on measured fuel and oxygen flowrates to the gas generator. Future work will include sa-mpling of the effluxcomposition for comparison to actual solid propellant output.

Althouigh not a ma-jor topic of cits paper, the operation of a ducted rocketramburner, fueled with the gas generttor, willi also be described.

FýUZL FrOPHUL*T ION

The solid propellant chosen for simsulation was compot.ed of a hydrocarbonfuel, an oxidizer, a binder, And various minor constituents such as combustioncatalysts. In the interest of simplicity, only the overall proportions of carbon,hydrogen, and oxygei. in the propellant formulation wore uiintulated, neglecting theminor con~stituents which comprise only approximately 5 percent of the composition.Thi.m :esulted in a simplified composition of:

Thin vWasF simulated with a mixture of TP-7 (a m).itary jet fuel), toluene, andgo-eous oxygen as shovn here:

3122 r~*1 1:H 4 . 346 C7 H8 4- .532 0,.

"tie c-hoice of liquid fuels was baned on the abilxtt1 to blend the two toobtjair, the correctr oveirall ratio of carbon and lkydrog'ýn, and because both wereicadi ly avail abl. For the f low rares anrl pressures re,1;ired, gase-ous oxygeni wast.hq monot pr-act ic-.%l choice for thO oJX1rd1ZeL

Ajj'o''-lfo pblcrolqas-i; iisiiu o n urlimiftel

SGENERATOR DE8 IGN

CONFIAURAT ION

The combustor design for the gas generator was based on representativeconfigurations of fuel-rich gas generators, used primarily for driving turbopumpsin liquid-fueled rocket engines and electrical generators in auxiliary powerunits. An excellent summary of basic configurations for liquid-fueled gasgenerators was published by NASA 2, and addresses the difficulties in achievingsufficient mixing for complete reaction of the fuel and oxidizer, and forobtaining uniform temperature profiles at the exit point. Because the internalflow velocities are relatively low, limited energy is available to aid in mixing;thus, most configurations include some means of mixing or stirring the reactantsin the gas generator combustor. Earlier designs, which were of cylindricalconfiguration, included internal mixing rings or baffles as described in Reference2, while later designs, such as those used in the H-1 and Atlas booster engines,used a bulk flow reversal in the gas generator for this purpose. More recentexamples of experimental gas generator designs described by Schoenman 3 and Benskyand Wong 4 revert back to the cylindrical configuration, due at least partially tothe ease of construction made possible by using cormtercial pipe and flanges. Thisconfiguration was chosen for the construction of the gas generator described inthis paper.

SIZING

Sizing of a gas generator combustor requires consideration of both thediameter, based on desired velocities through the combustor, and of the internalvolume to obtain sufficient residence time for the desired reactions to takeplace. These factors must be traded off against the practical considerations ofbuilding a vessel to contain the high pressures and temperatures associated withthe gas generator operation.

Bensky and Wong 4 suggest a maximum velocity of approximately 10 m/s (33ft/sec) through the gas generator for the flowing reactants before ignition isinitiated. A diameter of 50 mm (2 in) was determined to easily satisfy theseguidelines for the expected gas generator flow rates. Provisions were also madein the design to enlprge the internal diameter to 75 mm (3 in) by removing aheavy-walled cylindri.al liner.

The internal volume (actually, the leiigth) of the gas generator was sizedthrough the application of combustor "stay time" guidelines. As defined bySutton5 for use in liquid propellant rockets, the stay time describes the averagetime each molecule stays within the combustor volume, given as:

t3 - Vc /(wv)where;

V- chamber volumew - total propellant flow ratev - average specific volume of the gac in the chamber

According to Sutton, typical stay times for liquid propellant engines range from2 to 40 msec, while Reference 2 lists values of 5 to 10.5 msec for a variety ofliquid oxygen/liquid hydrocarbon fueled gas geneLators. An overall length of 356sun (14 in) was selected for the gas generator, which results in a stay time ofapproximately 25 msac, based on the expected flow rates and chemical equilibriumpredictions of the specific volume of the combustion products.

COf'qTRUCTION

Av shown in Figure 1, the gas generator construction was based on the use ofcoriaercial, stainless steel, three inch Schedule 40 pipe and 600 pound flanges,providing a considerable safety factor ovor the design operating pressure of 3500kPa (500 psi). The design was configured to accommodate a variety or fuelinjectors, internal mixing rings, and exit nozzles. Also, tho internal diametercould be increased by removing the atainless steel liner (which had a wallthickness ot 12.7 mm) The ignition source consisted of t hydrogen/oxygenigniter, located 21.9 mm (.875 in) from the top of the c:ombustor. Mixing rings,12.7 mm (.500 in) thick and with orifice diameters ranging from 25.4 to 38.1 mm(1 to 1.5 in), wore fabricated for installation between the flanges at the mid-

314

point of the combustor to enhance mixing and gas temperature uniformiLy at thecambustor exit. Only a single mixing ring, with a 25.4 mm (1.0 in) diameterorifice, was evaluated as of the writing of this report. Tie stainless steelexhaust nozzles were 19.1 = (.75 in) thick with diameters rarging from 6.35 to19.1 mm (.25 to .75 in). The nozzles, consisting of a cons:ant radius entrysection followed by a cylindrical throat, were each calibrated against a standardorifice plate to determine their discharge coefficients, with res lts ranging from.88 to .92. Gas generator operation was documented with 9.75 mi (.384 in) and12.7 mm (.500 in) nozzles, operating over a range of fuel flow conditions.

B'~d w Jg.Cra

rig. 1. Gas Generator Configuration

FUEL MMD 0XYGEN INJEC IL2H

It is logical to assume that the gas generator performance is greatlyinfluenced by the degree of mixing between the fuel, and oxidizer. What is notclear is the optimeim degree of evaporation and mixing prior to the onset ofcombustion of the indiv.*dual fuel droplets in the fuel--rich environment of the gasgenerator. It may well be that controlled evaporation of the fuel erhances thecombustion processes by limiting the local equivajlencoo ratio, thsis forming a"Pilot zone" of partially evaporated fuels surrounding the remaining fueldroplets. The fuel injector ploys a vital role in the atomization, ponetrationand dispersion of the fuel within the combustor, and therefore has a stxonginfluence on this process. Likewise, the method of inj*CLing oxygen into the gasgenerator is very important to the mixing process.

The appreoach taken in developing the fuel injector conufiguration for the gasgenerator invcolved surveying the available literature for both liquid rocketengine and gas generator designs, and attempting to employ the concept# foundthere in a manner consistent with practical fabrication tv'.hniques. Impinginginjectors were designed, fabricated, and tested ir. order to compare theirperformanc~e with commercial atomizing nozzleiA. These wore in the traditionallike-on-lik, fuel impingement configuration, with the addition of an oxygen lotflowing axially through the fuel impingement point. Unfortunately, they weretested over only a limited range of conditions, and comparison to other injectorswas not possible.

Two commercial prevsure-atomiziiug noztles were compared, one with a full conespray pattern and one wirth a hollow cone. Theme injectors were installed in thecontsir c.,f a fuel iaojeaLus plate, with orygen injected axially through four equallyspaced poL-to, 4.76 wA (.180 in) in diameter, distributed around the face of theinjector plate on a 35 mm (1.375 in) diameter cir-le. becmuse the hollow coneinjector was siz'sJ for relatively low fuel flow rates, two addlitional injectors

1|1

were employed for operation with higher flows. These were simple 1.32 tAm (.052in) orifice-type injectors located just below the igniter flange and oriented atopposite sides of the combustor to spray toward each other and mix with the fueland oxygen flow coming from the injector plate in the top of the gas generator.

rFUL AND OXYGEN SYSTEMS

The fuel and oxygen systems were designed to deliver the flow rates necessaryto provide a meaningful test envelope for the ductid rocket combustor withrealistic total pressures in the gas generator.

The liquid fuel blend was supplied from a tank pressurized to 6900 kPs (1000psi) with gaseous nitroQen, and the mass flow rate was measured with a K-flowCoriolis-type flowmoter. Oxygen was supplied from stan4ard high pressure tubers,and regula'ed t,7 2000 kPa (290 psi). The flow rate was measured with a turbineflowmeter. A s9ngle cylinder of hydrogen was used to fuel the hydrogen/oxygenigniter for the gas generator and the hydrogen/air torch for the ramburner withflow rates controlled by sonic orifices. Nitrogen, also oupplied from hrigh-rossura tubers, was utilized to purge the combustor and hydrogen lines, and toprevent contamination of the oxygen lines with fuel vapors between runs. Also,check valves were installed in the oxygen lines to help prevent the reverse flowof combustible products into the lines.

pyo'xy ROCKET COMJUST9O•.

The ducted rocket combustor consisted of two rectangular inlet arms, acylindrical combuntor fabricated from six-inch Schedule 40 staitnless steel pipe,and a watt.- cooled exhaust nozzle which providad aet area ratio of 60% of thecombustor area. As illustrated in Figure 2, the two parallel inlet arms were

connected to elbows to enter the lower quadrant of the coribustor at 45 degreeangles to its centerline. The inlet arms had an area of approximately 55 cm2 (8.5in ) each, And each contained a diverting vane at Ito exit to increase the airflow into th. domre area (forward area of thq cirhsjator) A 2-D converging-diverying choked :.ozzle was located at the entrance of each of the inlet arms tosimulate the missile inlet system and to ensure equal flow through both inlets.An aerogrid was inrta;lsd in both inlots between the diverging portions ot thechoke blocks and the 40 degr~ee elbows to help ensure uniform velocity profiles.A hydrogen/oxygeor torch and an ethylene igniter, located 36.8 cm (14.5 in) and 3cm (1.2 in), respectively, from the dome were used to ignite the effluent expelledfrom the gas gener'ator into the ramburner.

GAS 19,1.9 COMOUSTOn

R1NAU~ T OZZLY,SDIVERTING VANE'

CHOKED NOZZLr

AIROGOID

Fig. 2. l)uoted Rocket. Combuetor

9XRAfl111TAL FP.OCSOP.

Gas generator operation was preceded by establishing the proper vitiated asi

flow rate and te.'ljerat'ire hrn-lgh tho rAmharnsr. The 9gs 9anarator ,oxygen anitfuel flows wore preest. by actuatsng their tzeupective on/off valves and manually

adjusting the control valves from the control ianel. Aft.ez the fuel and oxyger)valve settings were established, the gas genoeraLor wao putged, anr: Lhehytrogyn/oxygon torch in the rams•,%tnor wao started to ensure that the gns

genesator counustioti products were butrned off betfre tilter aill the exhaust Synt lt.

116.

From this point, the gas generator operation was controlled by a Siemens S5-101Uprogrammable controller. The controller was employed to actuate air-operatedon/off ball valves via 24 VDC solenoid valves, and to activate the spark plugs forthe hydrogen torch and igniter. The program was designed to initiate a sequentialprocess when a momentary "run" switch was actuated. The sequence began with a 5second nitrogen purge of the gas generator via the oxygen lines. After a twosecond oxygen lead, the fuel flow began, along with the simultaneous firing of thehydrogen/oxygen igniter to initiate combustion. The gas generator tests normallylasted 20 seconds, during which time the efflux was burned in the ramjetcombustor. The torch in the ramburner was turned off after the onset ofcombustion to ensure accurate performance measurements. The gas generator wasautomatically purged at the end of the programmed sequence or whenever the "run"switch released.

Data was recorded continuously from the time of ignition to the end of eachtest with a VAXSTATION 3600 through a CAMAC A/D system. This system scanned allthe data channels in approximately one second, with each of the readingsrepresenting an average of 64 samples of each channel.

ASSULTS

A series of experiments was conducted to investigate the effects ot changesin internal volume, fuel injection, and combustion pressure on the gas g"neratcrperformance. This work was accomplished over a range of total fuel/oxidizer flowsto the gas generator, at approximately constant mixture ratios. While theseresults are somewhat preliminary, they point out interesting characteristics offuel-ri.h gas generator operation. Table I lists the variations in configuraticnand opa.'ting conditions tested as of the writing of this paper.

Table I. Gas Generator Configurations arid flowRates Tested

VABIABLE TE3T gONDITIONS

Internal Volume With and without liner

Internal Pressure Two exit nozzle diameters

Fuel In-ection Hollow cone spray nozzleCombination of hollow cone and

radial injection downstreamFull cone spray nozzle

Total Flow Rates .068 to .200 kg/*ec(.150 to .440 lb/aec)

The mixture ratio (oxidizer/fuel, by mass) required to simulate the oolidpropellant was calculated to be MR - .347, which war kept relatively constant forall the tests. The fuel injector utilized for all tests was the hollow conepressure atomizor, unless otherwise specified.

The characteristic velocity, c', was employed as the criterion for evaluetingthe performance of the gas generator. Although not an ii.dication of how w,il thegas generator's effluent resembl.es that of the solid prol:ellant, it is indicativeof the performance available for given combustion chamber conditions, providinga means for screening the different configurations under consideration. Thecharactettatic velocity is defined as:

<:" - ( 9 P , -. A , C ,j ) / I

wher.);

S- Gravitational constantP. - Gas generatoi pressureA, - N•,z~le throat, areaC,. - Nozzle d'icharge cuefficientm - Total mas& flow

117/

The ideal values of c* shown on the performance plots which follow are basedon thermochemical equilibrium calculations', using measured fuel and oxygen flowsto the gas generator.

In order to measure the effects of variations in combustor volume, the gasgenerator was tested with and without the internal stainless steel liner, whichresulted in volumes of 699 and 1573 cm3 (43 and 96 in 3 ), respectively. As shownin Figure 3, the c' performance was slightly higher with the liner in place. Thisconfiguration gives higher gas velocities and results in lower stay times in thegas generator. TYb increased velocity seems to overcome the negative effectsexpected from decroasing the residence time. The improved performance withincreased velocity is consiateit with the findings of Lawyer7 in research on highpressure preburners fox liquid rocket engines. Lawyer suggests that increasedvelocities rasult in higher gas-to-droplet rolative velocities, leading to fastervaporization. It would appear, then, that this configuration was limited by theinjection snd evaporation process more than by residence time.

CM ("Ys)11001

lo

Sc t

0 2 4 6 1) 12

Time (Sec)

W k. w Wfi' • ior Masi

fiq. 3. Effeoct of Chamber Volume Variations on Gem Generator PerformanceTotal Fl.ow Rate - .06t kg/Aoc

This theory is further supported by considoerng the effects of -ariations inchamber pressure by operating the gas generator with two different exit nozzles,one 12.7 mm (.500 in) and one 9.75 am (.304 in? at identical fuel and oxygen flowrates. As shown in figure 4, increasing the pressure, which also decreases theinternal velocity and increases the stay time, has a detrimental effect on the gasgenerator performance. The data again suggest& that, fox these flow rates, thegas generator performance is not limited by stay time effects, but by fuelvapor szation processes.

The hollow cone injectoL used in the tests deeszibed above was limitedto relativaly low flows rates. Ir. order to increase the fuel flow rates to therequired level, fuel was aloe inj*,ted radially through two simple orifineinjectors installed imediately drwnstream of the iguiter flange. figure 5compares tna paformance of the gas generator at identical flow rates, with andwithout the radial injectors flowing in combination with the hollow cone atomizer.As shown, the c perfr•rance is greatly improved by injecting a portion of thefool downstream. It oppuars that because only approximately 20% of the fuelentered through the primary atomizer, a mixtire ratio closer to stoichiometric wasprobably obtained upstream, which aided in the vaporization and reaction of theadditional fuel injected downstream. Enhanced mixing of the fuel, oxidizer, andcombustion products due to the cross--atroam !-njection may have also been a factor.It is interestJ.ng to note that this combination attained a higher percentage oftheoretical c than any other configuration.

Catar (m/i)1300.0

1200.0 FA : ; " • :•

1100.0 It

1000.01

900.0

800.0 _

7 0 0 .0 • - ..

600.0 , , I0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

Time (sec)

I - 860 KPa -'- 1380 KPa 0 Ideal

Fig. 4. Zffect of Chamber Pressure Variations on Gas Generator PerformanceTotal Flow Rate - .136 kg/sec

C.W (nyu) ________

10001 •

700 - --- -8001- I

4

800--. I I H

0 1 2 3 4 5 6 7 8 9 10 11 12

Time (sec)

- K Raaoh~cdon -~No ftsi I*c~on k"m

Fig. 5. Sffect of Fuel Injection Methods on Gas Generator PerformanceTotal Flow Rate - .068 kg/sec

Performance with the full cone spray nozzle was generally poor. Difficultywas experienced in achieving reliable ignition, and only intermittent combustionwas possible. This may have been the result of large droplet sizes, leading topoor evaporation, or to a local equivalence ratio which was too rich to supportcombustion near the point of ignition.

During all teats of the yas generator, the efflux was burned with heated airin the ducted rocket combustor described previously. Operation of the combustorwas stable and easily controlled, yielding combustion efficiencies appropriate forthe configuration tested, based on measured thrust and combustor pressure.

319

CONCLUSOU

Progress to date in building a liquid fueled gas generator for ducted rocketresearch applications has been very encouraging. After initial checkout tests inwhich the fuel and oxygen plumbing was developed, the gas generator operation hasbeen very reliable and easily controlled. Preliminary measurements, based on c*considerations, have revealed levels of performance consistent with data publishedelsewhere for fuel rich gas generators. Some additional work is required t6develop the fuel injection methods as the flow rates through the gas generator areincreased to levels sufficient to cover the expected flight envelope of a ductedrocket powered missile. Future testing will include sampling and analysis of thegas generator effluent composition for comparison to available data for thecombustion of the solid propellants being simulated.

The development of the gas generator has taken place in conjunction withramjet performance testing of a realistic ducted rocket configuration, and in theprocess, the viability of the concept has been at least partially demonstrated.

1. Buckley, P.L., Fisher, S.A., and Hillen, L.W., "A Simulator for a HydrocarbonRamrocket Fuel Gas Generator - First Phase Development," Aeronautical ResearchLaboratory, Propulsion Report 181, May 1989.

2. "Liquid Propellant Gas Generators," NASA SP-8081, 1972.

3. Schoenman, L. "Fuel/Oxidizer-Rich High Pressure Preburners, " NASA CR-165404,1981.

4. Bensky, M.S. and Wong, G.S., "Development of LOX/JP-4 Gas Generators forIntegrated Aircraft Auxiliary Power Systems," AIAA 77-890, AIAA/SAE 13thPropulsion Conference, 1977.

5. Sutton, G.P., "Rocket Propulsion Elements," p. 200, John Wiley and Sons,1986.

6. Cruise, C.R., "Theoretical Computations of Equilibrium CompoRitions,Thermodynamic Properties, and Performance Characteristics of Propellant Systems,"Naval Weapons Center, NWC TP 6037, April 1979.

7. Lawver, B.R., "Testing of Fuel/Oxidizer-Rich High Pressure Preburners," NASACR-1656U9, May 1982.

320

liYPERVELOCITY BY EXTENDED PROPELLANT BURN

F. A. VassalloCalspan Corporation

Buffalo, N.Y.

ABSTRACT

Under an internal research and development study Calspan has shown that projectile velocities in excessof 12,000 It/sec (3.4 km/sec) can be produced using a unique powder gun approach, the details of which arepresented in the paper. Briefly, the approach exploits the concept of downbore propellant loading .-. maintainbase pressure during projectile acceleration. By contrast with conventional powder guns, downbore loadingpermits a majority of the propellant to burn as it is uncovered by the projectile. Its major advantages leadingto projectile hypervelocities are:

1. Combustion gases are burned in the vicinity of the projectile base and thus need not flow to "catchup" with the projectile.

2. Because propellant is uncove-ed as volume is increasing, the effect of travelling charge is realizedwithout drawbacks associated with propellant acceleration.

3. The relationship between increasing volume and charge may be tailored to maintain essentially constantbase pressure throughout the entire burn period. Thus, moderate base pressures can be used to obtaindesired velocities.

4. The continuous addition of gas as a driver of the projectile overcomes the limitations on attainablevelocity usually associated with conventional gun systems because less kinetic energy is contained inthe propellant gas.

Results of a 30 mm firing test program using a 12 in. long charge tube have demonstrated the feasibility ofthe downbore loading approach at velocities in excess of 6250 It/sec. Analysis of data suggests that extensionto much higher velocities depends chiefly on charge tube length.

IN T R&ODUCTION

It has long been believed that the velocity limit for conventional power guns is about 10,000 it/sec (3km/sec). As this limit is approached, less and less propellant energy is transferred to the projectile while moreand more resides in the propellant gas kinetic energy. Thus, only very small projectile masses can be acceleratedto this limiting velocity.

The Strategic Defense Initiative (50I) has been emphasizing development of systems to launch projectilesat velocities in excess of 3 km/sec with projectile weights above I kg. This is desired to reduce time-of-flightand to ensure sufficient impact kinet;c energy and momentum for target defeat. With the apparent hiritationsof conventional powder guns, the majority of the SDI gun effort has been directed toward e'ectromagneticsystems which, theoretically, can attain the desired velocities, although at great developmenal cost.

There is reason to believe that powder guns can attain velocities in excess of 3 kin/sec and with significantprojectile mass 1I we deviate from the conventional chamber loading approach and utilize down-barrel propellantloading. Such an approach is illustrated in Fig. I whereby a majority of the propellant is placed along the tubebore rather than into a fixed chamber.

By contrast with conventional powder gas guns whereby the main propellant charge is burned in a chamberupstream of the projectile, the extended burn concept permits a majority of propellant combusion to occur downthe bore as propellant is uncovered by the projecti!e. Several advantages accrue from this propellant placementscheme.

I. Combustion gases are burned in the vicinity of the projectile bate and thus need not flow to "catchup" with the projectile.

2. Because propellent is uncovered as volume in increasing, a much greater propellant load may beemployed than tha; of a conventional gun system.

3. The realtionship between increasing voluimme and charge may be designed to mnaintain essentiallyconstant base pressure on the projectile throughout the entire burn period. Thus, mioderate basepressures can be used to obtain desired velocities.

4. Given that propellani gas generation rate is sufficient to maintain pressure, unlimited burn rates maybe used without overpressurm:.ation of the launch tube.

Approved for public release; distribution unhlnitmd

321

3. The continuous addition of gas as a driver of the projectile overcomes limitations on attainablevelocity usually associated with conventional gun systems because less kinetic energy is contained inthe propellant gas.

6. Development of multiple fully combustible stacked charges would allow mechanization for reloadingusing a revolver approach.

This scheme is well suited for both land and space-based systems. It does not require large power suppliesor their development, expensive gun components, exotic materials, high intensity electrical fields or massiveexpensive electrical switching apparatus.

Under a brief internal research and development effort Calspan has demonstrated the feasibility of thisapproach toward attainment of hypervelocities. The effort involved both analysis and experiment under which ithas been shown that:

1. r vnbore propellant initiation and combustion can occur at rates needed to maintain constant ornearly constant projectile base pressure as propellant is uncovered.

2. Pressure levels and velocities produced are in basic agreement wth computer predictions.

3. Predictions based upon an extension of test observations suggest that projectile velocities in excessof 3.5 km/sec can be produced in a gun of practical size.

EXPERIMENTAL APPARATUS

The aim of Calspan's internal, research, and development effort was to establish to what degree constantbase pressure could be maintained during projectile acceleration in a launch tube. A further objective was touse resulting experimental data to develop a working computer code for design purposes.

GUN COMPONENTS

For configurational simplicity, the research effort utilized a one foot long 30mm rigid matrix charge tubeenclsed by an outer pressure supporting chamber as shown in Fig. 2. This was attached to a smooth bore30mm barrel. The projectile was seated near the breech end of the charge tube, which contained a numberof propellant charges along its length. Propellant charge initiation was by a small booster charge placed intothe breech cavity and ignited using an electric match.

Pressure instrumentation was placed at the breech and along the chamber and barrel as shown. Downbarrel make wires and velocity screens were used to determine projectile position and velocity. The basic guncomponents are shown in the photographs of Fig. 3.

CHARGE TUBE

The charge tube was made from 4340 steel heat treated to Rc 38. 1 welve 0.25 inch diametei radial holeswere bored at each of twenty-six equally spaced axial stations along its length. The holes were filled usingloose propellant retained using strips of 2 mil plastic tape applied to the inner (bore) and outer tube surfaces asis shown in Fig. 4. With this configuration, a volume of 7.65 cubic inches was available in the charge tube forpropellant.

PROJECTILES

Preliminary test firings suggested the need to use rather substantial projectiles in oider to avoid projectilebreakup during launch. Nylon banded aluminum projectiles were found inadequate at chamber pressures above15 kpsi. After a number of iterations with projectile concepisp conligurdtions showsi in Fig. 5 were found to

survive the firing conditions without breakup. Here, structural integrity is obtained by threading the miid steel

band to the aluminum core. The sizes where chosen to obtain the desired projectile weights.

TEST RESULTS

The gun components were installed in the Calspan ballistic range. Pressures were recorded using PCB

Model 119A transducers and Nicolet Model 270 digital recorders. Early firings using partial loads of various

commercial propellants indicated that flake or waler type propellants having very small web would yield best

performance. Hence, commercial Hercules Bullseye propellant was chosen as the final energy source, it being

the quickest readily available propellant. This propellant has the essential characteristics of M9 propellant with

an 0.003 inch web. Computations indicated that pressures of at most 44000 psi would be produced with full

loading of the charge tube (94.5 gims) and booster chamber (10.0 gins). Figure 6 illusirates the pressuresobserved when firing this charge with a 40 gm projectile. When we compare these pressures with those generated

by the booster charge alone (Fig. 7), it is clear that considerable combustion o! the downbore charge is occuring

during projectile transit of the charge tube. Furthermore, the observed premur' level ii in good agreement with

the above estimate based upon immediate coinbusuon. The pressure plateau is also suggestive of desired

322

combustion behavior. The projectile velocity in this shot was interpreted to be more than 6300 ft/sec basedupon barrel transit time.

In an effort to attain higher velocities, the capacity of the charge tube was increased by machining 0.2inch wide grooves C.1 inch deep at the outer diameter along the tube at each station (Fig. 8). This modificationincreased its capacity by about 20% with a computed maximum pressure level of about 54,000 psi. The chargetube was loaded using a funnel loading procedure in which the unloaded tube was pushed through a propellantbed into the chamber housing. With this approach, the outer tape retainer was eliminated and a full chargetube loading of 106 gins was obtained. Figures 9 and 10 illustrate pressures obtained with this configurationwhen f[ring 70.5 and 45.3 gm projectiles. Inspection of Fig. 9 clearly reveals the constancy of pressure associatedwith this concept. Measurements indicate the base pressure to have maintained its maximum constant valuevirtually throughout the period of projectile transit of the charge tube. Further, the maximum of 50,000 psi isin good agreement wit'h above pressure expectations based upon rather immediate ccrnbustion of the downborecharges. The measured pressure at the barrel entrance demonstrates the severe loss in base pressure associatedwith acceleration of the combustion gases to projectile velocity once the projectile has left the charge tubeand mass addition is virtually complete. Note also that the chamber pressure maintains its maximum levelconsiderably after the projectile has left the charge tube. The added time-at-pressure is believed to be associatedwith the shock wave travel period needed to influence pressures at the booster chamber position. Thus, thispressure is in no way reflective of the base pressure on the projectile once the projectile has left the chargetube. The rapid increases in pressure at the 3.75 inch and 11.75 stations to essentially the chamber pressurelevel demonstrate the base prcssure constancy of the extended burn concept for a one loot charge tube. The

projectile velocity in this test was measured at 5700 ft/sec. The estimated velocity at exit of the charge tubeis nearly 4700 ft/sec. Hence, the gas expansion in the barrel appears to contribute approximately 50% moreprojectile energy. It is clear that further enhancement of velocity is best accomplished by increase in chargetube length and loading rather than added barrel length. Increase in charge tube length/charge shouid result inroughly linear increase in projectile kinetic energy.

Test results obtained for the 45.3 gin projectile shown in Fig. 10 are not so well defined. Althoughchamber pressures attain the appropriate levels, there appears to be significant delay in tube pressurization. Onthe other hand, the indicated pressure point exposure times suggest a velocity in excess of 5000 ft/sec. at exit ofthe charge tube. This magnitude of velocity could only have been attained with essentially full pressurization ofthe charge tube. Thus, possible pressure measurement errors are "rplied. The measured muzzle velocity of625)U t/sec is, however, somewhat lowpr than expected, supporting the reduced pressures recorded. At thiswriting no more precise explanation of the data can be made but it is encouraging that full pressues wereeventually attained with this projectile weight. Ignition techniques may need to be improved in order to obtaindesired pressure response with this projectile weight.

COMPUTER ANALYSIS

In order to indestigate performance beyond the test range, a simplified ballistics code was programmed foruse on a Franklin Ace 1200 computer. The code modeled the basic features of the concept; -- imely, boosterand charge tube combustion and pressurization during projected launch. Propellant burning and gas generationwas assurmed to follow the standard pressure law with ignition of all propellant surfaces assumed to occurinstantaneously on exposure to presure. The downbore charges and their associated volume were specified byintrodcing a ficticious diameter to generate an annular volume having a loading density equivalent to that ofthe charge tube. During projectile trans't of the charge tube, the propellant gas velocity was taken to be zero,so that the base pressure was equal to the averdge tuJbe press'-cre. Following exit of th- charge tmhe, trie basepressure was taken to follow the Lagrange approximation. Propellant gas temperatures were computed usingthe principle of (onservation of energy.

Computed ballist:cs for the two experimental firings under maxi•iu•n charge are given in Fig. II. Thesecases were computed using piopellant characteri-,tics for M9 propellant with a web of 0.003 inches. The resultsshow nearly constant base pressures during prnjectile transit of the charge tube. This is in basic agreementwith test dota. The level of base pressure appears to be somewhat lower for the lighter projectile indicatinginsufficiently high burn rates for the propellint. This may partially account for the lower pressures observedfor this weight of projectile in the firing tests. As noted in the firing tests the base pressure drops precipitouslyonce the projectile uncorks the charge tuhe, and the majority of the projectile velocity is produced in the chargetube. For the lighter projectile, a velocity of over 500D ft/sec is produced in one foot of charge tube travelwith potential velocities over 7)00 ft/:,ec indicated. The projectile energy appears to drop by ony 10% froii thatof the heavier 'ower velocity projectile which suggets that ballistic efficiency can be imintained at highvelocities with this concept.

The potential pcrlorminancm' advantages of the vxtended burn concept over a conventional cartridge isillustrated in Fig. 12. Computations show that when miaximum pressure is uninted, the eflectiveness o1 additionalcharge weight diminishes rapidly in a conventional (.drtridge such that a hminting velocity is quickly approached.This is a consequence (,f the changes in cartridge volume and propellant burn rate needed to hlnit the mnaxinumnpressure level. The extended burn concpt is less restricted by volurme/rate considerations and cdn iaintdin ainore nearly constant tnergy-to-chmrge ratio at hypervelocity. The funittng velocity vs charge for c:omumiercial

Buliseye propellant suggests that velocities in excess of 11,000 it/sec can he produced using a five loot 30minucharge tube with pressures beiow 55 kpsi. The charge tube lengith could he reduced in approximate proportion

to allowable maximum pressure. Increased burn rates at higher allowable pressures would further enhanceperformance as would use of finer grained/faster propellants.

CONCLUSIONS

The brief experimental analytical research investigation conducted at Calspan has demonstrated the feasibilityof downbore extended burn concept toward attainment of projectile hypervelocity. Measureinents taken usinga uriique test apparatus have shown that ignition and combustion of downbore charges can occur at rates necessaryto maintain relatively constant base pressure throughout the burn period. Analysis indicates that attainment ofvelocities iii excess of 12,000 ft/sec depends chiefly on the charge tube length employed. Performance wouldbenefit fromn development of laster burning downbore charges having higher loadinig density.

BOOSTER PROJECTILE STACKED DOWNBORE CHARGES BARRELBREECH CHARGE ANDl SEAL CONTAINED WITHIN OIGID MATRICES EXTENSION

a. AS LOADED BEFORE IGNITION

hb. COMBUSTION PHASE (CHARGES CONSUMED AS EXPOSED)

c. EXPANSION PHASE (ALL CHARGES EXPENDED)

Fig. 1. Extended Burn Concept and Sequential Events

CHARGE TUBE:3.75 Inch PROPELLANT CHARGES 11.75 Inch

PRESSURE PRESSURE IN RIGID MATRIX PRESSURE PRESSUREPORT PROJECTILE PORT PORT PORT

INTER-SEALS

- .. . .. . . .- ........ -.- -

BREECH \/BARRELBOOSTER CHAMBER

CHARGE

Fig. 2. Extended Burn Experimental Apparatus

CHAMBER (REAR VIEW) CHAMBER (FRONT VIEW)

BREECH

Fig. 3. Gun Components

CHARGE TUBE (UNLOADED) CHARGE TUBE LOADED FOR FIRING"O" RINGS INSTALLED

END VrEW SHOWING PROJECTILE INSERTION OF CHARGE TUBE INTOPLACEMENT IN CHARGE TUBE GUN CHAMBER

Fig. 4. Charge Tube and Loading Technique

326

APPROXI1MATELY APPROXIIAATELY?0 OM "nIOACTILE 45 GM I ROJECTILE

FIRis& DIECTO 10181TVC

Fig. 6. Projectile Configurations

/ P•.AOJEClTUBE o 4,0 0M$

1.14 1.14664.5 0-M "

tO -M /"w 0.2 0.988 "'-3' T

ALMWIM .28327440 04

1018 SEEL 10OSSTEE

PRJETLEa 4 MSAR.

ENTIANE - M81sog -. MWS ER sr CrKlROoE -o Li0ht PrjeMeUen

3CHMBERimCeg ".z .

0 0.6 1.0 1.0 2,0 2.5 3.0 3.5 4.0TIME - MS5C

Fig. 6. Mossursa Preseuft.. for Light Projec t 'e UsingMedlium Chairge

eo

PROJECTILE - 74.00MS

BOOSTER CHARGE OMS

40

BOOSTER20 CHAMBER

ENRAN4CE

0 is__ __ _

00.5 1.0 1. 2.0 2.6

TIME - MSEC

Fig. 7. Pressures With 10 GM Booster Charge Only

ig 8. Modiflied Charge Tube fur IncroackadLoading Capacity

CHAGE UB1 a100.5 0MB

117 INC MUZZLE EXIT VELOCITY11,75 ICH W- 700 FTISEC

~40 -IBA

3176 IN99K .1 172CI -

oj 1'0 1 202.fTIM& MIILC

F14. V irttergr 581116110k *at* fof tlaevy Ptlotilot~ UsIrvgMaximinum Chargo

40 Plo.aeC IlLt - 45.3 0MB

euMUZZLE EXII V&LOC,1'1'

11 71 INOV )

( 4t4AM~fI L ArUICLLNT11AHr'E

0 -. - -. -

D 0.6 1Es2.0 2.ATIML -M1119C

Pig. 10, entorloi 11#1ilics~k Date for Light ProJlotilI UsingMaiximnum Charge,

S0

• .- - 'f VM"OCWY. 45 OM PROJECTILE

/E&LOC(Tv, 70 oM (vJkU(iIJa•

40 ' .. BA PRI0URCE,S / ... ,M (WP •ROCTILE

20 /

0 06 1,0 1.6 2.0 2.'TIMe - MIIC

pIg, 11, Computed Uuillistio Using Maximum (Ohrg.

LMT pOR uuu.sYu /YPROPEILLANT /

(LOADIN DENSITY00W GMC()

/ / "* VELOCITY AT END,000 7 OF CHARFIM TUBE

So / F OR BULLIFYI (MO)

7000 FIING // E XIT VELOCITYDATA POIr/NT ~knVFORt BULL&KY11

000- PROJECTILE - 70 OM 1BORE DIA. e 30 MM

/ MAX. PRESS. K1O KPOID/ CONVENTIONAL Mg PROPELLANT

I CARTRIOO[,

W O1) ) m _ ._ _ j ..........-.........-....-

0 100 200 300 400 g0o g00 700

CIIAR0E WT. - OMB

Fig. 12. Velocity Cnmparlwons as a Function of Charge

Jil

Molecular Structure Of Low Temperature FormOf Triaminoguanidiniun Nitrate (TAGN)

A.J. BracutiUS Army ARDEC

Picatinny Arsenal, New Jersey 07806-5000

Abstract

A rigid-body notion study of triaminoguanidinium nitrate (TAGN) wasinitiated originally in order to determine if niltrate ion libration miqht beresponsible for the accelerated burning rate observed whon this salt isformulated into a propellant with a cyclic nitamine. P low temperaturepolymorph of TAGH was discovered by this laboratory during the course ofthis rigid-body motion study. Continuation of this study in the lowertemperature regimes, however, necessitated understanding the molecunlarstructure of the low temperature form of TAGN. To do this a single-crystalx-ray diffraction structure determination of the low temperature form wasaccomplished. The crystal structure was determined including space gtoup,lattice parameters, atomic coordinates and temperature factors, hydrogenbonding and density.

Introduction

When trisminoguanidinium nitrate (TAGN) is added to a solid propellantcontaining RDX, the burning rate of that propellant is increasedsubstantially. The mechanism for this burning rate acceleration is not yetunderstood. In fact, there is much speculation as to whether thisproperty in a macroscopic, microrcopic or molecular phenomenon. As a resulta Room temperature X-ray crystallographic study of TAGN was initiated bythis laboratory (ref 1) in attempt to ascertain if this behavior could beexplained at the molecular level.

The resulting crystal structure determination revealed that the ions inTAGH formed a layered structure with stronger bondinS forces within thelayers than betweeti the layers. furthermore, tho planar TAG cation wascoplanar within the layer, but the planar nitrate anion was normal to theplane of the layer with the central nitrogen atom and only one oxygen atomin the layer.

The molecular arrangement of TAGH explained the friability of thecrystals but did it oxploit the burning rate behavior? One could certainlyargue that easy crystal breakup might enhance burning. ThiL would imply,however, that any friable oxidizer would have the same effect onpropellants containing RDX. To the contrary, enhanced burning is notobserved with all friable oxidizers. This suggests that something peculiarto TAUN itself is responsible for this effect.

It was also found that the thermal amplitudes of the oxygen atoms of thenitrate ion were excessively large. This indicated disorder in the nitrateion which could either be static (positional) or dynamic (rigid body

Approved fur puhl .c relsanse dlmtrlbution is unlimited.

331

motion). On the basis of both shortened N--O bond lengths and the largethermal amplitudes, it was suggested that the nitrate '.on could be ahindered rigid rotor (ref 2). A simplified analogy Is & vinwheel thatoscillates rather than freely rotates. A similar sugqe.t".ýu w'. iaso made ina subsequent neutron diffraction study (ref 3;..

The TAGN crystal structure was later redetern.ned usi:g posIIA':i!,disordered nitrate ions in an effort to ascertain if quioscen• e'tr'w ýswould also account for the observed excessive thermal arplitudes. M:hxzgi'the results were not definitive, they suggested tha. nitrate ion iitimotion was probably the better explanation for the disorder observsC .temperature (ref 4).

In a later low temperature study (ref 5) it was shown that TAGNundergoes a phase transition at approximately -10½C. If a density changeoccurs with this phase transition, propellants formulated with TAGNcould be adversely affected at low temperatures. This paper describes themolecular and crysjal structure of the low temperature polymorph of TAGNdetermined at -104 C and its associated density.

Data Collection

A colorless acicular crystal of TAGN approximately 0.01 x 0.15 x 0.30 mmwas mounted with its long axis parallel with the phi axis of thegoniometer. All data were collected at -104(1) C (169 K) using Mo V0x-radiation (X - 0.71073 A) on an Enraf-izonius CAD computer controlled kappaaxis diffractometer equipped with a graphite crystal incident-beamsonochroaater.

The cell constants and orientation matrix for x-ray diffraction datacollection were obtained from least-squares refinement, using the settingangles of 25 reflections in the range of 4 < 0 < 8.

A total of 2622 relections were collected, of which 2140 we-e uniqueand systematically absent. Lorentz and polarization correctionL were made tothe data. Although no correction was made for absorption, a szondaryextinction correction was made to each xeflection (ref 6).

The heavy atom positions were determined with direct metnods (ref 7)using 188 reflections. The hydrogen atoms were located with differenceFourier Fourier techniques. The structure was refined with full-matrix leastsquares until convergence was attained using 994 reflections with intensityvalues 2 times their standard deviations. The non-hydrogen atoms wererefined anisotropically and the hydrogen atoms were refined with fixedisotropic thermal parameters.

Discussion

At -104 C, TAGN exists in the orthorhombic space group Pbca with a =

33.058 (10), b - 12.573 (2) and c - 6.573 (4) A (Table 1). At -100 C TAGN

3.12

transforms to the orthorhombic space group Pbcm with unit cell dimensionsmeasured at 23 C of a - 8.366(2), b - 12.649(2) and c - 6.556 A. The densityof the low temperature form is 1.63 g/cc and the density of the roomtemperature form is 1.60 g/cc (ref 4).

The low temperature polymorph unit-cell volume is almost quadruple thatof the room temperature form, but their densities are quite similar. In theroom temperature polymorph, the asymmetric group consists of one TAG cationand one nitrate anion. In the unit cell four asymmetric groups aredistributed over four special positions on mirror planes in the Pbcm spacegroup. In the low temperature form, there are eight asymmetric groups, eachof which contains two TAG cations and two nitrate anions, which aredistributed over eight general lattice positions in the Pbca space group.Thus, at low temporature the larger unit cell contains sixteen formulaweights of TAGN while at room temperature the smaller unit cell containsonly four. As a result, the densities of the two forms do not differ muchfrom each other.

Description of the TAG cations

In this low temperature form of TAGN there are two crystallographicallyunique TAGH groups (Table 2) which will be referred to as TAGN(I) andTAGN(II). The bond lengths in Table 3 reveal that there are no significantdifferences between the respective heavy atom bond-lengths of the two TAGcations and these values are consistent with those reported for the roomtemperature polymorph (rets 1, 2 and 4). In TAG(I) the mean C--N and N--Nbond lengths are 1.326(8) and 1.417(8) A, an'l in TAG(II) these bond lengthsare 1.324(9) and 1.417(8) A. The mean N-C-N bond angle for both TAG(l) andTAG(2) is 120.(6) (Table 4).

Both TAG cations are not strictly planar as they are in the roomtemperature polymorph where the heavy atoms mnd the imino hydrogen atoms areconstrained by symmetry to positions on mirror planes parallel to (001). Theconformation of both quasi-planar cations, however, is appoximately the sameand approaches C3h molecular symmetry as depicted in Figure 1.

Description of the nitrate anions

The M-O bonds of N(I) listed in Table 2. are equivalent with a meanvalue of 1.253(6) which is consistent with the value listed for the nitrateanion N-O bond in the International Tables For X-Ray Crystallography (ref8). By contrast, in N(2) the N-O bonds are not equivalent with values of1.269(7), 1.223(7) and 1.239(7) A for N(14)-O(4), N(14)-O(5) and N(14)-O(6),respectively.

In the room temperature phase the nitrate anion is both planar anddisordered with the oxygen atoms displaying extremely large thermalamplitudes. In the low temperature form, however, only one nitrate anion,N(II), is planar and disordered as revealed by the large oxygen atom thermalamplitudes shown in Table 2. The other nitrate anion, N(I), is also planarbut is not disordered since its oxygen atoms display smaller thermalamplitudes. The nitrate ions are depicted in Figure 2.

333

Description of structure

In general, the structure of the low temperature form resembles thestructure at room tu-perature except for very small differences caused byvariation in thermal motion. The low temperature structure can best beunderstood by comparing it with the structure of the room temperature form.

Room temperature

At room temperature, the structure consists of layers of TAG cations andnitrate anions positioned in mirror planes normal to the c axis which arelocated at +/- 0.25 c. The planar TAG cation is coplanar with the mirrorplane and the planar but disordered nitrate anion is bisected by the mirrorplane with central nitrogen atom and one oxygen atom located on the mirrorplane. Within the layer, each ion has three oppositely chargednearest-neighbor ions. This causes the plane of the TAG anion to beoriented normal to the plane of nitrate anion. In addition to this ionicbonding scheme, each TAG cation is also hydrogen bonded through its iminohydrogen atoms to two of the nitrate ions.

In the interlayer packing, TAG cations are stacked in adjacent layersalmost directly over one another along the c axis separated by approximately0.5 c ( 3.7 A). The nitrate ions are also stacked over one another alongthe c axis but in this case they are in alternate layers separated by c (6.5 A). In addition, each TAG cation is hydrogen bonded through its aminohydrogen atoms to one nitrate anion the adjacent lower layer and one in theadjacent higher layer. This structure is presented in Figure 3.

Low temperature

At -104' C, TAGN as shown in Figure 4 still exists as a layeredstructure, but now the quasi-planes of the TAG cations are no longercoplanar with the layers and the nitrate anions are not normal to thelayers. This can best be described as a puckered layer structure. Theplanar nitrate anions N(I) and N(II) are approximately coplanar with eachother while the quasi-planar TAG(I) and TAG(II) cations are tilted about 20with respect to each other. The quasi-plane of TAG(1) is approximatelynormal to the plane of N(l) but not to that of N(II) and, conversely, thequasi-plane of TAG(II) is appoximately normal to the plane of N1(1) but notto that of NIT).

Within ta-a layer both TAG cations are surrounded by three nitrateanions. TAG(I) is hydrogen bonded through its imino hydrogen atoms to onenearest-neighbor N(I) and two nearest-neighbor N(II) nitrate anions (Table5). Also, each TAG(II) is hydrogen bonded through its imino hydrogen atomsto one nearest-neighbor N(II) and two nearest neighbor N(I) nitrate anions.Between layers, TAG(I) through its amino hydrogen atoms is hydrogen bondeato three N(II) anions and each TAG(II) through its hydrogen atoms is bondedto three N(I) anions. This results in an infinite three-dimensional networkof hydrogen bonded puckered layers of TAGN.

334

A comparison of the low temperature unit cell (Figure 3) with four unitcells of the room temperature form stacked on top of each other on the bcplane (Figure 4) reveals a conspicuous similarity between the structures.The major difference between the two structures is the nitrate ion disorder.If all the nitrate ions in the low temperature form were disordered, Figure3 and Figure 4 would appear to be the same.

Very small changes in the TAGN structure cause the transition from Pbcmto Pbca symmetry observed at -10 C. This behavior typical of a second ordertransition agrees with the heat capacity data which also showed a secondorder transition at this temperature (ref 5).

Conclusions

1. The phase transition at -10'C is a second order transition.

2. At -10 C half of the nitrate ions librate as rigid bodies and halfare quiescent with their individual atoms vibrating independently withsmaller thermal amplitudes.

3. The densities of the two forms of TAGN are approximately the same.Therefore, it does not pose a mechanical problem in propellant formulationswith respect to temperature cycling.

335

References

1. A.J. Bracuti, "l,2,3-Triaminoguanidinium Nitrate," Acta Cryst., B35,PP. 760-761, 1979.

2. A.J, Bracuti, "The Crystal Structure Determination OfTriaminoguanidinium Nitrate: A Burning Rate Modifier For NitraminePropellants," Tech.Report ARLCD-78050, Picatinny Arsenal, Dover, NJ, 1979.

3. C. Choi & E. Prince, 1,2,3,-Triaminoguanidinium Nitrate By NeutronDiffraction,6 Acta Cryst. B35, pp. 761-763, 1979.

4. A.J. Bracuti, 01,2,3-Triamninoguanidinium Nitrate: An AlternateExplanation Of The Nitrate Anion Disorder," Technical Report ARAED-TR 8703,ARDEC, Picatinny Arsenal, New Jersey, December, 1987.

5. A.J. Bracuti & J.K. Salo, aDiscovery of Lo remperature Form OfTAGN,w Technical Report ARAED-TR-88009, ARDEC, Picatinny Arsenal, July,1988.

6. W.H. Zachariasen, "The Secordary Extinction Correction ,"ActaCryst., 16, 1139-1144, 1963.

7. B.A. Frenz, OComputing Crystalloqraphy", Edited by N. Shrenk,T.Olthof-HazelKamp, H. van Konigsveld and G.C. Bassi, Delft University,Delft, Holland, 1978.

8. International Tables For X-Ray Crystallography, vol. III, p. 270,Birmingham, England: Kynoch Press 91968).

3.36

Table 1. Crystal Data

Molecular formula CH9N703Formula weight 167.13Crystal system OrthorhombicSpace group Pbcaa 33.058(10)b 12.573(2)C 6.541(4) Av 2718.7 RZ 16D 1. 63 g/cm'A(Mo Ko, 0.71073 Au (Mo KL) 1.6 cm-'F(OOO) 1408T 169(1) KR 0.053Rw 0.061

337

Table 2. Atomic coordinates and equivalent temperaturefactors and their esd's

Atom x/a B(e/v)

C(1) 0.0717(2) 0.2618(5) 0.2354(8) 1.2(1)1(1) 0,0972(l) 0.1815(4) 0.2146(7) 1.5(1)N(2) 0.1384(1) 0.2045(4) 0.1756(8) 1.5(1)N(3) 0.0841(1) 0.3612(4) 0.2215(8) 1.7(1)N(4) 0.0533(1) 0.4445(4) 0.2402(8) 1.6(1)N(5) 0.0326(1) 0.2415(4) 0.2475(7) 1.4(1)N(6) 0.0199(1) 0.1342(4) 0.2858(7) 1.8(1)H(1) 0.091(2) 0.120(5) 0.210(9) 3.0H(2) 0.146(2) 0.162(5) 0.065(9) 3.0H(3) 0.155(2) 0.167(5) 0.270(9) 3.0H(4) 0.106(2) 0.384(5) 0.193(9) 3.0H(5) 0.064(2) 0.490(5) 0.345(9) 3.0H(6) 0.057(2) 0.470(5) 0.120(9) 3.0H(7) 0.019(2) 0.290(5) 0.285(9) 3.0H(S) 0.011(2) 0.130(5) 0.406(9) 3.0H(9) 0.002(2) 0.138(5) 0.172(9) 3.0C(2) 0.3200(2) 0.2569(5) 0.1893(8) 1.4(1)N(7) 0.3457(1) 0.1758(4) 0.2109(7) 1.4(1)N(S) 0.3872(1) 0.2024(4) 0.2331(8) 1.7(1)N(9) 0.3328(1) 0.3551(4) 0.2082(8) 1.5(1)N(10) 0.3046(1) 0.4377(4) 0.1784(8) 1.9(1)N(11) 0.2814(1) 0.2367(4) 0.1521(8) 1.r(1)N(12) 0.2687(1) 0.1292(4) 0.1380(8) 1.9(1)H(10) 0.340(2) 0.100(9) 0.175(9) 3.0H(11) 0.399(2) 0.I10(5) 0.125(9) 3.0H(12) 0.395(2) 0.167(5) 0.342(9) 3.0H(13) 0.360(2) 0.369(5) 0.204(9) 3.0H(14) 0.309(2) 0.476(5) 0.058(9) 3.0H(15) 0.307(2) 0.489(5) 0.275(9) 3.0H(16) 0.263(2) 0.284(5) 0.145(9) 3.0H(17) 0.250(2) 0.117(5) 0.247(9) 3.0"1(18) 0.254(2) 0.118(5) 0.026(9) 3.0N(13) 0.1894(1) 0.0965(4) 0.6435(7) 1.40(1) 0.2114(1) 0.1408(4) 0.5121(6) 2.30(2) 0.1590(1) 0.0453(3) 0.5852(6) 1.8(9)0(3) 0.1973(1) 0.1049(4) 0.8297(6) 2.3(1)N(14) 0.4413(1) 0.0997(4) 0.7174(8) 1.6(1)0(4) 0.4103(1) 0.0459(4) 0.6676(9) 3.6(1)0(5) 0.4505(1) 0.1134(4) 0.8966(7) 2.9(l)O(6) 0.4616(1) 0.1380(4) 0.5763(7) 3.1(1)

338

Table 3. Bond lengths in angstroms with esd's in parentheses

Bond Lengjth Bond Length

C(l)--N(l) 1.323(9) C(2) -- N(7) 1.335(9)C(l)--N(3) 1.319(8) C(2)--N(9) 1.312(9)C(l)--N(5) 1.340(8) C(2)--N(Il) 1.325(8)N(l)--N(2) 1.416(8) N(7)--N(8) 1.420(8)N(3)--N(4) 1.419(8) N(9)--N(10) 1.410(7)N(5)--N(6) 1.415(8) N(Il)--N(12) 1.418(8)N(l)--H(l) 0.80(8) N(7)--H(10) 1.00(8%1N(3)--H(4) 0.7q(4) N(9)--H(13) 0.93 (7)N(5)--H('7) 0.76(7) N(ll)--H(16) 0.86(7)N(2)--H(2) 0.94(7) N(8)--H(12) 0.86(7)N(2)--H(3) 0.94(7) N(8)--H(12) 0.87(7)N(4)--H(5) 0.94(7) N(10)--H(14) 0.95(7)N(4)--H(6) 0.85(7) N(10)--H(15) 0.90(7)N(6)--H(8) 0.85(7) N(12)--H(17) 0.96(7)N(6)--H(9) 0.94(7) N(12)--H(18) 0.89(7)N(1'3)--0(1) 1.257(7) N(14)--0(4) 1.269(7)N(13)--0(2) 1.253(6) N(14)--0(5) 1.223(7)N(13)--0(3', 1.250(6) N(14)--0(6) 1.239(7)

Table 4. Bond angles in degrees and their esd.'s

Atoms Angle Atoms Angle

N(l)-C(l)-N(3) 121.2(6) ?4(7) -C(2) -N(9) 120.2 (6)N(l)-C(l)-N(5) 119.3(6) N(7)-C(2)-N(11) 119.2(7)?I(3)-C(l)-N(5) 119.5(6) N(9)-C(2)-N(1l) .120.7ý6)C(1)-N(1)-N(2) 118.5(6) C(2) -H(7) -N(8) 116. 5(65)C(l)-Ii(3)-N(4) 119.1(6) C(2)-N(9)-N(10) 119)C(1)-N(5)-N(6) 118.5(6) C(2)-N(I1)-N(12) 118.7(6)

C(l)-N(3)-H(4) 130. (6) C(2)-MI9)-HI(13) 120. (5)

4(1) -14(2) -H(2) 106. (4) 14(7) ..N(8) -HIi() LOC.0)N(1)-N(2)-H(3) 109. (4) N(7)-N(8)-H(120)It (5)

N(5)-N(6)-H(8) 108.(4) N(11)-N(12)-H(l , i.(4

N(2)-N(2)-H(8) 943.(6) H(11) -N(82)-h(l L, .2.(8.

H(1)-N(13)-0(2) 118.9(6) 0(14)-N(14)-0(5' 12..(6

0(l)-N(13)-0(3) 120.5(6) 0(4) -N(14)-0(5) 121 t.4() (0(2)-N(13)-0(3) 120.5(6) 0(4)-N(14)-0(6) 121.6(6

0(2)-N(13)-0(3) ~ ~ ~ ~ ~ 12.() 05(N1)06)1166

Table 5. Details of hydogen bonding with bond lengths in angstroms, bond

angles in degrees arid their esd's in parentheses.

N(x)-H(.YL...0.(z) N(x)-H(y) P.yl ... O• ) •.y. .... Oj Angle

Imino TAG(1) H bondsN(1)-H(l)...0(4) 0.80(8) 2.10(8) 2.886(8) 167.(4)N(3)-H(4)...O(2) 0.79(7) 2.10(8) 2.882(7) 170.(4)N(5)-H(7)...O(6) 0.76(7) 2.37(7) 2.959(8) 135.(4)

Amino TAG(l) H BondsN(4)-H(6)...0(5) 0.85(7) 2.33(8) 3.098(8) 150.(4)N(6)-H(9) ... 0(6) 0.94(7) 2.11(7) 3.054(8) 161. (4)N(6)-H(8) ... O(5) 0.85(7) 2.38(7) 3.107(8) 143.(4)

Imino TAG(2) H BondsN(7)-H(10)...0(2) 1.00(8) 1.92(8) 2.903(8) 167.(4)N(9)-H(13) ... 0(4) 0.93(7) 1.98(7) 2.861(8) 157.(4)N(Il)-H(18)..O(1) 0.86(7) 2.12(7) 2.926(8) 158.(4)

AMINO TAG(2) H BondsN(10)-H(14)..0(3) 0.95(7) 2.18(8) 3.102(8) 163.(4)N(12)-H(18)..O(I) 0.89(7) 2.29(7) 3.120(8) 155.(4)N(12)-H(17)..0(3) 0.96(7) 2.17(7) 3.098(8) 162.(4)

3-1(

TAG(I)TAG PII)

I. Ho)ttcular c()Iifor-4aflons of TAG(I) and TAc4 II) cotunaWitt) Atua rnumbarirg gicholmse..

(2

Fiqur ~. olo~ulsv conformations ofI nitrat* onlits, $0I) soid MOII),vith vlumkbe)ing sahesmes

Figure 3. Unit call of low tomperature ?TAG lookingduon the a &wvi

.341

D

rigure 4. Unit call of room temjirAturo TAGN lookingdown the c axis.

144

Heats of Explosion, Detonation and Reaction Products:Their Estimation and Relation to the First Law of Thermodynamics

Edward E. Baroody and Susan T. PetersNaval Ordnance Station, Indian Head

ABSTRACT

This paper reviews a method to calculate Heats of Explosion (HOEs).Similarly, a method is provided to calculate Heats of Detonation (HDETI) whichcan be used to theoretically evaluate explosive performance. This calculationmethod also gives predictions of the reaction products for energetic materialswhich are burning or detonating under various conditions. since experimentalenergetic materials are often unavailable in quantities sufficient toexperimentally determine the HOE, HDET or reaction products, much theoreticalprocedures can be very useful. Later, when sufficent material can be had, thetheoretical values can be checked by determining the heat released when thematerial is burned or detonated in a bomb calorimeter.

After a discussion of the thermodynamics operative iii the calorimeter andour calculation method, we will show how our calculations can be used to explainvariations in HDETe meaplired under different conditions. Calculated values willbe presented along with experimental data.

BACKGROUND

Heats of explosion and detonation are valuable figures for ranking thepotential performance of energetic materials. Other researchers may requireknowledge of the reaction products of such materials under varying conditions toaosse environmental or other effects. Accurate prodiction of these values canaid the designer of new formulations and devices. The method of calculationdescribed in this paper allows the accurate prediction of these valuas as wellas providing a guide to the performance of materials under various conditions.Furthermore, in the case of HDFTo, the experimental procedures for determiningthese values are troublesome and rather experimental in themselves. Causing thereliable, reproducible detonatJoi of miniscule quantities of explosive andhaving a calorimeter react on a time scale sufficiently short to capture theevent are not straight forward. In the case of explosives then, having aconsistent calculation method which can show the qualitative, if not exactquantitative, behavior of the material could be very valuable.

PEP 1 is a thermochemistry code with a long history of 2 use by the rocketcommunity. It should be noted that PEP, unlike NASA-Lewis and BLAKE3 which usea frei energy minimization, uses an equilibrium constant scheme for calculatingthe final thermodynamic state. The limitations in PEP's method of calcu).ationare similar to those of the NASA-Lewis code; the ideal gas equation of state isasmumed, but all possible product species are considered in the calculation.For rocket conditions (pressures below 5,000 psi), the assumption of idealbehavior is a sate one. At the pressures one encounters in a gun (more than5C,000 psi), the BLAKE zode is more useful since it allows a more sophisticatedtreatment of the behavior of the gases: a choice of equation of state can bemade. This increased capability comes at the expense of limiting the number ofallowab)s product species.

After a number of years running PEP codes on energetic materials, it wasnoticed that the heat of reaction and product compositions calculated for thematerials reacting at a chamher pressure of 1000 psi exhausting to oneatmosphe:e matched the HOEs measured in our laboratory. We must stress that theoriginal idea tor the calculation method described in this paper came from an

"The opinions or assertions made in this paper are those of the author and arenot to be ',onstrued as offiiAul or reflecting the views of the Department of theNavy or the Naval Gervice at large."

Approved for public r leuxe; diptribiation in unlimited.

345

empirical observation of the correlation of these calculations to our verycareful measurements.

Table I shows the results of PEP calculations for rocket • ci' aconstant 1000 psi chamber pressure and various exhaust presswir, andnitroqlycerine (NG) as propellants. The data presented in .ih. at ofreaction, temperature and composition of the major reaction *'

examine the data for TNT first. As we go down the table, ti..with. decreasing exhaust pressure.

Table I Combustion Products Formed for TN? and Nc

Press. AHr Temp CO2 CO C(S) R2 1120 N2 minorcomponents

(psi) (cal/q) (K) moles/100q

chamber1000 0 1991 0.003 2.630 0.423 1.071 0.005 0.652 0,025

exhaust14.7 765 1045 0.24! 2.050 0.770 0.974 0.098 0.660 0.0151.0 699 864 0.543 1.427 1.103 0.954 0.129 0.660 0.0000.5 930 820 0.615 1.281 1.178 0.954 0.131 0.660 0.0080.1 993 758 0.772 0.965 1.339 0.958 0.132 0.660 0.005

0.05 1020 730 C.835 0.840 1.402 0.960 0.131 0.660 0.0040.01 1075 672 0.968 0.573 1.536 0.964 0.130 0.660 0.0030.005 1092 648 1.021 0.469 1.589 0.964 0.130 0.660 0.0030.001 1142 595 1.124 0.255 )..700 0.958 0.137 0.660 0.0020.0001 1198 512 1.196 0.048 1.835 0.893 0.201 0.660 0.003

Ms

Press. 4Mr TUP CO02 CO 02 H2 H2 0 U2 minorcomponents

,psi) (cAI/g) (K) olee/100g

1000 0 3293 0.946 0.374 0.243 0.050 0.970 0.628 0.259

exhaust14.7 1554 2224 1.276 0.045 0.125 0.007 1.083 0.656 0.0411.0 1602 1513 1.321 0.000 0.110 0.000 1.-00 0.660 0.0020.5 1602 1344 1.321 0.000 0.110 0.000 1.100 0,660 0.0010.1 1602 1006 1.321 0.000 0.110 0.000 1.101 0.660 0.0000.05 1602 683 1.321 0.000 0.110 0.000 1.101 0.660 0.0000.01 1602 641 1.321 0.000 0.110 0.000 1.101 0.660 0.0000.005 1602 554 1.321 0.000 0.110 0.000 1.101 0.660 0.0000.001 1602 387 1.321 0.000 0.110 0.000 1.101 0.660 0.0000.0001 1336 298 1.1"/4 0.000 0.096 0.000 0.954 0.572 0.252

Working down the heat of explosion column we can see that the total heat ofexplosion increases significantly. The temperature of the reaction productsdrops. The reaction products shift from predominantly carbon monoxide (CO) topredominantly carbon dioxide (C0 2 ) and solid carbon (C). If we now look to thedata for NG, we see a rather different situation: the total heat of reactionrises only slightly with decreased exhaust pressurei the temperature dropsprecipitously and the composition of the reaction products barely changes. Thusif we were to make up a schematic of KG-burning rocket& similar to figure 2, wewould sea different behavior. The rocket exhausting to near vacuum wouldrelease less heat than that exhausting to atmospheric and more work 4ould bedone, but the sum of the work and heat would be nearly constant over thesechanging conditions.

346

The question arises: why does the heat of reaction for a rocket operatingat 1000 psi, exhausting to one atmosphere match the heat of explosion measuredin a calorimeter? We need to examine some basic thermodynamics to find ananswer to this query.

Typically in the propellant and explosive community, the HOE is usuallydefined to be the heat released when the material is ignited in a bombcalorimeter under an inert atmosphere (such as helium or nitrogen) at a pressureof 450 psi. HD!? is the heat released when an explosive is detonated in oneatmosphere of an inert gas. According to the International Union of Pure andApplied Chemists' (IUPAC) conventions, these values would be negative since theheat is released by the system. Tradition calls these values positive, so letus amend the definitions to allow them to be consistent with IUPAC practice:the HOE and HDET are the absolute values of the heat released in the relevantreactions.

For a chemical reaction, the change in enthalpy, &H, is the heat absorbedat constant pressure when no work is done other than PAV work. Ourdefinitions of HOE or HDET are clearly not describing constant pressurereactions. Thus, for our calculations, we must keep in mind that we are notdiscussing an enthelpy of reaction. If we look to thermodynamics, we find thatthe conditions of reaction at constant volume better fit the definition ofinternal energy, aU.

Equation (1) is a statement of the first law of thermodynamics. It saysthat the change in internal energy between two states of a system is equal tothe sun of the work done on the system, w, and the heat absorbed by the system,q. The state of a system may be defined by its chemical composition, pressureand volume. While the amount of work done by a system or the heat released by asystem in going from one state to the next may vary depending on the path takento make the change, the sum is independent of the path. If one path releasesmore heat, less work will be done by the syster.

U - w + q (1)

So HOE is really the &U for the reaction of the material under theconditions in the bomb calorimetero Since too volume change is possible, w - 0,then by equation (1), HOE is q. how the enthalpy change for a chemicallyreacting system may be exnressed as

AH - AU 4 A(PV) (2)

Combining equations (1) and (2) and using the ideal gas equation, we can write

S- w + q + an(RT) (3)

Equation (3) shows that the energy available from the reacting energeticmaterials gets apportioned into work done by the system (moving a rocket,destroying a target) and heat (reaction products which are at a highertemperature than the reactants). In the special case of the calorimeter, w = 0and so

AH - q + Pn(RT) (4)

The enthalpy of reaction, loosely called heat of reaction, can becalculated in terms of the standard heats of formation of the reactants andproducts. The practice of calling enthalpies of reaction "heats" confuses theissue of "heat of explosion". One is an enthalpy while the other is an internalenergy. This distinction must be kept in mind while dealing with thesecalculations.

347

AHr(298K) - ZlWIf(298K) (products) - tnAnHf(298K) (reactants) (5)

Almfr•( K) is the standard heat of formation of either a reactant or productspecies at 298 K and one atmosphere of pressure, and n is the number of moles ofthat species present. The difference, Ar(298K), is called the standard heat ofreaction, the heat released when reactants at 2 8 K and one atmosphere pressureare reacted and the products brought to 298 K and one atmosphere. Since thenumber of moles and heats of formation of the reactants will be constant for thereaction of a given energetic material, we can see that the heat of reaction mayvary if different experimental conditions cause different products to be formed.

PEP, a computerized thermodynamics calculation code, can provide the valuesfor equation (5). PEP takes into account the chemical species present, theinitial and final pressures of interest in a specific system and calculates theproduct species and the composition of the products. From these values and theknown heats of formation of the product and reactant species, the energyreleased can be determined.

PEP's output can be used to apportion the total energy available from thereaction of the energetic material, AU, into work, w, which can accomplish themission, and residual heat, q, which merely makes the reaction products hot.Designers of explosives and rocket and gun propellants are interested in morethan the total energy, AU, available from the reaction of their formulations.To make a useful material, the energy available must be able to do the workrequired to make the munition successful. Materials whose DUs are divided intosmall amounts of work and large amounts of heat are not good propellants orexplosives. Hot reaction products at the expense of work are not desirable.(Hot reaction products in themselves are not wanted since they erode nozzles andgun barrels and cause higher thermal signatures and flash.) The idealpropellant or explosive would release its reaction products at 298 K, thusgiving all of its available energy as work to propel the rocket or destroy atarget. The PEP code is used by the rocket propellant formulator to determinehow the energy available from his formulation will be split between work andheat ynder a given set of conditions. (Similar codes exist for gun propellants,Blake being one of the most widely used.) The output includes the heat contentof the reaction products, which is q. The work done can be determined bysubtracting the standard enthalpy at the conditions in the chamber from that inthe exhaust. These enthalpies are included in the description of the conditionsin both the chamber and the exhaust. In our modification to the code, thissubtraction is shown on an extra line of output along with the calculatedenthalpy of explosion or detonation.

For exploring the relationships between energy released in a reaction andthe conditions under which the reaction takes place, PEP can be used tocalculate the pcrformance of a simple material like trinitrotoluene, TNT. If weare to use the PEP code to calculate the equilibrium composition, temperatureand work done by the reaction of TlT, we must specify the conditions under whichit reacts. we can do this by providing the operating and exhaust pressures of arocket motor. These values allow the code to calculate the equilibriumcomposition and then to divide &U between work and heat.

Figure la shows a schematic of a rocket motor whose propellant is TNT andthe breakdown of the energy available from the reaction of the TNT at 1000 psichamber pressure exhausting to one atmosphere. The temperatures in the chamberand at the exhaust were calculated to be 1991 and 1045 K respectively.Operating under these conditions, -500 cal/g of TNT do work in propelling therocket. The exhaust products in cooling from 1045 to 298 K release -265 cal/g.We can add these values to determine that bU is -765 cal/g.

t4 PaulOO0pei T-1991 I

w--500 cal/g wm-1015 co g

L "T~-1 K Exhaust Ta.613KfýW

q--265 cal/g qw-12co4/II

Is 4 -14.7 psI T-298

la Exhausting to 14.7 psi lb Exhausting to 0.0017 psi

Fgqure 1. Partition of TNT energy in rockets operating at different conditions

If we force the TNT to do more work, w, we can increase AU. If wepostulate a rocket operating at a lover exhaust pressure, more work is done onthat rocket than on the original rocket with the higher exhaust pressure.Figure 2b shows such a rocket, exhausting to 0.0017 psi. Now the exhausttemperature is 613K, leaving -212 cal/g to be released by the cooling gases.More work is done by this rocket, some -1015 cal/g. The total 4U is -1127cal/g.

If we were to look at the NG case, we would not see anything like thisshifting. The heats of reaction and composition of the reaction products stayrather constant as more work is done by the system, thus the heat released mustdecrease. Examination of the temperature reveals just that. The temperature ofthe reaction products drops rapidly with decreasing exhaust pressure. Therejust is not much energy left over to heat up those exhaust gases.

Why is there this dramatic difference in behavior between two seeeingiysimilar energetic materials? If we look to equation (3), we recognize that thebehavior of v does not change from the TNT to the NO case. The work done by thesystem on the rocket increases with decreasing exhaust pressure similarly inboth cases. The conditions at which the reaction take place force this changein work done by the system. The change in total energy released is mirrored Inthe heat lost. This can be seen in the more drastic drop in temperature of thereaction products for the NG case compared to the TNT case. It is worthremembering here how we determine the total heats of reaction with the PEP code.Equation (4) is used, where we use the heats of formation of the reactants andproducts, multiplied by the number of moles calculated by the code to exist atthe equilibrium conditions to determine the heat of reaction at thoseconditions. Sice the reactants do not change with changing conditionn, it mustbe that the shifting product compositions bring about any increase in heat of

349

reaction. As the exhaust pressure drops, the reaction products shift to morestable species, releasing more energy which can be used to do work. Thus, inresponse to the demand of the rocket to do more york, the TNT reaction shifts tomore stable products, yielding the energy required to do that work. It must beforced, by the conditions, to yield the maximum energy.

What is it that makes TNT behave so differently from NG? We can find theansver in their oxygen balances (08S). if we talk about the reaction of ageneralized explosive

CxHyOzNuFvAlV .-------> X(C0 2 ) + y/2(H2 0) + v(HP) + w/2(A1 2 0 3 )

then08 - (2z + v -3w - 4x - y1(8001

molecular weight

While the definition of oxygen balance includes such species as fluorineand aluminum, we will confine our discussion to the simpler carbon, hydrogen,oxygen and nitrogen compounds. A negative oxygen balance indicates that acompound does not contain enough oxygen to go completely to CO 2 and water (H20)products. Some carbon monoxide (CO), elemental carbon (C) and hydrogen (H2)will be formed. Additional oxygen is required if the reaction is to go to themsot stable products, i.e., CO2 and H20. If we write the balanced chemicalequation for the combustion of a material with a positive OB, we find that allthe carbon reacts to form CO2 and all the hydrogen goes to 120 with a bit ofexcess oxygen in the products. NG has a positive 08.

The positive 08 of NO allows its reaction products to go to their moststable composition whatever the reaction conditions. The moat stable reactionproducts, CO2 and H20, allow the greatest energy release by the reaction. Ifone gets the most energy possible from a reaction at any conditions, changingthe conditions to yield more work done by the system will result in a loweringof the heat released by the system. This is exactly what we see in thecombustion of NG ina varying rocket motors. The positive OB allows NG's reactionto proceed to a very stable set of products, which yield the maximum energy,despite the conditions under which the reaction takes place. We call such amaterial a "quick" explosive.

If we look again at the Table I data for the TNT case, we see that thenegative OB is reflected in the composition of the reaction products. Thereaction does not proceed quickly to the most stable reaction products, Co 2 andM20, because there just is not enough oxygen to go around. Instead we are leftwith a mixture of C0 2 , CO, C, H2 0 and H2. These exist in an equilibrium whichshifts depending upon the conditions. As we saw before, lowering the exhaustpressure forces the products toward the C0 2 , ruleasing more energy which is usedto do work upon the rocket. The conditions must demand the shift in thereaction products to those which provide more energy to do work on the rocket ifthe TNT is to yield its maximum energy change. For this reason, we call TNT a"lasy" explosive.

THE CALCIJIATIONS

To actually run the HOE calculations, we do a rocket problem with thechamber pressure set at 1000 psi exhausting to one atmosphere, 14.7 psi. Theheat content, q, of the expanded gases is calcuated and shown on the printout.The enthalpy diffaeence between the chamber and the exhaust is calculatedthrough a knowledge of the combustion products, their percentages and standardenthalpies of formation. This enthalpy difference io used as an approximationto the work done, added to the heat content and called the enthalpy ofexplosion. A routine has been written by one of the authors which performsthese nalculations automatically and adds a line to the output showing this"heat of explosion".

3350

If we nov look at our calculation, we can see that it does not really giveus the internal energy change. We do get out the q value as the heat content ofthe expanded gases, but the enthalpy difference between chamber and exhaust isnot exactly the work done in that step. Equation (4) shows us that the enthalpydifference for that step would be the work done plus an(RT). (Remember that q is0 because this is an adiabatic expansion.) The PEP code output gives us thenumber of moles of gas in the chamber and in the exhaust so we can subtractAn(RT) (setting T to 298 K since the calculation is based upon standardenthalpies of formation) from the enthalpy difference to give the work done onthe rocket by the expanding gases. Adding this work to the heat content wouldgive the internal energy change for the overall change from reactants atatmospheric pressure and 298 K to products at those same conditions.

Our observations were that use of the uncorrected enthalpy gave the bettermatch to the experimental values. Subtracting An(RT) from the total would leadto fairly consistent underestimation of the HOEs. Sources of non-ideality existwhich could be invoked to explain this dilemna. The calorimeter is ignited witha hot wire which adds several calories to the system. No standard exists forcorrecting for this in the calculation of HOE from the temperature rise of thewater bath.

Table II compares experimentally determined values to the results of HOEcalculations using both our method and the BLAME code. Where no indication ismade for a reference, the values are those measured in our lab. BLAKE-determined values are shown for freeze-out temperatures of 1500 and 2500 K. Formost materials, there is not a significant difference between the two methods ofcalculation.

Table II Comparison of Calculated and Experimental HOEs

Hat'I Expt'l PEP BLAKE (Freeze temp)1500 K 2500 K

TNT 665 766 775 733NG 1600, 1510 1554 1624 1599HMX 1358 1337 1361 1362RDX 1368 1349 1366 1367TETYL 932 909 892 881PENTOLITE 978 979 1045 1045NC (12.6%N) 968 930 959 960

One of the primary values of our method of calculation is that one need notchoose a freeze-out temperature. There has been a choice made: the chamber andexhaust pressures at which the two compositions are calculated. The freezingaccounted for in BLAKE's HOE calculation and the heat loss to the water comeinto play here to keep the pressure lower than that calculated in the POInt run.Thus we would expect that the two states which our method calculates, at 1000psi and atmospheric, come close to representing the actual conditions in thebomb after and before burninq the material.

HEATS OF DETONATION

In a way similar to that which led to our development of the calculationparameters for HOEs, it was noticed that determini''- -e sum of the work doneand heat released for a rocket operating at 1000 psi exhausting to 0.0017 psiwould yield a good match to Ornellas' experimentally determined heats ofdetonation. Table III shows calculated versus experimentally determined HDETsfor a number of materials. Ornellas determined the HDETs given in the tableusing gold cylinders to confined the explosive in the calorimeter. For lowoxygen balance explosives, this confinement yielded higher HDETs than otherworkers have reported for the same materials unconfined. We interpret this tomean that the confined "lazy" explosive was forced to do more work, thusyielding a higher HDET.

• ' ,, , a! I II351

Heats of detonation measured under different conditions would likely yieldsomewhat different values than Ornellas'. To match those values, a differentchoice of exhaust pressure would have to be made. Once that exhaust pressurewas determined through fitting experimental data, that value could be used inthe program to predict other values of the HDET measured under thoseexperimental conditions.

Table III Comparison of Calculated and Experimental HOETs

Nat'l Expt'l Ornellas calcid PEP-calc'd

TNT 1093 1133 1127MG 1486 1590 1603H]MX 1479 1514 1525RDX 1452 1488 1534FEFO 1279 1364 1349PETH 1495 1516 1574lMB 1653 1653 1666TATB 1018 ;022 1028

In Ornellas' experiments, he not onl' meafured the HDET, but determined thecomposition of reaction products. In Tabie III, the column labelled "Ornellas'calculated" is the HDETs he calculated using equation (5) and the moles ofreaction products he had measured.

VARYING REACTION PRODUCTS

It is possible to calculate different reaction products which will yieldthe same HOEs or HOETS. Table IV shows data for HKX burned at 1000 psi and 3277K and allowed to expand until the products are at 2000 K and 1000 K. In otherwords, differing amounts of work are done by the expanding gases, leavingdifferent amounts of heat to be given up when the gases cool. The sums of thenumber of moles of each product multiplied by the heat of formation for thatspecies less the heat of formation of HMX times the number of moles of HMX inlO0g (which are the HOEs) are the same, but the composition of the gasesdiffers. Thus, we can see that the gas composition will vary depending on theamount of work that the gases do in expanding, yet the HOE will be the same.Another example of this is shown in Table V which gives the composition of thedetonation products of HMX both as measured by Ornellas and calculated by ourmethod. The HDETs match, but the compositions differ!

Table IV HOE at Varying Temperatures

nMHf(products) (kcal/10Og)Product @ 2000 K @ 1000 K

CO2 0.431 (-94.051) = -40.536 0.737 (-94.051) = -69.316CO 0.919 (-26.417) - -24.277 0.613 (-26.417) = -16.194112 0.431 (0.0) . 0.0 0.734 (0.0) = 0.0H2 0 0.919 (-68.317) = -62.783 0.614 (-68.317) = -41.947No 1.351 (0.0) - 0.0 1.240 (0.0) = 0.0ml norproducts = 0.051 -0.139

totals (kcal/lOOg) -127.647 -127.647nflHf(HMX) 6.054 6.054(n - number moles in 100 g)

HOE - -[EnaHf(298K)(products) - EnLHf(298K)(reactants)] = 1337 cal/g

352

This argument would not apply to the ideal calorimeter since there is nowork done. There, the entire difference in internal energy takes the form ofheat lost to the bath. This assumes that no work of any sort is done in thecalorimeter which is not converted to heat which can then be lost to the bath.

Table V Comparison of HMX HDET Calculation with Varying Products

n6lHf(products) (kcal/lOOg)Product PEP Products Ornellas' Expt'l Products

CO2 0.948 (-94.051) - -89.160 0.648 (-94.051) - -60.945CO 0.003 (-26.417) - -0.079 0.358 (-26.417) - -9.378C 0.359 (0.0) = 0.0 0.320 (0.0) W 0.0

H2 0.447 (0.0) - 0.0 0.101 (0.0) - 0.0H2 0 0.803 (-68.317) = -54.860 1.070 (-68.317) - -73.099N, 1.351 (0.0) - 0.0 1.240 (0.0) - 0.0ainorproducts - 0.851 = -1.849

totals (kcal/lOOg) -144.950 -145.350n,,Hf(HHX) 6.054 6.054(n = number moles in 100 g)

HDET - -[EnIHf(298K)(products) - 2nAHf(298K)(reactants)]

- 1510.04 cal/g by PEP calculation

- 1514.04 cal/g by Ornellas' calculation from expt'l data

Note: The above PEP calculation was not run with an exhaust pressure of 0.0017psi, but rather the exhaust pressure was varied until a match to Ornellas' HDETwas found (exhaust pressure = 0.0025 psi). The point here is to illustrate thatthe same value of HDET (or HOE) can be determined from different productcompositions.

It is worth restating this phenomenon: the same heat of explosion may bedetermined though different combustion products are formed. In actuality, thedesign of calorimeters allows for non-idealities which will be shown invariations in combustion products. The combustion products expand against theinert pressurizing gas, mixing only to a limited extent •uring the combustionevent. Cohen et al at the Ballistic Research Laboratoryg examined very lowpressure HOE data and showed that this effect is quite important. Within therealm of standard test pressures, the effect is not great, but enough to helpaccount for variability between facilities. The sample size and particle size,degree of confinement and size and shape of the sample cell will all contributeto such variations.

Closer examination of the situation in which the combustion gases expand todifferent degrees before cooling at constant pressure reveals it to be similarto the concept of freezing. When combustion products begin to cool, theircomposition shifts to be at equilibrium at the successively lower temperatures.There comes a point at which the kinetic rates for the reactions causing theshifting composition are too slow to maintain pace with the heat loss. At thispoint, the reaction freezes, with the composition at the fre.ezing temperatureremaining unchanged as the gases continue to cool to ambient temperature. Wecan relate freezing of the composition to the amount of work done by the gases.More work done by the gases forces the freeze-out temperature down. Thus thefreezing temperature is not only a function of the explosive material, but theconditions under which it reacts.

UNDERWATER EXPLOS VIES

For underwater explosions, the pressure ot the water surrounding thereacting material forces the explosive to do more work in expanding its reactionproducts. Aecoqnizing this situation as another mode of confining theexplosive, we believe we can use this method of calculation to predict HDZ?. forunderwater explosives. Of course such prediction will have to *wait appropriateexperimental data to alloy us to find the beat input pressures for the code.

CONC•USIONS

We have presented a method of calculating NOEs, HD!?- and ,esction productsbave4 upon a rocket performance computer code, PEP. Won-r,"tlne use of thei'•.". provides insight into the factors causing variability in reportedCJ., i,-i.tal results. While the thermodynamic calcu!ations used in aikinq the

:r .on are at odds with the classic ideal case of calorJmetry, they do offerS,.�. a into the non-ideal actuality of the situation and perhaps an

appz.ýiation of just how non-ideal sums of our experiments are.

Predictions made with this method show good agreement for many typ4s ofenergetic materials. Application or the code to the prediction of underwaterexplosive performance seems to be its major potential. Another use may be inthe prediction of the reaction products from the disposal of energeticmaterials.

REFERENCES

2. OTheoretical Computation of Equilibrium Compositions, TheraodynamicProperties, and Performance Characteristics of Propellant Systems, D. R.Cruise, KNC TP 6037, April 1979

2. "Computer Program for Calculation of Complex Chemical EquilibriumCompositions", S. Gordon and Bonnie McBride, NASA BP-273, Lewis Research Center,1971

3. "BLA.K - A Thermodynamics Code Based on TIGER: Users' Guide and Manual*,ARBRL-TR-0241, July 1982

4. OCalorimetric Determinations of the Heat and Products of Detonation forExplosives: October 1961 to April 19820, D. L. Ornllaa, Lavrenco LivermoreNational Laboratory, April 5, 1982

5. "Heats of Explosion at Low Pressures*, A. Cohen, M. Miller, et al, BDL-TP-3024, August 1989

354

BALL POWD(R* PROPELLANT APPLICATIONS TOLARGE CALIBER IIMUNITION

Or. Ernest J. KirschkeAntonio F. Gonzalez

Olin Corporation, Ordnance OlvisionSt. Marks. FL

ABSTRACT

Recent advances in BALL POWDER* propellant manufaýturing technology have made possible theuse of this type of propellant in large caliber munition.

The tec ,ical advances in BALL POWDER* propellant theory relevant to large caliberapplications is discussed. The advantages and limitations of large caliber BALL POWDER*propellants are presented, as well as test results in. tank and artillery systems.

INTOOUCT ION

$ALL. POWIER propellant was commercially introduced In 1933, for use in selected ,mallcaliber systems. Since Its Introduction now BALL POWDER* propellant designs have been successfullyadded fir a wide range of s41il and meoJii, caliber systems, a-, well as mortars.

Because of limitations in the particle sizes that L.,.d be manufactured, BALL POWDER*propellants were Pxcluded from use in large caliber, This changed in 1987. when an economicallyviable process dos developed for mass prodJction of BALI. POWDERP propellants suitable to tank andP tillery applications. Today research and development efforts are aimed at the Introduction ofhi2h perfoimance, low cost, low vulnerability BALL PO, )-R propellants for tank and artillerysystews.

PERFOIF04ANCF THEORY AND APPLICATIONS OVERVIEW

BALL POWDERe propellfm, has been cr acterized as a granular propellant composed of smallspheroidal grains. Double base formulatior,- (nitroglyce-in and nitrocellulose) are most common,although single and triple base formulatisi, are available.

Progressiwity is controlled with deterrents or burning rate modifier!., which are impregnatedon the outer shell of the grains. 1 ,2 A typical cross sectional analysis of a double base deterredBALL POWDERe Propellant grain is illustrated in Figure .

S20)I2 -... NITROGLYCERIN

LJ-[ DETERRENT

w

z

I-

0 0

GAAIN O IAINSIUIRACI OCENTER

A.L 0 10 20 30 40 50 60 70 80 90 100

PENETRATION DEPTH 17. OF GRAIN RAOIUSI

Figure 1. Cross Sectional Lhemical Profile Of A BALL POWDER' Propellant Grain.

"Afpprnved for public release; distribution is unligited."

355

Grain size (web) determires the BALL POWDER* propallant surface to volipe ratio &Ad the burnout tir'ý.

Proper selectimn if the BALL POWDCI;* propellant particle size Is critical to. performance in anyapplicaticm. The jar icle )Is* requir-iment for a given systas is dopendoint on a ftuuber of factors:operating 'tIlotsre, rioptilant auid pro~ectile weights and bore dlimotor. In general terms. thelarger tht calite.r of the system, the ia~rgor the particle size required, Figure 2 illuttrateS thispoirtt noting that systems such as uie mortars do not fit this trond.

I

1MM~~~~~~ )UM J1M4IM 155 2M WPISTOLJJJ1

Bigur 2.BL. ODWOPoElatPatce ieReurviet

ADVAETAGNS AN IITTOSOFBL. ODR POELAT -

GRARRE SIEATN AND1EROSIO

M AM POWDE) Pr.elat haeben 1d 1otd1nrpd iean i pefrlac usystems~~~~~~~~~~ beas c t owrbrrlhain Jn eoinchrcersis

33')ii A"

The reduced barrel bcating and barrel srosion rates demonstrated with BALI. POWDER*D pr.3pellantIs directly relsited to tho chem icll profile thAtcACterlsitic or this propellant. The deterrent layerin SIALI. POWEROt prupalla~tit result in lower flame temperature and cool burning gases during theinitisi stages of combustion, and throuuh the time when maximum gun chamber pressure is ach~ieved.During these ttagwt of tho Interior ballistic cycle, BALI'. POWDER* propellan~ts can reduce the

avrae un gai temperature by sip to 1000*K, as compared to propellants of uniform chiiamcoelposi ion.

During the later stages of the interior ballistic cycle, the undeterred region of the BALI.POWVDft* propellaot grain,, ?s raschod and the propellant burnt. faster and at a higher flametemperature. It. is also during this stage tP'at the projectile is rapidly accclerating, causingrapid adiabatic ex;)ansion and cooling of the proipellant gases. As a result gun gas temperaturesfurther decreae. dpspitt the increaset1 propellant flame temperature.

Figures ,and 4 illustrate this po'nt, in the 155 Artillery M199 Cannon. Mean gas gunttffiperaturqs are tiaed on ISBHYG-11 models.

3400

CE 1200-----------

~ 2$00~. 31AIE1

7. 2400uJ II'-WJ 2200 - BALL POWDER*

2000 --

1800 ---

0 10 20 30 40 50 60 70 80 90 100

(SURFACE) DEPTH IN GRAIN BY WEIGHT PERCENT (CENTER)

figure 3. Flame Temperature Prcfiles of Extruded And BALL POWDER*Propellants For 155nv Howitzer M203AI Charge

230 ER GUN CODE IBHVS--4 MODELS

/2NO , CONDITION AT p MAX

2100 M31A1EI

2000ttwtA1'~( IGNITLR GAS TEMPERATURE

(fV)1900

S.................................",

BALL POWDER'17001

Im0 1 2 3 5 6 7 9 10 11 12

lIME (1MI)

Figure 4. Predicted Mean Gun Gas Temperature Vs. Time Of Extruded And

BALL POWDER* Propellants In The lSSmm Howitzer M199 Cannon (M203A1 Charge)

VULNERABILITY

Granular BALL POWDER* propellants for medium and large caliber &mmunition have demonstrated avery low level of response to shaped charge attach. Various tests hove been conducted by the BRLand ARDEC to evaluate the vulnerability of BALI. POWDER* prnpellants, especially formulationstailored for the 25rm Bushmaster, the 120mm Tank KE round and the 155mm Unicharge. In all casesgranular BALL POWDER* propellants have demonstrated a high degree of insensitivity to shaped chargeattach. The shock velocity data in Figure 5 is representative of the results observed.

358

7MS

6-

0*_j 0 4M30

> U

U'0 9 SALL POWDERP

Z~ 2i INERT Irlcel

I l i.J_ J. L L L 1____0 10 20 30 40 50 60 70 80 90 100 110

DISTANCE (mm)

*O.R.L. SHOCK VELOCITY TEST RESULTS

Figure 5. Response Of Various Propellants To Shaped Charge Attack

The high degree of insensitivity of BALL POWDERe propellants to shaped charge jets may be dueto the high concentration of deterrent on the outer surface of the grains and to the granulation ofthe propellant. Further experimentation Is ongoing to improve the understanding of the BALL POWDER'propellants response to shaped charge jets.

LOADING AND ENERGY DENSITY

The small granulation, spheroidal shape and smooth surfaces of BALL POWDERe propellantsresult in free flowing grdnular charges of loading densities averaging 1.0 gT/cc (.036 lbs./in3 ).

The free flowing characteristics facilitate loading and allow for increased charge weights intortuous volLmes, such as mortar in,'rements and regions around finned projectiles.

Proper ignition system diign is critical in achieving optinum performance with granular BALLPOWDEI* propellant charges. The high packing density and small granulation of this propellant mustbe considered when matching ignition and propellant charges, as localized ignition may yield nonuniform ignition and pressure differentials.

The impetus level of BALL POWDER* propellants currently designed for tank and artilleryammunitlon ,-ange from 925 to 1090 kj/kg I31O,OOO lo 365,000 ft-lbsf/lbm). The 1090 kj/kg (365,000ft-lbsf/lN) represents the upper limit presently attainable with large caliber double hasepropel ants, as low energy deterrents comprise 4 to 8% by weight of the fcrnulation.

35()

Energy densities of up to 1090 J/cc (13,140 ft-lbsf/in3 ) can be achieved with granular BALLPOWDER* propellants. This level of energy density meets or exceeds the requirements of most of thecurrent tank and artillery ammunition.

For applications where a higher energy density is required, compacted BALL POWDER* propellantcan be used, icreasing the energy density by up to 35% over that achievable with a loose granularconfiguration.

BALI.TSTIC TEST RESULTS

BALL POWDER' propellants in both loose and compacted forms are being evaluated in largecaliber systems, including 120= tank KE and trainer rounds and 155mm artillery.

The BALL POWDER* propellant charge designed for the 120mm tank KE trainer round (M865), is aloose granular charge. Ballistic test results indicate that this BALL POWDER* propellant chargecan meet the requirements of the 14865 round of ammunition. Test data is presented in Table I.

Table I. Test Results of Developmental BALL POWDER* propellantCharge for the M865 Round

Conditioning Pchamber Vmuzzle Action Time - AP

TeMp. (-C) (bars) (m/s) (msec) (bars)

+21" 4326 1688 13.4 128

Specifications (4950 1700 ± 10 <35 <345

+63* 5103 1776 11.0 4

Specifications <5910 - <35 <345

-460 3775 1577 26.2 10

Specifications - - <35 <345

M125 primer with modified vent pattern used.

lSm= Artillery is another large caliber system in which BALL POWDER' propellant hasdemonstrated promising ballistic test results. Evaluations in both the M199 cannon and thedevelopmental 23L (1400 in )/52 caliber cannon, demonstrate that BALL POWDER' propellant is aviable candidate for 155mm artillery charges, including the M203AI and unicharge.

The charges evaluated cowisist of annular sections of compacted BAI.L POWDER* propellant withan ignition train of single perf, slotted, M30 stick through the center hole of the BALL POWDER*propellant compacts. Various BALL POWDER* propellant charges have been evaluated, maintaining theM30 core constant at I Kg (2.25 lbs.).

Figure 6 depicts ballistic results of the two BALL POWDER' propellant charge weights whichbraý.keted the performance requirement of the M203A1 charge.

360

3000 30.0LLS.

t 2800___

BA26.25 LL3S. 4E REMENT =27 0 FPS0 BALL P01 DER

O 2600W -

.JI• 2400 - -......

2 M549S PROJECTILE2200 r I 1 1

35 40 45 50 55 60

CHAMBER PRESSURE (kpsi)

Figure 6. Demonstrated Ambient Performance Of BALL POWDEROPropellant In The 155mm Howitzer M199 Cannon

As shown in Figure 6, this particular charge configuration meets the velocity requirement of825 m/s (2710 fps) at approximately 325 MPa (47,000 psi). Based on these results it is calculatedthat a fully optimized charge of this type can reduce the ambient operating pressure in this systemto under 295 MPa (43,000) while meeting all other performance requirements.

It Is also calculated that the M2O3AI and unicharge performance requirements can be met witha loose granular BALL POWDER* propellant charge.

CONCLUSIONS

Technology advances in the manufacturing process now allow for cost effective productionof BALL POWDERe propellants of the grain sizes required for large caliber ammunition.

Loose and compacted BALL POWDER* propellant charges have been successfully evaluated intank and artillery systems.

Granular BALI. POWDER* propellants have been demonstrated nearly insensitive to shapEdcharge attack.

Reduced barrel heating and extended barrel life can be expected with the use of BALI.POWDER* propellant charges in large caliber ammunition.

M61

REFERENCES

1. A. Gonzalez, D. Worthington, Olin Corporation, St. Marks, FL, and W. Aungst and J. Newberry,ORL, "BAL. POWDOR0 Propellants And Temperature Compensation, A Combustion Study", 26th JANNAFCombustion Meeting, CPIA 529, Vol. 1, October 1989

2. R. 0. Anderson, R. T. Puhalla, BRL, "Parametric Study of Temperature Insensitivity of BallPropellants' 26th JANNAF Combustion Meeting, CPIA 529, Vol. III, October, 1989.

362

PROGRESS REPORT ON LIQUID PROPELA~NT INJECTOR/COMBUSTOR TESTS'

R. E. Rychnovsky, R. W. Carling, S. K. Griffiths,S. R. Vosen and R. F. RenziSandia National LaboratoriesLivermore, CA 94551-0969

ABSTRACT

A liquid propellant injector/combustor (LP VIC) fixture that produces pressure oscillations similar to those observed in liquidpropellant gun firings has been designed, built and tested. The oscillations are due to combustion of the LP. Methods to reducethese oscillations are a high priority in the development of an LP gun for future Army artillery.Several different methods to reduce or eliminate the oscillations have been considered and tested. Experiments were conductedwith three liners in the combustion chamber. Two showed little reduction in oscillations but the third, a flexible liner, eliminatedthe pressure oscillations. Results of these tests and other oscillation reduction methods will be presented.

INTRODUCTION

High frequency pressure oscillations have been consistently observed in the combustion chamber of regenerative liquidpropellant guns (RLPG) during firing. Their origin and long-term effects on gun performance and gun life are unknown, buttwo detrimental effects can be postulated. First, the oscillations may create a turbulent boundary layer that enhances conductiveheating of the barrel, increasing wear and erosion, thereby reducing the barrel life. Second, the oscillations may also couple intoprojectiles and cause failure of critical components that are sensitive to these frequencies. To support the RLPG development,we are investigating the origin of the oscillations and ways to significantly reduce them.

At Sandia National Laboratories, we have designed and developed a liquid propellant injector combustor (LP I/C) (Figure 1) tosimulate the test conditions in a gun and devise methods to reduce oscillations. This experimenul program is supported by amodeling effort to select the experimental parameters and aid in the understanding of the results. Our approach is to first reduceoscillations in our test device without regard to how the technique may be applied to the RLPG. Then the results will bemodified (if necessary) and applied to rcduction of oscillations in the full scale RLPG.

--BURST

DISKSGAS RESERVO,-iTRIGGER BAFFL -

PISTON HOUSINGJ '

METAL HONEYCOMB

LIQLID RESERVOIR

COMBUSTION CHAMBER

Fig. 1. An artist's conception of the Sandia liquid propellant injector/combustor is shown.

To date, 21 combustion tests have bezfn completed with the LP I/C. Pressure oscillations in this device during combustion aresimilar to those observed in the combustion chamber of the RLPG during firings. That is, the amplitude of ýhe oscillations are

This work is supported by the Department of Defense Office of Munitions and the Department of Energy through a Memorandum of Understanding.Approved for public release. Distribution unlimited.

Best Available Cony'y 363

similar (about plus or minus 30 percent of the base pressure) an:' .:t Llequencies are similar when geometric adjustments areincluded. However, the excited oscillations in our LP I/C testei wre in narrow frequency bands whereas the oscillations in the155 mm gun are dispersed over a wide frequency range.

An important result was recently achieved when a test was conducted with virtually no pressure oscillations. In this test, aflexible, shock-absorbing liner significantly reduced the pressure oscillations in the combustion chamber. A description of thesetests and their results are included in this paper.

TEST APPARATUS AND DESCRIPTON

The LP tester is a piston intensifier design where a moderate pressure (20 to 40 MPa) gas pressurizes LP in its chamber to 200to 400 MPa (Figure 1). The high pressure LP is injected into a combustion chamber that has been pressurized with hot gas froman igniter. The injection orifice and exit orifice on the combustion chamber are sized to provide the desired combustionpressure. Reference I and 2 describe the details of this test apparatus. Tests have been conducted with 35 to 65 cc of LP. Thediameter of the injection orifice has been varied from 5 to 7 mm. The combustion chamber is cylindrical with the interior wallconverging smoothly to the exit nozzle. LP is injected in a circular stream along the combustion chamber's centerline.

A typical injection pressure in the LP chamber was set at 200 MPa (30 ksi) and the resulting peak steady-state pressure in thecombustion chamber was 130 MPa (19 ksi) with a 5amm nozzle and 55 MPa (8 ksi) with a 3 mm nozzle. Figure 2 compares thepressure in the LP chamber with the pressure in the combustion chamber. Note the oscillations on the pressure trace taken fromthe combustion chamber. The LP I/C is instrumented with five high-frequency pressure transducers (four in the combustionchamber and one in the LP chamber), three low-frequency pressure transducers, an accelerometer on the piston that injects LP,and two fiber optics in the combustion chamber. The fiber optics provide optical access to the combustion chamber. For theexperiments described in this paper, the total visible emission collected from the fibers impinges on a photomultiplier, resultingin a voltage that is proportional to the total visible light emission.

240 1

LP chamber200 / "-"\

= 160 -- ,. Combustion

LP injection / chamber120

U)( /U&. 80 - 1

40 ..- "'.... .... .. ." -LP combustion

0 4 8 12 16 24 28

Time, ms

Fig. 2. Representative pressure traces from the LP I/C. The upper curve is taken from the liquid propellant chamber, thebottom from the combustion chamber.

In addition to baseline combustion tests into a two inch diameter chamber, tests with a lead overlay over the chamber and withfour different combustion chamber liners to suppress oscillations were conducted.

Best Availab' Cp

TEST RESULTS

The results seen in Figure 2 clearly show oscillations in the combustion chamber. A fast fourier transform (FFT) analysisshowed that most of the response was at a frequency of 27 kHz. To assure ourselves that the oscillations weren't the results ofmechanicaln g of the sduc rs we installed one transducer in a "blind" hole. The mounting configuration for thistransducer was detical to the active transducers except it did not communicate directly with the combustion chamber, a .060'thick, solid wall eeprated the transducer from the combustion chamber so it should read zero pressure. Initial tests were madewith a small amount of grease on the face of this transducer. It was found that the grease filled the small cavity and the pressurein the combustion chamber was transmitted through the chamber wall through the grease and was recorded by the pressuretransducer (although at much reduced pressure levels). However, when the blind hole had no grease, the pressure in thecombustion chamber was well isolated from the transducer and a small amplitude, broad band, high-frequency response wasrecorded This result demonstted that spurious signals (e.g., from response of the pressure transducer to mechanical inputs)are not a significant component of the pressure measurements in the combustion chamber. A comparison of pressuremeasurements taken in the blind hole with those in the combustion chamber indicates that pressure transducer measurements armaccurate for frequencies less than 60 kHz.

A spectral analysis of data from a base-line configuration test was performed to characterize the oscillations. Specifically,combustion-chamber presure and fiber optic probe data at several locations were compared. A dominant frequency componentat 27 kHz has been noted in all of the diagnostics. While no other dominant frequency components in the range of I kHz to 500kHz were measured with the fibet optics, several position deperdent frequency components occur in the pressure measurement.These frequencies are: 4.8. 13.5, 27.0, 50.0, and 55.0 kHz. A detailed analysis of the data, including phase differencesbetwecn thc signals, is currently being conducted in order to assign specific acoustic modes to these frequencies. It appears thatthe four lowest frequencies are associated with the first longitudinal, first tangential, first radial, and second radial modes.

The origin of the pressure oscillations have been thought to be acoustic, but some observers have suggested that they were aresponse to mechanical vibrations of the hardware. To evaluate this suggestion, a lead overlay was clamped over thecombustion chamber to try and dampen any mechanical response the pressure transducers were measuring. Two tests wereconducted with this configuration with no noticeable reduction in oscillations. These results support the conclusion that theoscillations are related to combustion and acoustic phenomena rather that a response to some mechanical vibration of the testchamber.

Once we had convinced ourselves that the pressure oscillations were originating from the combustion process, we set out tocvmluat mnethods to undersnd the om-illations and means to reduce them. Methods to redt"e the oscillations fall into two maincategories: (1) lower the unburned mass of propellant in the combustion chamber to eliminate the energy source that drives theoscillations and (2) modify the combustion chamber to suppress wave propagation (that is, to absorb energy). Experiments incategory one are underway. They center around changes to the injector design to break the jet into droplets, promote rapidcontrolled burning of the LP and to reduce accumulation of unbumred propellant in the combustion chamber.

To address the second method of reducing oscillations, a quarter-wave liner (Figure 3), a Helmholtz resonator (Figure 4). and aflexible liner have been tested. The 1/4 - wave liner was a thick cylindrical shell with an inside diameter of 1.33 inches. One-quarter inch diameter radial holes cover about 35% of the outside surface. In concept, the pressure wave is to be reflected at thebottom of these holes, mect the next wave 180 degrees out of phase and thus cancel the oncoming wave. The depth of the holesdetermine the frequency of the wave that is suppressed, so thLi device is effective for only a narrow frequency range

• LP LINER

lig. 3 A photgraph of the quarter-wave liner used in the experiments. The quarter inch holes in the liner are designed to trappretssure oscillations as they occur in the combuslion chamber. "he liner is 3.5 inches long, has an outside diameter of2 inches, and an inside diameter of 1.33 inches.

365

Another test included a Helmholtz resonaw. in the combustion chamber. The Heintholtz resonator is similar to the quarter-waveliner but is modified to trap oscillations over a wider band of frequencies. This thick-walled liner has a series of holes andcirvurfertntiai slots with an inside diameter of 1.2 inches.

Fig. 4. A4photograph of the Helmholtz resonator used in the experiments. The resonator is designed with holes and slots in aneffort to dampened the pressure oscillations over a larger bandwidth than the quarter-wave liner. The resonator is 3.5inches long, has an* outside diameter of 2 inches, and an inside diameter of 1.2 inches.

The results using a quarter-wave liner and a Helmholtz resonator were similar. That is, pressure oscillations were still observed,,but the frequency was shifted to 45 kl-z compared to 27 kHz without the liners. This shift in frequency is proportional to the,change in inside diameter of the combustion chamber with and without a liner. A straightforward experiment to support thisconclusion included a solid liner (inside diameter of 1.33 inches) in the combustion chamber that did not have any holes or slotsexcept thse to provide access for the igniter and diagnostics. The results from this test were similar to those using the quarter-wave and Helmholtz liners. That is, the dominant frequency was at 45 kHz. This result using the solid liner without holessbowed that the quarter-wave liner and Helmholtz resonator did not perform as hoped. A redesign of the Helmrholz resonator ismlerway. It offers the most promise of the two configurations because of its wider frequency bandwidth.

Finally, a flexible liner with an inside diameter of 1.25 inches was tested. This insert was a thick walled, steel-reinforced rubberhose. This material and design were suggested by Walt Pasco of GE Pittsfield. It has x¢veral properties that make it a goodcandidate to reduce pressure oscillations. For example, it has good acoustic impedance match with the combustion products sostress waves are transmitted into the flexible liner. The results from a test using the flexible liner are shown in Figure 5. Thefigure shows that the pressure osciliazions observed during combustion of liquid propellant in the LP I/C have been virtuallyeliminated by the flexible liner. The rubber liner acts to absorb the energy of th.- pressure oscillations resulting in a smootherpressure trace.

366

I I I

125

100

S75

* 50 _LP Injection

25 -IgnlterI ;- urns

o1 L, 1 1

0 5 10 15 20 25"Time, ms

Fig. 5. Pressure Uwe taken from the con'bustion chamber during an experiment using a flexible liner. Compare this result

with that in Figure 2.

SUMMARY

A liquid propellant injector/combustor is successfully mimicking the pressure oscillations observed in regenerative liquidpropellant guns. Mw origin of these oscillations ame related to the combustion process and not due to mechanical vibrations ofthe fixture itself. Three different liner configurations have been tested in die combustion chamber in an effort to reduce theoScilla•,onL These are quahir-wave, Helmholtz resonator, and flexible liners. The quarter-wa% e and Helmholtz liners had littleeffect on reducing the amplitude of the oscillations. However, the flexible liner successfully eliminated oscillations in thecombustion chamber. The flexibility of this liner sewves to absorb the energy in the oscillaons through matrial deformation.Currently the flexible liner will not survive repeated frings in a gun and is not suitable for incorporation into a real anillerysystem However, it does give us ressurance that oscillaions can be eliminated during LP combustion.

REIERENCES

1. R. W. Carling, R. E. Rychnovsky, and S. K. Crriffiths, "A Liquid Propellant Injector/Combustor," Proceedings of theJANNAF Propulsion Meeting, GCeveland, OHK May 23-25,1989.

2. R. W. Carling, R. E. Rychnovsky, and S. K. Griffiths, "Combustion of Liquid Propellant in a Righ-PressureInjector/Combusto.," Proceedings of the 26th JANNAF Combustion Meeting, Pasadena, CA, October 23-27, 1989, SandiaNational Laboratories Report, SAND89-"719, December 1989.

367

FLUID DYNAMICCOMBUST1ON INTERACTIlONS AS DRIV M "HANJSM OF PRPSJSURBOSCILATIONS IN A REGEldERAIrVE UQUID PRM "ANT GUN

K. C. Schalow sad E. CirnawkResearch Deparnment

Naval Weapons Centr. China Lake, Calif.Mind

N. S. NosseirAerospace Enginering and Engincerig Mechanic

San Diego State University. Son Digo. Casif

ABSTACT

Fluid dynamic/combustion interactions, which can Iced to the development of coherent flow ftructwPe slid porekicheat release, we suggested as possible driving mechaism of high-frequflc) presews oecillsdons in wegemnratIve liqludpropellant guns (RLPG). The domiunm role of coherent stuctxwas in drivig oecillliuone ho beoa deanoewvaad inranijet dtmp combustors and combustors with bluff-body flameholders. The physical tundestswding of Ow drivingmechanism wu used to passively control the pressure oscillatiom. To focus attention on flu;d dynaic/combustlio.interactions in guns, vortex dynamics and passive conutol wee studied in a simulted RLP0 using low velocity skiflow. Vortex development at the dowutreare end of the cenr bolt was visualizad by the -tnok.wire tc~hiqus.Different designs of the downstream end of the bolt were iasted to break up the large-scale voreices into Ano seea on.bulence. This approach was successfully used in previous wtes to suppreas ramet combustion m'swijtim

High-frequency pressure oscillations have been observed in all calibers of regenerative liquid propellent gun(RLPG). While a number of investigations into the origin and control of thes oscillations havc bmn courducted. adetailed understanding of the driving mechanism is still needed in order to be able to aujweus the oscillatins. Thspaper focuses on one possible driving mechanism related to fluid dynamic/combustion inarascons. which 6ut l"ad tothe development of coherent flow ssturtmes and periodic het farlease.

The idea. that the coherent structures or voraices play a dominant role in driving osillatiois. has lgaied puwir#.support during the past 7 years of research in ramjet combustion instahblities. For th propulsion sysem, tm physl.cal understanding of the driving mechanism was developed and. subsequently used to passively oontrol t"e pre"eeoscilletions in dump- and flitmeholder-stabillred cornt'stor flows, by A jdri.vs esignT$ of the flaMehold.r g,-_,M.f,•y

Although the interior dynamics of ramjet combustors are quite different dta in liquid propellant guns. it is aUS.gested that fluid dynamic/combustion interactions should be given increased considerations as a mnsar d4iving rnmcha.nism in guns. It is the objective of Uis paper, to bring to the attention of the gun community, recant resea.ch inramjet combustion instabilities and passive control methods to suppress them. in addition. a preliminary xperimemtalinvestigation will be discussed to study vortex dynamics in a simulated RLPO geometry. The deSign of the experinw.Wa set up is based on the RLPCI shown in Fig. I. In this concept, the center bolt is stationary, ond th oute pistonmoves rearward to inject a liquid monopropellant as an arnular sheet into the combustion clamber. IA the presntexperiments, the liquid propellant was replaced by a low-velocity airstreamn, nd the vortex develop"met at the downstream end of the center bolt was visualized by a smoke-wIre. Different designs of the downstrea•n e of the bolt weretested to break up the large-scale vortices into fine-scale turbulence. This appro•ch was succassfully used to auppasaramjet combustion instabilities. It is reaizad tha the simulated test conditions do not fepie"sfn the complex intwnalballistics of the RLPC; nevertheless. these initial experiments help to focus aroevtion to fluid dytamlctrcombustloninLeractLions as potential driving mechanism of combustion instabilities in guns.

RAMJET COMBUSTION INSTAB[lJTIES

Recent NWC research in ramjet combustion instabilities will be summarized. The work was part of maONR/NAVAIR Reseach Initiative (1984-1988) which Is reviewed in deotal in Reference 1. The NWC work was rslaeWdto dump combustor, and was recently extended to bluff body stabilized combustor flows.

In dump combustors, vortices we formed in the shea layer between the Idgh and low speed streams a the rearwardfacing step (dump) (Fig. 2). The vortex formation is stabilized in the presence of aoustic pressure oslUations. Ingeneral, the high speed stream cosuists of an aburmt mixture of air mad fuel. while the low speed stream Is composedlargely of hot combustion products forming the flameholdinA recirculation zone behind the dump plan. The vortexstructure has a significant influence on the combustion process. hL the early phase of the vortes development, withthe unburnt mixture on one side of an inteface and the hot combustion products on the other side. intense (flneascal)mixing and burning sie limited. When the vortex roll-up continues followed by interaction between vortices, or is

Approved for Public Release; distribution is unlimited.

169

obstructd by side walls. a large muont of interface between twe air/fuel mixture and the hot products is generated,leading to fire-scale twbulanr enhancennam and sudden heat release. This process is repeaed during each cycle of thepressue oaillatkms regultn in periodic he release. When a proper phase felationship betwee the perodic heatreleae and preasu oscillaions cxims (Rayleigh critieion) high waplitude pressme osnillations we excied.

IGNITER

BOLT(FIXED)

PRqOPELLANT

PP.OPELLAT

COMBUSTIONCHAMBER

" INJECTION

PISTON

Fig. 1. Sc,"•-m.ic of Regenerative Liquid Propcllynt Own (RLPG).

REARWARDFACING STEP•

HOT COMBUSTION PRODUCTS

S, INTERFACE.

UINSURIITMIXTUR& OF

REACTANTS VORTEX bRAID

Fig. 2. Vorter_ Developmrlcnt in Dump Comnbusto.

In order w explore the role of vorte), dynamics in the divi-ig of ramjct com iuotion imtbilitiec, th.e newvn,4 .amt ding of &hear layer dynia.ics wu 'nc sidered. This woik was initiated \/ the discvery of ]h-ge-scale

structures by Brown and Roshko2 and has beI..j repently reviewed by Ho an,, Humre. Based on iaburaoy low-

.3/0

Renokls number studies of teae flows. it was shown that th shear layer develop% intability waves in is initiruon. When the amntiild waves reach a certain argy level they roll up into vortices.

WheI acoustic waves biterac with the show layer, different vortex sizes c. : stabilized depending on thematching between the seoustc frequency ard the shear layer instability frequencu The size of the vortices vaill besmallest whan the acous froqxincy equals the initial vortex shedding frequency; will be largest wha the cousticfrequency is now the prefared.mode frequency, the chmactristic frequency at the end of the potemial caxe.

Experiments in amnlar diffs ;ion flames pq 3vided insight into the effect of the vortex dynsmics on combustionusing Planar Lase Induced fluorescence (PUF) imaging as flow visUAization tecMique.4.5 An instantaneous pictue (7wec) of the acoustically excited flmo is shown in Figure 3. The difference in btack/white tones indicat differeninýnsities of OH-fluoreacence. it may be seen that the flame consisted of large-scale sucwes. sinilar to those-bserved in nonreacting shear flows. The comubution was initiated (highest OH levels) at the circumference of lrge-scale structznes where secondary small-scale eddies we growing on the circumference of the large-scale vortexinitiating the process of transition to turbulent flew. The h-aids connecting adjacent vortices Wad low OH levels as aresult of local qurtcing due to high smraining ras in these regions (Fig. 4). From these experiments it is clear thatthe comb istion is associated with the flow s•nictures generated by oustic forcing of the sther layer. Due to the fluiddynamic/combustion interaction, the heat release was periowc and p.kets of high-temperature flow were convecteddownstream from the burne exit.

7o 1.3

4.-

PLIFIMAINGOF OH

' - 0. 44 I 0

, I FU

VITUE -A1.0.1

i 10 20 3 40 50DISTANCE|, mu

Fig. 3. Vor'ex Corriution in Annular Diffusion Flame.

Th. PUF %:,'jalizuion tas were ilso performed in 2e NWC coaxial dump combuslor. They confirmed thi evenat rexis:iC Lomi.stor conlinions, whe higih amplitude mrbuiiom oscillati-ni were usociated with p•iodic flow struc-tures grxcirlad tnrrug intersct~oij bctwen flrlw inctubilities and ;harnmr ewoustics. 5

Viese results iHg•esl iat a c .-)l[ ,slor flow which i4 do•ainted by vortex flow in the flameh,,ding region is&,,aoieted with pe-odic hert re]eivk . •f the hf-s! rtelase is i. phase with the pressure oscillqtions, driving occurs asstated by the Rayle,gh -ile, ion. To dewrnine the coonditions at which the periodic heat release is a driving fotce, it isneesssry uw hay', a conyloeI intowlaige of t)e spatial and iervn al distributions of heat id-sse mid acoustic Vesure.

bked o0 Lte deibcd dfiving mechat&sm, passive shear-flow control methods we.re developed to lier the€ornbustion p oeest, meiC•e goA!, to av.;i, Pirodic hea! release independent of v-ryv"g air and f- t mass flows.

For the N'WC N(,amp cor~bust, rkl..'irand e'ctiug sh.az flow dynarics of dif'-rent dump ,;:s•slv Weminvestigated. R..ttUl'S of a multapl - sMe, gy)rnelry ind Ltpered ;orrugafions wec ditcusAd in the fo'l,"oing.

3•71

AVERAGE

PUFIMAGINGOF CH

20 40 6 i

Fig. 4. Petiodic Hemt iee ne Aoousdcally Excited Flame.

A wulti-tep dIunp having several backward facing steps (Fig. 5) endamma fine -scal turbulene~ mid prevents large-scale structure development as shown in nortreacting mid reacting experiments. Therefore periodic heat releaseassocisatd with large-scale structures is avoided. In Wh KWC multi-step duMp combustor, the pressure oscillations werered"ce by a factor up to six, rolative to the standard snudden dump.6

78.2 - 4.3 cm

1.936.3 ý1 -5.012. 01= 7.62

H2102 . PILOT IGNITER

0.9 mAI ih (kqgl) -1.8

0'036 ' rnUEL, TOTAL. (kg/) 0.080

0.35, +1 1.3275 - Pc (kPa) -480 (70 psis)

Fig. S. Multi-Step Dumnp Comnbustor.

A dump with tapered corrugations enthances fine-scalt mining ilvouol developrneni of axial vortices (Fig. 6). Thisdwrip design is based on the tapered-elliptic nozzle. for which the effect of axial vortices on nonzeacting av4 reactingshear flows was studied in detail.7 With the new dump geomnety, icvssure oscillations were sigieificsfltly reducedrelativ, to the stanard sudden dwanp.

.17

Fig. 6. Dump With Tapered Corrugations.

Recently, passive control of JCAsNe oscillations was extended to blvffbody stabilized combustor flow~s.Plaineholders with tapered cotrugottior- extended the operational iegine wit smooth combustion reatve to standaddisk flameboldcr (Figs. 7& mid b).

The roomn success in muppreaamg ramjct combustion oas Hations i, hasa oat Ebe physical wwkastaiing of tINCdrivi4l vieckwaniaan, whichb is relaWtodh fluid dyiauvic/coambustion Wentormios. To lloc~a useazte m or. their pitsiAbeTole in driving pressure oxiillazk'as in *'=n, vorleu dynamics and psasive control wmr studid in a binw!MaW XiAFgeumietry end ane described in the follwing.

VOR=h DYNAMIbO IN A SDAUAI.TJ) R LW3

An experimentsl investigation was initiaitd to stuody the znehailan responsible for thm &mnu'&i of bhki-frcqiuaacy pwacurc oscillations in the combiaruon chnberb of a rsgeneratlve liqui propellangt. Tb. r arge-t-rikvortices, which we formed b/' -,he flow of dwe hq'M peieflim ever the downstream and of a *wwtionAl fixi4 OAIoiwre a highly proIteble source for driving the preasufe oscillatkou. Thies vortrius md COVzt wiawnrkq wii t6ebackgfr~tnd turbulent flow domhw lut e mixing and the coenbustic procearts ainsidiv ;be mabiu.asior ciawt.La. Theperiodic shet~ing of these rorden from &he piam'is end can cav a periodc rtiese of keat waeg'. Tbi in cumproduces oscilladions of both the temperamue savi jxait se irAsde &ti co*2IustiA c~xmuin.

f'lie genetic flowfilck :ov the ';qvid Vpetlenz ;nn -*. at of ~in .&rm- je.1 with w: vtisymaTiervi cmun-body (Fig.8). The- objctivr~s of *ht ;grult Verimevi~ woe Ut WO'ai d- Fthr.S.#% -X )arge-scale -wohrew tw'atirps ii st~ltflowricld. midl to explur4. s'ae~rods tr, býcW ap (neir vi,,:herc!.ee

The osea~s.I'tV-0ity ur*e4 ;I thowr in' P.. S. Air ericxi a conii.mI wi#ttJin% avt:n~I~w wO -uiow Cnro'agbpm~'ar["& Platt. 7he jlaift riUPVritg (,%'L. :~ tam The a.r fkwow dwwr aw kr &n zmoti nozzle 16ith aChkV*~'A COM,!r~~ bI 3Y. 'M,: k Clk' ofiJ.;A.- ( PIS. t; *0, iN'e of a TaOu-bodi 'hIavifl r, circitv arow'i-utfetka

In oida io masli -p the winwviu. rdie-tem-e me swpui~ces sn*-d -% tMI Wake ulgon of .ýW 'AMNt ioody, 01114 !%MWaholty qe(4lnaie', %.G!O tMaol. iomytt 'uajv, (VTA. VAi) iwered am~gsomn tTg ~'Sq . ar%ý Csu'-si body With mult6iSl

f ermnKC`@tiai) pdi (Fit 964) 1*110 0.".e ltý.rlv usal to b'--4 P~ew~ cAI.ar~c. L7 stjwonroic flowl ovu & slot to

371

2.0FLAPPIBLI!T AND STABILIJY LIMITS5

1.S

1.40

1.0

0.

.1WI

30

1741

IFLNWSILIJ' AM SID8IIUTY LNIIS

1.7

1.0

I I.a1.4 0 6 . . 0 o 90 #. s so101j 6 G

1.3 CLC1 it9

FI.7b prainl einsWthSmmhm ouhC mbsjotFr luotodr ii se.2Cr~aim

7Ucnerbd cnb mvdinteou m ie ietint hmete itonex bten h o. illad9clowsuein ace of th et;bd.s1 e.e9c irn~ww e ,r M I. n 1 ,3 A 1opItess et prfrmd it x.1d .. S n3wfs lcdd og % anxbd n o nwsno t vsue

viislio~a1- fo nl . 5 Stl Sigais& ie o4 rten wwdw st m v oetriimn woth lo ,(suti:focngb teIo q alv wt ue u tail -.v frmo .k oftel- ~ c t odm Ai

Ii~lv. iyckuito wihte-e s rvn i a & L opaeL- h ~t~ uigt, lvltin

Pig. 7bit, Opemlratisnl peions. wivi Smoilc bolt upd no the pamssedtion oue twoarowsdof W ith T hpe i.; -F AMiinThea- cetnter bo)(vrd" can be oved~ insth sinunier dretiona u to hnfeth vortices wa* beaee 4t"he )4,mzle ealtad h

(Pitu 12). The fowtle-d Stisll9 pho th ratios ubW wide mo:1, hc srath mlwsl Viw nrast io. Anc lKo4 ofe turblt raiwe, UP7 1 mlw ssn-oustVc frovkin bNvhet Iorsia e sca tublras usd . tallza thefrmatdion taif t ishe w s5gh~. otuies~e .Ii

AIR

SETUNO -9A NOZLCHAMBER - NZL

~A d2

dl

A-A

PERFORATED PLATESUPPORTING

FOUR SPEAKERS

Pit V. Expetrmental Set-Up For Simsulstd kLP).

EE.

is) CIRCULAR CUDlER BODY ISASU*49ff (b) STIPPID C&NSRA BODY

(a) CiNHIER NODY WIM# TAWIAW CQRRUOA1I*Nl (4d cimlahe Isoiy W1114 0#11i

Pit 10, mPwrowfl of Citcwkla C~M~ Rud0Y.

pis I I rh..flow uf (C,4. ROW Wilk Tow.it (wniugetM's

Fig. 12. Center Body With Multiple Steps.

SUMMARY

Fluid dynamic/combustion interactions can lead to periodic heat release and driving of pressure oscillations asdemonstrated in ramjet combustor tests. In that propulsion system, the physical understanding of the drivingmechanism was used to passively control the pressure oscillations. In simulated nonreacting RLPG tests, developmentof coherent vortices was observed. Several design configurations of the bolt were used to break up the vortices intofine-scale turbulence. Similar techniques were successfully used to suppress oscillations in the ramjet combustor.Although the specific gun interior ballistics were not addressed in this paper, it is suggested that increased attentionshould be given to fluid dynamic/combustion interactions as a driving mechanism of gun pressure oscillations.

REFERENCES

I. K. C. Schadow and E. Gutmark. "Large-Scale Structures and Passive Control." in Pressure Oscillations in Ramjets.CPIA Publication No. 535. pp. 33-57, October 1988.

2. G. L. Brown and A. Roshko. "On Density Effects and Larger Structure in Turbulent Mixing Layers," Journal ofFluid Mechanics, Vol. 64, Pt. 4, pp. 775-816, 1974.

3. C. M. Ho and P. Huerre. "Perturbed Free Shear Layers." Annual Review of Fluid Mechanics, Vol. 16, pp. 365-424,1984.

4. E. Gutmark, T. P. Parr, D. M. Parr, and K. C. Schadow. "Planar Imaging oa Vortex Dynamics in Flames," Journalof Heat Transfer, Vol. III. No. 1, pp. 148-155. February 1989.

5. K. C. Schadow, E. Gutmark, T. P. Parr, D. M. Parr. K. J. Wilson, and J. E. Crump. "Large-Scale CoherentStructures as Drivers of Combustion Instability," Combustion Science & Technology, Vol. 64, No. 4-6, pp. 167-186,1989

6. K. C. Schadow, E. Gutmark, K. J. Wilson, and R. A. Smith. "Multi-Step Dump Combustor Design to ReduceCombustion Instabilities," Journal of Propulsion and Power, Vol. 6, No. 4, pp. 407-411, July-August 1990.

7. E. Gutmark, K. C. Schadow. T. P. Parr, D. M. Parr, and K. J. Wilson. "Combustion Enhancement by AxialVortices," Journal of Propulsion and Power, Vol. 5, No. 5. pp. 555-560, Sept.-Oct. 1989.

8. E. Gutmark, K. C. Schadow, M. N. R. Nina, and G. P. A. Pita. "Suppression of "Buzz" Instability by GeometricalDesign of the Flarneholder," 26th AIAA Joint Propulsion Conference, Orlando, Florida, 16-18 July 1990. (AIAA PaperNo. 90-1966)

9. R. A. Smith, E. Gutmark, and K. C. Schadow. "Mitigation of Pressure Oscillations Induced By Supersonic FlowOver Slender Cavities," to be presented at the AIAA 13th Aeroacoustics Conference, 22-24 October 1990.

3est Available Copy37:N

DIAGNOSTICS OF IGNITION/COMBUSTJCN IN ABULK-LOADED LP GUN

Robert L. TalleyVeritay Technology, Inc.East Amherst, New York

ABSTRACT

Diagnostic, bulk-toaded gun firing tests were conducted using pyrotechnic igniters and the liquidmonopropetlant Otto Fuel II, or water as a situLant, to examine the phenomenological evolution of theinterior ballistic ignition and combustion processes. Particular attention was given to the early ignitioncombustion coupling behavior and to combustion instabilities. These activities were part of an experimentaldiagnostics effort to explore the potential and advance innovative solutions to achieve welt-behavedcoemustion performance in a bulk-Loaded, Liquid or gelled propellant, cannon-caliber gun system.

A windowed 25-rn test gun and transparent chamtber test fixture were employed together with high-speed photographic techniques to obtain flow and visualization data on selected aspects of these combustionphenomena.

Front ignition was eaployed in an overall approach to reduce the hydrod)namic coupling effects onthe liquid-propetlant coamustion. This enabled the evolution of combustion features to be examined underless chaotic conditions. The features of selected system compon*nts that influence cotbustion behavior wereexplored via tests in which various component characteristics and geometries were changed from test-to-test.

The diagnostic firing tests showed that quite stable cotbustion of Otto Fuel i was achieved in abulk-loaded chamber with an appropriately chosen geometry involving the relative diameters of the chanterand the igniter orifice. These parameters, in turn, correlate with the nature of the igniter-generated,gas-cloud development that was observed in water.

INTRODUCTION

It remains a challenge to understand the evolution of the interior ballistic ignition and cotbustionprocesses in butk-loaded Liquid propellant guns. The need for such understanding is still important, notonly for potential applications directly to bulk-loaded guns in the cannon-caliber and smaller scales, tobulk-type ignition systems for large guns, and to model development, but also for potential applications toetectrothermal-chemicat gun phenomena.

Diagnostic testing provides an effective means to characterize Liquid propellant gun phenomenotogy,including the features associated with normal gun performance, as well as those with occasional erraticbehavior. Techniques of particular interest here were those for direct visualization of the spatial andtemporal behavior of ignition, cotbustion, fluid flow, projectile motion, and their mutual couplings undertypical liquid propellant gun firing conditions.

The diagnostic studies reported here were conducted as part of a Phase 11 SBIR program sponsored bythe U.S. Army Armament Research, Development, and Engineering Center, and included two main topics:ignition-contustion coupling and comtustion instability (references 1, 2, 3).

The ignition-conbustion coupling portion of these studies emphasized diagnostic testing to bettercharacterize the phenomenotogical evolution of the interior ballistic comnustion processes in bulk-Loaded,Liquid propellant guns, particularly in the cannon-caliber size range. The emphiasis in the instabilityportion was partly on the overall comtustion evolution within the gun chambter, partly on identifying theLocation, time, severity and cause of ignition or combustion pulses within the chanter, and mainly onexploring means of minimizing these effects within the ballistic cycle.

A principal interest in both studies was to determine how the pressure-time behavior in a liquidpropellant gun chanmer correlated with other underlying processes revealed by the use of diagnostictechniques.

Past studies and tests have tended to address the ignition-combustion coupling, and to some extentthe resulting instabilities, from the standpoint of ignition. This viewpoint assumes that if the combustioncycle is started incorrectly, i.e., outside some ill-defined Limits, the remainder of the cycle will tend tobehave marginally--or more Likely--poorly, especially in the sense of repeatable combustion behavior. Onthe basis of rather extensive testing of bulk-Loaded Liquid propellant test guns at Veritay, we believe thatother coupling and contmustion development features are also important.

This work was performed under Contract DAAA21-88-C-0143 with the U.S. Army Armament Research, Developmentand Engineering Center at Picatinny Arsenal, New Jersey.


est/ Av&aibe C op) 379

In reproducibility testing, especially, pressure-time results in a liquid propellant gun chamberoften appear nearly identical at early times for individual tests, yet show quite different results at latertimes in the ballistic cycle. In such cases, it seems that ignition has already occurred at an early timein the cycle, but the coupling and combustion development which follow are somehow different.

The focus here was to explore conditions and techniques with the goat of understanding andovercoming such types of performances.

APPROACH

In this experimental diagnostics effort, it was expected that the greatest gains could be made via asystematic approach in which direct individual and serial changes were made in selected system components,characteristics, geometries, etc. Specific items explored included:

front ignition, to reduce randomness in Liquid propellant combustion rates associated withviolent fluid motions in the chamber;

chamber diameters and lengths;

chamber with cylindrical stages of increasing diameters to achieve more controlledprogressivity of liquid propellant burnings; and

single charber stage in the form of a truncated cone rather than separate cylindricalstages to achieve geometrically controlled progressivity of Liquid propellant burning.

EXPERIMENTAL FEATURES

To implement this approach, diagnostic tests were conducted using the liquid monoprztoetlH t OttoFuel II in the 25-mm singLe-shot firing fixture shown schematically in figure 1. Additional gas clouddevelopment tests were run using water in another 25-mm transparent chamber fixture, shown in figure 2,equipped with replaceable optically transparent plastic chambers. The pyrotechnic igniters used in thesefront ignition tests in both fixtures employed an Atlas M102 Electric MatchR primer and a 3 1 7mg boostercharge of Hercules Canister Powder UniqueR. Both were mounted in the base o: each 25-mm projectile as shownin figure 3.

Sapphire Crystal 90- WindowsBooster WidoBotr Wno Window Pressure

25mmBarrel

Breech Pressure Transducer Fill/DumpPort Port Port

Figure 1. 25-mm Single-Shot Firing Fixture


A range of cylindrical chanbers with different lengths and diawe~ers were tested in both fixtures.These included simple cylindrical shapes as well as composite chambers consisting of stepped cylinders oftwo and three different dimeters. These various combustion chamber geometries were implemented in the25-rm singte-shot firing fixture by using simtple, one-piece, plastic inserts in the basic chamber. Theactual chamber used for testing in such cases was the cavity witnin each plastic insert. With thisarrangement, different chamber lengths and diameters could readily he used by introducing appropriately

The firing test fixture was instrumented routinely with PCO piezoelectric pressure transducers andassociated charge amplifiers. Projec~tile velocities were determined from time-of-flight measurements of asteel projectile stug passing through magnetic coil sensors located a known distance apart. The timeinterval measures were made using a universal timer.

Tie Bolt

Projectile

B

rrel

25mm Breech Plastic Chamber

Mounting Blocks

Figure 2. 25-sn Transparent Chamber Fixture

Electric

-- oe t "t g Powder

': • lEpoxy

Figure .. Projectile With front Igniter

High-speed motion pictures of gas cloud development in water in the 25-ms transparent fixture weretaken using a HYCAN, 16-mr rotating-prism framing camera with a Macro zoom Lens. Backlighting for thesehigh-speed framing photos was provided by a combination of two electronic flash units triggered sequentialLyto provide an intense Light pulse over a duration of about 3.0 milliseconds.

A simple, inexpensive, image processing technique was used for converting the qualitative data fromphotographs to a quantitative form. This technique was applied to sequential frames of the high speedmotion pictures which recorded the gas cloud development sequences in both of the preceding types of tests.

The diagnostic tests, themselves, centered on assessing the ignition-combrustion coupling; first, byconducting firing tests with Otto Fuel 11 to learn if stable combustion could be obtained within a matrix ofchamber Lengths and diameters (possibly with contributions from other instability reduction techniques notedearlier); and second, by characterizing the nature and evolution of the igniter-generated gas cloud injectedinto water fitted chambers that corresponded to those in the Liquid propellant firing tests which resultedin both stable and unstable combustion. The goal was to identify parameters in the inert tests whichcorrelated with the actual firing test results.

EXPERIMENTAL RESULTS

EFFECTS OF CHAMBER CONFIGURATION

The range of chamber diameters and lengths explored in front ignition firing tests with Otto Fuel 11

is shown in figure 4. The individual points indicate the loading conditions for each test.

I * *,,- . .

LP: oTTO II

FRONT IGNITION:,,, IGNITER: ELEC. MATCH

4,G.OSTER; UNIQUER 320MG

D =15.9MM CHAMBER DIA.: 19.8MM

0 UrORIF1 i~ DIA.: I.32IAM

°0D 27.9MMILI

S 1 A I *

CHARGE/MASS CHAMBER L/D

Figure 4. ExperinentaL Loading Consistency Figure 5. Variation of Chamber Pressure Within Front Ignition Tests Configuration

Figure 5 shows the variation of peak chamberpressure obtained in a 19.8mm (0.780 inch) diameter chamber as the chamber Lentgth was varied from 38.1mm(1.50 inch) to 88.9mm (3.50 inch). The equation of the straight line, whic& -p, Ir•ents a ieast squares fitthrough these peak chamber pressures, is, in MPa,

P eak" 12.92 (1±0. 11) (-L1+ 351.8 (1±0. 014)P D

38es2 , pf\aao\e CO°V

382

in this expression the standard deviations of the slope and intercept coefficients are expressed as a

fraction of the respective mean values outside each parenthesis. The single observation, unbiased standard

deviation of the observations from the curve is 2.94 MPa, which together with the mean peak pressure of 378

MPa at L/D = 2.0 gives a relative peak pressure variation of + 0.8 percent as an approximation

representative of these data.

The interesting feature is the consistency and stability of peak pressures over the range of Length-

to-diameter ratio (LID) from 1.92 to 4.48. The complete pressure-time traces were nearly identical for

these tests. Because of equipment difficulties, muzzle velocities were not obtained for this test series.

The observed peak pressure and muzzle velocity results obtained in 23.9mm (0.940 inch) diameterchamber, as the length was changed from 38.1mm (1.50 inch) to 88.9m, (3.50 inch), are shown in figure 6.The straight line fits to these data for peak pressure, Ppeak, and muzzle velocity, V, respectively, are:

Peak- 34.8(1±0.50) (-)+434.2(1+0.iI) (in MPa)P D

L

V- 47.0(1 ±+0.08) (-L)+765.1(1 00.015) (in mis).D

The peak pressure variation (as above) for these d~ta is + 8.6 pecent, and thc velocity variation is + 0.5percent, each relativ. -, ,car, o nd veLocity values, respectively, at L/D = 2.0. These peakpressures show a significant increase with increasing chamber L/D. The consistency and stability of peakpressure values over the range of chamber L/D ratios examined is poorer than observed for the smallerchamber diameter in figure 5. in this case the projectile velocities increase slowly with chamber LID, andvary only a small amount from the Line fit through the set of velocity values.

VELOCITY ~ -VELOCITY I

Ur~ PRESSURE . PRlESSURE

LP: OTTO IIFRONT IGNITION: LP: OTTO IIIGNITER: ELEC. MATCH FRONT IGNITION:

BOOSTER: UNIQUER 320MG IGNITER: ELEC. MATCHCHAMBER DIA.: 23.9MM CHMSTER: UNIQUER 320MGORIFICE DIA.: 1.32MM CHAMBER DIA.: 27.9MM

ORIFICE DIA.: 1.32MMC -- . _ * L. i . ... S I tS 1 ___!~ .L........ L. U...

CHAMBER L/D CHAMBER L/D

Figure 6. Variation of Chamber Pressure and Muzzle Figur,: V".riation o .'' Z;aber Pressure and MuzzleVelocity With Configuration Velocity with Chamber Configuration

Figure 7 shows the results for corresponding test series using a 27.9mm (1.10 inch) diameter chamberwith chamber lengths which varied over the range from 31.8mm (1.25 inch) to 50.Smm (2.00 inch). Thecorresponding straight line fits to these data are:

3etAva'abýl CopyS• " 383

LPpeak- 3 4 9 . 0 (1±0.20) (--T)+ 88.8 (inMPa)

L

V- 185(1±0.18) (-L)+ 629(1±0.09) (inm/s).D

The peak pressure and velocity variations relative to associated mean values (as above) at L/D = 2.0, are,respectively, ± 7.6 percent and + 2.2 percent.

Here the peak pressures show a large change with chamber L/D. The consistency and stability of thepeak pressures over the L/D range is poor. Considerable variation was observed in the shapes of thepressure-time curves (not shown). The projectile velocity also appears to show a significant increase, withincrease in chamber L/D. The variation in relative velocity values is less than for peak pressure, asexpected.

Tests involving a chamber with two cylindrical stages were conducted to determine the length of asmall diameter first stage (with stable combustion), required to smoothly ignite a second stage. This wasto achieve stable burning, yet achieve more rapid liquid propellant combustion in the second stage, becauseof its larger diameter. Further, this approach was expected to increase the duration of high pressureproduced in the composite chamber. The goat of this approach was to obtain geometrically controlledprogressive burning of the liquid propellant in a bulk-loaded test gun. With the use of additional stages,an extended duration of high pressure might be produced in the chamber and used to accelerate projectiles toa high velocity.

Test results (not shown) indicate that either a first stage L/D of about 1.6 for a first tubediameter of 15.9mm, or an L/O of 2.G for a first tube diameter of 19.8mn, were sufficiently small toinitiate progressive burning in a second stage (23.9mm diameter by 25.4mm tong) chamber, and then in a thirdstage (27.9mrn diameter) chamber. These same L/D values were also sufficiently large to cause the secondstage burning to occur just after the peak pressure in the composite chamber, and thereby to extend theduration of the high pressure without increasing the value of the peak pressure.

While this scheme operated as expected, the amount by which the duration of the high pressure in thecomposite chamber could be extended did not appear to be sufficient to be very useful for achieving highprojectile velocities in the front ignition case. Further exploration of this approach may be desirable inthe rear ignition case.

EFFECTS OF IGNITER ORIFICE DIAMETER

Front ignition tests were conducted in the 25mm, single-shot firing fixture to examine under firingconditions whether combustion would remain stable when the igniter orifice size was scaled from 1.32mn for achamber diameter of 19.8mm (which gave stable and consistent combustion), to an igniter size of 2.37rmm for a27.9min diameter chamber.

These test results indicated peak chamber pressures of 582 (1 + 0.0075) MPa for the larger orificeand chamber, indicating that in this case, at least, the scaling was quite effective. In contrast, thesmaller 1.32mm orifice performs poorly in the 27.8mm diameter chamber as shown in figure 7.

GAS CLOUD DEVELOPMENT TESTS

A series of gas cloud development tests was conducted in the 25mm transparent fixture over a rangeof chamber Lengths and diameters corresponding to several of those used in the liquid propellant tests.Here, water was used as an liquid propellant simulant, and the igniters were identical to those used in theLiquid propellant firing tests.

A sequence of pictures shown in figure 8 are representative of how a gas cloud typically develops ina water-filled chamber when hot gas issues from a pyrotechnic igniter. This particular sequence wasassembled using the image processing technique noted earlier. The sequence was originally captured on high-speed photographs, taken with the 16rrm HYCAM camera in the framing mode, using backlighting. The pictureswere then transferred to video tape via a video camera, and then to a frame-grabber for image processing ina personal computer. The individual pictures in the sequence correspond to every other frame on theoriginal film strip and are about 135 microseconds apart.

The gas cloud outlines in water, quantified by a technique using the frame grabber and a light pen,are shown in figures 9, 10, and 11 for three different chamber diameters of 19.8nm, 23.9mn, and 27.9nm,respectively. In these figures the gas cloud outlines at successive times are aligned about a common

SBest Available Cop.

(a) Frame 3

(b) Frame 5

.-

(c) Frame 7

f ,jure 8. Gas Ctoud Oevelopnnt in watei for ý3,9- mrn 0iamtei, Cnaatr-.

0.5 ..... * .... ..... .... . ......... Ii . ...... 4.......I. ...... .... -4

0.4 ....

0.2 ....

............0.1 ...... .. ... ... . ... ...

-0.0 .... 4 .... i ......

-02

0.0 0.2 0.4 0.8 0.8 1.0 1.2 1.4 1.6 1.8 2.0Axial Distance (in)

Figure 10. Gas Cloud Profiles In Water for 19.8-wn Disivter Chaaiber

0 .8 .. . .. . .. ....... .... . . ... .. . . .. . . .

0 4 i ...... .

.............. .. ... ... .. ..

-0 ......... . ..

-0.60.0 0.2 0.4 0.6 08 1.0 1.2 1.4 1.6 1.8 2.0

Axial Distance (in)

Figure 10. Gos Cloud Profiles in Water for 23.9-Mm Diameter Chamber

0 . ..... ..... .. ....... ...... ..... ....,. . ..

chamber axis and superimposed to more clearly show the cloud behavior.

In figure 9, the gas cloud just did expend to the chamber watt and develop along it.

In figure 10 and 11, the gas clouds did not initially expand to the full diameter of the chamber,and they did not do so as the clouds expanded length-wise along the chamber axis until after they reachedthe end of the chamber. There was a variabte amount of clear liquid between the rough edge of each gascloud and the chamber watt over the tength of the cloud at any particular early time. The clear liquidlayer was greater and more variable for the larger diameter chaember in figure 11, than for the smallerchamiber shown in figure 10.

The gas cloud development tests shown have been repeated several times, and although the cloudshapes may very slightly from one test to another, the general cloud expansion behavior relative to thechamber walt remains consistent.

Comparing these gas cloud development test results in water with the results obtained during liquidpropellant test firings, indicates a strong correlation between consistent peak pressures in the testfirings, and the expansion of the gas cloud to the chambter walt. This preliminary correlation is shown infigure 12, for both peak pressure and muzzie velocity.

Tests for which the gas cloud developed in water, bat did not quite reach the chamber watt,correspond to firing tests that show considerable peak pressure variations -- even though the correspondingmuzzle velocity variatiois were snall. Finally, tests for which the gas cloud developed in water and left alarge am•unt of clear liquid adjacent to the wall, correspond to firing tests that show a large variation inboth peak pressure and muzzle velocity.

10

0 D = 23.9 MM

D PEAK PRESSURE

ZN0 6

-J

rZ >

LP: OTTO II2 FRONT IGNITION: VELOCITY %

IGNITER: ELEC. MATCH VLCTBOOSTER: UNIQUER 320 MG \D = 23.9 MM%

ORIFICE DIAMETER: 1.32 MM .

0_.• 0 1 - I I --- - -

0.5 .6 .7 .8 .9 1.0

CLOUp DIAMETER-CHAMBER DIAMETER, 0

Figure 12. Correlation of Firing Test Results with Gas Cloud Develomet in Water

a187

If this preliminary gas cloud development correlation can be shown to hold for liquid propellants orgelled propellants other than Otto Fuel II, it may provide a relatively simple experimental means forestimating at least one set of conditions which may allow stable combustion to be achieved in a buLk-LoadedLiquid propellant gun. The confirmattion of such estimates would require, of course, the conduct of actualfiring tests.

CONCLUSIONS

This experimental diagnostics study of ignition-combustion coupling behavior and of certain featuresof cotbustion instability has shown (in the approximate scale of a 25-nmi, liquid-propellant gun) that eachof these features are influenced by the igniter-chanrser geometry as well as by the dynamics of thecobtustion phenomena itself. It was found by usingy the tow-impetus, Liquid monopropeltant Otto Fuel II withpyrotechnic igniters that pressure-time characteristics can be generated over a range of acceptable tounacceptable interior ballistics behaviors.

while ignition is known to be important in property starting the interior-ballistic combustionprocess, the nature of how the propellant itself burns after ignition is still a most essential factor inachieving a desirable pressure-time evolution inside a bulk-loaded tiquid-propettant gun.

The study revealed that combustion stability of a liquid propellant in a chanter dependssignificantly on the relative diameters of the igniter o'ifice and the propellant containing chamter. For agiven igniter orifice size (and a booster mass sufficient to cause the liquid propellant to ignite), thepropellant chamber must be smatter than a retaucJ dihmeter to achieve stable combustion.

Over the range of chanter diameters investigated, it was possible to achieve stable coi•bustion in alarger chamber by scaling the orifice diameter to a larger value. The exact scaling law is not currentlyknown.

Progressivity in the burning of Otto Fuel 1I with front ignition was achieved geometrically by usingstepped chambers coaprised of cylindrical stages with increasing diameters. In such chamters with only twoor three cylindrical stages, the stable combustion condition (and size) of the first stage tended todetermine the peak pressure achieved; subsequent stages when property sequenced tended to maintain the highpressure over a longer duration.

REFERENCES

1. Talley, R.L., and Owczarczak, J., IlIgnition-Comaustion Coupling in Gelled Propellants," DraftContractor Report (Veritay Technology No. A10-01-90), ARDEC, Picatinny Arsenal, NJ, Jne 1990.

2. Talley, R.L., "Diagnostics of Ignition-Combustion Coupling in a Liquid-Propellant Gun," DraftContractor Report (Veritay Technology Mo. A10-02-90), ARDEC, Picatinny Arsenal, NJ, August 1990.

3. Talley, R.L., "Liquid Propellant Gun Diagnostics," Draft Contractor Report (Veritay Technology No.A10-03-90), ARDEC, Picatinny Arsenal, NJ, August 1990.

THE EFFECT OF ACOUSTIC DAMPENING DEVICESIN REGENERATIVE LIQUID PROPELLANT GUNS

N.E. Bayer, J. DeSpirito, J.D. Knepton, and G.P. ReevesU.S. Army Ballistic Research Laboratory

Aberdeen Proving Ground, Maryland 21005-5066

ABSTRACT

The effects of acoustic dampening devices on the pressure oscillations observed in a 30-mm regenerative

liquid propellant gun wmre analysed. Acoustic quarter-wave cavities and radial baffles were installed in

the forward section of the combustion chamber and attached to the face of the injection piston. The flow

exit geometry and the liquid propellant injection sheet thickness were also varied to determine their effect

on the pressure oscillations. The use of the forward chamber cavities reduced the average RMS pressure 16

to 39 percent in the plane closest to the cavities. The spectral density in the 23 to 27-kdz frequency

range was reduced 42 to 97 percent. The dominant frequency was shifted to the 30 to 40-kH2 range when the

forward cavities were used. The use of baffles reduced the RMS pressure 37 to 44 percent and the

peak-to-peak pressure 31 to 41 percent. No significant affects could be determined from changes in the

sheet thickness or the flow exit geometry due to the amount of scatter in the test data. The RKS And

peak-to-peak pressure and the spectral density level were generally higher closer to injector orifice.

INTRODUCTION

High frequency pressure oscillations have been observed in regenerative liquid propellant gun (RLPG)

tests. These pressure oscillations have been observed in calibers ranging from 25-mm through 155-mm17

Although no problems have been attributed to the existence of these waves, serious questions about their

effect on components in a fielded system have bean raised. Potential problems include increased gun tube

erosion due to increased heat transfer and detrimental effects on serisitive projectiles, fuzes, and payloads

due to mechanical vibrations.

5-8

Watson et el. pceviously reported Ln the nature of the pressure oscillations in a 30-mm Concept VI

RLPG. They concluded that there was an acoustic nature to the pressure oscillations that were observed in

the 30-mm RLPG fixture. The primary acoustic mode was believed to be the first radial-first tangential

combined mode, occurring at approximately 24 kHz. This conclusion was based on the eigenvaluos calculated

for an annular and circular combustion chamuber9

and an assumed speed of sound of 1000 m/s.'71 0

This paper describes the results of tests using acoustic dampening devices and changei in flow orifice

configurations to examsine the nature of the pressure oscillations in a 30-mm Concept VI RL.G. A series of

eleven tests were perfoimed with various combinations of baffles and acoustic cavities in the combustion

chamiber of the gun fixture. The effects of changes in the liquid propellant (LP) injection sheet thickness

and the flow exit geometry were also investigated.

EXPERIMENTAL

PROPEI.LANT

Two propellants now being used in RLPG testing are designated Liquid Gun Propellant (LGP) 1845 and LGP

1846. These propellants are mixtures of hydroxylarioniun nitrate (HAN), triethanol amnonium nitrate (TEtA)

and water. Table I presents the composition and thermochemical properties of these two propellants.

LGF 1845 was used in the tests to be described.

Approved for public release; distribution is unlimited

3Y)

Table I. Composition and Thermochemical Properties of LGP 1845 and LGP 1846.

Propellant Composition (wt. 1) Flame Temp. Impetus Co-volumo Frozen Garma

RAN TEAM B20 (K) (JI/) (cm31g) (--)

LGP 1645 63.2 20.0 16.8 2592 934.2 0.707 1.2178LGP 1846 60.8 19.2 20.0 2469 898.3 0.677 1.2225

The Concept VI configuration 1'

6 is an older annular sheet injection concept than the one now being

tested in large caliber.4 The new configuration, labeled Concept VIC, is similar, however, it ccnsists of

two moving pistons with hydraulic dampening on the inner, control piston. In the Concept VI configuration

tested at the Ballistic Research Labo~atory (BRL), the propellant is also injected into the combustion

chamber in the form of an annular sheet. Figure I illustrates the basic hardware configuration. In the

upper half of the figure, the injection piston is in the forward, before firing position. The lower half of

the figure shows the injection piston position after firing. The injection piston is a thin shell cylinder

supported from deformation by a lubricating film and the chamber wall. Initially, the reservoir is sealed

to prevent leakage of LP into the combustion chamber and to allow prepressurization of the LP reservoir. At

ignition, the pressure developed from the igniter venting into the combustion chamber forces the injection

piston to the rear. At the onset of piston motion, the chamber-to-reser-voir seal is broken and propellant

is injected into the combustion chamber. Because of the differential area (combustion chamber side/LP

reservoir side) of the injection piston, the pressure in the LP reservoir is higher thai that in the

combustion chamber. As a result, the liquid propellant is injected from the reservoir through an annular

orifice formed by the inner diameter of the piston and the outer diameter of the center control rod. The

injected nropellant than mixes with the igniter gases and burns. Contours machined along the outer diameter

of the center control rod determine the injection area. A dampening taper on the center control rod retards

the motion af the piston toward the end of stroke by increasing the LP reservoir pressure.

Gage Locations (cm)A C J PROJECTILE

Before Firing I _ __

6.27 4.37 "0.36

-INJECTION PISTON • 3.76-• • 1.04 I

LP RESERVOIR >CornbeusonIChtamber /

..- .. ..... C.ENTE R CONThOL .ROQ) ......-------J spacer

Fing

9.96 .--. , I--CHMBSER WALL

After Firing

Figure 1. Concept VI Ragenerative Liquid Propellant Gun.

z rhamber spacer originally was installed to provide a stop for the piston during prepressurization end

prevent hardware damage if a piston reversal occurred during firing. The spacer was a three-bladed

39"A

structure that was fixed with respect to the chamber and located downstream froc the piston, at the forward

end of the chawber. The design of this spacer was changed for the baseline tests of this series in order to

have a relatively open, circular combustion chamber forward of the injection piston. The new design

consisted of a simple ring mounted at the forward face of the combustion chamber. The ring spacer is shown

in detail in Fig. 2 and installed in the chamber in Fig. 1. Three pins were welded to the ring so that it

would perform the functions of the original spacer.

As part of the test series, the spacer ring was replaced with either one containing acoustic

quarter-wave cavities, or a six-blade radial baffle in an attempt to reduce the amplitude of the pressure

oscillations generated during combustion of the LP. Figures 3 and 4 shows these two configurations. In

addition, cavities and baffles were designed into the face of the injection piston. The piston face cavity

design is shown in Fig. S. The piston face baffle design was the same as shown in Fig. 4.

7.6 --

0 Pistonsupport

A-.- -A ~pinsfF0(3)_) ,o

SECTION A-A

Figure 2. Combustion Chamber Spacer Ring

The design of the cavities was based on the observed dominant frequencies of the pressure oscillations

in the RLPG. In previous tests this frequency was usually between 23 and 26 kat arnd was believed to be a

first radial-first tangential (IR-IT) combined mode. 5,7,8.10 In an attempt to reduce the amplitude in this

frequency range, cavities were machined on the outer perimeter of the ring, at the combustion chamber wall.12

This location was chosen because it would be the location of a pressure anti-node of a lR-1T mode. The

depth of the cevities, Figs. 3 and 5, was made 10.4 mm, assuming the speed of sound in the combustion

chamber was 1000 m/s. Cavities were also machined on the outer edge of the piston face using the same

procedure. it was believed that the cavities located on the piston face might be more effective than those

at the forward part of the chamber because they were closer to injection and, possibly, the combustiorn zono.

-7.6• --76-- --1Pistornsupport

r (3) ,o

/ - r)A-T

-- - SEC'IION A A-

riguce 3. Spacer king with Acoustic Cavities.

7.6

_ 0.762

60'6PL

G 64__ 3.0

SECTION A-A

qA--- 3 Al

Figure 4. Sir-blade Radial Baffle.

S... ?-----76 -•

1 05

0.48

04 SECTION A-A

Figure 5. Injection Piston Face with Acoustic Cavities.

Radial baffles are usually effective in dampening pure and combined tangential modes. Therefore, it

was believed that they would be effective in the RLR. Six-blade radial baffles, Fig 4, were tested in the

forward section of the chamber and also attached to the face of the piston. Although an odd number of

blades is most effective,12 design Gonsiderations led to the use of six blades. A six-blade baffle could

allow the existence of a standing third tangential mode12, however this mode (17-17.5 kHz) was not observed.

As shown in Fig. 4, the blades of the baffles were tapered toward the injector to prevent obstruction of the

flow and possible flame-holding. The use of a three-blade, radial baffle did not produce any significant

reduction in the amplitudq of the oscillations in earlier tests.(

In addition to various cavity and baffle combinations, the LP injection sheet thickness and the flow

exit geometry were varied. Four LP injection sheet thicknesses were tested, varying from 1.12 s to

3.00 mm. The sheet thickness is the gap between the center control rod and the inner diameter or the

piston. The thickness was changed by changing the outer diameter profile of the center control rod.

Figure 6 shows the outer diameter profile of the center control rod with the dimensions used to obtain the

four sheet thicknesses. Two flow exit geometries were tested with the difference being the outer face of

the injection piston. The two confi( =rations are shown in Fig. 7. The purpose of the changes in LP sheot

thickness was to datermine the effect of changes in droplet size and jet length. The purpose of the changes

in flow exit geometry was to determine the effect of potential changes in the exit flaw field, such as

vortices and recirculation zones. Th•ese changes could not be experimentally quantified. The flat-face

piston configuration had resulted in a significant decrease in the pressure obcillation amplitude in earlier8

tests.

392

A_ - B •--- -

L__,

3.635J 3635 J 36251

LP SHEET THICKNESS A Ccm

0.112 2 104 4.953 3.4290 175 2 413 4.699 3.3020239 3277 3 556 31750300 4 128 "350 3048

Figure 6. Center Control Rod with DiLansions end Sheet ThicknAes.

20.3 -' 0mI-

..........- 3- ............. 36 7 - --... ......... .- ... ...

1 _ _ , _/

(a) (b)

Ft•gure 7. Inlection 1i'rton with (a) Conic.i Face and (b) Flat Face.

I NSTRUM•NTATION

The prescure was measured aL various !.uc)ations in the combustion chamber using Kistior SO7C4

plazoelectric pressure transduceis. In ad'ition, l,•ninosity rnens,irzennts wore made in Test NoL. 415-46

through No. 415-51 using a High pressjro optical probe 13'14 located in the J90 position. Table I1 presfw.s

the locations of the gage positions In the coirbustion chamber. Thq front of t.he c ritbur i:. 6efired he;ia an

the face of the spacer rin4. a.i &hown in Fi 6 . 1. In adc~itioit to these nops-irernents, njicrlý.'vev

inteiferoninetr- was used to ,e-i;re pzojectile notion and an optical tracking device wcs used tý -oast, rA thOR

nIJic;tion pistcn motion

Table Il. Gae. Locatinns in the 30-mm Con'ý:ipt VI PJ.LI.

Label L~cation fyom Ra~tal Location'Front of Chambor CCW fromc Top (facing mrizzel)

(cm) (dog.)

J60 0.36 60J120 120J240 2403270 270C30 k.37 30C120 120CZ40 240Ago 6.27 90A330 330

AMLYI E4PED

A frequency spectrum analysib wAs ptrtcrmed on the data from every gage locat~ion in each. of the eleven

tests in the series. Figure 8 illustrates the procedure ftmz the 360 gage location in our baseline ruund,

Test No. 415-46. The raw preissure-time data, Fig. 8a, was band-passed filtered between 2.5 and 80 k~z. 1he

2.5-kBi L~wor limit. eliminated the baseline preassura. The acquisition system limited the m~aximrum frequency

of raw data to 80 kat to sn~ure no frequencies were stored that werea above the Nyquist frequency. The

sampling rate of tho ý'aw signal was 200 kOz. The resultant pressure trace, Fig. 8b, &hows the pressure

oscillations about zero. An 8-taa window was then chnocen that eincumpassaed the pressure oscillations (7-15 .1s

in Fig. 8b) and a fast fouriar transfor~m (FFT), Fig. 8c, and an s'jto spectrum, Fig. 8d. were then calculated

on this wintdow. The maxiraum poak-to-peak and RWC pressure level were calculaeed in this z.eme window from

the trace 1i .& 8b. Ouz analysis concentrated on the peak-to-peek pressure, the PMS pressure. and the

spectral density value o! the auto spectrum. We did not concentrate on the magnitude value of the FFT

because we were initereatmid in the effects on the acoust~ic resonance frequencies that would be indicated by

the auto spoctrumi analysis. The d.-icminrian frequency at &a~ age location wee csiorded with the reapeutivu

-pectral cbonfity value. Cross spectrums were also computed between gages locdted in the same plane to

determine the common freqruencies.

RESULTS AND DISCUSSI:ON

Table III libt~s the herd.4are configural-ion used i.', each of the teats in the serios along wiilh the

results of the freqluenicy .'niaiysiad. Only the results for tfim 3-plane and tho C-piane will be described.

Test, to test comparisons will he rrade <nily betmireen pages in the) caea plane. Tlie igniter propellant listed

in Table III was IMC 4350, a smail arms sm'jk,.ars powder. lhs5 variation in the ignitsr charge was due to

compensation fne tho changes In chamiber volunme in some configurations. The m~issing test n-,mbers indicate

tests that wore not used m.n the mneiyzis because Lhere was .Ln a~noe.!-y urmi~lated to the pressure 05,'Allatinns

or Lhiev were perfo~rmed with et di 'fetont teat. objective

It. is difficxml~t draw conclum~orns by looking at the toal6 results of all igages and rounds because t~here

was aig~nif.%.cant rmcattse- in Lhe netei bitween rounds thit, -were performed under the same configuration. For

oxatrcie, in Test Mfos. 415-47 itnd 415-48, tl.ore waa a 50 percont differiince in the measured PM' pressure

level. at the C30- Sagsi location. There% was a 43 pprcant, difference In the meazuroe'~ i~emk-to-poak preesure

level at the J240 10CALAOon. There were similar d~fforences in thqe alculated spectral density Phnd, ',.; a

Issser extent., the observed doisinent frequency. Another diffic-u.'ty is that morb than onic changA was

someticits eadg in t-he same teat., m~aking it difficult to compare the rerulta5 of the c'cnfiguratio., change,

F'ct exam-~o1,, &1I four zhpet thicknes-ses were not, tes.ted with the other paremoeters the. same. Theae c~anges

were usually ziecossery hovi-tuse (if prohlens with the 7eapectIve pie'ýes of hardware, i.e., the --)Jalr' and

piston, brti a ti.To coratvA)inlt x~til,'~ of t':e te!.t aer.xet waL perfocmed un~der.-

Table III. Hardware Configurations and Frequency Analysis Results in Test Series.

Test No. Igniter Piston Forward LP Sheet Gage Peak-Peak RMS Dominant SpectralCharge 7ace Chamber Thickness Loc. Frequency Density 2

(g) Config. Config. (mM) (MPs) (MPa) (kBz) (KPa/lz)

415-46 3.6 Flat Ring 1.75 J60 22.8 194 24.6 8.03 x 10-

J120 18.2 154 26.7 5.36J240 18.4 168 26.7 3.77C30 1,3.1 413 26.6 43.63C!20 21.7 230 26.7 15.54C240 14.9 144 26.7 5.51

415-47 3.5 Flat Cavities 1.75 JT6 10.2 105 37.Z 0.59J120 13.8 149 38.2 1.81J240 25.9 291 37.3 16.69C3M 44.9 4R9 27.3 25.31Ci20 22.6 243 34.2 3.63C240 24.7 306 28.3 8.18

415-48 3.5 Flat Cavities 1.75 J60 7.3 84 25.4 0.19J120 15.1 161 32.2 3.70J240 13.3 167 29.5 1.71C30 22.5 317 32.4 6 55C120 220 316 26.1 4.45C240 24.3 356 24.9 9.72

415-49 3.6 Ftat Cavities 2.39 J60J120 14.1 182 28.9 2.97J240 14.6 156 34.5 4.57C30 18.0 267 34.2 4.42

C120 23.0 300 27.9 3.46C240 15.3 230 51.8 1.39

415-50 3.6 Conical Cavities 2.39 J60 13.4 150 26.7 2.06J120 21.7 203 27.5 9.92J240 17.6 180 27.1 4.33C30 22.8 263 28.0 4.24C120 22.2 249 26.8 7.52C240 21.3 2,1 26.i, 5.96

415"51 3.6 Conical Cuvities 3.00 J60J120 14.1 161 30.0 1.21J240 16.8 198 34.3 2.67C30 24.5 358 27.7 7.19C120 23.2 294 30.2 4.75C240 22.8 298 32.1 4.63

415-52 3.6 Conical Cavities 1.12 J61, 11.5 91 5.3 1.79J120 14.6 258 5.3 4,05JZ40 13.6 251 29.3 1.55C30 17.4 210 27.0 2.90C120 19.8 237 5.3 8,24C240

415-54 3.6 Conical Cavities 1.75 J60 17.2 139 24.9 4.6',with 3120 14.2 116 29.0 3.60Cavtties J240 22.9 179 28.3 3.36

C30 24.8 268 26.8 6.43C120 21.1 '34 28.0 4.45C240 21.8 243 27.8 6.29

415-59 3.2 Conical Baffles 1,75 J60 10.2 90 21.3 0.73J120 10.6 106 30.0 0.87J240 12.5 109 45.7 1.44C30 23.2 252 33.5 9.33C123 17.0 173 31.0 6.04C240

415-61 3.5 BaffLos Cavities 1.75 J60 9.5 100 35.8 0.40J120 19.7 198 26.4 3.36J240 20.1 200 31.9 6.18C30 21.1 230 35.9 5 61

C120 18.3 190 35.8 3.64C240

415-72 4.0 Eaffles Baffles z.)9 J60 11.8 117 23.1 0.84Ji30 12.3 115 25.1 1.80J240 11.3 i25 M9.3 1.32C3U 16.9 188 39.4 2.64C120-)40 l .6 1/2 39.0 4.89

395

mea

~i

5*- io.~.

(ci (b)

LU9-

"sat'. . . -I

to o % 0. so is

(c) (J )

Figure 8. Results of frequency analysis procedure, Test No. 415-46, gage location J60;

(a) raw pressure, (b) raw data with baseline pressure removed,

(c) frequency rpoctrum. (d) auto spectrum.

The resul.e of the telst series showed that there was no major reduction in the peak-to-peak or the RMS

pressure level of the ,.sillations. Howevex, some positive effects of the a.oustic dampeninS dovices were

cbsmrv,.d in the test uata. The effect of the forwai l cavities was observed by comparinb Test Nos. 415-47

and 415-48 with the baseline round, No. 415-46. Thr, average J-plene RMS pressure level was reduced 16 and

39 percent, respectively. The overa&e J-plsn- peak-to-peak prospure a-tualiy increased in Test Ho. 415-47

but decreased 20 pc.'ceist iii Test No. 415-48. These effects '40e. not coberved 19 the C-plane. suggesting

that the cavities may only be effective in their vicinity. More importantly, the spectral density in the 21

to 27-;.Hz range, for which the cavities w)ee designed, was reduced 42 to 97 perceiit. Tnis is nct readily

evidert iT, Tabla :1I, however, it is shown thai the dominant frequency in the J-plane was 3hifted up to the

30 to 40kHz range whaen the cavities waeo used.

The use of baffles at the forward pert ofi the ct,ambr, Teat No. 4i5-59, reduced the average .i-plane RMS

pressure 44 percent. IT, cddtt:on. iz wat, the firs, observod reduction ins the average yeak-to-peak pre5.euro,

41 percent in the J-plana The flict reduction In the average C-plane RWI and peak-to'peak p:e3iure was

oserved it. a t19t with in~erloaving haffJla, Test No. 41|-72. fhIL tebt wns perfurmed with haffles located

both in the foiward part tif the chamber arid also attached to the pistor face. The average WMS pr-ss$Jre was

reduced 37 percent and the average ptak-t•-jyeak. pressure was reduced 31 percen'..

No rondo, rutside the scatter in thi data, were observed by changing the injection s};eet thickners, as

showii in Tust hos. 415-50 to 415-5Z. which wore pe -formed ut.der the sa.ze condit ioil. In Test. Nk-. 41,- 52 .

3.)6

performed with the thinnest sheet thickness of 1.12 am, there was a sisnificant reduction in the dominant

frequency. A frequency of 5.3 kHt was vbaerved, which may be a first tangential mode. It should be noted

that the maximum pressure in this test was lower than normal due to the decreased flow rate resulting from

the thin sheet thickness. The maxim joessure in Test No. 52 was in the range at which the pressure

oscillations usually begin to appear, 80 to 100 MPs (see Fig. 8a).

No effects were observed from changing between the flat-faced or conical-faced pistons. The use of

cavities on the piston face, Test No. 415-54, did not appear to have any effect. In general, tbe RMS,

peak-to-peak, and spectrel density level were consistently higher in the C-plane. This plane is closer to

injection and may be where most of the combustion is taking place,

Cross spectru analyses were also performed among the gages in the same plane to determine the common

frequencies. Three cross spectruts were performed in each plane, providing date was available for each

location. The results of these analyses are shown in Table IV. The range of frequencies listed in Table IV

indicates the high and low value that was calculated from the cross spectrum.

Table IV. Results of Cross Spectrum Analysis,

Test No. Cross Spectrumn Dominant Frequency

J-plane C-plane

415-46 26.7-26.8 26.7

415-47 37.3-38.1 27.2-28.2415-48 31.4-32.1 25.1-32.4

415-49 34.6 34.1-34.6

415-50 26 6-33.0 27.2-28.9

415-51 34.4 27.5-30.3

415ý52 5.3 27.7

415-54 27.6-28.3 27.:-27.9

415-59 26.9-49.7 31.1

415-61 31.9-35.7 35.8

415-72 25.1-39.1 39.2

The analysis of this data is not complete. In addition to deterwining the phase relationships among

the pressure oscillations recorded in each plans, the change in the oscillations with timid will have to be

analyzed. Waterfall plots showing the magnitude and the frequency of the oscillations as a function of time

illustrate this point very well. Also, the stnrt-up and decay of the oscillations needs to be compared to

the location of the injection piston or the combustion chamber pressure to determine if there is a

correlation.

Haberl15 recently proposed the theory that the majority of the pressure oscillations in the 20 to

35-kHz range aro due to localized combustion. flaberl based his theory on the analysis of data from a 155-mm

RLP3, in which he assuibrd that the Sun chamber dimensions made the existence of the ]ower acoustic modes

above 12 to 15 kHz unlikely. However, as noted by Uaberl, the dimensions of a 30-M RLPG gun chamber are

such that some of the lower order acoustic modes may exist in the 20 to 35-kHz range. In our tests, the

rhift in the dominant frequencies above 30 kHz and the reduction of the oscillations in the 23 to Z7-kHz

range with the use of the forward chambet cavities suggest an acoustical nature of the oscillations.

The lo:alized combustion Lheory must be taken into account when analyzing the data because it is one

possible source of t.he osciliationb. Additional sources of the oscillations are flow induced os illatLions

(i.e., vortices and reciiculation zones) due tc the LP injection, movement induced oscillations due to

coupling of the flow with flexing of the injection piston and/or the center control rod, and stre.o-saLraln

waves traveling through the metal. It is possible that several of these excitation mechanilss are

responsible for the oscillations observed in the data, making the task of eliminating them difficult, if not

impossible.

16Sandia National laboratory (SHL) performed some recent work on LP gun prossure oscillatiorns. They

17have used an injector/combustor teat fixture rather than a gin fixture, however, the combustion pressure

31I7

wa LsMla~r to that in the RLPG. Pressure our.illetions were observed in this test fixture corresponding to

the acoustical modes of the chamer. Several dampening devices were used, including a chamer quarter-wave

cavity liner and a Helmholtz carvity liner. Nto significant reduction in the magnitude of the oscillations

wsobservad in these tests. Dowever, a significant reduction was observed when a rubber Liner was itied inwa: 16the combustion chamber. This rubber liner was first used by Giovanetti at General Electric Defonte

Systems Departmernt (GEOS`) during work on a LP ignition &yet=e and was based on an analysis by Pasko. ai~o

of GEDSD. Another test that showed significant reduction in the magnitude of the prissure uscilLa-ions was

one In which a flow splitter-plate device was Installed in front of the injector. These findings, at least

in the fixture at SOL, suggest that the osiuillations are of an acoustical nature. The test with the flow

splitter-plate shows an important relationship of the flow field on the pressure oscillations- The results

of these tests are encouraging, however, the tests at SPlL were perforoed with a simple circular orifice.

Similar tests should be evaluated in a RLPO configuration.

SUIOIAJY AND CONCLUSIONS

The effects of acoustic dampening devices '.n the pressure oscillations observed in a 30-on regeneral.ive

liquid propellant Suin were analyzed. The scatte- in the experimental date obscured the effect& of most of

the configuration changes, however, some effect* we.zý observed, The use of the forward chamer cavities

rediced the average RMS pressure 16 to 39 percent in the ilane closest to the cavities. The spectral

density in the 23 to 27-kBz frequency range was reduced 42 to 07 percent. The dominant frequency was

Shifted to the 30 to 40-kB* range when the cavities were used. The use of baffles reduced the RMS prescure

37 to 44 percant and the peak-to-peak pressure 31 to 41 percent. Nlo significant effects could be determined

from changes in the *heet thickness or the flow exit geometry due to the amount of scatter In the test data.

However, the thinnest shoet thickness reduced dominant frequency to 5.3 k~z, possibly a firtt tangentialmdod. The RM6 and peak-to-peak pressure and the spectral density Level werm &e"erally higher clQose to

Injector orifice.

FurLher analysis of the test data needs to be perforvmed. The phase relationships ans the pressure

oscillations observed in each plane should be determitied to chock if they acoustical in nature. In

addition, the variation of the magnitude and frequency of the pressure oscillation as a function of time

needs to be determined. The start-up and decay of the oscillations should be compared to the location o.

the injection piston or the combustion chamer pressure. The purpose of the last two analyses would be Wo

determine it there is a correlation between the oacillaticro and any event occuiring during the ballistic

cycle. Additional tasting should also be performed In a 1100O to evaluate the results observed in the $"niaNatiogisl Laboratories test fixture.

REFUJNCES

1. Mandzy, ;., Custvaen, P.O., and Magoon. I.. "Technical Not*s on Scaling Investigation of Concept Vi,U.S. Army Ballistic Research Laboratory, Aberdeen Proving Crouna, Md., Contractor Report BRL.-Clt-580,August 1087.

2. Pat@, R. and IKagoon, I. , "Preliminary Resuilts froms Ballistic Invosti,-ationo In 30-tm RaeginerativeLiquid Propellant Gun Firings," CI'IA Publication 432, Vol. 11, pp. 213-224, October 196S.

3. Mandty. J. , Magoon, I , Morrison, W.F., &nd Knopton, J.D., "Prelisinaz-y Rsport on Teot Firinas of a105-rn Regenerativoo fixture," CPIA Publication 383, Vol 11, pp. 181-172, Oct~ober 198).

'.. Magoon, I., Haber!, I , and Plurtse. E. "Preliminary Report on lest Firings of a 155-ras RegenerativoLiquid Propallant Gun," CPIA Publication j29, Vol. 111, pi,. 243-251, October 10960

5 Watsun, C A.. "?I-esuri, Oscillations In Rejonerative Liquid Propellant, guns.- Proceedings of the 1289JAJfWA? Propulsion MO*Lin8,# CPJA Publication 515, Vol. IV. pp 345-3ýA, Key 1989.

6. Watson, C.A., Knaptoti, JO., De~pir~tts J., and Boyer, P., 'A Study on high Frequeticy OscillationisUbserved In a .,0-m Regeteorat ivo Liquid wro;'llant Gun," CPIA Pukblicatilon 452. Vol 'I. pp 239-254,October 1065.

'199

7. Knepton. J.D. and Watson, C.A., -Pressure Oscillations During the Interior Ballistic Firing of

ResenarstLve Liquid Propellant Guns." BRL Report in publication.

8. Watson, C.A., Knapton, 3.D., and Boyer. N.. -Effects of Injector Euit Gecmetry on Pressure Oscillations

in 30-m- Regenerative Liquid Propellant Guns,- CPIA Publication 529, Vol. I11. pp. 439-448.

October 1g8g.

9. Cushman, P.., and Orachis, 0., -Linesr Wave Motlon i., the Concept VI Liquid Propellsnt. Test Fisture,-

Generel Electric Ordiance Systems iiuiiton, utyublished report. Kuvembor 1983.

10. Watson, C.A.. Knepton. Jb.. "Ad Coaes, S., "An E•si&Ato of Sound Speed During the Interior Ballistic

Firing of & Rv&neratLive Liqjid Piopellant Gun,' CPMo "licet.ion 476, Vol. III. pp. 395-404,

October 1987.11. Freedman, E., "Thermodynaxic Properties of MiLitary un Fropellente,' in Gun Propulsion Technology, Ed.

Etiefel, L.. frogress in Astronautics and AeronauLics, AIAA. got 10. pp 103-132, 1988,

12. "Liquid Propellant Rockt Combustion Insta bility, narre, r.T., r4.. AASA SP-194. 1712.

13. Klingenbert. G.. "Invasive Spectroscopic Technique for Measuring Tempeurture in HighJly Pressurised

Combustion Chambors," Journal Optical ,nginesang, Vol. 24, 0-. 4, pp. 62-700, August 1985.

14. Boyer, NP. , Watson, C.A., Knapton, J.D., and Coffee, .P.., "Temperatu Prlsuretrt During the

Cxbusetin of a Hydroxrylmmoo.•um Nitrate Based Liquid Propellant," CPIA Pu"licatlon 529. Vol. II,

pp. 430-448. OLtobor 19089.15. babecl, J., "'Pressure OocilLt•tons in 155-rn Regenerative Liquid Propellent Gun No, 1,' DLA-O-1060

US/Oermen Workshop on Liquid Propellant Technology, U.S. AXru Ballistic Research Laboretor-y. May 1990.

16 Rychnovsky, R.E., Sa1d4e National Laboratories, private cornmication to J.D. Knapton, August 1990.

17. Carling, R.W. , Rychuiovsky, R.I. , end Oriffiths, S.K., "Combustion of Liquid Propellant in a

High-Pressure Injector/Comustor," CPIA Publicetion 529, Vol. 1, pp. 349-354, October 199

18. Giovanetti, A.J., General Electric Defense System Department, private communication to J.D. Knepton.July 1990.

REGENERATIVE LIQUID PROPELLANT GUNAND

SELF PROPELLED HOWITZER INTEGRATION

W. E. JacobsmeyerGeneral Electric Company

Defense Systems DepartmentPittsfield, MA 01201

ABSTRACT

A U.S. Army and General Electric Defense Systems Department (GEDSD) joint activity to integrate a155mm liquid propellant gun into a self propelled howitzer is presented in this paper. This jointactivity provides an automated, remote controlled, high rate of fire, munition delivery platform.In addition, the platform provides the capability of firing multiple, successive rounds ontrajectories that impact the same target area at the same time. Furthermore, it verifies thefunctionality of an integrated Liquid Propellant Gun (LPG)/Self-Propelled Howitzer (SPH) weaponsystem and the integrated system will demonstrate its ability to satisfy future artilleryrequirements.

Project requirements, equipment, and services necessitate the integration of and coordination withactivities being performed by various U.S. Army agencies with the GEDSD effort to satisfy projectobjectives. The Army provides the vehicle, gun, and gun mount, whereas GEDSD provides the elevatingsystem, system controller, and the overall integration for the LPG/SPH system. This system, becauseit is not to be man-rated, incorporates remote control of gun positioning, projectile loading,propellant filling, and weapon firing. Specific electromechanical devices developed by GEDSD forthis system provide the desired automation within a fail-safe environment.

INTRODUCTION

In order to counter future anticipated threats, each future self-propelled artillery cannon platformmust deliver more munitions at greater ranges within a given time period utilizing insensitivepropellants. Therefore, future artillery weapon systems must be more responsive, accurate, lethal,sustainable, and supportable with reduced personnel and material resources than present systems.

A gun propulsion technology (identified by the U.S. Army) that may satisfy future artilleryrequirements is a liquid propellant gun (LPG). However, this technology has never been integratedinto a self-propelled howitzer (SPH), (Figure 1), with the attendant resolution of performance andphysical characteristics of gun and vehicle. This gun will provide higher muzzle velocity per

size; continuously variable muzzle velocity; increased ammunition loads for a given weight;and cost savings per charge size while using a propellant that is being qualified as an insensitivemunition.

Figure 1. M109 SPH

Since the LPG is still in the developmental stage, the prototype LPG/SPH combination is beingconfigured to ensure personnel safety consistent with demonstration objectives. Remote operation ofthe LPG/SPH will be provided to prohibit personnel exposure to rapidly moving parts and to guaranteepersonnel safety during gun loading and fi:..ntj. Therefore, this project includes designs andequipment that provide remote control of gun tube positioning, projectile loading, propellantloading, and firing. This eliminates the need for any personnel to be inside the vehicle duringtest firings.

This work performed under Independent Research and Development funds.Approved for public release; distribution is unlimited.

Best Available 401

The LPG/SPH integrated system will provide the following capabilities (please note that thedevelopmental projectile loader contains only three projectiles):

a. Minimum firing range of < 6 km,b. Maximum firing range of > 40 km (using assisted projectiles)c. Burst firing rate of 3 rounds in 15 seconds at 945 m/s,d. Maximtm rate of fire of 9 rds/min for 3 rounds,e. Sustained firing rate of 3 rds/min for 1 minute, andg. Firing 3 rounds which land simultaneously on a single target.

INTEGRATION PLAN

The objective of integrating an LPG into a SPH is to demonstrate the weaponization of liquidpropellant gun technology by providing an automated, remote controlled, high-rate of fire (includingtime-on-target (TOT) capability), munition delivery platform. This platform--with its supportingelectromechanical, hydraulic, and electronic devices--will verify the functionality of theintegrated LPG/SPH weapon system and demonstrate its ability to satisfy future artilleryrequirements.

Variations of three candidate self-propelled howitzers (M110, MI09, and AS90), resulted in ninedifferent configurations with different levels of capability and trade-offs, were performed toselect the most cost effective configuration. These analyses utilized Advanced Field ArtillerySystem requirements for main armament systems and implementing cost factors as decision criteria.The analysis results recommended an M109 self-propelled howitzer configuration which provides TOTcapability. This recommendation was accepted (by the Government) because it would be an adequatetest bei to generate data for the gun propulsion down-select decision.

Genera.L Electric Defense Systems Department (GEDSD) and the U.S. Army Research Development andEngineering Center (ARDEC) have defined this project to be a joint effort. This joint project isintegrating a developmental LPG (which is mounted and fired on a modified M115 towed howitzer at theGE LP Test Facility, Malta, NY) with an SPH (which is an M108 self-propelled howitzer upgraded to anM109). In general, the U.S. Army is loaning the vehicle, gun, and gun mount whereas GEDSD hassystem engineering responsibility in addition to providing the equipment integration, gun tubeelevating system, and system controller. Also, supplemental subsystems (including data acquisitionand power distribution) are being designed, fabricated and/or acquired, and installed into the SPHby GEDSD.

Various methods are being taken to minimize the interruption of developmental firing of the LPG.One of these methods is the fabrication and installation of major component mock-ups into the SPH todefinitize space claims and to establish interfacing cabling and hose routing and lengths. Inaddition, these cables and hoses will be fabricated and installed into the vehicle prior to theavailability of the LPG equipm3nt. When the LPG equipment becomes available, the mock-ups will bereplaced with the actual equipment.

Verification of the integrated system shall take place at the GE LP Test Site, Malta, NY. Aftersuccesoful completion of integrated system verification testing, the LPG will resume the interruptedschedule of firings. In 1991, the LPG/SPH will be transported to a Government test range (to beselected by the U.S. Army) for the gun propellant demonstration. This range will allow full-chargefirings at maximum range and demonstration of TOT fire missions.

Fabrication of items and other activities associated with the LPG/SPH integrated system will belocated and/or performed in four different states. Government facilities at Watervliet, NY;Chambersburg, PA; and Picatinny, NJ are being utilized in designing, fabricating, assembling, andtesting major components of the LPG/SPH integrated system. GEDSD facilities at Pittsfield, MA andMalta, NY will be utilized for equipment design, fabrication, installation, and verification. Theneed to minimize LPG firing interruption, the reduction of transportation times as well as thenumber of times an item of equipment must be handled were prime factors in determining thatWatervliet, NY would be the site for gun, gun mount, and vehicle assembly. Specifically, theparticipating Government agencies, their locations, and project contribution(s) are:

Fire Support Armaments Center, System Integration Division, Picatinny Arsenal, NJ - LP Gun,M108, and Government Agency funding/coordination/integration;

Artillery Armaments Division of ARDEC, Picatinny Arsenal, NJ - LP Gun Mount and M109 TurretModification Designs;

Watervliet Arsenal, Watervliet, NY - LP Gun Mount Fabrication;

Benet Laboratory, Watervliet, NY - LP Gun Mount Components; and

Letterkenny Army Depot, Chambersburg, PA - M109 Turret Fabrication

Best Available402

INTEGRATED SYSTEM DESCRIPTION

The LPG/SPH Integrated System (Figure 2) consists of three segments: Chassis, Cab, and SupportingSystems. The Chassis is a product-improved MI08 chassis with spades. The Cab includes the AdvancedArmament Turret and will contain the LPG Cannon and those equipments necessary to support the Cannonduring gun propulsion demonstration firings. Equipment (for this demonstration) located external tothe SPH is categorized as Supporting Systems. The above segments will operate as an integratedsystAm to demonstrate the stated liquid propulsion gun capabilities.

LPOSPHINTEGRATED SYSTEMTOP DOWN BREAKDOWN

INTEGRATED |ST[M

&MAIN T €OTO RONT| PRLP111Roy

A T|MIM IYTIL POINTING R.O|MTE LOADER

Figure 2. LPGISPH SysTEm

This developmental LPG/SPH configuration (Figure 3) will not be man-rated at the time ofdemonstration firings and, therefore, incorporates remote control of gun positioning, projectileloading, propellant loading, and weapon firing. Specific electromechanical, hydraulic, andelectronic devices being developed for this system provide the desired automation and control withina fail-safe environment: the system will gracefully degrade so that no failure will allow acatastrophic event (to personnel and/or equipment) to occur. Furthermore, these devices are beingdesigned to highlight the attributes of the LPG in the areas of a high rate of fire and of firingmultiple, successive rounds on trajectories that impact the same target area at the same time (theelapsed time between impacts of the first and the last round shall not exceed 3 seconds). Thesegoals are achieved by safety interlocks in the gw• firing circuit and designing gun tubepositioning, propellant filling, and projectile loading subsystems to satisfy the time line depictedin Figure 4.

Chassis - A M1O8 chassis is being used for the LPC/SPH integrated system. This chassis has anew engine, firing spades, and MI09A3 torsion bars. The power plant will be used formaneuvering the vehicle at the firing point and will not be operating during the gun propulsiondemonstration firings.

Cab - The LPG/SPH cab is an MIOS cab modified to resemble a M109A3 cab and includes the AdvancedArmament Turret. The significant modifications consist of: adding a MI09A6 bustle, raising theroof, and expanding the gun mount opening. It will contain the Main Armament, Control,Projectile Loader, Cannon rointing, and Energy Systems.

Supporting Systems - Exterior ancillary equipments necessary to support the LPG/SPH duringdemonstration firings include auxiliary power supply, instrumentation, special handlingequipment, and logistics elements.

OPERATION

LPG/SPH automatic operation includes opening the cannc., breech and inserting 155mm projectiles intothe cannon with a projectile handling mechanism. Following the removal of the projectile ramtmerdevice from the cannon and closing of the breech, liquid propellant is injected into the breechplug. In a three round burst fire scenario, one projectile is rammed into the gun tube and twoprojectiles are positioned in a storage clip. Positioning the gun tube in elevation for any firingmode (single or multiple rounds) will be accomplished (prior to projectile loading), if necessary,by the Cannon Pointing System. Interlocks are provided throughout the above actions and the turretwill be locked in traverse to ensure personnel and equipment safety during firing(s).

3est Available Copy

HydraulicPowerSupply LP Effluent

LP Fuel Tank

Breech Filo SystemElevation Power Controller CnlnentsDrives Distribution

Autoloadef

uCrtrController

Batteries

Gun Mountaind Igniter Power

Recoil System• Supply

SROElevation Di veAutoloader • C n r l e

Elevation DriveGun Control InterfaceComputer

NOTE: FIrLL SYSTEM LOCATOD IN BUSTLE FOR COMIMIDCE

Figure 3. LPG/SPH Cab Layout

3 ROUND TIME-ON-TARGET TIMELINE(RANGE OF 12 Kin)

ACTIVITY I. .E. SEC) 13 12 I S 7 6 4 3 2 1 0 1 3 4 6, I ? 0 11 12 13 1. 1. 10 17.10 It 20P... .SE ..... .PS IMM* I l t r r !L,..I - UjSoTI N2.4N1L

. ,LL0I NAI - .1" 441 .

PONM.INIOIRVCS -'I

ILL .L• 1 1 .• ? 0 LITS I M IT I 1.I

.-1-1- 0t1 ' I3 l112 I1CE 1-r

016 11 NCACT.tIIMEAI2 5

IQN SE • (,."I

PILL LIKSI* .1 LIZ•ESE O I.. I

CLWS VAVi 1ES.. ... .. . .. ISI I I l Si l / f i t

* RANME!TOTOOET: 12KM ItAC LP INTO INITIATOA 10 1 1 ro

*MINMUIIM TIME BETWAEEN ROUJNDS: 10 SEC -- W0 AT -ISNILI.? t

I PROJECTILE TYPE: M T 10 FILL U IINSyIVICSI I 1.01. .. .STLIN V/ II)

TUBE NIlE LP T1 E OF LOAGEJC100TAI00 .12 LROUNID ELEVAIIN VELOCITY OUANITIY FLIGHT PNIS1 I10 '

LIFT T I INT P2TJTTILS 10 0 * It

I 4r 465 MyS 2.9L 66 SEC NOJCIS Ial1.2 W tO i 221o.SC-- I , I ION I I3.2 46 SEI3 2 5 N MIS G.1 T3L 6 SECI III I

,•,,+m - • ,m,,..,, I NJECT ILT INT 54200 OOI 102 [ ! |FILL LI ISSEIYOIILIt LIIEPSI II"41 IllI

INISOT LOTPINT:OIKNTI4ITTINITI0O1 Il1 51•R~tJNO IJ4T *TII*RlPI.II 1S1 Sloo

Figure 4. Time Line

4011 Best Available COPY

The LPG/SPH will be capable of automatically firing three projectile types (i.e., M107, M864, andM549A1) at varying muzzle velocities from 300 to 945 m/s in single, burst, and time-on-target (TOT)firing modes. Also, the LPG/SPH will be capable of firing M712 projectiles. However, the M712projectiles will be hand loaded which limits firing of these projectiles to single-round firingmode.

Single firings will occur at a rate of no more than one every 15 minutes which allows time forpersonnel to enter the SPH for resupply of projectiles and propellant. In a fully automated mode,the LPG/SPH is designed to be fired with a time between firings of no more than 7.5 seconds and witha burst length of three rounds at maximum charge per round. Location of additional propellent andprojectiles will support a manual resupply time of 15 minutes between three-round burst firings.

The LPG/SPH incorporates components and/or procedures necessary to permit both automated and manual(external of vehicle) control of safe clearing of propellant from the gun and shutdown of theLPG/SPH in case of a cancel-fire, misfire or projectile sticker. Moreover, the LPG/SPH isconfigured for battery-powered controlled removal of propellant from the Cannon in the event ofprime electric power failure. Furthermore, loss of prime electric power will place the LPG/SPH in afail-safe condition.

MAJOR COMPONENT DESCRIPTIONS

MAIN ARMAMENTThe LPG will be capable of firing the family of 155mm projectiles at propulsion demonstration muzzlevelocities and consists of: Cannon, Cannon Support Assembly, Fill, Ignition, and Gun ControllerSubsystems.

The Cannon Subsystem (Figure 5) will have a recoiling mass of less than 8400 lbs and iscomprised of combustor, breech, and gun tube with muzzi'r brake and bore evacuator.

LPG Mount (which weighs approximately 6000 lbs) and trunnion suppc.rt bracket, which is theinterface between the mount and the SPH turret, comprise the Cannon Support Assembly.

The Fill Subsystem injects measured quantities of propellant into the cannon breech and igniterand consists of: fill devices (igniter and breech), propellant containment vessels, valves,sensors, hydraulic supply, and piping.

The Liquid Propellant Ignition Subsystem initiates the combustion process and its elements arethe energy source and booster train which generates the proper temperature and pressure for maincharge combustion.

The Gun Controller Subsystem provides the processing, signal transformations for controlling thefill, ignition, and breech functions. It consists of processor, interfacing electronics, andportable keyboard/monitor (for LPG maintenance).

The LPG, when integrated into the SPH, will satisfy all the performance requirements of the LPG/M115configuration. It has a maximum liquid propellant volume of 1.4.2 litrs and can provide muzzlevelocities of 300 to 985 m/s. In addition, it will be capable of firing at all M109 quadrantelevations and firing three rounds such that they will land on a target area at the same time.

S~OPENINGBREECH MEHAIM

COMBUSTION BORECHAMBER EVACUATOR

GUN TUBE

MUZZLEBRAKE

F.igure 5. Carinon Subsystem

3CJL 40JA 15

LPG/SPH CONTROL SYSTEMThe Control System provides the supervisory control of the SPH, LPC, Cannon Pointing, and ProjectileLoader for all firings including simultaneoua rounds on target. It consists of: System Controllerand communications. Since the LPC/SPH will not be man-rated and firing control will be remotelylocated. communication circuits will be provided for operation, maintenarne, and range safety. TheSystem Controller provides the capability for an operator to select and control either single,burst. or TOT firing sequences of the LPG/SPH. Specifically, the functions of the System Controllerare:

a. Providing capability to select operating mode of system, that is, manual test,semi-automatic, or automatic.

b. Providing capability to select firing mode, that is, single round, or multiple round rate offire.

c. Providing capability to select type of fire mission, that is: fire for effect or TOT.

d. Computing ballistic solutions which determine muzzle velocity(s), tube elevation(s), andtime(s) of fire and incorporate meteorological connections to satisfy selected fire mission.

e. Providing capability to select tube elevations.

f. Providing capability to select muzzle velocity for each round of the firing mode selected.

g. Providing capability to select firing rate for TOT scenarios.

Control of the gun firing will be accomplished from a control panel that is remote from the SPHsince the LPG is not man-rated. After system power-up and initialization, the operator willinitiate the firing process by selecting fire mode and fire mission and inputting firing parameters(e.g., target range and firing point). Subsequently, performance of ballistic computations willresult in transmitting muzzle velocity, elevation angle(a), and times between rounds to the LPG GunController. Fur,''ermore, it will initiate and monitor the gun initialization and projectile loadingprocess. Any h','urual conditions identified in these processes will result In the firing sequencebeing interrupted (or terminated) until the situation is corrected. Termination of the firingprocess will result in the download of any propellant that may be in the gun.

PROJECTILE WADERThe Projectile Loader (Figure 6) will be capable of positioning 155w. projectiles into the Cannon attube elevations from 0' to +45" and consists of: rammer, lift/translate, and storage components.Control of its operation is provided by the Cun Controller. The Autoloader Subsystem (which isdesigned to interface with the LPC/Mll5 configuration) will be adapted to attach to the LPC Mount.Utilized in multiple round firings, it will automatically transfer a projectile from the storageclip and seat it into the cannon chamber. The storage clip is capable of storing three projectilesof any one of the following types: M107, M864, and M549A1 at any one time.

STORAGE CLIP

RAMMER

LIJ'I.ThANSLATE

Figure 6. 6 rje(-Ole I.iader

.100

I ! I Iim

The subsystem will load projectiles into the cannon in less than 2.5 seconmds. This activityconsists of: lifting the round to the gun tube axis; translation of the round into the combustionchamber: ramming the round into the tube; retracting the ramer device from the combustion chamber;and lowering the rammer out of the breech closing space envelope. The transfer of the nextprojectile from the storage clip to the rammer occurs at elevation angles of 45" or less and isdesigned to take less than 2.25 seconds. Moreover, to optimize component sizing and energyrequirements, the projectile loader will be constrained to operate at no more than 45" of gun tubeelevation. Furthermore, the Projectile Loader will be physically attached to the LPG Mount so thatit changes elevation in conjunction with the Cannon to maintain alignment with the gun tube duringmultiple round firings.

The rammer will accelerate the projectile to a projectile velocity of 11 +/- 1 ft/sec at the guntube to properly seat the projectile into the tube. This velocity will be provided over the rangeof projectile types and gun tube elevations (up to and including 45").

Cannon Pointing is accomplished by manually traversing the turret for azimuth and electricallypositioning Whe tube in elevation. Elevation will be from -3" to +75" and traverse shall beconstrained to 0O +/- 15" (to facilitate cable/hose routing in this developmental system). Thecannon elevati .n mechanim (Figure 7) responds to commands from the Control System to position thegun tube to the proper elevation to conduct the firing missions Furthertore, firing will beinhibited by the Control System if the zube elevation is not within range safety limits.

Elevation of the gun tube will be accomplished by dual actuators that are attached to each side ofthe LPG Mount and the turret overhead. Each drive is desig.oed to apply 22,350 ft-lbs of torque atthe trunnion axis via a geared motor It includeR a hydropneumatic equilibration unit with a ballscrew tube hydraulic activator and an external nitrogen accumulotor. The hydropneuoaticequilibrator force acts in parallel with the elevution drive to remove the effect of unbalance ofthe 155mm carnnon from the vehicle. This subsystem is designed to position the gun tube to anyelevation within 3.0 milliradiins which will satisfy the gun propulsion demonstration criteria.Also, it will be capable of elevating or depressing the cannon 15 degrees in 2 seconds whichincludes elevation drives settling time.

VEHICLE ROOF

.... ; MANUAL•'''

~ ~ .7 ~ GEAR BOX

BALI SCREW ACTUATOR

TxESI UBE ~ ~HYOATTACHMENýT ., I FLUID "-- ,] EQOU ILI qR ATI0 NAtTAHMET U.p~uO N-!'ACCUMULATOA

70OGUN MOUNT .

EXTENSION TUBE HYD FLUID

Figure 7. Tube Ele',ating Drives

On-board electrical, hydraulic, and pneumatic power components comprise the Energy Sys':em. Includedin the electrical category are power harnesses, 're harnesses, power switches, circuit breakers,ond lighting. Hydraulic and pneumatic power supp.ies fo- the Main Armament anc. Projectile Loaderare also elements of the energy system.

The energy system provides isolated and unint-errptable electrical power to the Gun Controller andIgriter High Voltage Power Supply. Also, power will be provided to the Breech. Projectile Loader.and Elevating Subsystem drive motors. It will maintain the electronic power between 18 and 32 voltsDC and motor drive power between 18 and 32 volts DC. A single point ground system will be utilizedin the cab as well as isolation of low-power users from hiph-power users.

407

AUXILIARY POWE GENERATIO & r1TZLI•U=

Anticipated Government range safety requirements stipulate that the SPE power plant nor operateduring demonstration firings. Therefore, the required AC and DC elentriCAl power will be pt-videdby generation and distribution equipment located external to the SPF. This power generation,control, and distribution equipment will satisfy the electrical power req-iirements of the LPG.Control System, Projectile Loader. Cannon Pointing, data acquisition, and communication eq::ipment.

Gathering of gun and vehicle performance data is extremely important to the integration project.Data acquisition equipment will be located in an instrumentation van (at the ,overrwent Test Site)to collect and record firing data for off-line analysis and will consist of off-the-shelf sensors,signal conditioners, data recording equipment, data processing equipmetit, and a local control panel.This equipment will be configured not to adversely affect the LPG technology dfvelopment equipment.In particular, a failure in the instrumentation equipment will not cause a failure in the priMAryequipment.

Logistics for the integrated system consists of those items necessary to support the RLG/'SPH duringthe integration and demonstration period and include: consumables (i.e., spare parts, projectille:;,and propellant), special tools, test equipment, and maintenatice of the automotive eleutnts of t.hei,,':egrated system.

SPECIAL HANDLING EOUIPMEHTPeriodic inspections of the cannon breech components are necessitated by the developmental naturt ofthe present LPG. Also, access to the cannon breech when installed into the SPH is via an opening inthe rear of the P4109 cab that is 48 inches wide and 36 inches high. Therefore, special handlingequipment is being designed and will be provided for installing and removing the breech (whichwighs approximately 2200 lbs) from the LPG Cannon and the M109.

SAFETY CHARACTERISTICS

The LPG/SPH is configured to ensure personnel safety consistent with propulsion demonstrp.tionobjectives throughout its life cycle at GE LP Test Site, Malta and Governomnt test site(s). Remoteopezation is provided to prohibit personnel exposure to rapidly moving parts and to ensure persomelsafety during gun loading and firing. Furthermore, the gun firing circuit includes an interlock forrange safety control to grant permission to fire. Also, readoutb of gull tube position aind liquidpropellant quantity in the gun shall be provided to the Test Diractor (and the Range Safety Officerwhen the LPG/SPH is located at the Government test site). In addition, presence of personnel in theMalta test cell and/or in the SPH prohibits the generation of a firing pulse.

The LPG/SPH contains no single component or element which, if failed, can cause catastrophic orcritical hazards. Furthermore, fail-safe features (e.g., computer power battery backup forcontrolled gun safing) are being incorporated into the LPG/SPII to ensure the safety of personnelduring the installation, operation, maintenance, and repair of equipment within the SPH.Furthermore, remote control of equipment within the SPH is prohibited if there is any personnellocated in the SPH. In addition, video monitor capability is provided in the test cell and in theSPH to ensure absence of personnel prior tc gunt fill and firing. Aiso, hazardous voltages andcurrents are covered by protective guards or barriers.

Liquid propellant fill and download lines and fittings are protected to prevent damage duringrecoil. Propellant fill lines, pneumatic lines, hydraulic lines and eectrical connections withinthe cab are uniquely keyed to prevent misconnection. In addition, for system safety, prior tofiring another round in a multiple round fire mode, the following checks are made:

a. Proper operation of gun components.b. Fuel metering device is full.c. The igniter is dry.d. Muzzle velocity of previous round is within tolerance.e. Gun is in battery.

The chassis and cab will bo equipped with an automatic fire detection and extinguisher systemcapable of sensing and extinguishing hydrocarbon fires. Cab firws will be detected and extinguishedwithin 250 milliseconds of ignition and the detectors will provide coverage of 95% of thecompartment. Engine compartment fir.-s will be detected and extinguished within 15 seconds ofignition.

CONCLUS ION

The integration of the liquid propellant gun into a self-propelled howitzer and operation of thisweapon system will verify the functionality and waponization of this technology in a fieldenvironment: a major step will be accomplished in transitioning from a laboratory environment to anoperational environment. Demonstration firings of the LPC/SPH configuration will initially beperformed in the first quarter of 1991 at Halta, NY and firings at a Government test site areplanned to occur in the second quarter of 1991

40N

AUTHOR INDEX

Acharya, A. V-377 Bond, G. N. VI-21Adolph, H. V-41 Booth, D. W. 111-323, M-387Ahmed, S. A. 1-305 Bowman, G. T. Ili-411, V-329Alexander, R. V. 111-435 Bowman, H. L IV-133Af!an, B. D. V-247 Boyd, W. C. 11-241Andersen, L D. 111-475 Boyer, N. E. 1-389Andersen, M. E. V-483 Bracuti, A. J. 1-331, IV-493, IV-501Anderson, T. . IV-177 Brady, V. L. 111-315Andrews, Jr., E. H. IV-193 Brasfield, R. G. V-483Ankeney, D. P. V-115 Braun, J. D. V1-73Arnold, C. 1. 111-253 Brennan, P. R. 11-97Asaoka, L. K. 111-573, V-11, V-31 Bricker, A. C. IV-IAskins, R. E. III-11 Brown, A. R. 111-69Atkeson, P. L 111-515 Buckley, P. L 1-313Ayers, 0. E. V-239 Burdette, G. W. VI-135, VI-173Baczuk, R. J. 111-55 Burnett, J. D. VI-11Bair, D. R. IV-521 Burns, III, W. G. V-75Bak, M. J. IV-251 Burton, R. L 1-211Baldwin, J. C. IV-327 Butler, B. V-301Banerian, G. IV-289 Calhoon, D. F. 111-113, 111-447,Barber, T. J. IV-177 111-461,111-475, 111-487Barkman, E. J. VI-153 Camin, R. A. V-145Barnard, J. C. IV-417 Canfield, A. V-435Barney, E. A. VI-135 Cannizzo, L F. HII-361Baroody, E. E. 1-345 Carling, R. W. 1-363Baum, K. V-41 Caron, S. D. IV-61Beaudet, R. A. IV-511 Carpenter, P. S. 111-341, 111-351Beaudette, B. 111-487 Carr, II, C. E. V-449Bcckstead, M. W. 1-119, 1-135 Cagreiro, L. R. 11-11, V-359Beichel, R. V-549 Carver, J. G. 111-181, V-31Beiter, C. A. 1-135 Cassem, B. 1-285Bell, R. 111-475 Caton, J. L. 111-85Bement, D. A. PV-193 Caveny, L. H. 11-383Bennetts, S. A. 111-113, 111-475 Chan, M. L. 111-335, 111-341, 111-351Benstein, E. H. 1-285 Chang, I-S. 11-47Bernard, J. P. VI-31 Chang, L-M. IV-459Bertolino, J. D. 11-241 Chew, J. S. B. 1-103Beyer, R. B. V-145, V-287 Chew, W. M. V-47, V-247Biddle, R. A. IV-469 Chi, M-S. V-55Birkner, B. W. 11-407 Chiappetta, L. IV-177Bixon, E. R. IV-423 Chik, J. 11-117Block, S. M. V-197 Chima, R. V. IV-73Blomshield, F. S. 1-119, i-135, 111-193 Chinitz, W. 1-211, IV-221Bochme, D. R. 11-359 Chiu, D. S. JV-493, IV-501

409

Christensen, B. Y. M11-25, II-35 Doyle, D. 0. V-347Christensen, K. IV-233 Ducote, M. E. V-67Chung, T. D. VI-199 Dumbacher, D. 11-107Ciaramitaro, D. A. 11-351 Dunham, R. S. V.287Cirincione, R. IV-433, IV.447 Dunn, B. M. VI-81, VI-18iClark, G. M. V-225 Dykstra, J. E. V-123Clark, J. R. VI-73 Eaton, A. M. M-35Clayton, W. 111-555 Eggers, J. M. IV-193Clift, W. M. 11-359 Eigel, C. R. IV-53Coble, E. V. VI-43 Ekman, K. R. IV-511Cohen, N. S. Il-I-, IH-387 Eldei, D E. IV-365Colket, M. B. IV-41 Elkins, R. T. IV-209Combs, L P. V-531 Elliott, K. 1-285Connaughton, J. W. 111-141 Elmquist, A. R. IV-145Conomos, H. 111-131 Emmer, D. IV-209Cook, F. R 1-55 Emmons, H. T. 111-505, VI-43Cook, J. 11-107 Ennix, K. 11-233Coon, J. 111-153 Erdos, J. IV-87Cooper, J. W. Il1-515 Escher, W. J. D. 11-31Corporan, E. 1-313 Fackrell, F. N. 111-45Cost, T. L V-287 Faith, W. N. V-115Couch, B. D. 1-235 Fang, J. J. 11-127Counter, M. S. V-177 Ferraro, N. W. IV-511Cox, R. W. VI-43 Fin aa, S. R. 111-207Crump, J. E. 1-119, 1-135 FiY.-., ;. 11-199Culver, D. W. 11-189 Fini-rson, J. C. 111-193Cumbo, F. M. 1-111 Fisher, J. M. 111-285, V-157Cunnington, G. R. 111-261 Ford, R. N. 11-97Curran, F. M. 11-407 Foreman, K M. IV-221Cyran, F. B. V-577 Fortin, R. 111-233Dahlem, V. V-399 France, J. IV-339Dale, M. R. VI-95 Francis, E. C. 1-151, 1-161Dare, A. A. VI-199 Franklin, P. M. 111-323, 111-387Dash, S. M. 1-247 Frazer, C. L. IV-157Davidson, T. F. V-265 Froning, H. D. 11-21Dawley, S. K 111-411, V-329 Frut, G. VI-43Day, R. S. 111-223 Fujimura, C. VI-81Dean, D. L. V-279 Fuller, S. T. IV-125, VI-103DeCoursin, D. G. 1-203 Gage, M. L. V-541DeMay, S. 111-525 Galambos, P. 1-235Der, J. J. 11-333 Gallagher, K. E. IV-61DeSpirito, J. 1-389 Gallet, D. B. 111-447, 111-461Dever, J. 111-113 Garufi, R. IV-433Dhillon, M. 1-69 Garver, L. C. V-41Dill, C. 11-117 Gause, R. L. V-265, V-301Dill, R. A. V-407 Gehris, Jr., A. IV-327Dixon, P. IV-327 Gcigcr, C. J. 111-351Donaldson, W. IV-13 (erson, J. H. 111-113

410

Glaittli, S. R. 111-19 Jacks, T. 111-315Glod. J. E. V-449 Jacobsmeyer, W. E. 1-401Gonzalez, k. F. 1-355 Janicik, T. J. V-265Goodman, W. A. V-521 Jankovsky, R. S. V-571Gord, P. R. V-399 Jay, D. A. VI-95Gott, R. A. 1-171 Jenkins, R. C. IV-221Gottzmann, C. F. V-377 Jensen, G. E. 11-143, IV-365Grabowsky, A. IV.433, IV-447, IV-483 Johnsen, P. T. V-107Graham, K. 111-563 Johnson, S. 111-181Graham, P. H V-145 Johnson, T. C. 111-213, 111-447Graham, R. P. V-287 Johnson, III, W. L V-473, V-483Graham, W. H. 1I1-545, V-I Jones, A. IV-259Grelecki, C. IV-525 Jones, S. C. 1-191Gribben, E. S. VI-51 kacynski, K. J. 11-273Griffiths, S. K. 1-363 Kamholz, M. H. 111-165Guile, R. N. VI-125 Karanian, A. J. IV-145Gupta, M. K IV-447 Kaste, P. J. IV-469Guthrie, D. M. 11-177 Kay, I. W. V1. 125Gutmark, E. 1-369, WV-133, IV-383 Ka- i,,L, J. M. V-557, V-571Haas, L. W. 111-381 kim, S. S. IV-319Haas, M. !V-145 Kirchoff, M. 111-105Haloulakos, V. E. 11-291 Kirschke, E. J. 1-355Hamilton, R. S. 111-361 Knapp, R. L. 111-241Hanson-Parr, D. 111-193 Knapton, J. D. 1-389Harper, S. 1-295 Knight, R. C. IV.397Hawkins, D. K. 111-293 Kocnig, J. R. V-435, V-505Hayes, E. 111-181 Kracutle, K. J. 111-193Heerema, S. W. 1-151 Kraig, A. IV-145Henry, D. L IV-469 Kramer, R. D. 11-73Herrera, W. IV-525 Kublin, T. V. V-30iHerring, E. 111-555 Kwasny, J. 11-407Hcrriott, G. E. 111-371 Landsbaum, E. M 11-225Hewitt, P. W. IV-397 Landulr, L. G. 1-45Hill, C. S. V-367 Langan, K. 1. V-399Hinshaw, J. C. 111-361 Latr.cn, H. M. VI-103Hifne:, J. M. 111-315, V.19 La~ronico, J. E. V-473Hdxlges, J. C. III-573, V-23V Le, M. 111-423Ilo.schclc, G. K. IV-469 Lew, R. E. [-2.35Holzman, A. L. 11-143 Utirigang, .. L. 11-21, V-359Hom'.r, G. D. V-541 Lcisch, S. 0. VI-187Hooper, W. R. V-287 Lcv, B. A. V-215Huang, 1). T. 11.367 LcA-nc, J. 11. V-265H-Jli, P. IV-433, IV-447, IV-483 I.,evinsky, E. S. IV-1]-ltnw•nvk, F. IV-29 Lillcy. J. S. IV-233, IV-259, IV-289Ibrahim, A. M. V-319 Lilly, J.M. 111-371Irizaify, J. IV-511 L"magc, C. R. IV-407Ise, M. 11-107 Little, R. R. V-187Jacks, J. W. 111-55, 111-381 Liu, C. T. 111-423

411

Liu, E. K. S. V-287 Moore, R. D. IV-73Lombrozo, P. C. 11-169 Moran, C. M. 11-299, 11-349Loomis, W. C. V-287, V-301 Morgan, D. B. V-549Loundagin, J. A. VI-153 Morgan, M. E. 111-273Lu, P. IV-433, IV-483 Morgan, R. G. IV-397Lund, G.K. 111-361 Monis, J. T. 111-165Lundin, S. J. VI-115 Morrison, M. W. VI.-61Lynch, R. D. III-563 Mortimer, R. J. 111-131Macpherson, A. K. IV-493, IV-501 Motz, E. J. 111447Macpherson, P. A. IV-501 Moy, S. 1V-483Madsen, C. B. HII-273 Mrozinski, D. P. IV-103Mahorter, L 11-117 Mueller, J. M. 11-217Malone, G. A. V-557 Murfree, Jr., J.A. 111-573, V-239Marchol, P. J. 111-113, Il-475 Myers, J. W. 111-505Marrs, D. M V-115 Myers, R. S. 11-1Marsh, B. P. V-187 Nabity, J. A. VI-135, VI-153, VI-173Marteney, P. J. IV-41 Nance, P. D. I:I-293Martin, Jr., D. L V-167, V-177, V-187 Neal, T. W. 111-285Martin, J.A. A11-73 Ncbilak, J. A. 111-213Mascio, W. A. 1-33 Nejad, A. S. 1-305Mason, B. E. V-115 Neumann, R. D. IV-103Massmann, T. A. V-75 Newey, S. L V-225Mathes, H. B. 1-119, 1-135 Newman, R. W. V-123Matson, J. M. VI-135, VI-153 Nichols, R. V-301Maurice, L Q. IV-353 Norris, F. 111-537May, D. L. V-11, V-47 Nosseir, N. S. 1-369Maykut, A. R. VI-61 Nowakowski, M. 111-153Maynard, D. P. P/-511 Nuismcr, R. J. 111-273McCain, J. W. VI-11 Obcrth, M. H. 111-241McCarty, K. F. 11-359 Ochrle, S. 111-181McClary, W. D. 111-85 Oglevie, R. E. 11-11McDaniels, D. 11-117 Ohler, K. V-435McFarlane, . S. 11-217 Olson, A. M. 111-207, 111-253McGrath, D. K. 111-223, V-347 Oppeit, J. B. 1-227McLafferty, G. H. IV-157 Palaszewski, B. A. 11-251McLendon, S. [V.407 Panclla, E. A. VI-31McMasters, J. R. V-135 Panossian, H. V. 11-63McParland, G. 111-153 Parker, D. P. 111-505Mehta, G. K. 11-97 Pattcrson, J. E. V-197Mello, J. D. 111-461 Pavli, A. J. V-557Melvin, W. S. 111-555, V-31, V-47 Pcngelly, S. L. IV-233Meyers, G. IV-327 Perkins, F. M. 111-77Michaels, R. S. 111-141 Peschke, W. T. IV-177Miller, T. M. 11-291 Peters, S. T. 1-345Miller, W. H. IV-407 Pctley, D. H. 1-191Milton, R. W. V-255 Pcttcrs, D. P. IV-309Mitchell, J. K. V-85, V-89 Phillips, D. 11-199Mooney, W. E. IV-1 Phillipsen, P. C. 1-61

412

Pledger, K. L 111.545 Singh, B. 1-295Pressley, Jr., H. M. V-417 Smith, A. L 111-207Prc:ton, S. B. V-41 Smith, E. IV.125Pritz, W. 11-233 Smith, J. B. I11-233, V-513Quinn, R. W. V-577 Smith, J. R. IV-353Reed, Jr., R. R. 111.315, V.19 Smith, R. A. IV-133, IV.383Reeves, G. P. 1-389 Smith, R. L V-115Renzi, R. F. 1-363 Smith, W. R. IW-157. VI.81Resch, C. L V.-1'3 Spadaccini, L 1. IV-41Reynolds, H. M. 1-69 Stallings, D. W. IV-113Richter, T. IV-423 Stalnake,, R, A. 111-113Rindone, R. R. IV.417 Stcfi... vu P. L. rr.33 ,. "Rizkalla, 0. 1-211, IV-221 Stcrn, A. G.Roberts, W. 1-295 Stevens, J. It. IV-193Robertson, D. IV-423 Stcvcn.A)n, 111, W. 1f.Robinson, K. P. 111-1 Stokes, B. B. I!1-11Roble, N. IV-13 Stokce, E. H. V-497, V-505Rodgcrs, F. C. 11-159 Stonc, W. C. 155Roffc, G. IV-87 Strauss, B. IV-483Rose, J. M. 111.95 Stringer, M. E. V-399Rosenberg, S. D. V.541 Stulcn, R. H. 11-359Rounthui, S, F. 111-487 Sunlinitt, J. V!. IV-29Ross, N. IV-433 Svoh, (. J. 1-45, 111-1Roth, G. J. 111-85 Swiliart,.1. M. 1-1Rouch, Jr., L1. L. V-319 TAlcy, R. [. 1.379Rouse, 0. C. IV.279 Taylor, Jr., It. If. 111-11, VI-IRoy, M. J. 111-131 ll4iwnw. E. 1.. 111-165Runge, J. L 111.487 Thomphon, M. W. IV-157, IV-193Rusck, J. 1-103 Thompson, R. E" 1-151, 1-161Russell, W. D. IV-521 Thorn, L. B. V-255Rychnovsky, R. E. 1.363 Tignac, 1.. L. V-75Sackhcim. R. L. 1!7367 Tingc, S. 1.D. 11-225Sankovic, J. M. 1I-3')1 Torick, Jr., i(. J. 1-33Sanlorcnzo, F. IV-87 Truc, W. (. 111-253Sanscraintc, W, A. 11.307 Turer, A. &D. 111.335, 111-341Sansoy, S. 111-105 ",: Ji, 1). 1.. IV-73Sarmiento, C.J. J1,4(17 Van Wic, 1). NI, IV-15"lSchadow, K. C. 1-369, IV-133, IV-383 Vandivcr, T. .. 1-89Schmidt, W. M. 11-417 VanKiJcck, J. A. IfV!11- , 111.417, 111-475Schubert, Jr., It. J. 111-95 Vaisallo, F. A. 1-321Schuler, A. L. 11-169 Vei, P. W. 1-45Schultz, S. V-67 Vcnlo, 1). M, 11-317Schumtcr, J, R. !1-317 Voccks, G. 1; IV-.511Schwartz.hart, A. 11-281 v4'n I'ragcnou, G. 1.. V-385scahs, W.O0. IV-525 Vondlra, It, J. !1-38";

Secly, J. F. 1-33; Vo•en, S. !(. 1-363Simpson, Jr., J. W. 1-45 Wa1htct, J. 1'. !1-117Singer, C. 11.107 Wi1cAchc, W. II. W. IV-407

413

Walls, T. V1-173Walsh, J. J. 111-165Walsh, R. F. V-53 1Walsh, R. K V.279Wang. T-S. I1-133Ward, R. IV-327Warden, L.R. 111-505, VI-43Warren, L 111-555Wauom, S. R. 111-293, 111.305Wch:r, H. IV.259Weldon, V. 11-199Werling, R. E. 111-213West, L K. 1.61,11-307, 111.261Weyland, H. H. V-337Wharton, W. W. V.239White, M. F. 1-235Whitc, V/. L 111.55Whitesides, R. If. V.407Wilde&, L. B. IV.53Wiley, V. R, 11-169Wiley. S. R. 111-213Wilier, R. L. II1.223, IV.4 AWilliams, A. W. IV-46"Williurns, G. S. V.123Wilson, B. F. 111-141, V.99Wilson, K.J.. IV-133, IV-383WO0un, S. C. V-593Wisian, R. IV-365Withcrsflo)n, F. 1). 1-211Wolf, R. S. 11-177Woods, Jr., I1. V.247Wright, Jr., J. W. IIJ-.73, V-197. V.209)Yagcr, J. 11-339YOgIM, J. J, I-IIYasuhara, W. K. 111-461Yavrouian, A. If. 11-349Yeh, 11, Y. 111.423Yi, A. C. V.367Z-achuy, A.T. 11-89Zrnin4r, B. IV-327Zinn, M. W. IV.303

414

SOURCE INDEX

6595th Test and Evaluation Group 111-85Vandenberg Air Force Base, California

Advanced Technulogy, IncA)rporatcd V-215Arlington, Virgini.:

Aerojet PropuLsion on 1-45, 1-6!. 11-189, 11-241, II-.3R)7, 111-1, 111-113, 111-153,Sacrainc,:o, Calaur:.'. 111-207, 111-213, 111-241, 111-253, 111-261, 111-435, 111-447,

11.461,111-475, 111-487, IV-417. V-225, V-287. V-541, V-549

Acroj:t Propulsitmn 'iiation 1V-233'liuntsviJc, AlaNma

The Ac'4 oxpacc Corpration 11-47, 11-89, 11-225, 11-333L Scggundo. .:'alifocrnia

Air Force a-ademy 11-407Colorado Springs, CAolorado

Allif'd-Signal Aco r;:,,Iacc (C mpany 111-165Tempe, Arizona

American Rocket Company 11-177, 11-217Carnarillo, California

ANAECfIIC Research (.rpolwtion V-287San Diego, California

Army Armament Research. l)cvclophnent, and Fnginrering CA-nICr 1-33,1, IV-423, IV-433,Picatinny Arsenal, New Jersey IV-447, IV-483, IV-493, IV-501, IV-51 I, IV-525

Army Ballistic Re.arch Lab)ratory 1-389, IV-459, IV-469Aberdeen Proving Ground, Maryland

Army Missile Command 1-89, 111-141, 111-181, 111-285, 111-555,Redstone Ascnal, Alabama I1I-573, IV-233, IV-259, IV-289, V-I 1. V-31,

V-47, V-67, V-85, V-99, V-9)9, V-157, V-177,V-187, V-197. V-209. V-219, V-247, V-255, VI-61

Army Strategic l)elcnsc Command 1-55, 111-153, 111-253, 111-505,lluntsvillc, Alabamia Vi-I I, VI-43

Arnold Linginecri'ig I)*velophnleil Center 1-33, V.577Arnold Air Fo rce Blsic. "lcnncshcc

415

Astronautics Laboratory (AFSC) 1-33, 1-103, 11-233, 11-417, 111-105, 111-423Edwards Air Force Base, California

Athena Engineering Company V-287Huntsville, Alabama

Atlantic Research Corporation V-123Camden, Arkansas

Atlantic Research Corporation 11-159, 111-95. 111-411, III-505,111-563, IV-397,Gainesville, Virginia IV-407, V-145, V-265, V-279, V-287, V-329, VI-43

Atlantic Research Corporation 11-307, 11-339, 111-131, VI-51Niagara Falls, New York

Boeing Aerospace and Electronics 11-199, VI-SI, VI-187Seattle, Washington

Brigham Young University 1-119, 1-135Provo, Utah

California Institute of Technology 11-299, 11-349, IV-511Jet Propulsion LaboratoryPasadena, California

Calspan Corporation IV-113Arnold Air Force Base, Tcnncssee

Calspan Corporation 1-321Buffalo, New York

Cohen Professional Serv'iccs Ili-1I, 111-387Redlands, California

Comarco, Incorporated 1-119. 1-135Ridgecrest, California

E. I. du Pont de Nemours and Company IV-469Wilmington, Delaware

Electroformed Nickel, Incorporated V-557Corona, California

Fiber Materials, Incorporated 111-233, V-513Biddeford, Maine

FluiDync Engineering Corporation 1-203Minneapolis, Minnesota

416

Fluorochem, Incorporated V-41Azusa, California

Ford Aerospace Space Systems Division 11-317Palo Alto, California

General Applied Science Laboratories, Incorporated 1-211, IV-87, IV-221Ronkonkoma, New York

General Dynamics Corporation IV-1Dayton, Ohio

General Dynamics Corporation 1-235Fort Worth, Texas

General Dynamics Corporation 11-97, 11-169, 11-317, IV-1San Diego, California

General Electric Aircraft Engines IV-209Cincinnati, Ohio

General Electric Company 1-401, IV-521Pittsfield, Massachusetts

Grumman Corporate Research Center IV-221Bethpage, New York

GT-Devices, Incorporated 1-211Alexandria, Virginia

Hazard Research, Incorporated IV-525Rockaway, New Jersey

Hercules, Incorporated 111-25, 111-35, 111-55, 111-273, V-287, V-417Magna, Utah

Hercules, Incorporated 111-55, 111-381, IV-365, IV-407McGregor, Texas

Hercules, Incorporated 1-171Radford, Virginia

Hercules, Incorporated 111-371Rocket Center, West Virginia

Irvine Tcehnolog,-y Group, Incorporated 11-11Fullerton, California

417

The Johns Hopkins University 1-235, IV-157, IV-193, V-123Applied Physics LaboratoryLaurel, Maryland

Lehigh University IV-493, IV.501Bethlehem, Pennsylvania

Lockheed Engineering and Sciences Company 1-191Hampton, Virginia

Loral Defense Systems V- 135Akron, Ohio

LTV Missiles and Electronics Group 111-505, IV-397, VI-43Dallas, Texas

The Marquardt Company 1-295Van Nuys, California

Martin Marietta Laboratories V-319Baltimore, Maryland

Martin Marietta Space Systems V-593Denver, Colorado

McCronc Associates V-483Westmont, Illinois

McDonnell Douglas Space Systems Company 1121, 11-291Huntington Beach, California

McLafferty Consulting, Incorporated IV-157North Palm Beach, Florida

NASA Headquarters 11-31Washington, DC

NASA Johnson Space Center 11-24 1Houston, Texas

NASA Lingley Research Center 1-191, 11-73. 3,-. •Hampton, Virginia

NASA Lewis Research Centcr 11-251, 11-273, 11-317. ll-31;1Cleveland, Ohio 11-407. IV-29, IV-73, V-557. V-5"11

NASA Marshall Space Flight Center 11-1(07, Ii-117, 11-133. V-3(. . V-585Marshall Space Flight Center, Alabama

418

Natqonal Center for Advanced Technologies I-1Washington, DC

Naval Ordnance Station I-111, 1-345, IV-303, IV-319Indian Head, Maryland

Naval Surface Warfare Center 1-181Dahlgren, Virginia

Naval Surface Warfare Center V-41, VI-199Silver Spring, Maryland

Naval Weapons Center 1-119, 1-135, 1-369, 111-193, 111-315, 111-335, 111-341, 111-351,China Lake, California 111-525, 111-537, IV-125, IV-133, IV-327, IV-339, IV-383, V-19,

V-107, V-115, VI-31, VI-73, VI-103, VI-115, VI-135, VI-153, VI-173

Olin Corporation 1-355St. Marks, Florida

OSU Aeronautical and Astronautical Research Laboratories IV-209Columbus, Ohio

Rockwell International Corporation 11-63, 11-127, 11-281, 111-69, V-75Rocketdyne DivisionCanoga Park, California

Rockwell International Corporation V-367Space Systems DivisionDowney, California

San Diego State University 1-369San Diego, California

Sandia National Laboratories 1-363, 11-359Livermore, California

W. J. Schafer Associates ITArlington, Virginia

W. J. Schafer Associates V-52 1 , i1Calabasas, California

W. J. Schafer Associates V-531Chelmsford, Massachusetts

Science Applications Inlernational Corporation 1-247Fort Washington, Pcnnsylvanil

419

Science Applications International Corporation V-265Huntsville, Alabama

Science Applications International Corporation V-301San Diego, California

Science Applications International Corporation V-287Santa Aria, California

Science Applications International Corporation V-265Webster, Texas

Societe Europeenne de Propulsion VI-43Saint Medard en JallesFrance

Southern Research Institute V-435, V-497, V-505Birmingham, Alabama

Southwest Research Institute IV-525San Antonio, Texas

SPARTA, Incorporated VI-21Huntsville, Alabama

SRS Technologies 11-73, V-407Huntsville, Alabama

Stone Engineering Company 1-55, 111-141. IV-289, V-167, V-177, V-187Huntsville, Alabama

Strategic Defense Initiative Organization 11-383Washington, DC

Sundstrand Power Systems IV-259, IV-279San Diego, California

Svcrdrup Technology, Incorporated 1-33, -V-577Arnold Air Force Base, Tennessce

Sverdrup Technology, Incorporalcd V-557Brookpark, Ohio

Teledyne CAE 1-285Toledo, Ohio

420

Thiokol Corporation 111-19, 111-45, 111-77, 111-293,Brigham City, Utah II-305, 111-361, V-123, V-287, V-435

Thiokol Corporation 111-165, 111-223, 111-515, IV-469,Elkton, Maryland V-55, V-347, V-449, V-473, V-483

Thiokol Corporation 111-11, 111-323, 111-387, 111-545, V-1, VI-1, VI-I1Huntsville, Alabama

TRW Space & Technology Group 11-367Redondo Beach, California

Union Carbide Industrial Gases, Incorporated V-377Tonawanda, New York

United Technologies Corpoa' in 1-69, 1-151, 1-161, 11-143,Chemical Systems Division IV-61, IV-365, V-337, VI-187San Jose, California

United Technologies Corporation IV-145Pratt & WhitneyWest Palm Beach, Florida

United Technologies Research Center IV-41, 1V-145, IV-177, VI-125East Hartford, Connecticut

University of Texas 11-407Austin, Texas

USBI Company I1-1Huntsville, Alabama

Veritay Technology, Incorporated 1-379East Amherst, New York

Williams International IV-251Walled Lake, Michigan

Wright Research and Dcvd!np,ncnt Ccntcr 1-227. '-305. 1-313, 11-11, 11-21,Wright-Patterson Air Force Basc, 01-i IV 13, 1V-29, WV.53, IV-1l3, !V-157, IV-309,

IV..353, V-3,59, V-399. VI-81, VI-95

..,2

SIJBJECI7 INEX2-DINITROBUTANOL

MIXED FORMALS OF 2,2-DINITROPROPANOL AND 2,2-DINITROBUTANOL AS SOLIDPROPELLANT PLASTICIZERS V-41

2-DINITROPROPANOLMIXED FORMALS OF 2,2-DINITROPROPANOL AND 2,2-DINITROBUTANOL AS SOLID

PROPELLANT PLASTICIZERS V-41

3-03-0 ANALYSIS OF SICBM STAGE I BORE TO FIN TRANSITION WITH COMPARISON TO

THE AXISYMMETRIC ANALYSIS 111-193D

LARGE SIZE 30 CARBON/CARBON PREFORM DEMONSTRATION V-513

20HIGH PRESSURE MACH 10 TO 20 ELECTROTHERMAL HYPERSONIC WIND TUNNEL 1-211

22NSTATE-OF-THE-ART 22N (5 LBF) THRUSTER FINDS BROAD APPLICATION 11-339

250DESIGN OF A 250 LBF THRUST HYDRAZINE FUELED AIR TURBORANJET IV-233

320-2MODEL 320-2: A COMPACT ADVANCED UAV TURBOJET 1-285

5055MOISTURE EFFECTS ON ACROSS PLY MECHANICAL AND THERMAL CHARACTERISTICS OF

FM 5055 CARBON PHENOLIC UTILIZING RAPID ISOTHERMAL HEATING V-505

AAAMAAAH FLIGHTWEIGHT COMBUSTOR DURABILITY TESTS IV-365

ABLATIVEABLATIVE INSULATORS FOR RAMJET ENGINES IV-327DEVELOPMENT OF ADHESIVE FOR BONDING COMPOSITE PROPELLANTS TO ABLATIVE

MATERIAL 111-381ABLATIVES

THE RHEOLOGICAL ANALYSIS OF PHENOLIC RESIN USED IN ABLATIVES 111-69ACCELERATED

ACCELERATED PROCESSING OF HTPB PROPELLANTS WITH TMXDI V-67ACCOMPLISHMENTS

OVERVIEW OF NASA SOLID PROPULSION INTEGRITY PROGRAM (SPIP) BONDLINE WORKPACKAGE RESULTS AND ACCOMPLISHMENTS V-301

ACOUSTICTHE EFFECT OF ACOUSTIC DAMPENING DEVICES IN REGENERATIVE LIQUID PROPEL-

LANT GUNS 1-389ACROSS-PLY

WITH-PLY PERMEABILITY OF CARBON PHENOLIC COMPOSITES AS A FUNCTION OFACROSS-PLY COMPRESSIVE LOAD AND TEMPERATURE V-497

ADDITIVELARGE-SCALE MOTOR/HAZARD TESTING OF HIGH PERFORMANCE DENSE ADDITIVE

(Bi203) PROPELLANT (U) VI-73ADHESIVE

DEVELOPMENT OF ADHESIVE FOR BONDING COMPOSITE PROPELLANTS TO ABLATIVEMATERIAL 111-381

AERODYNAMICEVALUATION OF AN AEROGRID FUEL INJECTOR AND OTHER AERODYNAMIC GRID CON-

FIGURATIONS (U) VI-153AEROGRID

FVAIUATION OF AN AEROGRID FUEL INJECTOR AND OTHER AERODYNAMIC GRID CON-fIGURATIONS (U) V1-153

AFLASAFLAS: AN EXTERNAL INSULATION FOR ROCKET MOTORS V-123

AGEDSERVICE LIFE TESTING AND ANALYSIS OF FIELD AGED MLRS MOTORS V-187

AGINGAGING CHARACTERISTICS OF HIGH BURNING RATE CATOCENE CONTAINING PROPEL-

LANTS (U) VI-1AGING CHARACTERISTICS OF PROPELLANTS CONTAINING GAP AND AN 111-351EVOLUTION OF AN AGING PROGRAM - MINUTEMAN STAGE II SOLID ROCKET MOTOR

I-45 1-145AICBM

ADVANCED INTERCONTINENTAL BALLISTIC MISSILE (AICBM) TECHNOLOGIES 111-105AIR-LAUNCHED

PERFORMANCE TRADES FOR AIR-LAUNCIIED TURBOJET AND RAMJET CYCLE HIGH MACHCRUISE MISSILES WITII ROCKET OR TURBOJET BOOST IV- 1

AIR-TO-AIRPRELIMINARY DESIGN Of AN EXTENIDEI) RANGE AIR-1U-AIR MISSILE (U) VI-lbi

AIR-TO-SURFACFPLRIORMANCE ANALYSIS OF A LOW-DRAG, VARIATII.E GEOMLTRY NOZZLE, INTEGRAL

ROCKET RAMJET AIR-TO-SURFACE MISSILE (U) VI-115AIRBORNE

DEVELOPMENTS IN THE DESIGN Of AN Al(BORNE ROIARY AIR SEPARAIOR V-3llAI RBREATtl I NG

AIRIBRLATHING BOOSTER PERFORMANCE OPIIMIiAT ION USING A NICRO(COMPUI!.I( II-

AIRBREATHING (cont'd)NEW AND UPGRADED SUPERSONIC - HYPERSONIC AIRBREATHING ENGINE TEST FACILI-

TIES A. CASL IV-87AIRCRAFT

COOLING SYSTEM AND INSULATION CONCEPT FOR A MACH 5 TURBO-RAMJET AIRCRAFT 1-191ENDOTHERMIC FUELS FOR HIGH-SPEED AIRCRAFT IV-41

AIRFLOWAIRFLOW MOOEL TESTING TO DETERMINE THE DISTRIBUTION OF HOT GAS FLOW AND

O/F RATIO ACROSS THE SPACE SHUTTLE MAIN ENGINE MAIN INJECTOR ASSEMBLY 11-117ALAS

SYSTEM APPLICATIONS FOR AN ADVANCED LIQUID AXIAL STAGE, ALAS 111-447ALES

AIR LIQUEFACTION AND ENRICHMENT SYSTEM (ALES) WITH TURBOROCKET FOR OR-BITAL VEHICLES V-367

ALL-HYDROCARBONALL-HYDROCARBON ORBITAL LAUNCH VEHICLE V-359

ALSDEVELOPMENT OF SCAVENGED PROPELLANTS FOR ALS V-329

AMMONIAA SYSTEM LEVEL ANALYSIS OF THE AMMONIA ARCJET 11-233

AMMUNITIONBALL POWDER PROPELLANT APPLICATIONS TO LARGE CALIBER AMMUNITION 1-355

ANTISUBMARINEVERTICAL LAUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMARINE WAR-

FARE (ASW) V-135AQM-37C

SOLID ROCKET BOOSTER INTEGRATION WITH THE AQM-37C MISSILE TARGET 1-111ARCJET

A SYSTEM LEVEL ANALYSIS OF THE AMMONIA ARCJET 11-233ARCJET NOZZLE AREA RATIO EFFECTS 11-407INVESTIGATION OF THE ARCJET NEAR FIELD PLUME USING ELECTROSTATIC PROBES 11-391

ARTILLERYDEVELOPMENT OF A SAFE LIQUID PROPELLANT LOGISTIC SYSIEM FOR FIELD ARTIL-

LERY IV-11 1ASROC

VERTICAL LAUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMARINE WAR-FARE (ASW) V-135

ASWVERTICAL LAUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMARINE WAR-

FARE (ASW) V-135ATR

ATR - PAST, PRESENT, AND FUTURE IV-13ATTITUDE

DESIGN AND DEVELOPMENT OF A LOW COST LIGHTWEIGHT, FAST RESPONSE ATTITUDECONTROL SYSTEM CNG;INE USING CLF5/N2lH; PROPELLANTS 111-475

DEVELOPMENT OF AN ATTITUDE CONTROL MO1OR FOR THE ERINT-1 MISSILE 111-505AXISYMMETRIC

3-0 ANALYSIS OF SICBM STAGE I BORE TO FIN TRANSITION WITH COMPARISON TOTHE AX;SYMMETRIC ANALYSIS 11-19

PREMIXED, TURBULENT COMBUSTION OF AXISYMMETRIC SUDDEN EXPANSION FLOWS 1-305AZ IDE

CLYCIDYL A/IDE POLYMER (GAP) SOLID GAS GENLRATOR FOR DUCIEO AND HYBRIDROCKET MW)TORS V-47

BALL 15T ICADVANCED INTERCONII NENTAL BALI. ISTIC MISSILE (AICBM) TECIIN"O.OGIES 111-105BALLISTIC MODELING OF TWO-PIECE (;ARTRIDGES FOR A TANK GUN IV-459

BALLISfICSLITERATURE REVIEW OF PLATEAU BALLISTICS IN NONALUMINIZED SOLID PROPEL-

LANTS 11-397PREDICTION CF INFERIOR BALLISIICS AND PRESSURE WAVES PRESENT IN UNIGHARGE

DESIGN IV-447BEARINGS

DAMPING BEARINGS FOR CRYOGENIC TURBOPUMPS V-585912O3

LAKGE-SCALL MOTOR/HAZARD rESTING OF HIGH PERFORMANCE DENSI ADDIT:Vf(11203) PROPELLANT (U) VI-73

B INDERS

SEMI-COMMERCIAL TPE$ FOR LOVA GUN PROPELLANT BINDERS IV-469TOUGH PROPELLANTS FORiMED FROM TETRAFUNCTIONAL BINOERS I11-315

1I PROPEt.L ANTDEVELOPMENT Or A BIPROPEI.ANT PLUME GENCRAIOR SUBSYSTfEM (U) Vl- IMINIMUM SIGNATURE INSENSITIVE MUNITION GEL hIPROPELLANT PROPUI.SIC.,N V-247

BOND-iN-TfNSION1;[CLICIION OF BoDit-IN-IENSION SPTCINEN FHOR TIlE NASA SPIP BONDI *NUS WORK-

PACKAGE V-2ý IHONDFD

A COMPOSI E OVEITWRAPPEDI BONDED R.LLING 0IA.PII.AUM VXPULSION SYSTEM I11-1111BOHDING

DEVCI.O!'MfNT Of ADHlESIVE tOR BONDING COMPOSIlt P!''Pt.Ii ANTS TO APLATIVFMATi.RIAL I111-a1

63NrL IN1DIELECITIC MONITORING 01 BONILINE MAIFRIAI-5 (01( IMPHOVYD PROCCSS (A)1J(I4o V-_'I9

424

BONDLINE (cont'd)NASA SPIP BONDL INE WORK PACKAGE. OVLRVIEW Of THE SPM MODEL PROCESS AND

PROCESS FMEA TASK V-265OVERVIEW OF NASA SOLID PROPULSiCN INTEGRITY PROGRAM (SI'IP) BONDLINE WORK

PACKAGE RESULTS AND ACCOMPLISHMIENTS V-301PREDICTION OF BONOLINE STRENGTH BY UNIQUE SPECIMEN DESIGN AND CUMULATIVE

DAMAGE CONCEPTS V-14~5BONDL INES

SELECTION OF BOND-IN-TENSION SPECIMEN FOR THE NASA SPIP BONGLINES WORK-PACKAGE V-287

BOOSTPERFORMANCE TRADES FOR AIR-LAUNCHED TURBOJET AND RAMJET CYCLE HIGH MACH

CRUISE MISSILES WITH ROCKET OR TURBOJET BOOST IV- 1BOOSTER

AIRBREATIIING BOOSTER PERFORMANCE OPTIMIZATION USING A MICROCOMPUTER 1Il-11N3)E OF THERMAL PROTECTION SYSTEM FOR SPACE SHUTTLE SOLID ROCKET BODSIER Il-iSOLID ROCKET BOOSTEX INTEGRATION WITH THE AQM-37C MIS~SILE TARGET I-11lWEIGHT AND COST ANIALYSIS OF IDENTICAL STAGE BOOSTER SET FOR EXO ATMOSPH-

ERIC :NTERCEPTOR COMPONENT FLIGHT TESTS (Ui) Vf-21BOOSTERS

A UNIQUE HYBRID PROPULSION SYSTEM DESIGN FOR LARGE SPACE BOOSTERS 11-159DESIGN OPTIMIZATION OF GAS GENERATOR HYBRID PROPULSION !300STERS 11-199HYBRID PROPULSION BOOSTERS FOR SPACE LAUNCH VEHICLES 11-169SOLID-LIQUID STAGED COMBUSTION SPACE BOOSTERS 11-189

BORE3-D, ANALYSIS OF SICBM STAGE I BORE TO FIN TRANSITION WITH COMPARISON TO

THE AXISYMMETRIC ANALYSIS 111-19BRIDGE

FEASIBILITY DEMONSTRATION OF AN ELECTRO-OPTICAL SEMICONDUCTOR BRIDGESAFE/ARM/FIRE INITIATION SYSTEM 111-515

BULK-LOADEDDIAGNOSTICS OF IGNITION/COMBUSTION IN A BULK-LOADED LP GUN 1-379

BULLETDAMAGE EFFECTS CAUSED BY MULTIPLE BULLET IMPACT V- 255

BURNHYPERVELOC)TY BY EXTENDED PROPELLANT BURN 1-321SIIELF LIFE DETERMINATION DF HIGH BURN RATE NEPE PROPELLANTr 1;11181

BURN INCAGING CHARACTERISTICS Or HIGH BURNING RATE CATOCENE CONTAINING PROPEL-

[ANTS (U) VI -1BURNING RATE CATALYSIS OF INSENSITIVE MUNITIONS AN PRoPrLLANTS 111-555BURNING RATE ENHANCEMENT BY ELECTROMAGNETIC ALIGNMENT OF STAPLE3 IN SOLID

PROPELL.ANTS V-11

C-C,VARIABIL ITY IN C-C THROA.T FROSION FOR A S;ELECTION OF SOLIV POCKET MOIORS 1141-7

CALIBRATIONTHE CAF IBRATION OF REAL CAS TEST FACILITIES: PROBLEMS AND PROGRESS ON A

COMBUSTION ORIVEN SIIOCK TUNNEL EXPERIMENT IV-103CANISTER

DEVELOPMENT Aht) TE'SlING CF lHE SOLID PRIOPELLANT GAS GENERATOR POWERED

CANNPEACEAEEPER RAIL GARRISON CANISTER ERECTION SYSTEM 111-95

SYSIEM Il[QUIRE#4LNTS AND iNFEGRATION FOR A FULLY AUTOMATED REGENERATIVE-LIQUID PROPELLANT CANNON I V-52 1

CAR13ONCHA14ACTERIZATION OF LCr4-V[NSITY CARBON f IBER REINFORCýO COMPOSITE V-~483DESIGN A!i9 LVLVLLOPMENI Of LCAPBON F S(R WRAPPED NON-LOAD SHARkth.; ýAlP4iS 11 1-1487OEVELDPMFNF Of HIlL ORIRUi 1 LARSON rIGEk/fPOXY MOTOR CASE FOR THlL STARS/

SfARBIRD PROGRAMS 1-69MOI:STURE [FFtUTS ON ACHOSS PLY I4ECIANICAL AND. IIhERMAL CHAFLACIERIISYICS OF

FIA 5055 C;ARFSON PHENOL IC UTILIZING RAPID ISOTHERMAL HEATING V-505WITH-PLY PERMI[ABILITY Of CARBON PHENOLIC COMPOSITES AS A FUNCTION OF

ACROSS-PLY COMPRESSIVE. LOAD AND TIMf'ERAlURE V-419-GAP PION -CAHEBON

I1lIROAT FROSION CIIARACTLIRIZATI'ON Of *4 DIMf.N'.IONAL CARBON-CARISON AT IfIdIUPkI.SSURI S V-44L9

CAR IsON/GARIIONLAHGE SIZE 3D C;AfNHON/CARLTON PREIU14M DEMONSIRATION V-513

CARDON-Cl OTII(,ARF3UN-CIOI'I PIIENOL IC--CONTINUIL) UIVIIOPMLNI AND CHIARACTERIZATION V-435

CARI ItI 101 SIIALLISIIC MOD~l. INC 0f IWO-PI[Cf- CARIRIDGI.S fOR A TANK GUM 1 V-1159

CAS(DI VI I DIMI NT Of lIlT OHILUS 1 CARIIION I 101E.11/[.POXY HOT O GA~if I Oft IlIf SIAHS/

SIARBEIRID PRO)GRAMS 1-69S II N'1R I I L I GiIT MOT Oft CASA Z;I Stll :,;rii 0:, 1 ,ON I VAI UA II ON V -135SUIISGALL MDI OH CASI ANAI-OG FOR I A5 I UOK-011 I INSIMS IT IVI MUM III ON!, IIiT V. ?(19

(;AS ISI FI(MA II ON AND 1([ 5S1I)UAI g;TII Nr;TI IfIN (,ONIOS I I I IIIICKI I MO 101 CA!)I Si I I

GA!) I NGI OW PI'IOJI C I I II HOCK I I MU I O CA% IN!,: I IiAGMI Nu I1 11ON C;ON II'OI I I(;IINUI f(;'e V-/ i5

CATALYSISBURNING RATE CAIALYSIS CIF INSENSITIVE MUNITIONS AN PROPELLANTS 111-555

CATOCENEAGING CHARACTIc,IeiCS OF HIGH BURNING RATE CATOCENE CONTAINING PROPEL-

LANTS (U) WI-11CENTR IFUGE

STATIC AND CENTRIFUGE FORWARD DOME INTEANAL INSULATION CHAR AND 'dSTRU-MENTATION DATA 111-25

CHAMBERCOMPATIBILITY OF COMBUSTION CHAMBER LINERS WITH RP-1 V-541CONTROL OF CHAMBER IfEAT FLUX BY IF'JECTOR DESIGN 11-127

CHAMBERSA LIFE COMPARISON OF TUBE AND CHANNEL COOLING PASSAGES FOR T HRUST CHAN-

BFr'l V-571A TI -'#-UIIMENSIONAL TURBULENT HEAT TRANSFER ANALYSIS FOc ADVANCED TUBULAR

ET THRUST CHAMBERS I11-273..(W M..-HOD OF MAKING ADVANCED TUBE-BUNDLE ROCKET THRUS. CHAMBERS V-557

CHAPNELA LIFE COMPARISON OF TUBF AND CHANNEL COOLING PASSAGEL FOR THRUST CHAM-

BERS V-571CHAPARRAL

CHAPARRAL SERVICE LIFE EVALUATION V- 157CHAR

MODELING INTERNAL INSULATION CHAR AND EROSION IN FORWARD DONE ENVIRON-MENTS 111-35

STATIC AND CENTRIFUGE FORWARD DOME INTERNAL INSULATION CHAR AND INSTRU-MENTATION DATA !11-25

CHLOR INEA VERSATILE CHLORINE PENTAFLUORIDE/HYDRAZINE (CIF5/N2H'i) ALTITUDE ROCKET

ENGINE TEST FACILITY 111-213CLEAN4

CL!!*N PROPCL.' ANT FOR LARGE SOLID ROCKET MOTORS, 11 V-337CE ANER

QUALIFICATION OF A CLEANER PROPELLANT FOR T1 'iTAN RETRO V-34&7CLF 5/N2HI&

A VERSAT ILE CHLOPINE PENTAFLIJORIDE/HYDRAZINE (C.. *i/N2H14) A: TITUDE ROCKETENGINE IEST FACILITY 111-213

DESIGN AND DEVFLOPMENT OF A LOW COST LIGHTWEIGHT, FAST RESPONSE ATTITUDECONTROL SYSIEM ENGINE USING CLF5/N2H4 PROPELLANTS 111-475

DESIG.J AND DEVELOPMENT OF L.IGHTWEIGHT AXIAL PROPULSION ENGINE USING CLF5/N?H4 PROPFLLANTS 111-113

COATEDCHARAGCEKIZATION OF IRIDIUM COATFD RHENIUM USED IN IIIGH-TE1MPERATU:'E,

K?'~iATION-COOLEL) ROCKET THRUSTERS 11-359COLD-SAT

THE COLD-SAT EXPERIMENT rOR CRYOGENIC FLUID MANAGEMENT TECHNOLOGY 11-317COMBUST ION

COMPATIBILITY OF COMBUSTION C.HftMBER LINERS WITH RP-1 V-541DISTRIBUTED COMBUSTION RESPONSE mL"HANISMS 111-193FEASIBILITY STUDY FOR EMPLOYING SOLZ ROCKET COMBUSTION SIMULATORS FOR

SOLID ROCKET MOTOR NOZZLE TES]ING V-h417LIQUID FUEL RAMJET HvDROCARBON FUELS! C~js*:YIc.TIOt4 EVALUAl ION (U) VI-135'.ON-ACOiJSTIC COMBUSTION INSTABILITY IN IIYBRIt ROCKET MOTORS 11-177PASSIVE COMBUS11ION COkIROL IN , OUCTED RO-_KEl %WITF. SIDE MOUNTED INLETS IV-383PFAFMIXED, TURBULENT COMBUSTION OF AXISYMMETRIC SUDDEN EXPANSION FLOWS 1-305SGLID-LIOUID STAGED COMBUSTION SPACE BOOSTERS 11-189STUDY OF SUPERSONIC COMBUSTION PHENOMENOLOGY IN A SUBSCALE COMBUSTOR AT

SIMULATE:D FLIGHT MACHI NINREfkS FROM 7 To 10 I V-209THlE CAL IBRATION OF REAL GAS TEST FACILITIES: PRO)BLEMS AND PROGRESS ON A

COMBIUST ION DRIVEN SHOCK TUNNEL EXPFX114ENT IV-103COMBUS 1OH

AAAM I7LIGHTWL'Gi[ COMF;USIOR DURADILITY TESTS IV-365AN EXPERIMENTAL EVALOATION OF COMBUS70R LINER MATERIALS FOR SOLID FUELL

RAMJET TES-FING 1-227HYDROCARBON-flJ!LfH) SCRAN)ET COMBUJSTOR INVESTIGATION (U) VI-125PROPELLANT CCMBUS7,Of; YSING CHEMICAL KINETICS I V-50 1SlIJOY OF TýsPLRSON'C ýOM4NUSTION PHENOMENOLOGY IN A SUBSCALE COMBUSTOR AT

SIMULATED FLICk1T HACH NUMBERS FROM 7 TO 10 IV-209COI4PuS ITE

A COMPOSITE OVFrýWkAPPED BONDED ROLLING DIAPHRAGM EXPULSION SYSTEM 111-1141CIIARACTURI/ArION OF 1.0W-DENSITY CARBON F IBER REINFORCED COMPOSITE V-1483COMI'0SIt IF ICT(ID ROCKET INLET DFIIlONSTRA1 IUN IV-397GOMPOSIlf OVEITWRAP POSITIVI: EXPUL5ION PHOPELLANT TANKS 11-307DESIGN ANI) DEVFLOPMENT Of A PR!MARY STRUCTURE BASED UPON ADVANCEP CON--

POS#Ti. MATERIALS IýCHNOIOý;IfS 111 -'461DfEVELO:'-IALNI Of ADIIESIVE FOR BONDING COMPOSITE PROPELLANTS 10 ABLATIVE.

MATE H I AL. 111I-351FORMATION AND HI SIDUAI- STRENGTH IN COMPUSITI. POCKET MOTOR CASES 111-213INVISTIGAT INC THE- RATEL FFECT ON THlE CRACK CROWTI; BLi.AVIOR IN A COMPOSITE

.SOI 10) PFUPILL ANT 111-42JANNAF SCANDAFIDIZAT ION OF TENS ION, COjMPHI S5 l0t, AN[! SHEAR TEST METHODS TO

DL.IfHMINf MLCIIANIGAL '-AIATRIA,. PROV'ER) !F'',k Ci~ AMENT WOUND COMPOSITE.1 RUG T UN S I-

.12o

COMPOSITESWITH-PLY PERMEABILITY OF CARBON PHENOLIC COMPOSITES AS A FUNCTION OF

ACROSS-PLY COMPRESSIVE LOAD AND TEMPERATURE V-497COMPRESSION

JANNAF STANDARDIZATION OF TENSION, COMPRCSSION, AND SHEAR TEST METHOOS TODETERMINE MECHANICAL MATERIAL. PROPERTIES FOR FILAMENT WOUND COMPOSITESTRUCTURES 1-89

COMPRESSIVEWITH-PLY PERMEABILITY OF CARBON PHENOLIC COMPOSITES AS A FUNCTION OF

ACROSS-PLY COMPRESSIVE LOAD AND IEMPERATURE V-497COMPUTATIONAL

A COMPUTATIONAL INVESTIGATION OF THE OPERATING CHARACTERISTICS OF AHYPERSONIC INLET IV-157

ADVANCED COMPUTATIONAL MODELS FOR ANALYZING HIGH SPEED PROPULSIVE FLOW-FIELDS 1-247

CONTROLLEDDEVELOPMEN'f OF A PINTLE CONTROLLED VARIABLE THRUST SOLID ROCKET MOTOR 111-241

COOK-OFFSUBSCALE MOTOR CASE ANALOG FOR FAST COOK-OFF INSENSITIVE MUNITIONS TESTS V-209

COOLINGA LIFE COMPARISON OF TUBE AND CHANNEL COOLING PASSAGES FOR THRUST CHAM-

BERS V-571COOLING SYSTEM AND INSULATION CONCEPT FOR A MACH 5 TURBO-RAMJET AIRCRAFT 1-191

COREUNoCHARGE CENTER CORE IGNITION IV-433

CORROSIONSTINGER FLIGHT MOTOR CASE STRESS CORROSION EVALUATION V-85

COStDESIGN AND DEVELOPMENT OF A L.OW COST LIGHTWEIGHT, FAST RESPONSE ATTITUDE

CONTROL SYSTEM ENGINE USING CLFS/N2H4 PROPELLANTS 111-475DEVELOPMENT OF LOW COST PROPULSION REQUIREMENTS 11-97L IQUID PROPELLANT ROCKET ENGINE COMPONENT COST MODEL V-521WEIGHT AND COST ANALYSIS OF IDENTICAL STAGE BOOS1ER SET FOR EXO ATMOSPH-

ERIC INTERCEPTOR COMPONENT FLIGHT TESTS (U) VI-21CRACK

ItIVEST!GATING THE RATE EFFECT ON THE CRACK GROWTH BEIHAVIOR IN A COMPOSITESOLID PROPELLANT 111-423

CRUISEPERFORMANCE tRADES FOR AIR-LAUNCHED TURBOJET AND RANJET CYCLE HIGH MACH

CRUISE MISSILES WITH ROCKET OR TURBOJET BOOS IV-1CRYOGENIC

CRYOGENIIC HYDROGEN-INDUCED AIR LIOUtfAUI ION ILGHINOI OGIES 11-31DAMPING BEARINGS FOR CRYOGENIC TURBOPUMPS V-565THE COLD-SAT EXPERIMENT FOR CRYOGENIC FLUID MANAGEMENT TEC14NOLOGY 11-317

CRYOJETTHE CRYOJET REVISITED 1-295

CYCLEAF/NASA HIGH MACIH COWJINED CYCLE ENGINE DEVEI.OPMENTS IV-29EVALUATION OF GENERIC EXPENDABLE TURBOJET CYCLE DECKS IV-289PERFORMANCE TRADES FOR AIR-LAUNCHED TURBOJET AND RAMJET CYCLE HIGH MACH

CRUISE MISSILES WITH ROCKET OR TURBOJET LOS1 IV-t

DAMAGEDAMAGE EFFECTS CAUSED BY MUL1IPLL BULLET IMPACT V-I5PREDICTION OF BONOLINE STRENGTH BY UNIQUE SPE'CIMEN DESIGN AND CUMULATIVE

DAMAGE CONCEPTS V- 145DAMPENING

1I1E EIFECT O|f ACOUSTIC DAMPENING DEVICES IN REGENt LIQUID PROPEL-IANT GUNS 1-389

DAMPINGDAMPING BEARINGS FOR CRYOGENIC TURBOPUMPS V-585STRUCTURAl. DAMPING/ACOUSTIC ATTENUATION OPTIMIZATION VIA NON-OBSTRUCTIVE

PARTICLE DAMPING 11-63DAMPING/ACOUSTIC

STRUCTURAL DAMPING/ACOUSIIC ATTENUATION OPTIMIZATION VIA NON-OBSTRUCTIVEPARTICLE DAMPING If-63

DEMONSTRATIONSI. IGHTWEIGHT KINETIC VEHICLE FREE FLIGIIT/IIOVER DFMON•THATIONS V-75

DEMONSIRATORTIE JOINT EXPI dDBLE TURBINE ENGINE CON(EPT (JETEC) DEMONSTRATOR PROGRAM

(U) VI-95DENS"

tARG: -!CALE MOTOR/HAZARD TESTING OF HIGH PER(ORMANCE DENSE ADDITIVE(fHiU3) PROPELI.ANT (Ul VI-73

DEPLE I IONREACI ION KINETIC MODEL ING Of STABII IZER DEPI lETION IN M6 PROPELLANT IV-423

DES IGNA UNIQUE HYBRID PROPULSION SYSTEM DESIGN FOk LARGE SPACE BOOSTERS 11-159CONTROL OF CHAMBER HEAT fLUX BY INJCGTOR DESIGN I1-11?DESIGN AND DEVELOPMENT Of A LOW COT l IGHITWEIGIIT, IAST RESPONSE ATTITUDE

CONTROL SYSTEM ENGINE USING CLF51/N2H4 PROPLILANIS 111-475DESIGN AND DEVELOPMENT OF A PRIMARY STRUCTIURI BASED UPON ADVANCED COM-

"OSITE MATERIALS TECHNOLOGIES I1 -4161

'1.'!

DESIGN (cont'd)DESIGN AND DEVELOPMENT' OF CARBON FIBER WRAPPED NON-LOAD SHARING TANKS 111-487DiESIGN AND DEVELOPMENT OF LIGHTWEIGHT AXIAL PROPULSION ENGINE USING CLF5/

N2H4 PROPELLANTS 111-113GE..IGN OF A 250 ISV THRUST HYDRAZINE FUELED AIR TURBORANJET IV-233OL%,tV4 OPTIMIZATION OF GAS GENERATOR HYBRID PROPULSION BOOSTERS 11-199OEVFLO)PMENTb IN THE (DESIGN OF AN AIRBORNE ROTARY AIR SEPARATOR V-377FIBER 41PTIC ORDNANCE SYSTEMS: DESIGN, PERFORMANCE, AND TESTING 111-85LEAP z;OLIO PROPULSION DIVERT SUBSYSTEM DESIGN 11I1-165MECHANICi'L DESIGN Of SURFACE LAUNCHED TACTICAL MISSILES 1-181ORBIT IPANSFER VEHICLE PROPULSION DESIGN: TRADES AND COMPARISONS 11-291PREDICTION~ Of BONDLoNr STKrNGTH BY UNIQUE SPECIMEN DESIGN AND CVI 'LATIVE

DAMAGE LONCEPTS V-145PREDICTION Of INTERIOR BALLISTICS AND PRESSURE WAVES PRESENT IN Li, JIARGE

OEbiION OV-44I7PRELIMINARY 0(81CN AND PERFORMANCE Of AN ADVANCED RAMJET MISSILE CONCEPT

FOR THE CIRCA-2010 OUTER AIR BATTLE (U) VI-103PRELIMINARY OESo('N OF AN EXTENDED RANGE AIR-TO-AIR MISSILE ()VI-187THE IMPACT OF DESiGN OPTIMISM ON THE PERFORMANCE OF A RAMJC M'ISSILE IV-125TRAJECTORY/ENCROY MANAGEMENT DESIGN CONSIDERATIONS IN PROPULSION STUDIES

OF P&D1 BATTL.E SPACE 111I-435iETONAT ION

HEATS OF EXPLOSION, DETONATION AND REACTION PRODUCTS: THEIR ESTIM4ATIONAND RELATION TO THE FIRST LAW OF THERMODYNAMICS 1-345

IMPLICATIONS OF THE RESULTS OBTAINED IN THE GRUMMAN DE7ONATION SHOCKTUNNEL NOZZIE E FXPERIMENTS ON SCRAAJET PERFORMAN4CE OV-221

DIAPHRAGMA COMPOSITE OVCRWRAPPED 6ONDrD ROLLING DIAPHRAGM EXPULSION SYSTEM 111-141

01(1 ECIFRICOIILIECTRIC M4ONITORING 0f BONDLINC MATERIALS FOR IMPROVED PROCESS CONTROL V-279

O IVE'rft.IGHTWCIGHI DIVERT PROPULSION SYSTEM4 140T.FiI: TEST 1-61LIAP SOLID PROPULSION DIVERT SUBSYSTEM Ol~biON I11-165

DOMEAfl DOME INTERNAL INSULATION EROSION ON SMALL ICBN PR17-11SO AND FID STAGE

I SOL ID PROPELLANT ROCKET MOTORS 1II -45MODELINU INTERNAL INSULATION CHAR AND EROSION IN FORWARD DOME ENVIIRON-

MENTS III -35STATIC Awnf CENTRIrUGE FORWARD DOME INTEPRNAL INSULATION CHAR AND INSTRU-

MENTATION DATA 1 11-25DUC TED

COMPO.iTE DIICTLO ROCKET iNLLT U1.0UNbIIIAI IO1 IV-397GLYCIDYL AZIDE POLYMER (CAP) SOLID GAS GENERATOR FOP0 DUCTED AND HYBRID

ROCKET MO0TORS V- 47PASSIVE COMOGUST ION CONTROL IN A OUCTFU ROCKET WITH SIDE M4OUNTED INLETS IV-383THE DEVELOPMENT OF A I IQUID FUELED GAS GENEIRA(OR FOR DUCTfO ROCKET

RESEA14CH 1-113DUCTED-ROCKET/RANJE F

VARIABLE-FUEL.-FLOW DUCTED-ROCKET/RANJET PROGRAM STATUS IV-407DYNANIC/COMBUST ION

FLUID DYNAMIC/CCMBIJSAION INTERACTIONS AS DRIVING MECHANISM OF PrESSURrOSC I L IAl IONS I N A 91.Of i1:RAT IVC 1.I1QU ID PROPELLAN T GUN 1-30

EARTH-TO-01111 I IEVALUATION Of PROPI'SrU. kO(IKET CNOINES FOR EARTH-TV-ORBIT VI~h,(;l( 11-73IMPACT Of PROPIASIVE ADVANCFMrNI$ ON CAPABILITIES OF REU'jABLC: LAMTHi-fO-

ORBIT SliIIb 11-71EL ECTR IC

ANAI YSIS 0I EXPENDABLIF 5ULAR ELECTRIC ORBIT TRANSILM V1.1110.111 1 1-417SDlO EtI*CTRIC PROPULSION REQUIHEMENTS 11-383

CLECIRO-Ol' FICALFEASIBIL-ITY DEMONSTRATION or AN CLECTRO-OPTICAL SEMICONDUCTOTI BROClf

SAFE/ARM/FIRE INITIATION SYSTEM 1I1-515E.LECTROMAGNE TIC

BURNING RATE ENHANCEME.NT BY UL.ECTROMACNETIC ALIGNMENT Or 51AI'L(S IN i;,.l)tPkOPEILANTS V-1l

ELECTROSTATICINVESTIGATION OF THE AftCjLJ NEAR FIELD FLUME USING FrLTHJOIJATI(; PP0IiLS 11-391

ELECTROTHERMAL.HIGH PRESSURE MACH 10 T0 20 ELECTROTHERMAL HYPERSONIC WIND TUNNEL 1-211

EL.LIPTICPURFORNAhCL CHARACTERISTICS OF SHROUDED SUPERSONIC RECTANGULLAR AND

TAPERED ELLIPTIC JtTS IV-133ENDOATMOSPHER IC

ADVANCING (NDOATMOSPHERIG INTERCIPTOR SOLID ROCKET ISOTOR jLINM) TECHNOLOGY(U) VI- 11

ENDOTHERMI CENDOTHIERNIC FUELS FOR HIGHI-SPLED AIRCRAFT I V-41THE AIR FORCE ENDOTHLRMIC FUEL PROGRAM IV-51

ENLIROET ICENERGETIC MINIMUM SMOKE PROPEI.LANTS FOR IN [VALUAT ION V-P25PRIOPULSION SYSTEM HAZARDS AND ENERGETIC INGREDIENTS SYNTHESIS EVALUATION V-I

428

ENERGYAPPROACH FOR LOW EXPONENT AND TEMPERATURE SENSITIVITY IN HICH ENERGY

PROPELLANTS I111ENGINE

A VERSATILE CHLORINE PENTAFLUORIDE/HYORAZINE (CIF5/N2H4) ALTITUDE ROCKETENGINE TEST FACILITY 11.-214

AF/NASA HIGH MACH COMBINED CYCLE ENGINE DEVELOPMENTS IV-29AIRFLOW MODEL TESTING TO DETERMINE THE DISTRIBUTION OF HOT GAS FLOW AND

O/F RATIO ACROSS THE SPACE SHUTTLE MAIN ENGINE MAIN INJECTOR ASSEMBLY 11-117AN ADVANCED LOW-COST ENGINE FOR HIGH-THRUST APPLICATIONS 11-89DESIGN AND SEVLLOPMENT OF A LOW COST LIGHTWEIGHT, FART RESPONSL AlTITUDE

CONTROL SYSTEM ENOINE USING CLF5/N2H4 PROPELLANTS 111-475DESIGN AND DEVELOPMENT OF LIGHTWEIGHT AXIAL PROPULSION ENGINE USING CLF5/

N2H4 PROPELLANTS i11-113DEVELOPMENT Of A HIGH-ENTHALPY TEST CAPABILITY USING A LIQUID-PROPELLANT

ROCKET ENGINE IV-113GENERIC HIGH-SPEED ENGINE TESI RESULTS AT MACH 4.3 AND 5.0 IV-193LIQUID PROPELLANT ROCKET ENGINE COMPONENT COST MOOEL V-521NEW AN) UPCRAD(O SUPERSONIC * HYPERSONIC AIRBRCATHING ENGINE TEST FACILI-

TIES AT GASL IV-87SERVICE LIFE ASSESSNCNT PROGRAM FOR THE HARPOON J402-CA-400 ENGINE IV-303TEST REXULTS Of THE MODIFIED SPACE SHUTiLE MAIN ENGINE AT THE MARSHALL

SPACE FLIGHT CENTER TECHNC.OGY TEST BED FACILITY 11-107THL IMPACT Of MARGIN ON THE ADVANCED LAUNCH SYSTEM LIQUID PROPELLANT

ENGINE V-531THE JOINT LXPLNDABLE TURBINE ENGINE CONCEPT (JETEC) DEMONSTRATOR PROGRAM

(U) VI-95UPRATLD OMS ENGINE STATUS - SEA LEVEL TESTING RESULTS 11-241

ENGINrEABLATIVE INSULATORS FOR RAMJET ENGINES IV-327ALTITUDE SIMULATION TESTING FACILITY CONCEPTS FOR LARGE LIQUID-PROPELLANT

ROCKET ENGINFS V-577EVALUATION Of PROPOSED ROCKET ENGINES FOR EARTY-TO-ORRIT VEHICLES 11-73PREOIGTION OF TOIL THRUST PERFORMANCI AND THE FLOWFI(LD OF LIQUID ROCKET

ENGINES 11-133ENRICHMENT

AIR LIQUEFACTION AND ENRICHMENT SYSIEM (ALES) WITH TURSOROCKET FOR OR-BITAL VEHICLES V-367

N NVIHONMENTSMODELING INTERINAL INSULATION CHAR AND EROSION IN FORWARD DOME ENVIRON-

MENTS 111-35oUII IF"RIUM

THIE INFLUENCE OF PROPILLANr EQUILIBRIUM MODULUS ON kOCKLT MOTOR SERVICELIFE PREDICTIONS V-167

LRI.C' IONDEVELOPMENT AND TESTING OF THE SOLID PROPELLANT GAS CENERATOR POWERED

PEACEKEEPER RAIL GARRISON CANISTER ERECTION SYSTEM 111-95ERINT-I

DEVELOPMENT OF AN ATTITUDE CONTROL MOTOR FOR THE ERINT-1 MISSILE I11-5O0LRINT-1 bOLID ROCKET MOTOR RCQUIREMENT5 AND STATUS (U) VI-43

ERO3IONAFT DOME INTERNAL INSULATION EROSION ON SMALL ICOM PRE-FSD AND FSO STAGE

I 3OLID PROPELLANT ROCKET MOTORS IlI1-45MOOELING INTERNAL INSULATION CHAR AND EROSION IN FORWARD DOME ENVIRON-

MLNTS 111-35THROAT IROSION CHARACTERIZATION Of 4 DIMENSIONAL CARSON-CARSON AT HIGH

PRLSSURES V-449VARIADILITY IN C-C THROAT EROSION FOR A SELECTION Of SOLID ROCKET MOTORS 111-77

EUR(iPI ANA SUMMARY OF EUROPEAN AND JAPANESE HYPERSONIC FACILITY ACTIVITIES 1-203

EXPIOSIONHEATS OF EXPLOSION, DETONATION AND REACTION PRODUCTS: THEIR ESlIMATION

AND RELATION TO THE FIRST LAW OF THERMODYNAMICS 1-345EXPLOSIVCS

CHARACTERIZATION OF INSENSITIVE HIGH EXPLOSIVES DEVELOPED WITH PROPELLANT1.0CIIN.OOY 111-563

LASER INITIATION Of PROPELLANTS AND EXPLOSIVES V-11lEXPONENT

APPROACH FOR LOW EXPONENT AND TEMPERATURE SENSITIVITY IN HIGH ENERGYPROPELLANTS II1-11

EXPULSIONA COMPOSITE OVEHRI1APP[O BONDED ROLL ING DIAPHRAGM EXPULSION SYSTEM 111-141COMPOSITE OVERWAAP POSITIVE EXPULSION PROPELLANT TANKS 11-307

LXTINCTIONDEVELOPMENT OF AN HTPO PROPELLANT WITHl ENHANCED EXTINCTION CHARACTERIS-

TICS V-19

FACiLII IESNEW AND UPGRADED SUPERSO)NIC - HYPERSONIC AIRBREATHINU ENGINE TEST FACILI-

TIES AT GASL IV-67THE CAL IBRATION Of REAL GAS TEST FACILITIES: PROBLEMS AND PROGRESS ON A

COMP4USTION DRIVEN SHOCK TUNNEL EXPERIMLNT IV-103

429

FAC IL ITYA SUMMARY OF EUROPEAN AND JAPANESE HYPERSONIC FACILITY ACTIVITIES 1-203A VERSATILE CHLORINE PENTAFLUORIDE/HYDRAZINE (CIF5/N2H4) ALTITUDE ROCKET

ENGINE TEST FACILITY 111-213AL!ITUDE SIMULATION TESTING FACILITY CONCEPTS FOR LARGE LIQUID-PROPELLANT

ROCKE T ENGINES V-577TEST RESULTS OF THE MODIFIED SPACE SHUTTLE MAIN ENGINE AT THE MARSHALL

SPACE FLIGHT CENTER TECHNOLOGY TEST BED FACILITY 11-107FAN

NASA LEWIS SUPERSONIC THROUGHFLOW FAN PROGRAM IV-73FIBER

CHARACTERIZATION OF LOW-DENSITY CARBON FIBER REINFORCED COMPOSITE V-483DESIGN AND DEVELOPMENT OF CARBON FIBER WRAPPED NON-LOAD SHARING TANKS 111-487DEVELOPMENT OF THE FIBER OPTIC GUIDED SKIPPER ROCKET MOTOR (U) VI-31FIBER OPTIC ORDNANCE SYSTEMS: DESIGN, PERFORMANCE, AND TESTING 111-85

FIBER/EPOXYDEVELOPMENT OF THE OFIBUS 1 CARBON FIBER/EPOXY MOTOR CASE FOR THE STARS/

STARBIRD PROGRAMS 1-69FIBER-MATRIX

VARYING FILAMENT-WOUND NOZZLE PROPERTIES USING SELECTED FIBER-MATRIXLAYERING V-473

FILAMENTJANNAF STANDARDIZATION OF TENSION, COMPRESSION, AND SHEAR TEST MLTHOOS TO

DETERMINE MECHANICAL MATERIAL PROPERTIES FOR FILAMENT WOUND COMPOSITESTRUCTURES 1-89

FILAMENT-WOUNDVARYING FILAMENT-WOUND NOZZLE PROPERTIES USING SELECTED FIBER-MATRIX

LAYERING V-473F ILLETS

A LINEARIZED THEORY FOR UNSTEADY FLOW ALONG VANE-FORMED FILLETS DRIVEN BYSURFACE TENSION 11-333

IIN3-0 ANALYSIS OF SICBM STAGE I BORE TO FIN TRANSITION WITH COMPARISON TO

TOE AXISYMMETRIC ANALYSIS 111-19ILIGHT

STINGER FLIGHT MOTOR CASE STRESS CORROSION EVALUATION V-85STUDY OF SUPERSONIC COMBUSTION PHENOMENOLOGY IN A SUBSCALE COMBUSrOR AT

SIMULATED FLIGHT MACH NUMBERS FROM 7 10 10 IV-209TEST RESULTS OF THE MODIFIED SPACE SHUTTLE MAIN ENGINE AT THE MARSHALL

SPACE FLIGHT CENTER TECHNOLOGY TEST BED FACILITY 11-107WEIGHT AND COST ANALYSIS OF IDENTICAL STAGE BOOSTER SET FOIt EXO ATMOSPH-

ERIC INTERCEPTOR COMPONENT FLIGHI TESTS (U) VI-21FLIGHT/HOVLR

LIGHTWEIGHT KINETIC VEHICLE FREE FLIGHT/HOVER DEMONSTRATIONS V-75FLIGHTWEIGIHT

AAAM ILIGIITWEIGIIT COMBUSTOR DURABILITY TESTS IV-365FLICHTWEIGHT DIVERT PROPULSION SYSTEM HOT-FIRE TEST 1-61

FLOWA LINEARIZED THEORY FOR UNSTEADY FLOW ALONG VANE-FORMED FILLETS DRIVEN BY

SURFACE TENSION 11-333AIRFLOW MODEL TESTING 10 DETERMINE THE DISIRIBUTION OF HOT GAS FLOW AND

O/F RATIO ACROSS THE SPACE SHUTTLE MAIN ENGINE MAIN INJECTOi ASSEMBLY 11-117FLOW FIELD FNVIRONMENT IN THE NASA TWO-INCH TEST MOTOR V-407

FLOWFIELDPREDICTION OF THE rHRUST PERFORMANCE AND THE FLOWFIELD OF LIQUID ROCKET

ENGINES 11-133FLOWFIELDS

ADVANCLE COMPUTATIONAL MODELS FOR ANALYZING HIGH SPEED PROPULSIVE FLOW-FIELDS 1-247

FLOWSAN EFFICIENT, INTELLIGENT SOLUTION FOR VISCOUS FLOWS INSIDE SOLID ROCKET

MOTORS 11-47PREMIXED, TURBULENT COMBUSTION OF AXISYI4METRIC SUDDEN EXPANSION FLOWS 1-305

FLUIDFLUID DYNAMIC/COMBUSTION INTERACTIONS AS DRIVING MECHANISM OF PRESSURE

OSCILLATIONS IN A REGENERATIVE LIQUID PROPELLANT GUN 1-369THE COLD-SAT EXPERIMENT FOR CRYOGENIC FLUID MANAGEMENT TECHNOLOGY 11-317THE STS FLUID MANAGEMENT DEMONSIRATION TEST BED V-593

FLUXCONTROL OF CHAMBER HEAT FLUX BY INJECTOR DESIGN 11-127

FMEANASA SPIP BONDLINE WORK PACKAGE: OVERVIEW OF THE SRM MODEL PROCESS AND

PROCESS FMEA TASK V-265FORMALS

MIXED FORMALS OF 2,2-DINITROPROPANOI AND 2,2-DINITROBUTANOL AS SOLIDPROPELLANT PLASTICIZERS V-41

FORMEDTOUGH PHOPELLANTS FORMED FROM TETRAFUNCTIONAL BINDERS 111-315

FOURINVESTIGATION OF A FOUR POINT 1ES1 FOR THE STRUCTURAL ANALYSIS EVALUATION

OF PIIENOLIC MATERIALS 111-285FRAGMAT

FRAGMAT TEST MODIFICATIONS 111-537

4II0

FRAGMENTATIONLOW PROJECTILE ROCKET MOTOR CASING: FRAGMENTATION CONTROL TECHNOLOGY V-215

FSDAFT DONE INTERNAL INSULATION EROSION ON SMALL ICBM PRE-FSD ANM FSD STAGEI SOLID PROPELLANT ROCV[T MOTORS 111-45

FUELAN EXPERIMENTAL EVALUATION OF COMBUSTOR LINER KATERIALS FOR SY.ID FUEL

RAMJET TESTING 1-227DEVELOPMENT AND TESTING OF A HIGH-ENERGY LOW-COST SOLID RAMJET FUEL AT

THE NAVAL WEAPONS CENTER (U) VI-173EVALUATION OF AN AEROGRID FUEL INJECTOR AND OTHER AERODYNAMIC GRI) CON-

FIGURATIONS (U) VI-153LIQUID FUEL RAMJET HYDROCARBON FUELS: COMBUSTION EVALUATION (U) Vi-135THE AIR FORCE ENDOTHERMIC FUEL PROGRAM IV-53

FUELSENDOTHERMIC FUELS FOR HIGH-SPEED AIRCRAFT IV-41LIQUID FUEL RAMJET HYDROCARBON FUELS: COMBUSTION EVALUATION (U) VI-135

GAPAGING CHARACTERISTICS OF PROPELLANTS CONTAINING GAP AND AN 111-351GAP CHARACTERIZATION 111-361GLYCIDYL AZIDE POLYMER (GAP) SOLID GAS GENERATOR FOR DUCTED AND HYBRID

ROCKET MOTORS V-47GARRISON

DEVELOPMENT AND TESTING OF THE SOLID PROPELLANT GAS GENERATOR POWEREDPEACEKEEPER RAIL GARRISON CANISTER ERECTION SYSTEM 111-95

GASAIRFLOW MODEL TESTING TO DETERMINE THE DISTRIBUTION OF HOT GAS FLOW AND

O/F RATIO ACROSS THE SPACE SHUTTLE MAIN ENGINE MAIN INJECTOR ASSEMBLY 11-117DESIGN OPTIMIZATION OF GAS GENERATOR HYBRID PROPULSION BOOSTERS 11-199DEVELOPMENT AND TESTIG OF THE SOLID PROPELLANT GAS GENERATOR POWERED

PEACEKEEPER RAIL GARRISON CANISTER ERECTION SYSTEM 111-95GLYCIDYL AZIOE POLYMER (GAP) SOLID GAS GENERATOR FOR DUCTED AND HYBRID

ROCKET MOTORS V-47REAL GAS DATA REDUCTION METHODS FOR RAMJET TESIING IV-353STOICHIOMETRIC GAS GENERATOR - A STRATEGIC DEPARTURE V-549THE CALIBRATION OF REAL GAS TEST FACILIT!ES: PROBLEMS AND PROGRESS ON A

COMBUSTION DRIVEN SHOCK TUNNEL EXPERIMENT IV-103THE DEVELOPMENT OF A LIQUID FUELED GAS GENERATOR FOR DUCTED ROCKET

RESEARCH 1-313VARIABLE-FLOW GAS GENERATOR IV-339

GASt.NEW AND UPGRADED SUPERSONIC - Hw'PLRSONIC AIROREATHING ENGINE TEST FACILI-

TIES AT GASL IV-87GEL

MINIMUM SIGNATURE INSENSITIVE MUNITION GEL BIPROPELLANT PROPULSION V-247GENERATOR

DESIGN OPTIMIZATION OF GAS GENERATOR HYBRID PROPULSION BOOSTERS 11-199DEVELOPMENT AND TESTING OF THE SOLID PROPELLANT GAS GENERATOR POWERED

PEACEKEEPER RAIL GARRISON CANISTER ERECTION SYSTEM 111-95DEVELOPMENT OF A BIPROPELLANr PLUME GENERATOR SUBSYSTEM (U) VI-51GLYCIDYL AZIDE POLYMER (GAP) SOLID GAS GENERATOR FOR DUCTEU AND HYBRID

ROCKET MOTORS V-47STOICHIOMETRIC GAS GENERATOR - A STRATEGIC DEPARTURE V-549THE DEVELOPMENT OF A LIQUID FUELED GAS GENERATOR FOR DUCTED ROCKET

RESEARCH 1-313VARIABLE-FLOW GAS GENERATOR IV-339

GLYCIDYLGLYCIDYL AZIDE POLYMER (GAP) SOLID GAS GENERATOR FOR DUCTED AND HYBRID

ROCKET MOTORS V-47GRADIENT

PRESSURE GRADIENT EFFECTS IN SUPERSONIC SHEAR LAYERS IV-177GRID

EVALUATION OF AN AEROGRID FUEL INJECTOR AND OTHER AERODYNAMIC GRID CON-FIGURATIONS (U) VI-153

GRUMMANIMPLICATIONS OF THE RESULTS OBTAINED IN THE GRUMMAN DETONATION SHOCK

TUNNEL NOZZLE EXPERIMENTS ON SCRAMJET PERFORMANCE IV-221GUIDANCE

MINIATURIZED GUIDANCE PROPULSION FOR VOLUME-LIMITED GUN-LAUNCHED INTERCE-PTOR PROJECTILES 111-261

GUNA HYDRODYNAMIC CHEMISTRY MODEL OF THE LIQUID PROPELLANT GUN IV-493BALLISTIC MODELING OF TWO-PIECE CARTRIDGES FOR A TANK GUN IV-459DIAGNOSTICS OF IGNITION/COMBUSTION IN A BULK-LOADED LP GUN 1-379FLUID DYNAMIC/COMBUSTION INTERACTIONS AS DRIVING MECHANISM OF PRESSURE

OSCILLATIONS IN A REGENERATIVE LIQUID PROPELLANT GUN 1-369INSENSITIVE GUN PROPELLANT - MODIFIED SINGLE BASE PROPELLANT IV-483REGENERATIVE LIQUID PROPELLANT GUN AND SELF PROPELLED HOWITZER INTEGRA-

TION 1-401SEMI-COMMERCIAL TPEs FOR LOVA GUN PROPELLANT BINDERS IV-469UNDERWATER WARHEAD GUN SYSTEM (U) VI-199

431

GUN-LAUNCHEDMINIATURIZED GUIDANCE PROPULSION FOR VOLUME-LIMITED GUN-LAUNCHED INTERCE-

PTOR PROJECTILES 111-261GUNS

THE EFFECT OF ACOUSTIC DAMPENING DEVICES IN REGENERATIVE LIQUID PROPEL-LANT GUNS 1-389

HARPOONSERVICE LIFE ASSESSMENT PROGRAM FOR THE HARPOON J402-CA-4OO ENGINE IV-303

HAWKRELIABILITY ANALYSIS OF THE HAWK MOTOR SERVICE LIFE PREDICTIONS V-177

HAZARDHAZARD CLASSIFICATION OF LIQUID PROPELLANTS IV-525

HAZARDSPROPULSION SYSTEM HAZARDS AND ENERGETIC INGREDIENTS SYNTHESIS EVALUATION V-1

HEATA THREE-DIM4ENSIONAL TURBULENT HEAT TRANSFER ANALYSIS FOR ADVANCED TUBULAR

ROCKET THRUST CHAMBERS 11-273CONTROL OF CHAMBER HEAT FLUX BY INJECTOR DESIGN 11-127

HEATINGMOISTURE EFFECTS ON ACROSS PLY MECHANICAL AND THERMAL CHARACTERISTICS OF

FM 5055 CARBON PHENOLIC UTILIZING RAPID ISOTHERMAL HEATING V-505HEATS

HEATS OF EXPLOSION, DETONATION AND REACTION PRODUCTS: THEIR ESTIMATIONAND RELATION TO THE FIRST LAW OF THERMODYNAMICS 1-345

HELIUMHELIUM SATURATION OF LIQUID PROPELLANTS 11-349

HELLFIREJOINT ARMY/NAVY INSENSITIVE MUNITIONS HELLFIRE MOTOR TECHNOLOGY PROGRAM 111-573

HIGH-ENTHALPYDEVELOPMENT OF A HIGH-ENTHALPY TEST CAPABILITY US!WG A LIQUID-PROPELLANT

ROCKET ENGINE IV-113HIGH-PERFORMANCE

HIGH-PERFORMANCEo REDUCED-SMOKE 7POPELLANT 111-335HIGH-SPEED

ENOOTHERMIC FUELS FOR HIGH-SPEED AIRCRAFT IV-41GENERIC HIGH-SPEED ENGINE TEST RESULTS AT MACH 4.3 AND 5.0 IV-193

HIGH-TEMPERATURECHARACTERIZATION OF IRIDIUM COATED RHENIUM USED IN HIGH-TEMPERATURE,

RADIATION-COOLED ROCKET THRUSTERS 11-359HIGH-THRUST

AN ADVANCED, LOW-COST ENGINE FOR HIGH-THRUST APPLICATIONS 11-89HOT-FIRE

FLIGHTWEICHT DIVERT PROPULSION SYSTEM HOT-FIRE TEST 1-61HOWITZER

REGENERATIVE LIQUID PRnPELLANT GUN AND SELF PROPELLED HOWITZER INTEGRA-TION 1-401

HTPBACCELERATED PROCESSING OF HTPB PROPELLANTS WITO TMXDI V-67DEVELOPMENT OF AN HTPB PROPELLANT WITH ENHANCED EXTINCTION CHARACTERIS-

TICS V-19PROPELLANT PROCESS CONTROL OPTIMIZATION STUDY WITH HTPB PROPELLANT 111-371

HYBR IDA SYSTEM FOR THE TESTING OF HYBRID ROCKET MOTORS 11-217A UNIQUE HYBRID PROPULSION SYSTEP DESIGN FOR LARGE SPACE BOOSTERS 11-159DESIGN OPTIMIZATION OF GAS GENERATOR HYBRID PROPULSION BOOSTERS I;-199GLYCIDYL AZIDE POLYMER (CAP) SOLID GAS GENERATOR FOR DUCTED AND HYBRID

ROCKET MOTORS V-47HYBRID PROPULSION BOOSTERS FOR SPACE LAUNCH VEHICLES 11-169HYBRID PROPULSION TECHNOLOGY PROGRAM 11-143NON-ACOUSTIC COMBUSTION INSTABILITY IN HYBRID ROCKET MOTORS 11-177

HYDRAZINEDESIGN OF A 250 LBF THRUST HYDRAZINE FUELED AIR TURBORAMJET IV-233MONOPROPELLANT HYDRAZINE SPACECRAFT PROPULSION SYSTEMS - 30 YEARS OF

SAFE, RELIABLE, FLEXIBLE AND PREDICTABLE PERFORMANCE IJ-367HYDROCARBON

A PARAMETRIC STUDY OF A HYDROCARBON FUELED SCRAMJET IV-309LIQUID FUEL RAMJET HYDROCARBON FUELS: COMBUSTION EVALUATION (U) VI-135

HYDROCARBON-FUELEDHYDROCARBON-FUELED SCRAMJET COMBUSTOR INVESTIGATION (U) VI-125

HYDRODYNAMICA HYDRODYNAMIC CHEMISTRY MODEL OF THE LIQUID PROPELLANT GUN IV-493

HYDROGEN-INDUCEDCRYOGENIC HYDROGEN-INDUCED AIR LIQUEFACTION TECHNOLOGIES 11-31

HYPERSONICA COMPUTATIONAL INVESTIGATION OF THE OPERATING CHARACTERISTICS OF A

HYPERSONIC INLET $V..157A SUM1ARY OF EUROPEAN AND JAPANESE HYPERSONIC FACILITY ACTIVITIES 1-203HIGH PRESSURE MACH 10 TO 20 ELECTROTHERMAL HYPERSONIC WIND TUNNEL 1-211NEW AND UPGRADED SUPERSONIC - HYPERSONIC AIRBREATHING ENGINE TEST FACILI-

TIES AT GASL IV-87SHOCK WAVE/BOUNDARY LAYER INTERACTION CONTROL IN A GENERIC HYPERSONIC

INLET 1-235

432

HYPERVELOCITYHYPERVELOCITY BY EXTENDED PROPELLANT BURN 1-321

1-45EVOLUTION OF AN AGING PROGRAM - MINUTEMAN STAGE II SOLID ROCKET MOTOR

1-45 1-45ICBM

AFT DOME INTERNAL INSULATION EROSION ON SMALL ICBM PRE-FSD AND FSD STAGEI SOLID PROPELLANT ROCKET MOTORS 111-45

IDENTICALWEIGHT AND COST ANALYSIS OF IDENTICAL STAGE BOOSTER SET FOR EXO ATMOSPH-

ERIC INTERCEPTOR COMPONENT FLIGHT TESTS (U) VI-21IGNITION

SUPPRESSION OF IGNITION RISE RATE IN SOLID PROPELLANT ROCKET MOTORS V-10OUNICHARGE CENTER CORE IGNITION IV-433

IGNITION/COMBUSTIONDIAGNOSTICS OF IGNITION/COMBUSTION IN A BULK-LOADED LP GUN 1-379

IMARMY INSENSITIVE MUNITIONS (IN) PROGRAMS AND OBJECTIVES FOR ARMY MISSILE

SYSTEMS V-239ENERGETIC MINIMUM SMOKE PROPELLANTS FOR IN EVALUATION V-225

IMADAN OVERVIEW OF THE U.S. NAVY IMAD PROPELLANT TECHNOLOGY INITIATIVES 111-525

IMAD-301INSENSITIVE MINIMUM-SIGNATURE PROPELLANTS (IMAD-301 PROPELLANT) 111-341

IMPACTDAMAGE EFFECTS CAUSED BY MULTIPLE BULLET IMPACT V-255IMPACT OF MISSION REQUIREMENTS ON PROPULSION SUBSYSTEMS 111-253IMPACT OF PROPULSIVE ADVANCEMENTS ON CAPABILITIES OF REUSABLE EARTH-TO-

ORBIT SHIPS 11-21THE IMPACT OF DESIGN OPTIMISM ON THE PERFORMANCE OF A RAMJET MISSILE IV-125THE IMPACT OF MARGIN ON THE ADVANCED LAUNCH SYSTEM LIQUID PROPELLANT

ENGINE V-531IMPINGEMENT

IMPINGEMENT STARTING OF SMALL EXPENDABLE JURBOJETS IV-279INDUCTION

EXPERIMENTAL DEMONSTRATION OF THE OPERATION OF AN INSTALLED SUBMERGEDSUPERSONIC AIR INDUCTION SYSTEM (U) VI-81

INGREDIENTSMICROENCAPSULATED INGREDIENTS FOR PROPELLANT APPLICATIONS 111-411PROPULSION SYSTEM HAZARDS AND ENERGETIC INGREDIENTS SYNTHESIS v:J.'.JATION V-1

INITIATIONFEASIBILITY DEMONSTRATION OF AN ELECTRO-OPTICAL SEMICONDUCTOR BRIDGE

SAFE/ARM/FIRE INITIATION SYSTEM 111-515LASER INITIATION OF PROPELLANTS AND EXPLOSIVES V-115

INITIATIVESAN OVERVIEW OF THE U.S. NAVY IMAD PROPELLANT TECHNOLOGY INITIATIVES 111-525

INJECTORAIRFLOW MODEL TESTING TO DETERMINE THE DISTRIBUTION OF HOT GAS FLOW AND

O/F RATIO ACROSS THE SPACE SHUTTLE MAIN ENGINE MAIN INJECTOR ASSEMBLY 11-117CONTROL OF CHAMBER HEAT FLUX BY INJECTOR DESIGN 11-127EVALUATION OF AN AEROGRID FUEL INJECTOR AND OTHER AERODYNAMIC GRID CON-

FIGURATIONS (U) VI-153TITANIUM PLATELET INJECTOR DEVELOPMENT 111-207

INJECTOR/COMBUSTORPROGRESS REPORT ON LIQUID PROPELLANT INJECTOR/COMBUSTOR TESTS 1-363

INLETA COMPUTATIONAL INVESTIGATION OF THE OPERATING CHARACTERISTICS OF A

HYPERSONIC INLET IV-157ADVANCED INLET DATA ANALYSIS TECHNIQUES IV-145COMPOSITE DUCTED ROCKET INLET DEMONSTRATION IV-397SHOCK WAVE/BOUNDARY LAYER INTERACTION CONTROL IN A GENERIC HYPERSONIC

INLET 1-235INLETS

PASSIVE COMBUSTION CONTROL IN A DUCTED ROCKET WITH SIDE MOUNTED INLETS IV-383INSTABILIlY

NON-ACOUSTIC COMBUSTION INSTABILITY IN HYBRID ROCKET MOTORS 11-177INSTALLED

EXPERIMENTAL DEMONSTRATION OF THE OPERATION OF AN INSTALLED SUBMERGEDSUPERSONIC AIR INDUCTION SYSTEM (U) VI-a1

INSTRUMENTATIONSTATIC AND CENYRIFUGE FORWARD DOME INTERNAL INSULATION CHAR AND INSTRU-

MENTATION DATA 111-25INSULATION

AFLAS: AN EXTERNAL INSULATION FOR ROCKET MOTORS V-123AFT DOME INTERNAL INSULATION EROSION ON SMALL ICBM PRE-FSD AND FSE STAGE

I SOLID PROPELLANT ROCKET MOTORS 111-45COOLING SYSTEM AND INSULATION CONCEPT FOR A MACH 5 TURBO-RAMJET AIRCRAFT 1-191MOOELING INTERNAL INSULATION CHAR AND EROSION IN FORWARD DOME ENVIRON-

MENTS 111-35STATIC AND CENTRIFUGE FORWARD DOME INTJvt.L INSULATION CHAR AND INSTRU-

MENTATION DATA 111-25

413

INSULATORSABLATIVE INSULATORS FOR RAMJET ENGINES IV-327

INSULINERINSULINER TECHNOLOGY DEMONSTRATION PROGRAM A PROGRESS REPORT 111-55

INTEGRALPERFORMANCE ANALYSIS OF A LOW-DRAG. VARIABLE GEOMETRY NOZZLE, INTEGRAL

ROCKET RAMJET AIR-TO-SURFACE MISSILE (U) VI-115SERVICE LIFE MONITORING OF INTEGRAL ROCKET RAMJET (IRR) PROPULSION

SYSTEMS IV-319INTELLIGENT

AN EFFICIENT, INTELLIGENT SOLUTION FOR VISCOUS FLOWS INSIDE SOLID ROCKETMOTORS 11-47

INTERACTIONSFLUID DYNAMIC/COMBUSTION INTERACTIONS AS DRIVING MECHANISM OF PRESSURE

OSCILLATIONS IN A REGENERATIVE LIQUID PROPELLANT GUN 1-369INTERCEPTOR

ADVANCING ENDOATMOSPHERIC INTERCEPTOR SOLID ROCKET MOTOR (SRM) TECHNOLOGY(U) VI-11


WEIGHT AND COST ANALYSIS OF IDENTICAL STAGE BOOSTER SET FOR EXO ATMOSPH-ERIC INTERCEPTOR COMPONENT FLIGHT TESTS (U) VI-21

INTERCONTINENTALADVANCED INTERCONTINENTAL BALLISTIC MISSILE (AICBM) TECHNOLOGIES 111-105

IRIDIUMCHARACTERIZATION OF IRIDIUM COATED RHENIUM USED IN HIGH-TEMPERATURE,

RADIATION-COOLED ROCKET THRUSTERS 11-359IRR

SERVICE LIFE MONITORING OF INTEGRAL ROCKET RAMJET (IRR) PROPULSIONSYSTEMS IV-319

ISOTHERMALMOISfURE EFFECTS ON ACROSS PLY MECHANICAL AND THERMAL CHARACTERISTICS OF

FM 5055 CARBON PHENOLIC UTILIZING RAPID ISOTHERMAL HEATING V-505

J402-CA-400SERVICE LIFE ASSESSMENT PROGRAM FOR THE IIARPOON J402-CA-400 ENGINE IV-303

JAPANESEA SUMMARY OF EUROPEAN AND JAPANESE HYPERSONIC FACILITY ACTIVITIES 1-203

JETECTHE JOINT EXPENDABLE TURBINE ENGINE CONCEPT (JETEC) DEMONSTRATOR PROGRAM

(U) VI-95JETS

PERFORMANCE CHARACTERISTICS OF SHROUDED SUPERSONIC RECTANGULAR ANDTAPERED ELLIPTIC JETS IV-133

KEYNOTEKEYNOTE ADDRESS "AEROSPACE COMPETITIVENESS FOR THE TWENTY-FIRST CENTURY" I-1

KINETICLIGHTWEIGHT KINETIC VEHICLE FREE FLIGHT/HOVER DEMONSTRATIONS V-75REACTION K!NETIC MODELING OF STABILIZER DEPLETION IN M6 PROPELLANT IV-423

KINETICSPROPELLANT COMBUSIOR USING CHEMICAL KINETICS IV-501

KKVKKV PROPELLANT TANK DEMONSTRATION PROGRAM 111-131

LASERLASER INITIATION OF PROPELLANTS AND EXPLOSIVES V-115NOZZLE VECTOR ANGLE DETERMINATION USING A LASER MEASUREMENT SYSTEM 1-33

LAUNCHADVANCED LAUNCH VEHICLE CONFIGURATION AND PERFORMANCE TRADES V-399ADVANCED LAUNCH VEHICLE UPPER STAGES USING LIQUID PROPULSION AND METAL-

LIZED PROPELLANTS 11-251ALL-HYDROCARBON ORBITAL LAUNCH VEHICLE V-359HYBRID PROPULSION BOOSTERS FOR SPACE LAUNCH VEHICLES 11-169STINGER MINIMUM SIGNATURE LAUNCH MOTOR TEST PROGRAM V-89THE IMPACT OF MARGIN ON THE ADVANCED LAUNCH SYSTEM LIQUID PROPELLANT

ENGINE V-531VERTICAL LAUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMARINE WAR-

FARE (ASW) V-135LAUNCHED

MECHANICAL DESIGN OF SURFACE LAUNCHED TACTICAL MISSILES 1-181LAW

HEATS OF EXPLOSION, DETONATION AND REACTION PRODUCTS: THEIR ESTIMATIONAND RELATION 10 THE FIRST LAW OF THERMODYNAMICS I-345

LAYERINGVARYING FILAMENT-WOUND NOZZLE PROPERlIES USING SELECTED FIaER-MATRIX

IAYERING V-473LEAP

LEAP SOLID PROPULSION DIVERT SUBSYSTEM DESIGN 111-165LIFE

A LIFE COMPARISON OF TUBE AND CHANNEL COOLING PASSAGLS fOR THRUST CHAM-BERS V-571

CHAPARRAL SERVICE LIFE EVALUATION V-157

434

LIFE (cont'd)RELIABILITY ANALYSIS OF THE HAWK MOTOR SERVICE LIFE PREDICTIONS V-177SERVICE LIFE ANALYSIS OF THE PATRIOT ROCKET MOTOR V-197SERVICE LIFE ASSESSMENT PROGRAM FOR THE HARPOON J402-CA-400 ENGINE IV-303SERVICE LIFE MONITORING OF INTEGRAL ROCKET RAMJET (IRR) PROPULSION

SYSTEMS IV-319SERVICE LIFE TESTING AND ANALYSIS OF FIELD AGED M.RS MOTORS V-187SHELF LIFE DETERMINATION OF HIGH BURN RATE NEPE PROPELLANT 111-181THE INFLUENCE OF PROPELLANT EQUILIBRIUM MODULUS ON ROCKET MOTOR SERVICE

LIFE PREDICTIONS V-167LIGHTWEIGHT

DESIGN AND DEVELOPMENT OF A LOW COST LIGHTWEIGHT, FAST RESPONSE ATTITUDECONTROL SYSTEM ENGINE USING CLF5/N2H4 PROPELLANTS 111-475

DESIGN AND DEVELOPMENT OF LIGHTWEIGH1 AXIAL PROPULSION ENGINE USING CLF5/N2H4 PROPELLANTS 111-113

LIGHTWEIGHT KINETIC VEHICLE FREE FLIGHT/HOVER DEMONSTRATIONS V-75LINEARIZED


LINERAN EXPERIMENTAL EVALUATION OF COMBUSTOR LINER MATERIALS FOR SOLID FUEL

RAMJET TESTING 1-227LINER MATERIALS BY SONIC VELOCITY MEASUREMENTS V-319

LINERSCOMPATIBILITY OF COMBUSTION CHAMBER LINERS WITH RP-1 V-541

LIQUEFACTIONAIR LIQUEFACTION AND ENRICHMENT SYSTEM (ALES) WITH TURBOROCKET FOR OR-

BITAL VEHICLES V-367CRYOGENIC HYDROGEN-INDUCED AIR LIQUEFACTION TECHNOLOGIES 11-31

LIQUIDA HYDRODYNAMIC CHEMISIRY MODEL si THE LIQUID PROPELLANT GUN IV-493ADVANCED LAUNCH VEHICLE UPPER STAGES USING LIQUID PROPULSION AND METAL-

LIZED PROPELLANTS 11-253DEVELOPMENT OF A SAFE LIQUID PROPELLANT LOGISTIC SYSTEM FOR FIELD ARTIL-

LERY IV-511FLUID DYNAMIC/COMBUSTION INTERACTIONS AS DRIVING MECHANISM OF PRESSURE

OSCILLATIONS IN A REGENERATIVE LIQUID PROPELLANT GUN 1-369HAZARD CLASSIFICATION OF LIQUID PROPELLANTS IV-525HELIUM SATURATION OF LIQUID PROPELLANTS 11-349LIQUID FUEL RAMJET HYDROCARBON FUELS: COMBUSTION EVALUATION (U) VI-135LIQUID PROPELLANT ROCKET ENGINE COMPONENT COST MODEL V-521PRED!CTION OF THE THRUST PERFORMANCE AND THE FLOWFIELD OF LIQUID ROCKET

ENGINES 11-133PROGRESS REPORT ON LIQUID PROPELLANT INJECTOR/COMBUSTOR TESTS 1-363REGENERATIVE LIQUID PROPELLANT GUN AND SELF PROPELLED HOWITZER INTEGRA-

TION 1-401SYSTEM APPLICATIONS FOR AN ADVANCED LIQUID AXIAL STAGE, ALAS 111-447SYSTEM REQUIREMENTS AND INTEGRATION FOR A FULLY AUTOMATED REGENERATIVE

LIQUID PROPELLANT CANNON IV-521TECHNOLOGY FOR PRODUCIBLE TACTICAL LIQUID MONOPROPLLLANT PROPULSION V-99THE DEVELOPMENT OF A LIQUID FUELED GAS GENERATOR FOR DUCTED ROCKET

RESEARCH 1-313THE EFFECT OF ACOUSTIC DAMPENING DEVICES IN REGENERATIVE LIQUID PROPEL-

LANT GUNS 1-389THE IMPACT OF MARGIN ON THE ADVANCED LAUNCH SYSTEM LIQUID PRO'ELLANT

ENGINE V-531LIQUID-PROPELLANT

ALTITUDE SIMULATION TESTING FACILITY CONCEPTS FOR LARGE LIQUID-PROPELLANTROCKET ENGINES V-577

DEVELOPMENT OF A HIGH-ENTHALPY TEST CAPABILITY USING A LIQUID-PROPELLANTROCKET ENGINE IV-113

LOGISTICDEVELOPMENT OF A SAFE LIQUID PROPELLANT I.OGISTIC SYSTEM FOR FIELD ARTIL-

LERY IV-51 1LOVA

SEMI-COMMERCIAL TPEs FOR LOVA GUN PROPELLANT BINDERS IV-469LOW-COST

AN ADVANCED, LOW-COST ENGINE FOR HIGH-THRUST APPLICATIONS li-89DEVELOPMENT AND TESTING OF A HIGH-ENERGY LOW-COST SOLID RAMJET FUEL AT

THE NAVAL WEAPONS CENTER (U) VI-173LOW-OLNSIIY

CIIARACTERIZA7ION OF LOW-DENSITY CARBON FIBER REINFORCED COMPOSITE V-1483LOW-DRAG

PERFORMANCE ANALYSIS OF A LOW-DRAG, VARIABLE GEOMETRY NOZZLE, INTEGRALROCKET RAMJET AIR-TO-SURFACE MISSILE (U) VI-115

LPDIAGNOSTICS OF IGNITION/COMBUSTION IN A BULK-LOADED LP GUN 1-379

M6REACTION KINETIC MODELING OF STABiLIZER DEPLETION IN M6 PROPELLANT IV-423

M7PROCESS AND QUALITY IMPROVEMENTS FOR M7 PROPELLANT 1-171

4115

MACHAF/NASA HIGH MACH COMBINED CYCLE ENGINE DEVCLOPMENTS IV-29COOLING SYSTEM AND INSULATION CONCEPT FOR A MACH 5 TUiRBO-RAMJET AIRCRAFI 1-191GENERIC HIGH-SPEED ENGINE TEST RESULTS AT MACH 4.3 AND 5.0 IV-193HIGH PRESSURE MACH 10 TO 20 ELECTROTHERMAL HYPERSONIC WIND TUNNEL (-211PERFORMANCE TRADES FOR AIR-LAUNCHED TURBOJET AND RAMJET CYCLE HIGH MACH

CRUISE MISSILES WITH ROCKET OR TURBOJET BOOST IV-1STUDY OF SUPERSONIC COGMUSk ION PHENOMENOLOGY IN A SUBSCALE COMBUSTOR AT

SIMULATED FLIGHT MACH NUMBERS FROM 7 TO 10 IV-209MASTERPLAN

INSENSITIVE MUNITIONS MASTERPLAN AND ITS IMPLICATIONS FOR THE SDI COM-MUNITY 1-55

MECHANICALJANNAF STANDARDIZATION OF TENSION, COMPRESSION, A'4j SHEAR TEST METHODS To

DETERMINE MECHANICAL MATERIAL PROPERTIES FOR fILAMENT WOITN3 COMPOSITESTRUCTURES 1-89

MECHANICAL DESIGN OF SURFACE LAUNCHED TACTICAL. MISSILES 1-181MECHANICAL PROPERTIES OF THE PEACEKEEPER STAGE II PROPELLANT-60C40 SYSTEM:

AN EXAMINATION OF VARIABILITY 11-1MOISTURE EFFECTS ON ACROSS PLY MECHANICAL AND THERMAL CIhARACTERISTICS OF

FM 5055 CARBON PHENOLIC UTILIZING RAPID ISOTHERMAL HEATING V-505MECHANISM

FLUID DYNAMIC/COMBUSTION INTERACTIONS AS DRIVING MECHANISM OF PRESSUREOSCILLATIONS IN A REGENERATIVE LIQUIO PROPELLANT GUN 1-3691

MELT I NGNP If: A NEW LOW TEMPERATURE MELTING NiTROPLASTICIZER IV-417

METALLIZEDADVANCED LAUNCH VEHICLE UPPER STAGES USING LIQUID PROt'tllSION AND METAL-

LIZED PROPELLANTS 11-251MICROCOMPUTER

AIRBREATHING BOOSTER PERFORMANCE OPTIMIZATION USING A NICROCOM•PUTER I-11-MICROENCAPSULATED

MICROENCADSULATED INGREDIENTS FOR PROPELLANT APPLICATIONS 111-411M I N I ATUR I ZED


MINtMUMENERGETIC MINIMUM SMOKE PROPELLANTS FOR IM EVALUATION V-Ž22MINIMUM SIGNATURE INSENSITIVE MUNITION GEL BIPROPELLANT PROPULSION V-247PSAN EFFECTS ON MINIMUM SIGNATURE PROPELLANT PROPERTIES V-31STINGER MINIMUM SIGNATURE LAUNCH MOTOR TEST PROGRAM V-89

MINIMUM-SICNATUREINSENSITIVE MINIMUM-SIGNATURE PROPELLANTS (IMAD-301 PROPELLANT) 111-341

MINUTEMANEVOLUTION UF AN AGING PROGRAM - MINUTEMAN STAGE II SOLID ROCKET MOTOR

1-45 I -45MISSILE

ADVANCED INTERCONTINENTAL BALLISTIC MISSILE (AICBM) TECIINOLOGIES 111-105ARMY INSENSITIVE MUNITIONS (IN) PROGRAMS AND OBJECTIVES FOR ARMY MISSILE

SYSTEMS V-239DEVELOPMENT OF AN ATTITUDE CONTROL MOTOR FOR THE ERINT-1 MISSILE 111-505MISSILE INTEGRATED STA.;E TECHNOLOGY (MIST) PROGRAM 111-153PERFORMANCE ANALYSIS OF A LOW-DRAG, VARIABLE GEOMETRY NOZZLE, INTEGRAL

ROCKET PAMJET AIR-TO-SURFACE MISSILE (U) V-!115PRELIMINARY DESIGN AND PERFORNANCE OF AN ADVANCED RAMJET "!SSILE CONCEPT

FOR THE CIRCA-2010 OUTER AIR BATTLE (U) VI-103PRELIMINARY ESIGN OF AN EXTENDED RANGE AIR-TO-AIR MISSILE (U) VI-187SOLID ROCKET POOSTLR INTEGRATION WITH THE AQM-37C MISSILE TARGET 1-111THE IMPACT OF DESIGN OPTIMISM ON THE PERFORMANCE OF A RAMJET MISSILE IV-125VERTICAL L.AUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMRRIHE WAR-

FARE (ASW) V-135MISSILES

MECHANICAL DESIGN OF SURFACE LAUNCHED TACIICAL MISSILES 1-181PEIRFORMANCE IMPROVEMENTS FOR SHORT-RANGE TACTICAL MISSILES USING PULSE

MOTORS 111-305PERFORMANCE TRADES FOR AIR-LAUNCIIED TURBOJET AND RAMJET CYCLE HIGH MACH

CRUISE MISSIIES WITH ROCKET OR TURBOJET BOOST IV- IPULSE MOTOR PROPULSION FOR TACTICAL MISSILES 111-293WJ119-2 MODULAR TURBOJET PROPULSION SYSTEM FOR TACTICAL MISSILES !V-251

MISSIONIMPACT OF MISSION REQUIREMENTS ON PROPULSION SUBSYSTEMS 111-253

MI'TMISSILE INYFOPATED STAGE TECHNOLOGY (MIST) PROGRAM 111-153

MLRSSERVICE LIFE TESTING AND ANALYSIS OF FIE•O AGED MLRS MOTORS V-107

MODELA HYDRODYNAMIC CHEMISTRY MODEL OF THE LIQUIU PROPELLANT GUN ;V-bi93AIRFLOW MODEL. IESTING TC DETERMINE THE DISIRIBUTION OF HOT GAS FLOW ANU

O/F RATIO ACR1OSS iHE SPACE SHUrIL MAIN ENGINE MIN INJECTOR ASSEMBLY i-t117LIQUID PROPELLANT Qr•o(Er ENGINF C•W'UNENT COST MODEL V-521HOD1l 320-2- A COMPACT ADVANCCO IhAV IUR,'j.{ T 1-265NASA SPIP RONDLINE WO'K PACKAGE: OVEFVILY OF TH- SRM 0t0EI. PROCESS AND

PROC'_SS FMEA TASK V-5

4' .6

MODELSADVANCED COMPUTATIONAL MODELS FOR ANALYZING HIGH SPEED PROPULSIVE FLOW-

F IELDS 1-2147MODULE

DEVELOPMCNT OF THE NLOS TECHNICAL RISK REDUCTION SUSTAINER MODULE IV-259MODULUS

THE INFLUENCE OF PROPELLANT EQUILIBRIUM MODULUS ON ROCKET MOTOR SERVICELIFE PREDICT!GNS V- 167

MO I STUREMOISTURE EFFECTS ON ACROSS PLY MECHANICAL AND THERMAL CHARACTERISTICS OF

F14 5055 CARBON PHENOLCC UTILIZING RiAPID ISOTHERMAL HEATING V-505MIOLECUL AR

MOLECULAR STRUCTURE OF LOW TEMPERATURE FORM Of TRIAMINOGUANIDINIUMNITRATE (TAGN) 1-331

MON ITORIINGDIELECTRIC MONITORING OF BONDLINE MATERIALS FOfI. IMPROVED PROCESS CONTROL V-2-19SERVICE LIFE MONITOPiNG OF INTEGRAL ROCKET RAf4.IET (IRR) PROPULSION

SYSTEMS IV-319MONOPROI'ELLANT

MO0NOPROPELLANT HYDRAZINE SPACECRAFT PROPULSION SYSTEMS - 30 YEARS OFSAUE, RELIABLE, FLEXIBLE AND PRLOICTABLr PERFORMANCE 11-367

TECHNOLOGY FOR PRODUCIBLE TACTICAL LIQUID MONOPROPELLANT PROPULSION V-99MOTOR

ADVANC'hIG ENDOATMOSPHERIC INTERCEPTOR SOLID ROCKET MOTOR IS3RM) TECHNOLOGY(U) ViI-ll

DEVELOPMENT OF A PINTLE CONTROLLED VARIABLE THRUST SOLID ROCKET MOTOR 111-241DEVELOPMENT OF AN ATTITUDE CONTROL MOTOR FOR THE ERINT-1 MISSILE 111-505CEVELOPMENT OF THE FIBER OPTIC GUIDED SKIPPER ROCKET MOTOR (U) VI-31D)EVELOPMNET OF THE ORBUS 1 CARBON FIBER/EPOXY MOTOR CASE FOR THE STARS/

STARSIRD PROGRAMS 1-69LRIt4T-1 SOLID ROCKET MOTOR REQUIREMENTS AND STATUS (U) VI-Z43EVOLuTION 01- AN AGING PROGRAM - MINUTEMAN STAGE II SOLID ROCKET MOTOR

1-45 1-45FEAS:BILITY STUDY FOR EMPLOYING SOLID ROCKET COMBUSTION SIMULATORS FOR

SU,1ID ROCKET MOTOR NOZZLE TESTING V-1417fLOW FIELD ENVIRONMENT IN TIlE NASA (WO-INCH TEST MOTOR V-407F' MATiON AMD KESIDUAL STIRENGTh: IK COMPOSITE ROCKET MOTOR CASES 111-273

41 ARMY/NAVY INSENSITIVE MUNITIONS H~i.LFIRE MOTOR TECHNOLOGY PROGRAM 111-573Luo PIPOJECTILE ROCKET MOTOR C4SING: FRA~GMENTATION CONTROL TECHINOLOGY V-215PULSE MOTOR PROPULSION FOR TA4CTICAL. MISSILES i11-293rCLIAB!LITY ANALYSiS OF THE HAWK MOYOR SERVICE LIFE PREDICTIONS V-177SERVICE LIFE ANALYSiS Of THE PAIRIOT iýOCKET MOTOR V- 197STINGER FLIGHT MOTOR CAfE !>TRESS CORROSION EVALUATION V-155STINGER M#INIMUM SIGNAIURE LAUNCH MOTOR TEST PROGRAM V-89SUBSCALE MOTOR CASE ANALOG FOR FAST COOK-OFF INSENSITIVE MUNITIONS TESTS V-209THE INFLUENCE OF PkOPELLANT EQUILIBRIUM MODULUS ON ROCKET MOTOR SERVICE

LIFE PREDiCTIONS V- 167MOTOR/HAZARD

LARGE-SCALE MOTC.R/rIAZARD TESTING OF HI1GH PERFoRMANCE DENSE ADDITIVE(Bi2O3) PROPELLANT (U) VI-73

MOTORSA SYSTEM FOR THlE TESTING OF HYBRID ROCKET MOTORS 11-217AFLAS: AN EXTERNAL INSULATION FOR ROCKET M07ORS V- 123AFT DOME INlEIRNAL INSULATION EROSION ONt SMALL ICBM f'RE-FSO AND FS0 STiAGE

I SOLID PROPELLANT POCK~f MOTORS III1-45AN EFFICIENT, INTELLIGENT SOLUTION FOR VISCOUS FLOWG INSIDE SOLID ROCKET

MOTORS 11-47CLEAN PROPELLANT FOR LARGE SOLID ROCKFT NOTWiS, 11 V- 3.ý7GLYCIDYL AZIDE POLYMER (GAP) SOLID GAS GENERATOR FOPR DUCTED AND HYBRID

ROCKET MOTORS V-47NON-ACOUSTIC COMOUSTION INSTABILfrY IN HYURID ROCKET M4OTORS 11-117NONLINEAR STABILITY TESTING AND PULSING OF FUILL SGALE TACTICAL MOTORS 1-135PERFORMANCE IMPROVEMENTS FOR SHORT-RANIGE TACTICAL MISSILES USING PULSE

MOTORS 111-305SERVICE L!FE TESTING AND ANALYSIS OF FIELD ACED MLR-S MOTORS V-187SLAG ESTIMATION FOR SOLID ROCKrT MOTORS 1 1-225STABILITY TESTING OF FULL SCALEC TACoHCAL MOTORS 1-119SUPPRESSION OF IGNITION RISE RATE IN SOLID PROPELLANT ROCKET MOTORS V-lOTfVARIABILITY IN C-C THROAT EROSION FOR A SELECTION OF 5OLID ROCKET MOTORS ill-Ti7

MULTI PARAMErR ICNONCONTACTING MULTIPARAMETRIC SENSOR FOR SHAFT SPEED, TORQUE, AND POSI-

TI ON 11-281MUN. I T I ON

MINIMUM SIGNATURE INSENSITIVE MUNITION GEL BIPROPELIANT PROPULSION V-2417MUN ITI ONS

ARMY INSENSITIVE MUNITIONS (IM) PROGRAMS AND OBJECTIVES FOR ARMY MIFSILlESYSTEMS V-2ý'

BURNING RATE CATALYSIS OF INSENSITIVE MUNITIONS AN PROP'ELLANTS il1-555INSENSITIVE MUNITIONS MASTERPIAN AND ITS IMPLICATIONS FOR THE SDI COM-

NUN I TY 1-55JOINr ARMY/NAVY INSENSITIVE MUNITIONS HELU FIRE MOTOR TECHNOLOGY PROGRAM 111-573ROCKET PROPELLANTS FOR INSENSITIVE MUNITIONS RESEARCH 11 1-545SUDSCALE MOTOR CASE ANALOG FOR FAST COOK-OFF INSENSITIVE MtUNITIONS TEST-S V-209

417,

NASAFLOW FIELD ENVIRONMMcT IN IHE NASA TWO-INCHi TEST MOTOR V-407NASA LEWIS SUPERSONIC TRP'JUGIFLOW FAR PnO&4AN IV-73NASA SPIP BONDLINE WORK PACK(AGE: OVLRVIEW OF THU SRM M~ODEL PROCESS AND

PROCrESS FME:A TASK V-265OVERVIEW OF NASA SOLID POOPUL31ON INTEGRITY PROGRAM (SF15-) 8ONOLINU WORK

PACK~AGE RESULTS AND ACCONPLHSHMENIS V-301SELECTION OF BOND-IN-TERSION SPECIMEN F~f THEL NASA SPIP BONDLINES WORK-

PACK AGE V-287NAVAL

DEVELOP'MENT AND TESTING Oý A HIGH-ENERGY LOW-COST SOLID RAMJET FUEl. ATTHE NAVAL WEAPONS CENTER (U) VI-173

NAVYAN OVERVIEW Of Thf. U.S. hAVY IMAD PROPELLANT TECHNOLOGY INITIATIVES1155

NOENuc or THCRMAL PRVTEC.TION SYSTEM FOR SPACE S3UTTLC SOLID ROCKET BOOSTER 1l-1

NF.PESHELF LIFE DETERMiNAl ION or HIGH BURN RATE NEPE PROPELLANT 111-181

AdI TRAI CIAOLEfvULA~l SrRUCTUrE OF LOW TEMP[RATURE FORM OF IMIA&IINOGUANIOINIUM

N(IRATF (TAGN) I1-331NI IROPLAST~tIIER

NP iI: A NEW LC.W TEMPERATURE MELTING NITROPLASTICIZIER :V-41""HLOS

DEVELOPMENT OF THE NLOS TECHNICAL RISK REDUCTION SUSIAINER MODULE IV-2',9NON-ACOUST IC

NON-ACOUSTIC COMBUSTION INSTABILITY IN HYBRID ROCKET MOTORS 11-17/NONALUM IN IZED

LITERATURE REVIEW OF PLATEAU BALLISTICS IN NONALUMINIXIf.) f$OLIL PR(OPEL-LANTS 111-387

NONCONTACT INGNONCONTACTING MULTIPARAMETRIC SENSOR FOR SHAFT SPEEO, TORQjUE, AIMJ POSI-

T ION 11-281NO NDE£TON ABL E

RUGGED NONDETONABLE PROPELLANTS V-!)5NONL INEAR

NONLiNEAR STABILITY TESTING AND PULSIND OF FULL SCALE IACTICAL MOTORS 1-135NORMAL

NORMAL STRESS TRANSDUCER BEHAVI R 1-161NOZZLE

ARCJET NOZZLE AREA RAT:O EFFECTS I; -h.(7FEASIBILITY STUDY FOR LMPI.OYING 5OLID ROCKET COMBUSTION SIMULATORS FOR

SOLID ROCKET MOTOR NOZZLE TESTING V-1417IMPLICATIONS OF THE RESULTS OBTAINED IN THE GRUMMAN DETONATION SIV*CK

TUNNEL NOZZLE EXPERIMENTS ON SCRAKJET PERFORMANCE IV-221NOZZLE VECTOR ANGLE DETERMINATION USIN4G A LASER MfASUkEMfNT SYSTEM I-isPERFORMANCE ANALYSIS OF A LOW-DRAG, VARIABLE GEOMETRY N4WzLE, IWTEGRAL

ROCKET RAMJET AIR-TO-SURFACE MISSILE (U) VI -115SOI/KEW UL'TRALIGHT WEIGHT NOZZLE STRUCTURUS 11v-23,VARYING FILAMENT-HOUND NOZZLE PROPERTIES USING SELECTED FIOF.H-MATRIX

LAYER IN N-4NP

NP II: A NEW LOW TEMPERATURE MELTING NIIROPLASTJCIIfR iv-I17

0/FAIRFLOW MODEL TESTING TO DETERMINE THE DISTRIBUTIO'N C-" 'I' ( Al I-Ow AND

0/F RATIO ACROSS THE SPACE SHUTTLE MAIN ENGINE MA .1A.'.U A3,E 013L .1-117OBSERVABLE S

HIGH PRESSURE, LOW RATE, LOW OBSERVABLES PFIOPELL1'NT ýl ýL(AK'Nl - i IN~I.REPORT I h-i?]

OHSUPRATED OHS ENGINE STATUS - SEA LEVEL TESTING RESUL,.ý 11-24u1

OPTICDEVELOPMENT Of' THE FIBER OPTIC GUIDED SKIPPER ROCKET .4f)' X,(i) VI. 31FIBER OPTIC ORDNANCE SYSTEMS: DESIGN, PERFORMANCE, .j, iLoTING

ORBITANALYSIS OF EXPENDABLE SOLAR ELECTRIC ORBIT TRANSFER VEHICELS 11-41u7ORBIT TRANSFER VEHICLE PROPULSION DESIGN: TRADES AND COMPARISONS 11-291

ORB ITALAIR LIQUEFACTION AND ENRICHMENT SYSTEM (ALES) WITH 1URBOROCKET FOR OR-

BITAL JEHICLES V-367ALL-HYDROCAPBON ORB ITAL LAUNCH VEFI ICLE( V-359

OR BUSDEVEI.O?MENI Of fIIE URESUS I CARBON FIBER/EPOXY MOTOR CASE FOR THE SIARS/

SI ARB IRD PROGRAMS 1-69ORDNANCE

I1 OPTIC ORDNANCE SYSTEMS: DESIGN, PERFOHMANCE. AND TESTING 111-85OSC IIL',AT IODNS

FIIJIO I)\NAMIC/COMBUST ION INTERACTIONS AS DRIVING MECHIANISM Of 1ISRSUIP.[OSCILILAT IGOS 'N A REGLNERATIVE LIQUID PROPELLANT GUN 1-309

OVERWHAF'COMPOSITE OVERWRA? PO3 ITI"/F LAPULSION P-RoPrLLiNI lANK5 I11-307

438

OVCRWRAPPEDA COMPOSITE OVERWRAPP(O BONDED ROLLING DIAPHRAGM EXPULSION SYSTEM 111-141

PARTICLE.STRUCTURAL OAW ONO/ACOUSTIC ATTENUATION OPTIMIZATION VIA RON-OSCItUCTIVE

PARTICLE DAMP ING II 63PASSAGES

A LIFE COMPARIISON 01 TUBE AND CHANNEL COOLING PASSAGES FOR THRUST CHAN-~SERb V-571

PASSIVEPASS IVE COMBUST ION CONTROL IN A DUCTED ROCKECT WITH SIDE MOJPITLUINLMLTS IV-18,

PATIR OTPROPULSION STUDY FOnl AN (XTLND(:D RANGE PATRIOT (U) VI *61SERVICE. LIFE ANALYSIS Of THE PATRIOT AOCKC1 MOTOR V-197

PC ACE K CCPEROF.VELOPHCNT AND 71511MG Of tHlL SOLID PROPILILANT GAS OCNE11RATOR POWLRIEO

OPEACEKEEPEII RAIl. GARRISON CANISTER ERLCTON SYSTEMN 111-95MECHANICAL PROPERTIES 01 THE1 PEACLKELPIR $?AOL If PROPLLLANT-0000 SYSTEM

AN EXAMINATION or VARIABILITY Ill-1PtNTAFLUOR I OC/IYORAZ INC

A VCRIIATILC CHLOAINI' PCNTAILUORIOE/I4YDRAZINt, (CIIF5/N2HI.) AtTITIJOL ROCKETINCOMNE TEST FACILITY 111-711

PCRI 0IMANCEADVANCI.0 LAUNCHI VI.HICA( CONVIOURATIDN AND PIRIOR01MANCI, IRA011 V-399AIRIIIAYHING 1110051,111 PIRFORMANCI, OPTIMI7AT 104 1.11100 A NICROCONPIJlt I1-11I 10CR OPTIC ORDN1ANCE. SYSTCMI:1 0151014 "IRFOAMAIICI AND TtSIIWG 111-85IMPIICAT IONS OF IN( rESU1,TS OBTAINID Im" THE GAUP*4AA OCTONAT ION SHO(CK

iUNNir NOlZI.E EXPIRIM1.11TS ON0 SCANJLT P11.11RFORNA110.1 IV.77LARCJI-SCAI.I. 04OTOR/IIA7ARD 7tSIIWj Of HIGH PI.111OA0MANCI UIN&I ADUITIVI.

191;103) PRlOPEiI.ANT (U) VI -7)MONOROPELLANT HYOiNAIINI. bPACIURAII PRO10PUtilION SY11TIM11 - 13n YtAPS Of

SAIl, N~l, IA11111 ftIEXIBLI. AND P1141101C1Abil PIff1OftIAAMC1 11-167PERI 014MANCL AN kLYSi it O A I OW-PRAG VA I A01 0 I 14 GIMFY N01LJI , INTLORAI

ROCKET MANJIT AIRt-1O.S1URIACC MIiSI U VI * II5PI.AfOIMANCt CIIARAC11IR1RTICS Of 11101N11OUDL SUIIP(SONIC MEICAPNOUI Al Aft,'

TAPI Pf.l) It 1,IPI IC if11I IV- 1)P ,IAI ORMANCL I 4P ROVLMI NT 10 I Ot6HOR 1-11AIANI 7 AC I ICAI. 19 11111. 151U%14MG PUt O

NO I1OR4I II I-5Q)PI HI0 OGiAN(I TIkADI S 1 011 AIRt-I AUNCHfIU TUPIOO11JI I ANU RLAMJII LY(A4. Iflu" MA(;"

CR4UISE 14I1pbII.cs WiliT IOCKII ON fUhnOJiT 111005v IV-IPNIi.OI(ION Of TII(l IIRAUSI PLRIOEL10MANCI. AND THIlH~ 1`t t1101)10OfLIUI ROCKIT

(NI 0141 II:1*V1411 IMIIEA$Y ()151J14N ANa) PIRIIFCMANCI Of AN ADVANrG1r OAAIJII 11416I1, £ONCLPI

0 VA1 fill CLlftA-W'OIO 001114 AIM bAI111 (U) VI' 505lift IMPAf;? 01 I)Ibi(jN U10II04ftM ON lift PINIUIINANLJ Ofl A 4AMJI I M1611OI11 IV-1111

PLRMIABII IlYWITHi-PtY i'li 141AtiII I TY 01 CAMHON PHION(W IC CW4POSI 111.11 AN A F UNGI ION Of

ACI1UUS-PLY COMPkIISMIVI I()AU ANIf) TIMPO.MAIUNIV-9pHlitNOt. IC

INV(SITICATION Of A F OUft I01IR1 1111 ION Till SlINUCIUMAS ANAL (315 OVAIUATIO14Of PHENOl It; MAYIRIAI 1 II e?785

M0114IUIS I1 L1`1lS ON A011011 P1 Y MICIOAlfirAI ANOI) TIANAI COIAMANAISIII, 0511 OIN !ý5 CAHOON P1111401 IC Vill I11101 MAPII) ISOIIIIIIMAI IIIAIING Ou

Tlif 11111 OCOOICAI ANAL Yb % Of PHIENOl IC HI SIN UXI I) INl AIII AlIVII 111-69WI OI-Pt V PIRMF.A1111.ITY Of CAHOON P014100(11 IC LcMPOSIII.11 At A F U1N(LIl IONf

ACltOSS-PL.Y COMPhISItIVI lUOAU AMa) IIMP.I4HAIU~f, V4.1`111.04LIC--CONTINUII)

CARSON-Cl 0111 PIllNO 140 I--C;ONI INUI U DI VIIUPFMI NI ANDI LiIAFIACIII II/Al ION V11

b 'UDY 0f1 SUPI lbUNIC C0141105 ION11 Pill H141Milt OGY IN A bV1Jl1CAI I COM1,0USI1N AlhiMULAII0 11.16111 MAClI NUNIIIF4b (RiON 7 T`10 11 I ,(vLJ

0t-VILOPMtNI Of A PONILL COINIIOiLtU VAMIASII 1IMS 711116 101 IDI MOKIT 1 401')Ml 111-141ePL1AST IC IZIll

MIYU 0 1olrAI S 01 ?,;A-00 NlI4'JPl1PAN01 AND ~'?IIN I HUI1UIA1dO1 At bOS IlIPRIOPELLANT Pt AbIiIC/I 1416 V-eu I

PLATI AllI lI TO AlUHlt H(.VIIUW 01 PI.AII AV IIAI 1 1111 ICI1 IN NONA MIN0414/ID U 0l11I) PD'.jPD I -

1 ANl [IT bIII l

11I (ANI VM1 I'IAII IlIl INJI CI(IH Of VI LOPM1 NI I

0I)V1'I OHI NI 0f A 111 PROtPI I I AN[ PI UMI 01 N ILAII 1014IIJIbYI.11IM (U) VI,INV[ S' IGAl ION 01 Tiff AIICJ1 I N( Ali lI IlIIj P1 0141 USI NG I II GIM011 [AI I K Plt(III fAI

1401 51011 I11 I ILS ON ACIIOb4 P1 Y MI CIIANIGAI AND fill f(MA( C.IIAkACII 1415 1 1I.., OfIMH 5055 ;AILISUPI PII11401. IC UT II 0111,16 1HAPO ID 60111iffMAi III AlI INOOI

PD1t YMIJI.GLCI. Y(I Y A/I DI P0L14VIP. (GCAP) 2hUS 11) ,Ah GI NII1%AlDI I 1041 DUC IIII ANO 11Y11IL410

1,10101(IJI~es V-4l II'OL YMI 145.

SOL 10 NSOCK( I PILOIUI.5 ION APPI' l;AI IONS I Off AI.VAN(l 1) POO eMI FIt' - 01

POWDERBALL POWDER PROPELLANT APPLiCAThJMS TO LARGE CALIBER AMMUNITION 1-355

POWEREDDEVEB OPMENT ANID TEST ING OF THE SOLID PROPEL' ANT GAS GENERATOR POWERED

PEACEKEEPER RAIL GARRISON CANISTER ERECTiON SYSTEM 111-95PRED ICTABLE

MONOPROPELLANT I4YDRAZINE SPACECRAFT PROPULSION SYSTEMS - 30 YEARS OFSAFE, RELIABLE, FLEXIBLE AND PREDICTABLE PEVIÒRMANCE 11-367

PREDICT IONSRELIABILITY ANALYSIS Of THE HAWK MOTOR SERVICE LIFE PREDICTIONS V- 177THE INFLUENCE Of PROPELLANT EQUILIBRIUM MOD0ULUS ON ROCKET MO0TOR SERVICE

LIFE PREDICTIONS V- 167PREF ORM

..ARGE SIZE 3D CARBON/CARBON PAEFORM DEMONSTRATION V-513PREMIXED

PREMIXED, TURBULENT COMBUSTION OF AXISYMMETRIC SUDDEN EXPANSION FLOWS 1-305PRESSURE

FLUIO OYNAM",.'OMIBUSTION INTERACTIONS AS DRIVING MECHANISM OF PRESSUREOý:CILLAYIC-HNS IN A REGENERATIVE LIQUID PROPELLANT GUN 1-369

HIGH FRESSIJAC, L')W RATE, LOW OBSERVABLES PROPELLANT DEVELOPMENT - FINALRcprRT 111-323

HIGHI PRESSURE "OACi 10 TO 20 ELECTROTHERMAL HYPERSONIC WIND TUNNEL 1-211PREDICi11 s)N If NTLRIOR BALLISTICS AND PRESSURr WAVES PRESENT IN UNICHARGE

L 5I6" IV-4417PRESFýUAL GARAUIENT EFFECTS IN SUPERSONIC SHEAR LAYERS IV-177

P'RESSURESTHROAT EROSION CHARACTERIZATION OF 14 DIMENSIONAL CARBON-CARBON AT HIGH

PRESSURES V-449PRIMAR't

DESIGN AND DEVELOPMENT OF A PRIMARY STRUCTURE BASED UPON ADVANCED COM-POSITE MATERIALS TECHNOLOGIES 111-461

PROBESINVESTIGAT'C:.o Or TH-E ARCJET NEAR F IELD PLUME 'JSING ELECTROSTATIC PROBES 11-391

PROCESS INGACCELERATED PROCESSING OF HTPB PROPELLANTS WITH TMXDI V6

PRODUC IBLETECHNOLOGY FOR PROJUCIBLE TACTICAL LIQUID MONOPROPCELLANT PROPULSION V-99

PR OGRAMEVOLUTION JF AN AGING PROGRAM - M14UTEMAN STAGE 11 SOLID ROCKET MOTOR

1-Is5 1-45HYBRID PROPULSION TECHNOLOGY PROGRAM 11-143INSULINER TECHNOLOGY DEMONSTRATION PROGRAM A PROCRESS REPORC 111-55JOINT ARMY/'NAIVV INSENSITIVE MUNITIONS HELLFIRE MOTOR TECHNOLOGY PROGRAM 111-573KKV PROPELLANT TANK DEM4ONSTRATION PROGRAM 111-31MISSILE INTEGRATEL) STAGE TECHNOLOGY (MIST) PROGRAM II i-153NASA LEWIS SUPERSONIC THROUGHFLOW FAN PROGRAM IV-73OVERVIEW OF NASA SOLID PROPULSION IKTEGRITY PROGRAM (SPIP) RONOLINE WORK

PACKAGE RESULTS AND ACCOMPLISHMENTS V-301SERVICE LIFE f~.SESSMENT PROGRAM FOR THE HARPOON J402Z-CA-400 ENGINE IV-303STINGER MINIMUM SIGNATURE LAUNCH MOTOR TEST PROGRAM V-89THlE AIR FORCE ENDOTHERMIC FULL PROGRAM IV-53THE JOINT EXPENDABLE TURBINE VNGINE CONCEPT (JETEC) DEMONSTRATOR PROGRAM

(U) VI-95VARIABLE-FUEL-FLOW DUCTED-ROCKET/RAMJET PROGRAM STATUS IV-407

PROGRAMSARMY INSENSITIVE MUNITIONS (IM) PROGRAMS AND OBJECTIVES FOR ARMY MISqILE

SYSTEMS V-239DEVELOPMENT OF THE ORBUS 1 CARBON FIBER/EPOXY MOT('1 CASE FOR THE STAIRS/

STARBIRD PROGRAMS 1-69PROJECT ILIE

LOW PROJECTILE ROCKET MOTOR CASING: rRAGMENTATION CONTROL TECHNOLOGY V-215PROJECT ILES

MINIATURIZED GUIDANCE P'ROPULSION FOR VOLUME-LIMITED GUN-I.AUNCHED INTERCE-PTOR PROJECTILES 111-261

PROP EL.L ANTA IHYDRODN*NAMIC CHEMISTRY MODEL OF THE LIQUID PROPELLANT GUN IV-493AFT (DOME INTERNAL INSULATION EROSION ON SMALL ICBM PRE-FSD AND F5D STAGE

I SOt.PJ PROPELLANT ROCKET 143tOR1S 111-45AN OVFRVItW OF *THE U.S. NAVIY 114:0 PROPELLANI TECHNOLOGY IN:TliAIVES 111-525BALL POWDER PH,)P:-LLt.NT APP4-L;IIAONS TO LARGE CALIBER A4ý14UNITfON 1-355CHARACTERIZ~ilON OF INSLNSIIIVE H!GH EXPLOSIVES DEVELO0ýD WiTH I't(0PEIl.'-NT

TE CH;4OLOGY 111-563CL.EAN PROPELLANT FOR LARGE SO..IU tAOCKE) MO)IDNs, if W-337G(JMtrOSITC OVEHWRAIt' POSITIVE CYPO' SION PRU?-LLANsi TANKS 11-307DEVELOFMENT AND TESTING OF THE SOLIC PRCýEI.t.ANT GAS GENLRATOP P.Ah:'RED

1'EACUKEEPER RAIL GARRISON CANISTER ERECTION SYSlEf' II -95DF-VELOPMENT OF A SAFE LIQIJID PROPELLAIDI LOGISTIC SYSTFM fot' I 0 A10I1-

ILRY IV*DEVCLOP14LNT OF AN HTPH PROPELLAN( %'171 1`10ANCED FXT INCTIC)N CIIAt'A~f`L8Is-

TI1CS V-19FLUID DYNAMIC/COMBUI)S' iN !NFERACTIWNS AS )RIVING MfC;;AhNiSM Of PIVLSCLRL

OSCIL'tA['ONS IN RGENCltAliVL LIQOID PROP.:LLANT CUN -6III SII-PifR' (XMA'CE , RF.tjCE0-SMOKE PROu'LLL-AN' I I3ý

'40(

PROPELLANT (cont'd)HIGH PRESSURE, LOW RATE, LOW OBSERVABLES PROPELLANT DEVELOPMENT - FINAL

REPORT 111-323HYPERVELOCITY BY EXTENDED PROPELLANT BURN 1-321INSENSITIVE GUN PROPELLANT - MODIFIED SINGLE BASE PROPELLANT IV-483INSENSITIVE MINIMUM-SIGNATURE PROPELLANTS (IMAD-301 PRCeELLANT) 111-341INVESTIGATING THE RATE EFFECT ON THE CRACK GROWTH REHhVIOR IN A COMPOSITE

SOLID PROPELLANT III-423KKV PROPELLANT TANK DEMONSTRATION PROGRAM 111-131LARGE-SCALE MOTOR/HAZARD TESTING OF HIGH PERFORMANCE DENSE ADDITIVE

(81203) PROPELLANT (U) VI-73LIQUID PROPELLANT ROCKET ENGINE COMPONENT COST MODEL V-521MICROENCAPSULATED INGREDIENTS FOR PROPELLANT APPLICATIONS 111-411MIXED FORMALS OF 2,2-DINITROPROPANOL AND 2,2-DIN!TROBUTANOL AS SOLID

PROPELLANT PLASTICIZERS V-41PROCESS AND QUALITY IMPROVEMENTS FOR M7 PROPELLANT 1-171PROGRESS REPORT ON LIQUID PROPELLANT INJECTOR/COMBUSTOR TESTS 1-363PROPELLANT COMBUSTOR USING CHEMICAL. KINETICS IV-501FROPELLANT PROCESS CONTROL OPTIMIZATION STUOY WITH HYPB PROPELLANT 111-371PSAN EFFECTS ON MINIMUM SIGNATURE PROPELLANT PROPERTIES V-31QUALIFICATION OF A CLEANER PROPELLANT FOR THE TITAN RETRO V-347REACTION KINETIC MODELING OF STABILIZER DEPLETION IN 146 PROPELLANT IV-423REGENERATIVE LIQUID PROPELLANT GUN AND SELF PROPELLED HOWITZER INTEGRA-

TION 1-401SEMI-COMMERCIAL TPEs FOR LOVA GUN PROPELLANT BINDERS IV-469SHELF LIFE DETERMINATION OF HIGH BURN RATE NEPE PROPELLANT 111-181SUPPRESSION OF IGNITION RISE RATE IN SOLID PROPELLANT ROCKET MOTORS V-107SYSILM REQUIREMENTS AND INTEGRATION FOR A FULLY AU7OMATED REGENERATIVE

LIQUID PROPELLANT CANNON IV-521THE EFFECT CF ACOUSTIC DAMPENING DEVICES IN REGEMERATIVE LIQUID PROPEL-

LANT GUNS 1-389THE IMPACT OF MARGIN ON THE ADVANCED LAUNCH SYSTEM LIQUID PROPELLANT

ENGINE V-531THE INFLUENCE OF PROPELLANT EQUILIBRIUM MODULUS ON ROCKET MOTOR SERVICE

LIFE PREDIC1IONS V-167PROPELLANT-BOND

MECHANICAL PROPERTIES OF THE PEACEKEEPER STAGE II PROPELLANT-BOND SYSTEM:AN EXAMINATION OF VARIABILITY Il1-1

PROPELLANY/MATERIALLONG-TERM PROPELLANT/MATERIAL COMPAFIBILITY 11-299

PROPELLANTSACCELERATED PROCESSING Oi HTPB PROPELLANTS WITH TMXDI V-67ADVANCED LAUNCH VEHICLE UPPER STAGES USING LIQUID PROPULSION AND METAL-

LIZED PROPF'.LANTS 11-251AGING CHARAC T E.,ISTICS OF HIGH BURNINC RATE CATOCENE CONTAINING PROPEL-

LANTS (U) VI-1AGING CHARACTERISTICS OF PROPELLANTS CONTAINING GAP AND AN 111-351APPROACH FOR LOW cXPONENT AND TEMPERATURE SENSITIVITY IN HIGH ENERGY

PROPELLANTS II1-11BURNING RATE CATALYSIS OF INSENSITIVE MUNITIONS AN PROPELLANTS 111-55521IRNING RATE ENHANCEMENT BY ELECTROMAGNETIC ALIGNMENT OF STAPLES IN SOLID

PhZ"ELLANTS V-11DESIGN AND DEVELOPMENT OF A LOW COST LIGHTWEIGHT, FAST RESPONSE ATTITUDE

CONTROL SYSTEM ENG;NF USING CLF5/N2H4 PROPELLANTS 111-475DESIGN AND DEVELOPMENT OF LIGHTWEIGHT AXIAL PROPULSION ENGINE USING CLF5/

N2H4 PROPELLANTS 111-113DEVELOPMENT OF ADHESIVE FOR BONDING COMPOSITE PROPELLANTS TO ABLATIVE

MATERIAL 111-381DEVELOPMENT OF SCAVENGED PROPELLANTS FOR ALS V-329ENERGETIC MINIMUM SMOKE PROPELLANTS FOR IN EVALUATION V-225HAZARD CLASSIFICATION OF LIQUID PROPELLANTS IV-525HELIUM SATURATION OF LIQUID PROPELLANTS 11-349INSENSITIVE MINIMUM-SIGNATURE PROPLLI.ANTS (INAD-301 PROPELLANT) 111-341LASER INITIATION OF PROPELLANTS AND EXPLOSIVES V-115LITERATURE REV!EW OF PLATEAU BALLISTICS !N NONALUNINIZEG SOLID PROPEL-

LANTS 111-387FROPELLANTS 111-223ROCKET PROPELLANTS FOR INSENSITIVE MUNIIIGNS R"ESEARCH 111-545fiLIGCED NONDFTONAgLE PROPELLANT . V-55TOUGH PROP'ELLANFI FNRKED fROM TETRAFUNCTIONAL BINDERS 111-315

PROPrLL EDPfGENCPAlIVE L..IQUI) PROPELLAN" GUN AND SELF PROPELLED HOWITZER INTEGRA-

"iON 1-401PROPUI.SION

A UNIQUf HYBRID PROPULSION SYSTEM DESIGN FOR LARGE SPACE BOOSTERS 11-159ADVANCEO LAI'NCH VEI;CLE UF`PR STAGES USING LIQUID PROPULSION AND METAL-

L IZO PROPELLANTS 11-251DESIGN AND DLVELOPMEN! OF LIGHTWEIGHT AXIAL PROPULSION ENGINE USING CLF5/

N2H4 PROPFLLANTS 111-113DESIGN OPTIMIZATIO$f OF GAS GENERATOR HYBRID PROPULSION DOOSTERS 11-199DEVCiOPMENT GF LOW COST PROPULSION REQUIREMENTS 11-97Fi.IGHTWEICH; DIVERT PROPULSION SYSTEM HOT-FIRE TEST 1-61HYBRID PROI'PULSION BOOTLRS FOR SPACE LAUNCH VEHICLES 11-169

441

PROPULSION (cont'd)HYBRID PROPULSION TECHNOLOGY PROGRAM I 1-143IMPACT OF MISSION REQUIREMENTS ON PROPULSION SUBSYSTEMS 111-253

LEAP SOLID PROPULSION DIVERT SUBSYSTEM DESIGN 111-165MINIATURIZED GUIDANCE PROPULSION FOR VOLUME-LIMITED GUN-LAUNCHEO INTERCE-

PTOR PROJECTILES 111-261MINIMUM SIGNATURE INSENSITIVE MUNITION GEL BIPROPELLANT PROPULSION V-247MONOPROPELLANT NYURAZINE SPACECRAFT PROPULSION SYSTEMS - 30 YEARS OF

SAFE, RELIABLE, FLEXIBLE AND PREDICTABLE PERFORMANCE 11-367,jtBIT TRANSiER VEHICLE PROPULSION DESIGN: TRADES AND COMPARISONS 11-291OVERVIEW OF NASA SOLID PROPULSION INTEGRITY PROGRAM (SPIP) BONDLINE WORK

PACKAGE RESUL.$ ANU ACCOMPLISHMENTS V-301PROPULSION STUDY FOR AN EXTENDED RANGE PATRIOT (U) VI-61PROPULSION bYSTEM HAZARDS AND ENERGETIC INGREDIENTS SYNTHESIS EVALUATION V-1PULSE MOTOR PROPULSION FOR TACTICAL MISSILES 111-293SOlO ELECTRIC PROPULSION REQUIREMENTS 11-383SERVICE LIFE MONITORING OF INTEGRAL ROCKET RAMJET (IRR) PROPULSION

SYSTEMS IV-319SOLID ROCKET PROPULSION APPLICATIONS FOR ADVANCED POLYMERS 1-103TECHNOLOGY FOR PROOUCIBLE TACTICAL LIQUID MONOPROPELLANT PROPULSION V-99TRAJECTORY/ENERGY MANAGEMENT DESIGN CONSIDERATIONS IN PROPULSION STUDIES

OF PSDI BATTLE SPACE 111-435WJ119-2 MODULAR TURBOJET PROPULSION SYSTEM FOR TAC1ICAL MISSILES IV-251

PSANPSAN EFFECTS ON MINIMUM SIGNATURE PROPELLANT PROPERTIES V-31

PSDITRAJECTORY/ENERGY MANAGEMENT DESIGN CONSIDERATIONS IN PROPULSION STUDIES

OF PSDI BATTLE SPACE II[-435PULSE

PERFORMANCE IMPROVEMENTS FOR SHORT-RANGE TACTICAL MISSILES USING PULSEMOTORS 111-305

PULSE MOTOR PROPULSION FOR TACTICAL MISSILES 111-293PULSING

NONLINEAR STABILITY TESTING AND PULSING OF FULL SCALE TACTICAL MOTORS 1-135

QUALIFICATIONQUALIFICATION OF A CLEANER PPOPELLANT FOR THE TITAN RETRO V-347

QUALITYPROCESS AND QUALITY IMPROVEMENTS FOR 97 PROPELLANT 1-171

RADIATION-COOLEDCHARACTER!ZATION OF IRIDIUM COATED RHENIUM USED IN HIGH-TEMPERATURE,

RADIATION-COOLED ROCKET THRUSTERS 11-359RAIL

DEVELOPMENT AND TESTING OF THE SOLID PROPELLANT GAS (PENERATOR POWEREDPEACEKEEPER RAIL GARRISON CANISTER ERECTION ZYSTEM 111-95

RAMBURNERHIGH SPEED TURBORAMJET RAMBURNER COMPONENT TEST RESULTS IV-61

RAMJETABLATIVE INSULATORS FOR RAMJET ENGINES IV-327AN EXPERIMENTAL EVALUATION OF COMBUSTOR LINER MATERIALS FOR SOLID FUEL

RAMJET TESTING 1-227DEVELOPMENT AND TESTING OF A HIGH-ENERGY LOW-COST SOLID RAMJET FUEL AT

THE NAVAL WEAPONS CENTER (U) VI-113LIQUID FUEL RAMJET HYDROCARBON FUELS: COMBUSTION EVALUATION (U) VI-135PERFORMANCE ANALYSIS OF A LOW-DRAG, VARIABLE GEOMETRY NOZZLE, INTEGRAL

ROCKET RAKJEr AIR-TO-SURFACE MISSILE (U) VI-115PERFORMANCE TRADES FOR AIR-LAUNCHED TURBOJET AND RAMJET CYCLE HIGH MACH

CRUISE NISSILES WITH ROCKET OR TURBOJET BOOST IV-1PRELIMINARY DESIGN AND PERFORMANCE OF AN ADVANCED RAMJET MISSILE CONCEPT

FOR THE CIRCA-2010 OUTER AIR BATTLE (U) VI-103REAL GAS DATA REDUCTION METHODS FOR RAMJET TESTING IV-353SERVICE LIFE MONITORING OF INTEGRAL ROCKET RAMJET (IRR) PROPUISION

SYSTEMS IV-319THE IMPACT OF DESIGN OPTIMISM ON THE PERFORMANCE OF A RAMJET MISSILE IV-125

RATEAGING CHARACTERISTICS OF HIGH BURNING RATE CATOCENE CONTAINING PROPEL-

LANTS (U) VI-1BURNING RATE CATALYSIS OF INSENSITIVE MUNITIONS AN PROPELLANTS 111-555BURNING RATE ENHANCEMENT BY ELECTROMAGNETIC ALIGNMENT OF STAPLES IN SOLID

PROPELLANTS V-11HIGH PRESSURE, LOW RATE, LOW OBSERVABLES PROPELLANT DEVELOPMENT - FINAL

REPORT 111-323INVESTIGATING THE RATE EFFECT ON THE CRACK GROWTH BEHAVIOR IN A COMPOSITE

SOLID PROPELLANT 111-423SHEL.F LIFE DETERMINATION OF HIGH BURN RATE NEPE PROPELLANT III-1iSUPPRESSION OF IGNITION RISE RATE IN SOLID PROPELLANT ROCKET MOTORS V-107

RATIOAIRFLOW MODEL TESTING TO DETERMINE THE DISTRIBUTION OF HOT GAS FLOW AND

O/F RATIO ACROSS THE iPACE SHUT TLE MAIN ENGINE MAIN INJECTOR ASSEMBLY 11-117ARCJET NOZZLE AREA RATIO EFFECTS 11-407

REACTIONHEATS OF EXPLOSION, OCTONATION AND REACTION PRODUCTS: THEIR ESTIMATION

442

REACTION (cont'd)AND RELATION TO THE FIRST LAW OF THERMOOYNAMICS 1-345

REACTION KINETIC MODELING OF STABILIZER DEPLETION IN 16 PROPELLANT IV-423REDUCED-SMOKE

HICII-PERFORNANCE, REDUCED-SMOKE PROPELLANT 111-335REGENERATIVE

FLUID DYKAMIC/COMBUSTION INTERACTIONS AS DRIVING MECHANISM OF PRESSUREOSCILLATIONS IN A REGENERATIVE LIQUID PROPELLANT GUN 1-369

REGENERATIVE LIQUID PROPELLANT GUN AND SELF PROPELLED HOWITZER INTEGRA-TION 1-401

SYSTEM REQUIREMENTS AND INTEGRATION FOR A FULLY AUTOMATED REGENERATIVELIQUID PROPELLANT CANNON IV-521

THE EFFECT OF ACOUSTIC DAMPENING DEVICES IN REGENERATIVE LIQUID PROPEL-LANT GUNS 1-389

RE INFORCEDCHARACTERIZATION OF LOW-DENSITY VARBON FIBER REINFORCED COMPOSITE V-4l83

RESIDUALFORMATION AND RESIDUAL STRENGTH IN COMPOSITE ROCKET MOTOR CASES 111-273

RESINTHE RHEOLOGICAL ANALYSIS OF PHENOLIC RESIN USED IN ABLATIVES 111-69

RETROQUALIFICATION OF A CLEANER PROPELLANT FOR THE TITAN RETRO V-347

REUSABLEIMPACT OF PROPULSIVE ADVANCEMENTS ON CAPABILITIES OF REUSABLE EARTH-TO-

ORBIT SHIPS 11-21RHENIUM

CHARAC'tERIZATION OF IRIDIUM COATED RHENIUM USED IN HIGH-TEMPERATURE,RADIATION-COOLED ROCKET THRUSTERS 11-359

RHECOLOGICALTHE RHEOLOGICAL ANA.LYSIS OF PHENOLIC RESIN USED IN ABLATIVES 111-69

R ISKDEVELOPMENT OF THE NLOS TECHNICAL RISK REDUCTION SUSTAINER MODULE IV-259

ROCKETA SYSTEM ýGl THE TESTING OF HYBRID ROCKET MOTORS 11-217A THREE-DIMENSIONAL TURBULENT HEAT TRANSFER ANALYSIS FOR ADVANCED TUBULAR

ROCKET THRUST CHAMBERS 11-273A VERSATILE CHLORINE PENTAFLUORIDE/HYDRAZINE (CIF5/N2H4) ALTITUDE ROCKET

ENGINE TEST FACILITY 111-213ADVANCING ENDOATMOSPHERIC INTERCEPTOR SOLID ROCKET MOTOR (SRM) TECHNOLOGY

(U) VI-11AFLAS: AN EXTERNAL INSULATION FOR ROCKET MOTORS V-123AFT DOME INTERNAL INSULATION EROSION ON SMALL ICBM PRE-FSD AND FSD STAGE

I SOLID PROPELLANT ROCKET MOTORS 111-45ALTITUDE SIMULATION TESTING FACILITY CONCEPTS FOR LARGE LIQUID-PROPELLANT

ROCKET ENGINES V-577AN EFFICIENT, INTELLIGENT SOLUTION FOR VISCOUS FLOWS INSIDE SOLID ROCKET

MOTORS 11-47CHARACTERIZATION OF IRIDIUM COATED RHENIUM USED IN HIGH-TEMPERATURE,

RADIATION-COOLED ROCKET THRUSTERS 11-359CLEAN PROPELLANT FOR LARGE SOLID ROCKET MOTORS, II V-337COMPOSITE DUCTED POCKET INLET DEMONSTRATION IV-397DEVELOPMENT OF A HICH-ENTHALPY TEST CAPABILITY USING A LIQUID-PROPELLANT

ROCKET ENGINE IV-113DEVELOPMENT OF A PINTLE CONTROLLED VARIABLE THRUST SOLID ROCKET MOTOR 111-241DEVELOPMENT OF THE FIBER OPTIC GUIDED SKIPPER ROCKET MOTOR (U) VI-31ERINT-1 SOLID ROCKET MOTOR REQUIREMENTS AND STATUS (U) VI-43EVALUATION OF PROPOSED ROCKET ENGINES FOR EARTH-TO-ORBIT VEHICLES 11-73EVOLUTION OF AN AGING PROGRAM - MINUTEMAN STACF 11 SOLID ROCKET MOTOR

1-45 1-45FEASIBILITY STUDY FOR EMPLOYING SOLiD ROCKET tXA,•UGTION SIMULATORS FOR

SOLID ROCKET MOTOR NOZZLE TESTING V-417FORMATION AND RESIDUAL STRENGTH IN COMPOSITE ROCKET MOTOR CASES 111-273GLYCIDYL AZIDE POLYMER (CAP) SOLID GAS GENERATOR FOR DUCTED AND HYBRID

ROCKET MOTORS V-47LIQUID PROPELLANT ROCKET ENGINE COMPONENT COST MODEL V-5221LOW PROJECTILE ROCKET MOTOR CASING: FRAGMENTATION CONTROL TECHNOLOGY V-215NOE OF THERMAL PROTECTION SYSTEM FOR SPACE SHUTTLE SOLID FOCKET BOOSTER I1-1NEW METHOD OF MAKING ADVANCED TUBE-BUNDLE ROCKET THRUST CHAMBERS V-557NON-ACOUSTIC COMBUSTION INSTABILITY IN IIYBRID ROCKET MOTORS 11-177PASSIVE COMBUSTION CONTROL IN A DUCTED ROCKET WITH SIDE MOUNTED INLETS IV-363PERFORMANCE ANALYSIS OF A LOW-DRAG, VARIABLE GEOMETRY NOZZLE, INTEGRAL

ROCKET RAMJET AIR-TO-SURFACE MISSILE (U) VI-115PERFORMANCE TRADES FOR AIR-LAUNCHED TURBOJET AND RAMJET CYCLE HIGH MACH

CRUISE MISSILES WITH ROCKET OR TURBOJET BOOST IV-1PREDICTION OF THE THRUST PERFORMANCE AND THE FLOWFIELD OF LIQUID ROCKET

ENGINES I 1-133ROCKET PROPELLANTS FOR INSENSITIVE MUNITIONS RESEARCH 111-545SERVICE LIFE ANALYSIS OF THE PATRIOT ROCKET MOTOR V-197SERVICE LIFE MONITORING OF INTEGRAL ROCKET RAMJET (IRR) PROPULSION

SYSTEMS IV-319SLAG ESTIMATION FOR SOLID POCfET MOTORS 11-225SOLID ROCKET BOOSTER INTEGPATION WITH THE AQM-37C MISSILE TARGET 1-111SOLID ROCKET PROPULSION APPLICATIONS FOR ADVANCED POLYVERS 1-103

443

ROCKET (cont'd)SUPPRESSION OF IGNITION RISE RATE IN SOLID PROPELLANT ROCKET MOTORS V-107THE DEVELOPMENT OF A LIQUID FUELED GAS GENERATOR FOR DUCTED ROCKET

RESEARCH 1-313THE INFLUENCE OF PROPELLANT EQUILIBRIUM MOOULUS ON ROCKET MOTOR SERVICE

LIFE PREDICTIONS V-167VARIABILITY IN C-C THROAT EROSION FOR A SELECTION OF SOLID ROCKET MOTORS 111-77

ROLLINGA COMPOSITE OVERWRAPPED BONDED ROLLING DIAPHRAGM EXPULSION SYSTEM 111-141

ROTARYDEVELOPMENTS IN THE DESIGN OF AN AIRBORNE ROTARY AIR SEPARATOR V-377

RP-1COMPATIBILITY OF COMBUSTION CHAMBER LINERS WITH RP-1 V-541

RUGGEDRUGGED NONDETONABLE PROPELLANTS V-55

SAFE/ARM/FIREFEASIBILITY DEMONSTRATION OF AN ELECTRO-OPTICAL SEMICONDUCTOR BRIDGE

SA0E/ARM/FIRE INITIATION SYSTEM 111-515SATURATION

HELIUM SATURATION OF LIQUID PROPELLANTS 11-349SCALE

NONLINEAR STABILITY TESTING AND PULSING OF FULL SCALE TACTICAL MOTORS 1-135STABILITY TESTING OF FULL SCALE TACTICAL MOTORS 1-119

SCAVENGEDDEVELOPMENT OF SCAVENGED PROPELLANTS FOR ALS V-329

SCRAMJETA PARAMETRIC STUDY OF 3 HYDROCARBON FUELED SCRAMJET IV-309HYDROCARBON-FUELED SCR.AMJET COMBUSTOR INVESTIGATION (U) VI-125IMPLICATIONS OF THE RESULTS OBTAINED IN THE GRUMMAN DETONATION SHOCK

TUNNEL NOZZLE EXPERIMENTS ON SCRAMJET PERFORMANCE IV-221SDI

INSENSITIVE MUNITIONS MASTERPLAN AND ITS IMPLICATIONS FOR THE SD, COM-MUNITY 1-55

SDI/KEWSDI/KEW ULTRALIGHT WEIGHT NOZZLE STRUCTURES 111-233

SDIOSDOO ELECTRIC PROPULSION REQUIREMENTS 11-383

SEMI-COMMERCIALSEMI-COMMERCIAL TPE$ FOR LOVA GUN PROPELLANT BINDERS IV-469

SEMICONDUCTORFEASIBIIlTY DEMONSTRATION OF AN ELECTRO-OPTICAL SEMICONDUCTOR BRIDGE

SAFE/ARM/FIRE INITIATION SYSTEM 111-515SENSITIVITY

APPROACH FOR LOW EXPONENT AND TEMPERATURE SENSITIVITY IN HIGH ENERGYPROPELLANTS II1-11

SENSORNONCONTACTING MULTIPARAMETRIC SENSOR FOR SHAFT SPEED, TORQUE, AND POSI-

TION 11-281SEPARATOR

DEVELOPMENTS IN THE DESIGN OF AN AIRBORNE ROTARY AIR SEPARATOR V-377SERVICE

CHAPARRAL SERVICE LIFE EVALUATION V-157RELIABILITY ANALYSIS OF THE HAWK MOTOR SERVICE LIFE PREDICTIONS V-177SERVICE LIFE ANALYSIS OF THE PATRIOT ROCKET MOTOR V-197SERVICE LIFE ASSESSMENT PROGRAM FOR THE HARPOON J402-CA-400 ENGINE IV-303SERVICE LIFE MONITORING OF INTEGRAL ROCKET RAMJET (IRR) PROPULSION

SYSTEMS IV-319SERVICE LIFE TESTING AND ANALYSIS OF FIELD AGED MLRS MOTORS V-187THE INFLUENCE OF PROPELLANT EQUILIBRIUM MODULUS ON ROCKET MOTOR SERVICE

LIFE PREDICTIONS V-167SHAFT

NONCONTACTING MULTIPARAMETRIC SENSOR FOR SHAFT SPEED, TORQUE, AND POSI-TION 11-281

SHEARJANNAF STANDARDIZATION OF TENSION, COMPRESSION, AND SHEAR TEST METHODS TO


PRESSURE GRADIENT EFFECTS IN SUPERSONIC SHEAR LAYERS IV-177SHEAR STRESS TRANSDUCER CONCEPTS 1-151

SHELFSHELF LIFE DETERMINATIQN OF HIGH BURN RATE NEPE PROPELLANT 111-181

SHIPSIMPACT OF PROPULSIVE ADVANCEMENTS ON CAPABILITIES OF REUSABLE EARTH-TO-

ORBIT SHIPS 11-21SHOCK

IMPI.ICATIONS OF THE RESULTS OBTAINED IN THE GRUMMAN DETONATION SHOCKTUNNEL NOZZLE EXPERIMENTS ON SCRAMJET PERFORMANCE IV-221

SHOCK WAVE/BOUNDARY LAYER INTERACTION CONTROL IN A GENERIC HYPERSONICINLET 1-235

THE CALIBRATION OF REAL GAS TEST FACILITIES: PHOBLEAw AND PROGRESS ON ACOMBUSTION DRIVEN SHOCK TUNNEL EXPERIMENT IV-103

444

SHORT-RANGEPERFORMANCE iMPROVEMENTS FOR SHORT-RANGE TACTICAL MISSILES USING PULSE

MOTORS 111-305SHROUDED

PERFORMANCE CHARACTERISTICS OF SHROUDED SUPERSONIC RECTANGULAR ANDTAPERED ELLIPTIC JETS IV-133

SHUTTLZAIRFLOW MOOEL TESTING TO DETERMINE THE DISOC "'ON OF H4T GAS FLOW AND

O/F RATIO ACROSS THE SPACE SHUTTLE MAIO rh.- MA14 INJECTOR ASSEMBLY 11-117HOE OF THERMAL PROTECTION SYSTEM FOR SPA E S' I ROCKET BOOSTER I1-iTEST RESULTS OF THE MODIFIED SPACE SHUTTLE %T THE MARSHALL

SPACE FLIGHT CENTER TECHNOLOGY TEST jfr 11-107SICBm

3-D ANALYSIS OF SICBM STAGE I BORE TO 1 ,lh COMPARiSON TOTHE AXISYMMETRIC ANALYSIS 111-19

SIGNATUREMINIMUM SIGNATURE INSENSITIVE MUNITION GEL b.,R' V1" AhT PROPULSION V-247PSAN EFFECTS ON MINIMUM SIGNATURE PROPELLANT 'O'3PLITIES V-31STINGER MINIMUM SIGNATURE LAUNCH MOTOR TEST PROGRAO V-89

SIMULATIONALTITUDE SIMULATION TESTING FACILITY CONCEPTS FOR LARGE L'OUID-PROPELLANT

ROCKET ENGINES V-577SIMULATORS

FEASIBILITY STUDY FOR EMPLOYING SOLID RCCVET COMBUSTION SIMULATORS FORSOLID ROCKET MOTOR NOZZLE TESTING V-417

SKIPPERDEVELOPMENT OF THE FIBER OPTIC GUIDED SKIPPER ROCKET Mf)TOR (U) VI-31

SLAGSLAG ESTIMATION FOR SOLID ROCKET MOTORS 11-225

SMOKEENERGETIC MINIMUM SMOKE PROPELLANTS FOR IM EVALUATION V-225

SOLARANALYSIS OF EXPENDABLE SOLAR ELECTRIC ORBIT TRANSFeR VEHICLES 11-417

SOLIDADVANCING ENDOATMOSPHERIC INTERCEPTOR SOLID ROCKET MOTOR (SRM) TECHNOLOGY

(U) VI-11AFT DOME INTERNAL INSULATION EROSION ON SMALL ICBM PRE-FSD AND FSD STAGE

I SOLID PROPELLANT ROCKET MOTORS II I-45AN EFFICIENT, INTELLIGENT SOLUTION FOR VISCOUS FLOWS INSIDE SOLID ROCKET

MOTORS 11-47AN EXPERIMENTAL EVALUATION OF COMBUSTOR LINER MATERIALS FOR SOLID FUEL

RAMJET TESTING 1-227BURNING RATE ENHANCEMENT BY ELECTROMAGNETIC ALIGNMENT OF STAPLES IN SOLID

PROPELLANTS V-11CLEAN PROPELLANT FOR LARGE SOLID ROCKET MOTORS, II V-337DEVELOPMENT AND TESTING OF A HIGH-ENERGY LOW-COST SOLID RAMJET FUEL AT

THE NAVAL WEAPONS CENTER (U) V1-173DEVELOPMENT AND TESTING OF THE SOLID PROPELLANT GAS GENERATOR POWERED

PEACEKEEPER RAIL GARRISON CANISTER ERECTION SYSTEM 111-95DEVELOPMENT OF A PINTLE CONTROLLED VARIABLE THRUST SOLID ROCKET MOTOR i11-241ERINT-1 SOLID ROCKET MOTOR REQUIREMENTS AND STATUS (U) VI-43EVOLUTION OF AN AGING PROGRAM - MINUTEMAN STAGE II SOLID ROCKET MOTOR

1-45 1-45FEASIBILITY STUDY FOR EMPLOYING SOLID ROCKET COMBUSTION SIMULATORS FOR

SOLID ROCKET MOTOR NOZZLE TESTING V-417GLYCIDYL AZIDE POLYMER (GAP) SOLID GAS GENERATOR FOR CI•',ED AND HYBRID

ROCKET MOTORS V-47INVESTIGATING THE RATE EFFECT ON THE CRACK GROWTH BEHAVIOR IN A COMPOSITE

SOLID PROPELLANT 111-423LEAP SOLID PROPULSION DIVERT SUBSYSTEM DESIGN 111-165LITERATURE REVIEW OF PLATEAU BALLISTICS IN NONALUMINIZED SOLID PROPEL-

LANTS 111-381MIXED FORMALS OF 2,2-DINITROPROPANOL AND 2,2-DINITROBUTANOL AS SOLID

PROPELLANT PLASTICIZERS V-41NDE OF THERMAL PROTECTION SYSTEM FOR SPACE SHUTTLE SOLID ROCKET BOOSTER ;1-1OVERVIEW OF NASA SOLID PROPULSION INTEGRITY PROGRAM (SPIP) BONDLINE WORK

PACKAGE RESULTS AND ACCOMPLISHMENTS V-301SLAG ESTIMATION FOR SOLID ROCKET MOTORS 11-225SOLID ROCKET BOOSTER INTEGRATION WITH THE AQM-37C MISSILE TARGET 1-111SOLID ROCKET PROPULSION APPLICATIONS FOR ADVANCED POLYMERS 1-103SUPPRESSION OF IGNITION R;AE RATE IN SOLID PROPELLANT ROCKET MOTORS V-lu?VARIABILITY IN C-C THROAT EhI',SION FOR A SELECTION OF SOLID ROCKET MOTORS lli-77

SOLID-LIQUIDSOLID-LIQUID STAGED COMBUSTION SPACE BOOSTERS 11-189

SOLUTIONAN EFFICIENT, INTELLIGENT SOLUTION FOP VISCOUS FLOWS INSIDE SOLID ROCKET

MOTORS 11-47SONIC

LINER MATERIALS BY SONIC VELOCITY MEASUREMENTS V-319SPACE

A UNIQUE HYBRID PROPULSION SYSTEM DESIGN FOR LARGE SPACE BOOSTERS 11-159AIRFLOW MODEL TESTING TO DETERMINE THE DISTRIBUTION OF HOT GAS FLOW AND

O/F RATIO ACROSS THE SPACE SHUTTLE MAIN ENGINE MAIN INJECTOR ASSEMBLY 11-117

445

SPACE (cont'd)HYBRID PROPULSION BOOSTERS FOR SPACE LAUNCH VEHICLES 11-169NOE OF THERMAL PROTECTION SYSTEM FOR SPACE SHUTTLE SOLID ROCKET BOOSTER I1-1SOLID-LIQUID STAGED COMBUSTION SPACE BOOSTERS 11-189TEST RESULTS Of THE MODIFIED SPACE SHUTTLE MAIN ENGINE AT THE MARSHALL

SPACE FLIGHT CENTER TECHNOLOGY TEST BED FACILITY 11-107TRAJECTORY/ENERGY MANAGEMENT DESIGN CONSIDERATIONS IN PROPULSION STUDIES

OF PSOI BATTLE SPACE 111-435SPACECRAFT

MONOPROPELLANT HYDRAZINE SPACECRAFT PROPULSION SYSTEMS - 30 YEARS OFSAFE, RELIABLE, FLEXIBLE AND PREDICTABLE PERFORMANCE 11-367

SPECIMENPREDICTION Of BONDLINE STRENGTH BY UNIQUE SPECIMEN DESIGN AND CUMULATIVE

DAMAGE CONCEPTS V-145SELECTION OF BOND-IN-TENSION SPECIMEN FOR THE NASA SPIP BONDLINES WORK-

PACKAGE V-287SPEED

ADVANCED COMPUTATIONAL MODELS FOR ANALYZING HIGH SPEED PROPULSIVE FLOW-FIELDS I-247

HIGH SPEED TURBORANJET RANBURNER COMPONENT TEST RESULTS IV-61NONCONTACTING MULTIPARAMETRIC SENSOR FOR SHAFT SPEED, TORQUE, AND POSI-

TION 11-281SPIP

NASA SPIP BONDLINE WORK PACKAGE: OVERVIEW OF THE SRH MODEL PROCESS ANDPROCESS FMEA TASK V-265

OVERVIEW OF NASA SOLID PROPULSION INTEGRITY PROGRAM (SPIP) BONDLINE WORKPACKAGE RESULTS AND ACCOMPLISHMENTS V-301

SELECTION OF BOND-IN-TENSION SPECIMEN FOR THE NASA SPIP BONDLINES WORK-PACKAGE V-287

SRMADVANCING ENDOATMOSPHERIC INTERCEPTOR SOLID ROCKET MOTOR (SRM) TECHNOLOGY(U) VI-11

NASA SPIP 8.3NOLINE WORK PACKAGE: OVERVIEW O0 THE SRM MOOEL PROCESS ANDPROCESS FMEA TASK V-265

STABILITYNONLINEAR. STABILITY TESTING AND PULSING OF FULL SCALE TACTICAL MOTORS 1-135STABILITY TESTING OF FULL SCALE TACTICAL MOTORS 1-119

STAGE3-D ANALYSIS OF SICBM STAGE I BORE TO FIN TRANSITION WITH COMPARISON TO

THE AXISYMMETRIC ANALYSIS 111-19AFT DOME INfERNAL INSULATION EROSION ON SMALL ICBM PRE-FSD AND FSO STAGE

I SOLID PROPELLANT ROCKET MOTORS 111-45EVOLUTION OF AN AGING PROGRAM - MINUTEMAN STAGE II SOLID ROCKET MOTOR

1-45 1-45MECHANICAL PROPERTIES OF THE PEACEKEEPER STAGE 11 PROPELLANT-BOND SYSTEM:

AN EXAMINATION OF VARIABILITY Il1-1MISSILE INTEGRATED STACE TECHNOLOGY (MIST) PROGRAM 111-153SYSTEM APPLICATIONS FOR AN ADVANCED LIQUID AXIAL STAGE, ALAS 111-447WEIGHT AND COST ANALYSIS OF IDENTICAL STAGE BOOSTER SET FOR EXO ATMOSPH-

ERIC INTERCEPTOR COMPONENT FLIGHT TESTS (U) VI-21STAGES

ADVANCED LAUNCH VEHICLE UPPER STAGES USING LIQUID PROPULSION kt:D METAL-LIZED PROPELLANTS 11-251

STANDARDIZATIONJANNAF STANDARDIZATION OF TENSION, COMPRESSION, AND SHEAR TEST METHOJS TO


STANDOFFVERTICAL LAUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMARINE WAR-

FARE (ASW) V-135STAPLES

BURNING RATE ENHANCEMENT BY ELECTROMAGNETIC ALIGNMENT OF STAPLES IN SOLIDPROPELLANTS V-11

STARS/STARBIRDDEVELOPMENT OF THE ORBUS 1 CARBON FIBER/EPOXY MOTOR CASE FOR THE STARS/

STARBIRD PROGRAMS 1-69STATE-OF-THE-ART

STAlE-OF-THE-ART 22N (5 LOF) THRUSTER FINDS BROAD APPLICATION 11-339STINGER

STINGER FLIGHT MOTOR CASE STRESS CORROSION EVALIUATION V-85SlINGER MINIMUM SIGNATURE LAUNCH MOTOR TEST PROGRAM V-89

STOICHIOMETRICSTOICHIOMETRIC GAS GENERATOR - A STRATEGIC DEPARTURE V-549

STRATEGICSTOICHIOMETRIC GAS GENERATOR - A STRATEGIC DEPARTURE V-549

STRESSNORMAL STRESS TRANSDUCER BEHAVIOR 1-161SHEAR STRESS TRANSDUCER CONCEPTS 1-151STINGER FLIGHT MOTOR CASE STRESS CORROSION EVALUATION V-85

STRUCTURALINVESTIGA1ION OF A FOUR POINT TEST FOR THE STRUCTURAL ANALYSIS EVALUATION

OF PHENOLIC MATERIALS 111-285STRUCTURAL DAMPING/ACOUSTIC ATTENUATION OPTIMIZATION VIA NON-OBSTRUCTIVE

446

STRUCTURAL (c-ont'd)PARTICLE DAMPING 11-63

STRUCTURESJANNAF STANDARDIZATION Of TENSION. COMPRESSION, AND SHEAR TEST METHODS TO


SOI/KEW ULTRALIGHT WEIGHT NOZZLE STRUCTURES 111-233STS

THE STS FLUID MANAGEEMENT DEMONSTRATION TEST BED V-593SUBMERGED

EXPERIMENTAL DEMONSTRATION OF THE OPERATION OF AN INSTALLED SUBMERGEDSUPERSONIC AIR INDUCTION SYSTEM (U) VI-81

SUBSCALESTUDY OF SUPERSONIC COMBUSTION PHENOMENOLOGY IN A SUBSCALE COMBUSTOR AT

SIMULATED FLIGHT M4ACH NUMBERS FROM 7 TO 10 IV-209SUBSCALE MOTOR CASE ANALOG FOR FAST COOK-OFF INSENSITIVE MUNITIONS TESTS V-209

SUPERSONICEXPERIMENTAL DEMONSTRATION OF THE OPERATION OF AN INSTALLED SUBMERGED

SUPERSONIC AIR INDUCTION SYSTEM (U) VI-81NASA LEWIS 3UPERSONIC THROUGHFLOW FAN PROGRAM IV-73NEW AND UPGRADED SUPERSONIC - HYPERSONIC AIRBREATHING ENGINE TEST FACILI-

TIES AT GASL IV-87PERFORMANCE CHARACTERISTICS OF SHROUDED SUPERSONIC RECTANGULAR AND

TAPERED ELLIPTIC JETS IV-133PRESSURE GRADIENT EFFECTS IN SUPERSONIC SHEAR LAYERS IV-177STUDY OF SUPERSONIC COMBUSTION PHENOMENOLOGY IN A SUBSCALE COMBUSTOR AT

SIMULATED FLIGHT MACH NUMBERS FROM 7 TO 10 IV-209SURFACE


MECHANICAL DESIGN OF SURFACE LAUNCHED TACTICAL MISSILES 1-181SUSTAINER

DEVELOPMENT OF THE NLOS TECHNICAL RISK REDUCTION SUSTAINER MODIAE IV-259

TACTICALMECHANICAL DESIGN OF SURFACE LAUNCHED TACTICAL MISSILES 1-181NONLINEAR STABILITY TESTING AND PULSING OF FULL SCALE TACTICAL MOTORS 1-135PERFORMANCE IMPROVEMENTS FOR SHORT-RANGE TACTICAL MISSILES USING PULSE

MOTORS II 1-305PULSE MOTOR PROPULSION FOR TACTICAL MISSILES 111-293STABILITY TESTING OF FULL SCALE TACTICAL MOTORS 1-119TECHNOLOGY FOR PROOUCI;3LE TACTICAL LIQUID MONOPROPELLANT PROPULSION V-99WJ119-2 MODULAR TURBOJET PROPULSION SYSTEM FOR TACTICAL MISSILES IV-251

TAGNMOLECULAR STRUCTURE OF LOW TEMPERATURE FORM OF TRIAMINOGUANIDINIUM

NITRATE (TAGN) 1-331TANK

BALLISTIC MODELING OF TWO-PIECE CARTRIDGES FOR A TANK GUN IV-459KKV PROPELLANT TANK DEMONSTRATION PROGRAM 111-131

TANKSCOMPOSITE OVERWRAP POSITIVE EXPULSION PROPELLANT TANKS 11-307DESIGN AND DEVELOPMENT OF CARBON FIBER WRAPPED NON-LOAD SHARING TANKS 111-4167

TAPEREDPERFORMANCE CHARACTERISTICS OF SHROUDED SUPERSONIC RECTANGULAR AND

TAPERED ELLIPTIC JETS IV-133TARGET

SOLID ROCKET BOOSTER INTEGRATION WITH THE AQM-37C MISSILE TARGET 1-111TEMPERATURE

APPROACH FOR LOW EXPONENT AND TEMPERATURE SENSITIVITY IN HIGH ENERGYPROPELLANTS III-i1

MOLECULAR STRUCTURE OF LOW TEMPERATURE FORM OF TRIAMINOGUANIDINIUMNITRATE (TAGN) 1-331

NP I1: A NEW LOW TEMPERATURE MELTING NITROPLASTICIZER IV-1417WITH-PLY PERMEABILITY OF CARBON PIIENOLIC COMPOSITES AS A FUNCTION OF

ACROSS-PLY COMPRESSIVE LOAD AND TEMPERATURE V-497TENSION


JANNAF STANOARDIZATION OF TENSION, COMPRESSION, AND SHEAR TEST METHODS TODETERMINE MECHANICAL MATERIAL PROPERTIES FOR FILAMENT WOUND COMPOSITESTRUCTURES 1-89

TETRAFUNCT IONALTOUGH PROPELLANTS FORMED FROM TETRAFUNCTIONAL BINDERS 111-315

THEORvA LINEARIZED lHEORY FOR UNSTEADY FLOW ALONG VANE-FORMED FILLETS DRIVEN BY

SURFACE TENSION 11-333THERMAL

MOISTURE EFFECTS ON ACROSS PLY MECHANICAL AND THERMAL CHARACTERISTICS OFFM 5055 CARBON PHENOLIC UTILIZN', RAPID ISOTHERMAL HEATING V-SO5

NDE OF THERMAL PROTECTION SYSTEM FOR SPACE SHUTTLE SOLID ROCKET BOOSTER I1-1THERMODYNAMICS

IIEAIS OF EXPLOSION, DETONATION AND REACTION PRODUCTS: THEIR ESTIMATIONAND RELATION TO THE FIIRST LAW OF TIIERMOOYNAMICS I-315

447

THREE-DIMENS IONALA THREE-DIMENSIONAL TURBULENT HEAT TRANSFER ANALYSIS FOR ADVANCED TUBULAR

ROCKET THRUST CHAMBERS 11-273THROAT

THROAT EROSION CHARACTERIZATION OF 4 DIMENSIONAL CARBON-CARBON AT HIGHPRESSURES V-449

VARIABILITY IN C-C THROAT EROSION FOR A SELECTION OF SOLID ROCKET MOTORS 111-77THROUGHFLOW

NASA LEWIS SUPERSONIC THROUGHFLOW FAN PROGRAM IV-73THRUST

A LIFE COMPARISON OF TUBE AND CHANINEL COOLING PASSAGES FOR THRUST CHAM-BERS V-571

A THREE-DIMENSIONAL TURBULENT HEAT TRANSFER ANALYSIS FOR ADVANCED TUBULARROCKET THRUST CHAMBERS 11-273

DESIGN OF A 250 LBF THRUST HYDRAZINE FUELED AIR TURBORANJET IV-233DEVELOPM4ENT Of A PINTLE CONTROLLED VARIABLE THRUST SOLID ROCKET MOTOR 111-241NEW METHOD OF MAKING ADVANCED TUBE-BUNDLE ROCKET THRUST CHAMBERS V-557PREDICTION OF THE THRUST PERFORMANCE AND THE FLOWFIELD OF LIQUID ROCKET

ENGINES 11-133THRUSTER

STATE-OF-THE-ART 22N (5 LBF) THRUSTER FINDS BROAD APPLICATION 11-339THRUSTERS

CHARACTERIZATION OF IRIDIUM COATED RHENIUM USED IN HIGH-TEMPERATURE,RADIATION-COOLED ROCKET THRUSTERS 11-359

TITANQUALIFICATION OF A CLEANER PROPELLANT FOR THE TITAN RETRO V-34.7

TITANIUMTITANIUM PLATELET INJECTOR DEVELOPMENT 111-207

TMXDIACCELERATED PROCESSING Or HTPB PROPELLANTS WITH TMXDI V-67

TORQUENONCONYACTING MULTIPARAMETRIC SENSOR FOR SHAFT SPEED, TORQUE, AND POSI-

TION 11-281TOUGH

TOUGH PROPELLANTS FORMED FROM TETRAFUNCTIONAL BINDERS 111-315TPES

SEMI-COMMERCIAL TPE% FOR LOVA GUN PROPELLANT BINDERS IV-469TRADES

ADiVANCED LAUNCH VEHICLE CONFIGURATION AND PERFORMANCE TRADES V-399ORBIT TRANSFER VEHICLE PROPULSION DESIGN: TRADES AND COMPARISONS 11-291PERFORMANCE TRADES FOR AIR-LAUNCHED TURBOJET AND RAMJET CYCLE HIGH MACH

CRUISE MISSILES WITH ROCKET OR TURBOJEt BOOST I-TRAJECTORY/ENERGY

TRAJECTORY/ENERGY MANAGEMENT DESIGN CONSIDERATIONS IN PROPUILSION STUDIESOF P501 BATTLE SPACE 111-435

TRANSDUCERNORMAL STRESS TRANSDUCER BEHAVIOR 1-161SHEAR STRESS TRANSDUCER CONCEPTS 1-151

TRIANINOOUANIDINIUMMOL.ECULAR STRUCTURE OF LOW TEMPERATURE FORM Of TRIANINOGUANIDINIUM

NI-(RATE (TAGN) 1-331TUBE-BUNDLE

NEW METHOD OF MAKING ADVANCED TUBE-BUNDLE ROCKET THRUST CHAMBERS V-557TUNNEL

HIGH PRESSURE MACH 10 TO 20 rLECTROTHERMAL HYPERSONIC WIND TUNNEL 1-211IMPLICATIONS OF THE RESULTS OBTAINED IN THE GRUMMEAN DETONATION SHOCK

TUNNEL NOZZLE EXPERIMENTS ON SCRANJET PERFORMANCE IV-221THE CALIBRATION OF REAL GAS TEST FACILITIES: PROBLEMS AND PROGRESS ON A

COMBUSTION DRIVEN SHOCK TUNNEL EXPERIMENT IV-103TURBINE

THE. JOINT EXPENDABLE TURBINE ENGINE CONCEPT (JETEC) DEMONSTRATOR PROGRAM(U) VI-95

TUkDO-NAMJE TCOOLING SYSTEM AND INSULATION CONCEPT FOR A M4ACH 5 TURBO-RAMJET AIRCRAFT 1-191

TURBOJETEVALUATION OF GENERIC EXPENDABLE TURBOJET CYCLE PECOS IV-289NOVEL 320-2: A COM4PACI ADVANCED UAV TURBOJET 1-285PERFORMANCE TRADES FOR A4R-LAUNCHED TURBOJET AND RAM4JET CYCLI HiGH MAClI

CRUISE MISSILES WITH ROCKET OR TURBOJET BOOST IV- 1WJ119-2 MODULAR TURBOJET PROPULSION SYSTEM FOR TACTICAL NII16S(LES IV-251

TUk8HIUOJTSIMPINGEMENT STARTING 0F SMALL EXPENOASLE TURBOJETS IV-279

TURBOPUMPSDAMPING RIEARINOS (OR CRYOUENIC TURlBOPUMPS V-5,65

TURBORANJEIDESIGN OF A ;e5O 1SBF THRUST IIYDRAZINE FUELED AIR TURBVA4AfJ(T IV-233HIGII SPEED TURBORANJIT RAMBURNI.14 COMPONENT TEST RESULTS IV-61

TURBOROCKE TAiR LIQUEF ACTION ANO ENKICIIMLNI SY~tEM (ALES) WITH TIIRHOROCKIII IO* OR-

BITAI VEHICLIS V-361TUNUHIFUNT

A TIIR(EL-IMI NSIONAL lUIIIIUt(INI lIRA? IRANSI IR ANAL YSIV 10OR AI)VANHLIU l'LIJI.AmROCKIT THRUST CIIAMIlfR3i II 1-23

448

TURBULENT (cont'd)PREMIXED, TURBULENT COMBUSTION OF AXISYMMETRIC SUDDEN EXPANSION FLOWS 1-305

UAVMOOEL 320-2: A COMPACT ADVANCED UAV TURBOJET 1-285

ULTRALIGHTSDI/KEW ULTRALIGHT WEIGHT NOZZLE STRUCTURES lil--233

UNDERWATERUNDERWATER WARHEAD GUN SYSTEM (U) VI-199

UNICHARGEPREDICTION OF INTERIOR BALLISTICS AND PRESSURE WAVES PRESENT IN UNICHARGE

DESIGN IV-447UNICHARGE CENTER CORE IGNITION IV-433

UNSTEADYA LINEARIZED THEORY FOR UNSTEADY FLOW ALONG VANE-FORMED FILLETS DRIVEN BY

SURFACE TENSION 11-333UPGRADED

NEW AND UPGRADED SUPERSONIC - HYPERSOhýC AIRBREATHING ENGINE TEST FACIL;-TIES AT GASL IV-87

VANE-FORMEDA LINEARIZED THEORY FOR UNSTEADY V:LOW ALONG VANE-FORMED FILLETS DRIVEN BY

SURFACE TENSION 11-333VARIABLE-FLOW

VARIABLE-FLOW GAS GENERATOR IV-339VARIABLE-FUEL-FLOW

VARIABLE-FUEL-FLOW DUCTEO-RtC.KET/RAMJET PROGRAM STATUS IV-407VARYING

VARYING FILAMENT--WOUND NOZZLE PROPERTIES USING SELECTED FIBER-MATRIXLAYERING V-473

VEHICLEADVANCED LAUNCH VEHICLE CONFICURATION AkD PERFORMANCE TRADES V-399ADVANCED !-AUWH VEHICLE UPPER STAGES USiNC LIQUID PROPULSION AND METAL-

LIZED PROPELLAIZIS 11-25.ALL-HYDROC.,SON ORB!TAI. LAUNCH VEHICLE V-359LIGHTWEISHT KINETIC V0;ICLE FREE FLiGHIT/HOVER DEMONSTRATIONS V-T7ORBIT fItA!'SrER VEHICLE PROPULSION DESIGN: TRADES AND COMPARISONS 11-291

VEHICLESAIR L'QUEFACTION AND ENRICHMENT SYSTEM (ALES) WITH TURBOROCKET FOR OR-

BITAL VEH!(GLES V-367AIALYSI5 Of EXPENDABLE SOLAR ELECTRIC ORBIT TRANSFER VEHICLES 11-417F'V UATION OF PROPOSED ROCKET ENGINES FOR EARTH-TO-ORBIT VEHICLES 11-73HYIKtU PROPULSION BOOSTERS FOR SPACE LAUNCH VEHICLES 11-169

VELOCITYLI'rR MATERIALS BY SONIC VELOCITY MEASUREMENTS V-319

VERSATILEA VERSATILE CWVORINE PENTAFLUORIDE/HYDRAZINE (CIF5/N2H4) ALTITUDE ROCXET

ENGINE TEST FACILITY IIt?1.VERTICAL

VERTICAL LAUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMA8IMF. WAf-FARE (ASW) V-135

VISCOUSAN EFFICIENT, INTELt ICENT SOLUTION FOR VISCOUS FLOWV' INSIDE SOLID RyI KET

MOTORS ; 1-7VLA

VERTICAL LAUNCH ASkOC (VLA) MISSILE SYSTEMS SIANGOFV ANT:SUBMARINE WAR--FARE (ASW) V-135

VOLUME-LIMITEDMINIATURIZED GUIDANCE PROPULSION FOR VOLUME-LIIiTED (tiN-LAUNCHED IKIERCE-

PTOf JROJECTILES tl-261

WARFAREVERTICAL LAUNCH ASROC (VLA) MISSILE SYSTEMS STANDOFF ANTISUBMARINE WAR-

FARE (ASW) V-135WARHEAD

UNDERWATER WARHEAD GUN SYSTEM (U) VI-)99WAVE/BOUNDARY

SHOCK WAVE/BOUNDARY LAYER INTERACTION CONTROL IN A CENERIC HYPEPSONICINLET I-435

WAVESPREDICTION OF INTERIOR BALLISTICS AND PRESSURE WAkES PRESENT IN UNICHARGE

DESION IV-447WEAPONS

DEVELOPMENT AND TESTING OF A HIGH-ENERGY I.OW-COSr SOLID HANJEIT FUEL ATTYtE NAVAL WEAPONS CENTER (U) VI-173

WINDHIGH PRESSURE MACH 10 TO 20 ELECTROtHERNJA' HfPERSONIC WIND TUNNEL .- •11

WITH-PLYWITH-PLY PERMEABILITY Of CARBON PHENOLIC COMPOSITES AS A FUNCTION CW

ACROSS-PLY COMPRESSIVE LOAD AND IFMPIRATURE V-497WJII 19-

WJ119-2 MODULAR TURBOJET PROPULSION SYSTEM FOC TACIlCAC. MISSILES IV-251

14,9

WOUNDJANNAF STANDARDIZATION Of TENSION, COMPRESSION. AND SHEAR TEST METHODS YO

DETERMINE MECHANICAL MATERIAL PROPERTIES FOrt FILAMENr WOUND COMPOSITESTRUC;TUJRES 1-89

WRAPPEDBE'3IGN AND DEVELOPMENT OF CARBON FIBER WRAPPED NON-LOAD SHARING TANKS 111-487

450

MEETING ATTENDEES

ABEL, LARRY Micro CniratIullb-oma FiOOTH-, DAVID W. ThiolcolHunurvilleACHARYA, ARUN Union Cartbjdcronawanda BORCIIERDING, RON UT*C5Dt.San JoseAHMED, SAAD WRkDCAVPAFlI BOWE. JAMES Atlantic Rescarch/GainavilleAIELLO, PETER 3M/Fouter City BOWMAN, GARY Atlantic Reitearch/GainesvilleALFXANDER, RICHARD AeWeUSaaaunento BOW11AN, LES NWC/Coina L~akeALLEN. HENRY Stone En&'Huntsvilk- BOYD, D. EDDIE Hercuics/MagnaALOISE., JAMES R. Santa Bharbara RscM/Groleta BOYER, NATIIAN BRI-AberdeenALTMIAN, DR. DAVID RANN.TeIo, Alto BRACUTI, DR. A. J. ARDULVPicatiiinyANDERSON, DAVE HcetuIes(.,,agr'.a B3RADLEY, STEVEN 'Ibiokoi/Bnigt'am CityANDREPONT, WILBUR U71-C5 DAAxý &ter BRADY, VICKI LYNN NWIC/Cbira LakeANDREW, JOHN L. Hercut s/jNssb?& BRAUN, JOHN D). NWCICIbina LakeANDREWS, EARL. NASA-iLARCI-iampion UB.EINING, DENNIS 3M/St. PaulANKENEY, DFWEY NWC/China iAke BRuTIANsG, DR. SIIIRL A(ýAFSC)ffEdwardsARLAB, DUSHYANT Beech AiircaWWichi~a 31RENNAN, PAUL R. General Dynamics/Sari DiegoARCHIBALD, DR. THOMAS Fluorochem/Azusa BRICKER, ADAM Genera Dynamis>/DavionARNOLD, CiHARLES ASDC/H-untsvilft BRIDGES, STU) AL(AFSC)/EdwardsASAOKA, LEO M:C04J'MRedstui.e BROWN, ALAN Rockwelf/Canoga ParkATKESON, PETER Tnitol/Elkton BROWN, DON L AL(AFSC)/E.dwardsAUSLENDER. DR. AARON Rtx-we~l)Canoga Pi'rk BROWN, ERNEST )-Irl'ules/MagnqAUSTIN. RICHARD AEr)C/Arnold A1-_B BROWN, K,%W:N JHU-CPIAlA~urelBACHMAN, 13RYAN SwirdrupA-Iun-svili BRUMMEL, ELIOT FOX ARDFA'lorvwoodIIAC7ZIJK. ROBELRTJ. Ilvccules/Magna 13RYANT, DR. JAMES T. NWC/Chirla LakeRACLINI. J!'M Jntidynamia/Goodye-.r 3UFF, JAMES Mission RvgearctiJNa~shuaBAIRD, MAWOR JASON Al(ýAFSC)/Edwards IlUi-DANO, QUE NWC/China LakeBAK, MIKE Williams IntnMIa'/Walled Lake BURCH, DAN NWC/CraneBAKER, CLIFF Fiber Maicrials/Biddefoid BURNE-IT, JIMMY -ASr.)MuntsvillcBAKOWSKJ, CHRIS rOPH ZR Olini/Stariord BURROUGH-S, SUSAN L_ AAWS/RedstoneBANERLAN, DR. GORDON Slone Fing/IHuntsvtile BURT, BOB C,1ipan Corp/Arnold AFESBARKMA~N, JOHN NWCICh'na I-ake BU~tER, T. DANIEL LANLA-os AiamosBARNES, MICHAEL Atlantic ResearchlC'aintzvi4Ile CALHOUN, DAViD) Arrojel/SacrarrcntoBAROODY, EDWARD NOSAndian Head CAMPEN, liAL Acroici/FolsfvmBAUER, RONALD KasSer Aerioech/San Leandido CANFIELD, ALAN R. ThiokoVklngham CityBAUM, DR. KURT Fluorochem/Azusa CANN4IZZO, LOU Tbiokol/Bnrgham CityBAZIL, RAY H. NSWC,/Silver Spring CANNY, RUBEN Atlantic Rescarch/Niagra FallsBEAN, DON General DMam~iea/Oniario CARON, STEPHEN D. UTr-CSD/;an JoseBECKLEY, DON BP Chemicals/Santi. Ana CARR, CLYDE 1'. Tlijokol/ElktonBLCKMAN, CHARLES AL(A1-SC)l/Edwailds CARREIRO, LOU' ARDC/WPAFI3BEECIXOFI', BILL. WR)JC/WPAF3 CARRENO, DAfl NWC/China LakeBIIEM, JOHN W. JI'lj/Paden~a CARROLL, BILL Trojan/Spanish ForkBEIGH-LEY, ROBERT N. BP Chemic~atsa,;ta Ana CARVER, DR. JAMES MICOMIRedstoneI3FNNETTS, ST-cVE Aerojet/Sacramncnto CATON, CAIPT JEFFREY 6595 TEG/VandeinbcrgB77NSTEIN, ELI Teledyne CAElrol7do CHAFFIN, BEVERLY WRDG'WPAJ-BBERARLDI, MARTY MooVEas: Aurora CHAN, MAY LFE NWCe/Chiiia LakeBERKCJI'I2K, F`R&NK I) NASA-LeRC/Cle~vca')d CHANG, DR. L-ANG-MANN BIRiAbcrdeenBERTOLINO, JON Acrojcl'Sacramento Cl!A'~J, I-Still I Aerospace/El SegundoB3EYER, MARK ,j. HQ AUSC/Andi-cs AFB CHAPMAN, DR ROBEkT 1). Fluorcxlicm/A/.u,,aPBEYER. ROD Atlantic ,'esearch/Gain-_sviIle CHENi', TIMOTHY RockwclV/Canoga FarkRIXON, DR. ERIC it. AMCCOM/?icatinny CHERRY, STEVE TýW/Resiondo (leachBLANCHARD, DOUG1AS G. NWC/0iiin Lake CHEW, DR. WIL.LIAM M. M!COM/RedastoneBJLOCHER, LOREN M. Doeit ,,SUatl:. CI EW,.JAME-S S. B. AL(AFSC)/EdwardsBLOMSIIIELD, FRED NWC/I('in;i Like CHEWEY, PAUL I.MSC/SunnyvaleBlLUE, DUANE NWCI/2hina Lake CI II, MINN-SI ONO 'Miokol/FElkionBIOARDMAN, TERRY j hooko/BiA binghi City C1I IIN 117, W. GASlIJRc~nkonkomraBOLTTCHER, PAUL L. CvncraJ Dynamics/Pomona CHIIII, DONALDI S/l-'I TAK ARD)1C/PicdiinnyBOGGS, T'. L i MV.'/Chiria Loike UIow. L.OU Gjeneral [yiinv'~a.BOND, GARY N. Si'ARTA/iluntvillc CHRIS ILNý'N. BlLAINE fierculcs/Magna

451

CH-RISTENSEN, KIRK AerojeL/acrameitto DUNN, PAUL M. NUSC/NewportCHRISTIAN, TOM JHU-CPIA/LAurel DURHAM, LT. ERIN AL<AFSC)/EdwardsCLkRAMflAR0, DAVID NWC/ChinA Lake DURHAM, S. E. ConsultantCIRINCIONE, RICHARD ARDECI1'icatinny DWYIER, DON BP Chemicals/GardenaCLARSON, VICKIE AL(AFSC)/Edwards ECKMANN, JAMES B. General Dynamics/San DiegoCLAUSEN, CAPT. ROBERT DM0/Norton AFB EGAN, CHRISTOPHER NUSC/NewportCOATS, DOUGLAS E. Sftw & Fag Assoc/Carson City EGGLESTON, DEBBIE JHU-CPIAJL~aureICCJCCHIARO, JAMES I- JHU-CPIA/Laurel EIGEL, CHARLOl'TE W`RDC/WPAFBCOFFEY, III, LT. L C- AL(AFSC)/Eidwards ELDER, DUANE E. UT-CSD/San JoseCOHEN, NORMAN S. Cohen Prof Svcu/Redlarids ELKINS, DR. RON GE Aircraft/CincinnatiCOLLIGNON, KARl NAVAIR/Washington F!LL.IS, RUSS UT-CSD,/San JoseCOMFORT, TED Hercules/Rockct Center ELMENDORF, HARRY W J Schaler/ArlingionCOMP'TON, DR. L- E. JPU/Pasadena F.LMQUIST, RICHARD UTRC1East HartfordCONNAL1 Y', MARC B. AL(AFSC)/Edwards ELWELL_, PAUL Mo-og/East AuroraCONNAUGFJTON, JOE Stone Eng/Huntsviulc EMMONS, THERYLE LIV/DallasCOOKE, CHARLIE Atlantic Resea-rch/Lancaster ENDICOIT7, DAVE Thiokol/Brigham CityCOOMBS, CHAD Atlantic Research/Gainesv6ille F.NNIX, KIMBERLY A- AL(4AFSC)/EdwardsCOON, JOHN Aecrojet/Sacramento ERDOS, JOHN GASL/RonkonkomaCOOPER, BO0B Ro)ckwell/Dayton ZýSCHER, WILLIAM NASA-HO/WashingtonCOOPER, ED Stone Eng/Huntsville FSTEY, DR. PAUL American Rocket/CamarilloCOPENHAFER, BOB Hercules/Magna ESTRELLA, GEORGE Rockwetl/Canoga ParkCOPPOLA, ED OO-ALC/1-fill AFB; ETniERIDGE, FREDERICK LPJLA-os AngelesCORPORA.N, EDWIN WRDCAWPAFB EVANS, DR. STEPHEN Rock-welliCanoga ParkCOUNTER, MICKEY MICOMJRedstone EVERDING, CHARLE.S Ball Corp/FairbornCRAWFORD, JOHN F. USASDC/Huntsville FACKREIL, FORREST N. Thiokol/Bnighamn CityCRESAP, BILL Ford Aeroispace/Newport Beach FAIR, DAVID ~ Army Prod Base Mod/PicatinnyCULVER, DON Aerojet/Sacramento FANG, JAMES Rockwcll/Canoga ParkCUMBO, FRANK McDonnell Uouglab~'Ureenbelt FARMEZ, JIM Lo-khccd/Pialo AljtoCUNNINOTON, RICK Aerojel/Nimb'is FARNER, KAREN A14L(.CjÈdwatdsCURRAN, FRANCIS NASA-LeRC/Cleveland FA.(R-U., BILL Muog/EIsht AuroraCYRAJJ, FRED Sverdrup/Arnold AF'3 FA~RROW, KiN.NETH4 Sv.rdnip/1iurtsvil;eDAHLEM, VAL WRDCIWPAFB FEATHERSTON, RICKY NWC/China LakeDAMON, BOB Olin/Redmond FELIX, ROSS UT-CSD/San JoseDANCEY, MARGARET UT-CSD/Sain Jo'.c FERRARO, NED W. JPL/PasadenaDARBY, ALAN K. RockcL/'Car~og-,s Park FINGERHUT, BOB LMSC/SunnyvaleDASH, STANFORD SAIC,'F1. Washington FINK, ARNIE Aerojet/Sacrzimcnio,DAVIDSON, FRED Alldntic Research/Gainesville FINK, LARRY Boeing/ScatileDAVIDSON, TOM Tliiokol/Ogden FISHER, JAMIE MICOMJRedstoii'DAVIS, FRANK Allied-Signal/Fairborn FISHER, MICHAkEL NWC/China LakeDAWILY. SCO(YI1 Atlantic Research/Gainesville FISHERKELLER, KERRY General Dynamics/San DiegoDeCOURSIN, DAVID Fluidyne/Minncapolis FLEMIG. ERNFST R. Thiokol.44.ur~tsvilkcDcGFOR6E, DREW, 0. AL,(AF`SC)/Edwards FLEMING, WAYNE Universal Pmopuhsion/Phoen~txDEMAY, SUSAN C. NWC/China Lake FOGLIEI'IA, JIM Wyle L-abs/NorcoDER, DR. JAMES Aerospace-1i-i Segundo F-OLSOM, MARK Whittaker Ordnance/Hl-loLterDETIEN, WALT Rocket Research/Redmond FONTENOT, JOHN S. NWC/China LakeDET~iING-, C.ARO-.LYN NWClChina Lake FORBES, WOODY IITRL/LanicasterDETTLING, RON COMARCO/Ridgecrtst I-ORD, BO0B General Dynarrics/San D.-guDILL, RICHARD A. SRS Tech/i-luntsville FOREMAN, KENNEI1-H M. Grumman/BethpageDILLINGF.R, ROBF-k-1 -i. NWC/Chins Lake FORSrER, JIM New Mexico Tech/SocorroDIRECI'OR, MAi-ýX A~arniic Research/Gainesville FORTIN, BOB Fiber Mateuiias./3&ddefordDONALD)SON,V.WAYNtE WRD3CAWPAF[I FRANCE, JOHN NWC/China LakeDONONi E, MIKE TaIley/Mesw FRANCIS, E. C. IJT-CSD/San JoseDOUI3, JOSEPIH Maitin MaristtaJDeiiver FRANKENBERGER, CH-UCK NWVC/Ch'ina LakeDOUKAK'.S, If ARRYz' . TIhiokol/Elkion FREDEI:-ICK, BOB Sverdrup/Arnold AFBDOWLE1, WARREN L. JP~I~aadena FREDERICK, FIlANKL~i~ General Drynaruics/Ont.rioDOWNS, DAV'U) S. ARDEC/Picatinny FRIANT, LEE Martin Marieiia/H1altim-orcDUCOT-E, MAJORIE E. MICOMfitedstonc FRONrNCG. DAVf: McDonnell Doi~glas.'Hiu.,tingtonDIJFAt'UrI, K. S. Vacco/EI Monte FULL-IER, SCOTI- NWC/Chuia LakeDUNCAN, SIEPHEN J. Indiana AAP/Charlestown GA(3R1IELSEN, EVAN Herctz-As/MagnaDJI INCAN4, TOM General Dynamics/'Pomona (,A IAfI, I RNEAL(AFSCgEdwards5DUWrAM, DR. ROBERT S. Anatech/San Diego GAL BREATH, JOE Rocket Rtuearch/RedmondDUNN, BRAX(TON M. Boe0ing,6Seattle GA!.£AGHER, KEVIN UTf-aWSitSr Jose

452,

GALLE'r. DAVID Aerojeu~auiamcito I O01.lMAN, ALI !:N L UT-CSD/San JoerGARCIA, RAY AerojeiSacrameit~io IiOOPER, BILL Hcrculca/MagnaGARR!SON, PHIL JPUiPasaderu HOPKINS, REID W Herculeit/iancastcrGAUSE., DR. RAYMOND L. SA1C/Huaitsv~ik tiORTON, DEBBIE AL(AFSC`),1idwardAGEHR IS, ALLE N NWCIChina Lake AOWARD, STACEY NWC/China LakcGEHRIS, RAMA 1). NWQCsbin~a Lake i NU, DR. GEORCýE JP[A'asidenaGIBONEY, ROD K~a&ir/I-vire H-UANG, DAVID 7.TRW/Redondo UceachGILLARD, TOM SverdrupArnol AFB I UI FERD. W L. LjT.CSD/Ssn Jcbz.GLADSTONE, VAL M. Pctf~ic Mil Tit Ctr/P3Iint Mugu IiUGGulNS, MIKE AL(AF3Cy/E~wardzGL AJJTLI, STEVE Thioko~l/righani City IlUI, PHILIP ARDEC/PicalinnyGLICK, BOB Taliey/1Mcsa I; LJNT, JAY Sverdrup/StennisGLICKSTEIN, MARVIN R. Prji' & WhilrieyfW. Palm Bch ';-IJSKE Y. CONRAD Tloley/MosaGOEDERS, DAVID C. IN-D onell Doughi%.Hun~tngton I LiVCHENS, I)ALL Thiokolfi IunizvdleGOLDBERG, BEN NASA-MSFC/MSFC INVERso, TONY Aerospace/El SegundoGONZALES, ROMAN Santa Baitbara Rsch/Groleta IVES, CRAIG General Dynamnics/an Diego,GONZALEZ, REYNALDO AcrojeL~acramento IWANCIOW, BERNIE Storic PEno/-untsville-GOODMAN, BILL W J schafer/Cala!,asas JACKS, JAM~ES W. Herculcs/Mc(;rego)rGOSHGARIAN, BERGE AL{AFSC)/Edwards JACOX, JAMES L. Herculca/Ma&naGOTT, RICK Hercules/Radford JAN ICIK, THOMAS SAIC/WebsterGRABOWSKY, AARON ARDEC/Picatinny JANSEN, LINDA Callery Chemical/PiltsburghGRAHAM, DR. ALFRED R_ Benet Li.buratory/Watervihct JASPI3RING. DAIN McDonnell IDougias/St LouisGRAH-AM, MARK A. Acrojct/Sacrarnento JAY, DAV'ID WR)C/WPM.4TjGRAH-AM, ROBERT ThiokoL,'Jrigham City JENSEN, DR GORDON UT-CSD,'San Jo',eGRASER, WARREN Thic~ko:/Elkton JE-NSLN, JE-FI: Marquardl/Van Nuy..GREENE, 1. PAUL LMSC/Sunnyvale jHS- A, W/hn ,kGRGGBARYTholoIE~unJOHNSON. DR. W L Thiokol/ElkionGRIBBEN, ED Atlantic Restatch/Niagra Falls JOHNSON, JIM Olin/ManonGRIFFIN, GREG UTC-Hamilton Std/Ljs Angeles JOHNSON,"I IMI Acro t/S~icrerp~nlnGU, ALSTON LCEE-AN AJitcd-Signavl/brrance JONES, RAYMUN' AL(AF'SC311:dwardsGUILE, ROY UTRC/E-ast Hartford JONES, ROBERXT 'Ibiokcl/IBrigham CitGURTLER, RROBERT L. Atlantic Research/Gainesville JONES, S'IijART Lockheed/1-IampiHABERMAN, GENE SPARTA/laguna Hills JOYNER, DR. TAYLOR 11 New Mexico Tr 'HAGSETH, PAUL General Dynamics/R on Worth JUDGE. ARINDLEt NAVAIR/Ww iiii:-~IIAINS, MICH-AEL UTRC/East Hanford KAISERMAN, MICHAIEL 3. Hughes Aircra't/HALL, ROBERT B. SAF/WXashingion KNFL, BRIAN Tcledyne/Hunt,HALL, SUE AL_{AFSC)/Eiiwards KAY, IRA W. UTRC/East 142-rifo-OHALOULAXOS, BILL McDonnell Douglas/I I'.ntington KAZAROFF, JOHN NASA- LvRC/ClevelandHAMvMER, BILL I-10 AF-SCiAndrews AFB KCELLER, RO)BERT F I ierculca/MagnaHAMMOND, WAYNE E. Acrojet/Sacramcnto KENNY. ANDkEW AL(AI"SC)/EdwardsHANNUM, JOHN NOS/Indian H-ead KETC! IUM, WiILILAM General Dynarrijs/Sari DiegoHANSEN, LES fThiokol/Brighamn City KIERNAN, RAY Defense Rsch/RockvilleHARLAMBAKIS. Cl-RIS Acrojet/EI Segundo K(IM, STEVEN S. NOS/Indiaii HeadHARRIS, MARK Talicy/'Mesa KING, DAVID AL(ARSNC)/EdwardsHARTKE. DICK Aerospace/Washington KINSEL, TOM AL(A1-SC`J/Edward&HARVEY, HARRY E. H-ercules/DeSoto KILABER, KE ITH SSPOfWashington, DCH-ASEROT, RICH ThiokoV/Marina del Rev KIEIN, LEYLA Boeing/SeatfleHASTY, SHARON J1HU CPIA/Laurcl KLUCZ, RAYMOND iWAFSC)/EdwarijsHAUN, DAN Talley/Mesd KNAPrTON, JOJ IN BRUtAberdcenHAWK. DR. C:LARK AL(AFSC)!Ed.wards KNOWLTON, DIC GRE6 Talley/MesaIHIEAT7H, .ILFFREY ASI.'WPAI-T KONG, JERRY NWC/Cliina LakeHEATI.EY. BILL 0. Miller Assoc,'Los Angeles KONSZA, RON GE.RESD/PhiladcfcphjaHEINEY, 0. I. RockwelL'C'anoga Park KOPPISCH, I-RIC AL(AFSC),T-wardsHEMMINGER, JOE NASA-LcRQCIlevcandi KORS, DAVID Acrojet/acrantentoHERRIOIT, EPHRAIM Hecrcules/Rocket Ccitter KOSOWSKI, BERNIE MACH 1/E-disonHERT7A.ER, BARRY L. Lockheed/Palo Alto KRAMER, DR. MIKE' NWC4'Chtna LakeHIRSCH, NORM WRDC/WPAFD KRIFI)E.L, MARY M, TRW/Redondo BeachHIlsER O Hercules/Huntsville KUBLIN, TI1-IODORF V NASA-MSFC/MS[-CHIITE, DALE A1EASC)/LdAvards <tTSCIIENREUTER, PETE GE Airciaft/lCincinnatiHITNI1R, J01HN NWC//CI1 t,1, Lake KWASN Y, JAM~lES USAFAJColorado Springs110, SON TIJANH. WSMO-Zinoinberg LANCASTER, J. THOMAS ASDC/H-untsvilleHODGES. Js,. N'iCO1%/Reds,,)nc LANI)EL., ROBERT JPL4PaaadenaHOFFTMAN, HABRM

JHU-CIIAA.,auieI LANI)RITH, CRAIG Aerojel/EI Segundo

453

L.ANDSBAUM, EUA± Aerc~paceAiot An~cie- [,~I JN, IM 'j~ I 1A E. NA ýAMS,-C/MSFzCLAR, 1~41XE TI[okol/Elkto S~ LTON, n-'MY U W. M1COM/ReCdi..IoICLARIMER, ME~RLIN W. AllarizI Reacarr/Cii .' TC UONALD. H I IIrj Ics.Mag na1k., IRUNG V. AEDCAriwL AFlI MIT.J 'LiLt KERMIT MWC0MiR,-dxoncLEDDEN, BILL NWC/0C1jnt WKt7 MvirirnkMAI F.I, NORM Ae--o~c/w~aramen!oL=E TUN M. j kghca Ajrc xIL1Cafwg P-) MOGRE, ROYCE D.. 'IrSA-IxRC/C1ev'e-indLERE, WAYNE "TflIy/MC~d MORAN. CLII.7oMjD jvlji.udenhLEINUANG, JOHN WJRDC4WMPIIBR MORA-SI. DOUGLAS Mo~igtAst Aurorta-E{VINSKY, DR. E~LY S. GenerAl iDyn~am.eic/a Thego MORGANMARK fI-rcule,¶Aagna

LEWIS, DAVID AL(AFSq!Tidwvrdj MORGAN, ROMERT A;Jvmc RereucKiainevoteLEWIS. SUSAN AL(AFSCY/F.Jwar M.ORRIS. JOF. Acmer'j~acram-ntoLILLLYJ -. JAY S- MlC0Pfld;o'un N)RR IS. JOH N1--I-RiUCK, C(flOCK Pnrhi & Whitn.-yrA; Pjirr, tkh MORRISON, DIC(K SWCA2'irwlzxihL[15OWSKI, BOB DIIA Ca~l/Sn )m- ~ MORRISON, MICHALjI. MvICOWM;RcdztuncIU11. CIJII.ThE14 (JIMPAY) AIJi*AFSCVT42wari:1 MORTIMIR. 9011 AzIl.nwic R ieicLI.1Niagra Fal's

L dO, DI(. EU1MUNE" AtrCieL'She.fancnto WMOSIER, STAN pral! & wbhtuiry/W 1,2m 3ch1.0, (,W)RGk: Io khcecJ/PI`.%) &;;o M017, HUGk1 J. ;xrj :cralnwt:01.03`773S, PAM tJT-C.sDCjli)ton P-,'7J/jS' S17'VEkIl 1. W7RD'-/WI'Ar-liinSFq±. LYMAN /A, I3Ltrcijk.Mapla >.! LLR, j. Alncaaai ltel/ý'aineri!!I (,U, 1.11. H. L Arft.1jel;"1'r'1;ita,)t MVYERS, JOHN? A-launJ: IRecicrc,hI/(ain(-.v)!;cLOIJNOA~iM, JIM NWC(2/.tii,j~o Ake myhI(S, f4()NYtSFll(H;ntsville!,OVING. 11 ýN St-rjinic APAx~.~IrP'* MYI.F.& CRAIGi A USA'I14AMiAt-bcr~fr:nLUIND, i'RiC tkoiog/Seattie NABITIY. J'M.SWAC/Chiu, l akeIINI) 3 IEVEN J.NWCl/h; -A i.akr NA)IL'OVSK'Y. 1`30?!' AcrojcSarccrma.-oI.USAADI, ARTIiIJI P. A f-?i.'n. NA!"(;l, FAIIL 1 litrkolfi'ikqamn CilyIAN~tCI1. (001 A!Ionl~c Resirch'6am~c-v Ilk. PNE~. 1lXIL ~ A14. 0i-,C).'t-.v.rds

.:-AC!-!fJ -?. eJI A.RIF.l- LI-V Miamjle.i~l7 l'gJi Nl4i i .!(,WNPi IL ERSON, PROF A. K Cvn.ci ;. .~ )cksri ~l'.i AM!- .I hodiLna~i

MAO'I)~N, CARL 3 Iercuw,./?A I~ NE.0,IINAo\)!.ckvvcll/Canop !'ijrkM.;iIOR'11¶:, LAi'.rNA.,Ml~',.I 'l;WLY. 5! AJON A(-roje/SacramcntoMANSERI, (";'kXY AeNcI;;F.IL.'MAN. ROBB!1 W. ARI-1API L .auji:MAI,14ZAIA TONY IM/St. PaIL' NiLWOIJIST, CHUC'(K Boinog.:aulf-MAk KAR lAi . f ,A,c~ N;-;J 3ii.6;Ii. Nl-WJ"ON, CI Yý,i'AI. M,1crials Scien "Al~lue BellMARS,1, ,A~j1A.k-A MCl/(ds, N6II)Yl- N. IIWL 'U . A14AI SC)Ji-dward46klAil'lN, D.ON Slot~n lrig/I untuyli~c NIC-'ILOI-S, JIM AL(AJ--SC)/I-Award%

A.CI. I NAS-iaI(CtrImpt(,c; NORRIS, FRANK NWC/China LakcMA5CIO, 13111.1. <I().iwrI NOWAKOWSKI, MICI1IAI : ASI)U/I IuntsvjlkeMALI I -.R S-j iJ'IILN E. lW'S/Ix C .rulces NUINN, ROSS Sverdrup/SlennisMArI`SON, LARRY lit W/R~cdondo lkicI. O'CONNE,1l., ALEIX Marlin Marictta/FokumnMAURtICI i, LOIJIIDES WRIC/`WIÀF`B O'MLARA, WILI . AM Olin/F.I. Marics

MctAMvIS, RO11 Sve'rdrup/Armdd AFII OCAMPO, CA.IT. (;iI IIMO/No. uon AFBlMcCAIN, J. WAYNE* 'I'livol/I funisvilk; OG1.L['Vlli, RON Spacc~raft Ouuurnl/1ul1clunnMcDONALD, T-IMOTI1 IY W. NAVA I H/Vaithngi on OKPAI.A, I--LIX Gecneral IDynjrnuca/Or.lar~oMrFARLAINJ, J. SCOTI American Rockcl/C~arnrillo 01 ANI) I, DRI D)ONALDI Universal Propulotsil/'llocnixMCGI(A'III, DAVID K.. Thiokol/11klcjn Ol-SON, ALVIN M Acrtpct/SacramcnioM.cIliMOi'.I, DAVE: Sverdr-upfArnrold All! Olf ' I.IT.l J01 IN WRDC[W:/'IAI-I1McMAS'IL:tS, Ai IRY Lo~ral I)ckcnw/Akrnn ORD1. kI(1' 'AI I)R 'I'lickol/Irighain CityMcIAIIIANI, ORIIFG Acrt-jeI/S'cramncn1-., OR)TWIE'I IfI1,1 DR PAl If Itockwcll/canop a i'~kMcI'I IERSON, MICEIfAE'l. Ac rjcl^/S AF! anu (nh) OSTIROWSKI, DRl P]K .It1 P' Energcluc Mcrl.%'lsAloxanoruiaMcSIÀDI)DN, I IUDjii Universal I'rjpul3ionA'I'nKcnLx (.)XI.I; Y, JIM I lvrculc-%/Mc~ircgorMFIIIA, (.OIAI- (jencral lDyuuanics/Sain lDiuo P'ACKARDI, ED) IDlcrite R~ch/ltockvillcMENDEZ7. CAuL1 AL{AI:S(:)/I :djw~irJ l'A6FI. )I:1 I'MOSSP/Magna

MLII4.lI. CAUIJ A{A'SC)/Idar~ I' 1, RICE 1(( AIt) 1I(rd Acr spa.cc/Ncwliort l1(ac.hMESSINA., DR1. NCAIL A. ICRI4Iv~oninuiulh Juriiction l'ANDITlIJN1I, )RA. Il ' 11 llJ-APIji.aorelMICIJAEI..S. SCO Ii Ml(OM/Red.4ume 'AINLII.A. L~. A. NI(:/Chinae LajkeMIGIIACCIO. JOE: SvrrdrupArnaJ Al I I'AN()S~AN, I 1(AC.)!' 10 ' k~vii ( ~aw 1' lirkMIII All., L.INDA 6,. NOS/Iidiooi I lcaid PAl'lA". *l.-'I IT l-N NW('i(;lifiji I i-MIILLER. I:I)JGENI-, Suiidslr;i-d.JiSn I)IvE. (AltK's, Ilk! J. Ic 1 ii! CAý'M WC IMII LIER, RAY NW(. hmi,j I Akc, FARKlISI 1, C. Al A14 I h!cuic,'.'lti'ifor

454

PASWllK. GREGIORY AC%)X4.1~Sa&.iwmero SANI ORENZO, ERNEST GAS[J/Ronkonkoma"'A 'ITERSON. JO!!L E. MICOM/Redslonc SANSCIRA!NTE, Wil i.ARD Atlantic ResearchlNiagra Falls.'AL LIMK STEW Toocle Army Depo4'rToeie SANSOY, LT SABR i At.(AFSC)Fdwsrd%P '.A t!ON, TERRY L. AerojeiiSacrmmenic SANIW.M. JOSEPH ALRDECAlicalinnyPLa(T-MPS. RAINDY AroetScramento SATcER, KELLEY Ford Aerospace/Newporl BeachPEIPN-zANN, MANFREL) RrckweIVCaa-ga Ave SAYLE.S, DAVAiD USASDC/HunusvilePE-N(,, DR. S7EVEN JPI.Paaadena SCHADOW, K! A'kS NWC/China LakePERKINS, FWtD M. Tbiolrol/Brnghairn SCHATZ, WILLIAM i. Ford Aerospace/Palo AJtoPESCHICE. BILL U'IMC/Fia Hartford SCHINDLER1, BOB Aercjeu1SacramnvenkPL'TERS, SUSAN NOS/indian Hem SCHMIDT, CAPT. WAYNE A14AFSCYlEdwardsP*~rERSON, JERRY HicrculeadT~agna SCHNIT1'GRUND, DR. GARY Rockwll/VCanoga ParkPETTERS, LT. DEAN W`RDC/WPAFR SCHULER, ALAN LEE General Dywmrics,/San DiegoPICKE IT, r-RANK. NWC/CJuna Lake SCH 1JST)R, J01- AN R -General Dynamics/San DiegoPIETZ. JOHN *faliky/7,ksa SCHWARTZ, DANIEL AL(AFSC)/FdwardsPIPER, LEE 31-1--CPLA/Laurcl SCHWARI7BIART, AARON Rockwell/Canog3 PsrkPIVIRO'TrO, TAVIMAS J. JPLR'.%sddeniL SEALS. WILLIALM ARDEC/PicatinnyPOULTER, LARRY W. Thiokol.'Brighaom City SEBERT, JAMES W. NOS/Indian IHeadI'OWEl.- STAN Caispan Corp/Arno!d AY-ýB SEDILI .0, LOU A4LAF-SC)/EdwardsPOWELL, IHOMAS Pratt & Whitney/Vi. Palm Bch SELPH, CURTIS AL(A FSC)IEdwardsPRESSLEY, HOMER M. Heccules/Mogna SEVER, MELVIN E. ICI Americas/Charlestuwr,PRESTON, DR. SCOrC Pluoro-hcm/Azusa SIIA.FFER, DONALD E. Thiokol/FlklonPRICE, WILLIAM A. Vi~rc.jWashinglon, DC SHELLEY, JANET AL(AFSQ/FdwardsPRITZ. WAYNE AL4AFSC)/Edwards SHERMAN, ANDY kilu anaet/PacoimnaPUGMIRE, DR. T. KEN T AiJIVFNor'w'z~d SHUMATE, TIM McL'ioanell Douglai/H-untingtcnPURVIS, JIM NV SMC/Vandcnbcrg SIMEN, ALLEN MarquardtlVan NuysQUINN, RANDAlL. W. AEDC/Arnold AFB SIMMONS, R. L. NOS/Indian HeadRAMIREZ, PAUL RockwellvDowney S!,NCLNJR, CHUCK UT-CSD/San JoseREARDON, JOE Thiokol/Eilkon SINGHI BRUI Marquardt/Van NuysKh-EDL, [DR RUSSELL NWC/Ci'ina LAke S .IQIVINSKY. DR. SANDRA If. AL(AFSC)/EdwardsREEDY, 110M JHU-CFIA/Laurel SMITH. ALFRED NWC/Cbina LakeREHI, KIM MICHAEL General Dynamic&/Pumoina SMITH, ANDY Aerojet/SacramentoREMEN, JOHN AL(AF5C)/Edwa.rds sivrfI-, ANDY L. LMSC/SunnyvaleRENNER, LOUIS 4cRM/Ridgecresi SMITH, JACK Fiber Materials/BliddefordREUSS, RON AIliccd.Signal/Los Anigeles SMITH, JOHN It. WRDC/WvPAFBREYES, LT PATRICK ASDAWPArS SMITH. KEVIN DSC/RenoREYNOLDS~, HUGH UT-CSD/San .3w- SMAITH1, VIRGIL. SverdrupiArnold- AM3-ICILH-TER, HERDER F NWC/Cbina 1--ke SMITII, WADE WRtDCAWPAI-'RIGHThEY, PAUL General Dynamics/San Diego SPADACCINI, LOU UTHRC/HAst HanrtkrdRINDONE. REN Acrojct/Sacrarnento SPENCER, ROGER ASD/W PAFB1RINKO, JOH N D. NSWC/Dahlgren SPENDLOVE, GREG6G ThiokolfHuntsvilleRITTENHOUSE, CHARLES Unidynamics/Phioenix SPRITZER, MIKE- Gencral Atoirmica!/San DiegoRIVKIN, ST7EVE .lP14isPakena Sl'ROW. WILLIAM AerojeIPacrArnenoRIZKALLA, 0. GAS, /Ronkonkome STARINSKY, I-EON RockwelL/Canoga ParkROBIBINS, JOHN NWC/China Lake S-IA!-LINCS, DAVID Caispan Corp/Arnold AFBROBBINS, STEVE Thickol/Lancaster STANGER, RON Hiercules/WilmningtonROBERTS, JIM NOS/Irdian H-ead STAUI3S, HARRY L. McI~onnell Douglas/HuntingtonROBINSON, KA11HY AcrojetSacramento STAYTON, L.EROY NWC/China LakeROBL.E, NEIL WRDC/WPAf-B STEIN, ED) Aiuaniic Re.¶carch/Niagra FailsROFER. CHEI~RYL LANI/Los Alanios STE1INIE, ME'L. 'IalleyiMes.-iROLON, CLAUDIA E. LMSctSunnyvale STEPHAliN. CAPI. JUT+ AFSTC/Kriiend AFBROPEI.FW/!KI, CAPT LARRY BI3O/Nooion AFB3 STEP'HENS, CHUCK L'I V/DallasROSE, jAMES M. Atla'iltc ResearchiCia iincsviIk: ST*EVEN'S, JAMES JIIU.AI'14,/I;,urelROSENBERG, SANP-. AerojetSacramento STIjVEF'SON, DR WIILLIAM Thiiokol/IlunisvilleHOSEiNTIIAL, SUS.) Acrojet/Sacramento STOCKE-R, I ARO!J: Allison GasllitdianapolisROSS. L-AWRENý L. UICSI)/San Jose SIOKF..j, DO Thiokol/HuntsvilleROTHI, L.T. GRI-'a- 6595 ThG'-'Nandenbcrg STOKELS, CEICIL. 1-hiokoL/HunisvilleROil-GERY, C;,Y Olin/Cheshire sTOKIES, FRIC SoRl/llir~ninghainROUSE. GtREGL-il Sundstrand/San Diego SI ONF. n111.1 Stone Eng/Huntsi"0cRUSSEL.L, BILL Genecral [lecurc/Piltsfield STORRUD, SIIANI:)N JAMFS Hceruies/MagnaRYCIINOVSKY, RAY SAN fl~ivcmitorc . lOYANOfFT, S'li 'I Al(ýAFSC)/EdwardsSACKHEIM, BO0B 1'RW/IRedondv Beach Sii~tAMA, LT1 .111.. AL(AFSC)/EduhardsSAIAZJ'ýR, MIKE NWC/China Lake STRAND), LEFON JPLA'asadcna

455

STULEN, RICHARD SANLIAscrniore WANG, TEN-SEE NASA-MSFC/1MSFCSULLINS, GARY JHU-APLI1-aureJ WARDEN. JOHN A.SDC/1-untsvilleSULLIVAN, DR. ROY M. NASA-MSFC/MSFC WARDLE, ROBERT "nickoV/Bnghiam CitySUMMERFIELD, M_ PCRL/Monmouth Junction WARDLOW, GREGORY A)AFSC)/EdwardsSUMMrIT, JIM WRDC/WPA!9zB WARR, FRANK W. Hercules/'MagnaSUTITON, GEORGE LLNLA~p&crmorc WARREN, LARRY MICOM/RedsioneSUTTON, TREVOR Alliad-Signal/I'empe WASSOM, DR. STEVEN R. Thiokol/Brigham CitySWEET, ELBERT NWC/hirna Lake WEBBER, DEBBIE AL(AF33C)/EdwardsSWENSON, IVAN Thbokol/Brigham City WEEKS, LARRY Advanced Technology/AdiingionSWIHART, JOHN KEYNOTE SPEAKCER WEISS, DR. RICHARD AL(AFSC)/EdwardsSWINT, MARION NASqA-MSECIMSFC WELDON, ViNCE Boeing/SealtlcTAKAYESU, BECKY Aerojct/EI Segundo WERLING. ROBERT S Acrojet/Sacramer.toTALIANCICH, CAPT. TONY AL(AFSC)/Edwards WEST, LARRY Acrojet/SacramentoTALLEY, ROBERT Veritay/East Amherst WÈSTBERG, KARL Acrospacc/ts AngelesTAYLOR. BOBBY Tbiokok/Iluntsvillc WETHERELL, DICK 1-fercules/MagnaTAYLOR, DON E. General Dynamics/Pomcona WE-171L4'JFR, B3. M. Sverdnip/Arnold AFBTAYLOR, KARL Genera., Dynamics/San Diego WEYLAND, HERM UT-CSD/San JoseTEED, SUZANNE Universal Propuision/Phoenix WHITE, CRXIUi Allied -Signal/?hoenixTH-ACHER, JIM Hercules/Magna WHITE, JAMES W. ThiokoL/Brigharn CityTHELEN, CHARLES Coma rco/R idgecrest WHITE, MIKE JHU-APIL~aurcITHOMAS, HOMER IHcrcule&I~vagnta Wl IITF. N I LL'S Atiantic Research/HuntsvilleTHOMAS, JIM ICI Fibenile/Huntsvillc wl ii Tr3 W. F_ H,:rcules/MagnaTHOMAS, RON Talley/Mesa WHI77EHEAD. JOHN LL1NjLAivermoreTHOMAS, WILLIAM ThiokoL/H-Ljntsvillc WICHMANN, IHORST] Marquardt/Vaii NuysTHOMASSY, LT. FERN BMO/Ed~son WILD3Si, CAPT LINDA WRLDC/WPAFB3THORSTED, KEN NSWC/Dahlgten WILEY, SAMUEL Aerojet/SacramentoTIGNAC, LOUIS RockwelV/Canoga Park WILKERSON, BERNARD AL(AFSC)/EdwardsTINGLE. (C'TT Aerroepnce/'El Sccgundc, WILKS, RODNEY 'Ibioliol/Brigham CityTOE WS, HANS Gi. Moog/East Aurora WILLER, DR. RODNEY T'Mokol/ElktonTOLENTINO, ARTIE Marquardt/Vain Nioys WILLIAMS, DICK Talley/MesaTOWNSEND, JERRY Whittaker Ordisanccfl Ituhhster WILLIAMS, GARY Atlantic Research/CamdenTRAl, YVONNE NSWC/Silver Spring WiLSON, BEN Mlf*OM/RedstoneTRUE, WILLIAM Acrojet/Sdcramenlo WILSON, STEVE M;-(.n Marietta/DenverTSE, FRANK NOS/Indian Head WILSON, TRACY D. JH-U-CPIA/LaurelTUFFIAS, BOB Ultramet/pacoima WISNESKIE, BRAD Ford AerospaceftNewpori BeachTURNER, ALAN D. NWC/China Lake WIS WELL, BOB AL4AY-SC)/EdwardsURBAS, JOHN E. IJOS/Indian Head WOOD, BILL West inghouse/SunnyvaleUSHER, ROB ANSER/Arlington WOODS, KEN NWC/China LakeVACEK, BOB AL(AFSC)/Edwards WOTEL, GERALD UT-CSD/San JoseVAN WIE, DAVID W. JI-U-APL/iaurtl WRIGHjT, JAMES MICOM/RedgioneVAN KLEECK, JULIE A- Acrojet/Sacrameniv WRIGHT, TROY NOS/Indian HeadVAN GRIETIiIJYSEN, V. WRDC/WrAI-B WRIGI IT, VALRI NOS/Indian HeadVANDERAHI, DAVID J. NWC/Chmna Lake YAGER, JACQUELINE Atlantic Research/Niagra FallbVANDERHYDE, NORM JPIj~asidena YAGLA, JON NSWC/DahlgrenVANDIVER, TERRY L MICOM/Rcdslone Y I, ALEX Rockwelk/DowneyVANEK, CHESTER F. LMSC/Sunnyvsic ZACHARY, ANTHONY Aerospace/1-n AngelesVASSALLO, FRANKLIN Calspanfllutfalo Z-ARLINGO, FRED NV/C/China LakeVEru, PHYLLIS Acrojel/Sacramcnio ZENTNER, DR. BRIAN NWClChina LakeVERNA. JOSEPH B. Herculet/Washingion, DC ZINN, MICHAEL NOS/Indian HeadVERNON, GREG N4WC/China LakeVIC'] OR, ANDREW Comarso/RidgecrestVICTOR, D)OROTHY NWQChina LakeVOECKS, GERALD JP.A'asadenaVON PRAGENAU, GEORGEi NAaij%-MSFCflMSFCVoss, KIM Hercules/MagnaWAESCHE, WOC-DWARD Atlantic Reitcarch/GaincsvilieWAUBE, BUD Moolp/Exit AuroraWALLS, TON4Y NWC/China LakeWALSH, JOHN J. Thiokol/ElkionWALSH, RAY W J Schaftr/CliclinsfordWALTRUP. DR. PAUL JHU-APL,/LsurelWANDER, JOE. AFE-SCfrynda~l

156

NITIAL DISTRIBUTION-Army -Navy- -Air I-o;ce-

ARDEC/PICATINNY ARSENAL NAVAL AIR SYS 6510 TESTO/LDWARDS AFBI DR. ANTHONY J. BEARDELL COMD!WASHINGTON, DC 1 TECH LIB--STOP 2381 ERIC R. BIXON 2 AIR-5004-TECH LIB1 DAVID FAIR 1 KARl COLLIGNON AEDC/,ANOLD AFS3 SMCAR-IMI-I I ARINDER JUDGE I MURRAY 0. KING

1 TIMOTHY W. MCDONALD 1 TECH LIBARMY BALLISTIC RESEARCHLABS/ABERDEEN NAVAL EXPLO ORD DSPL Al. (AFSC)/EDWARDS AFB

I ILANG-M.XNN CHLANOG bAC/INDIAN HEAD 1 DAVID R. BALLI WARREN W. HILLSTROM I IB DIV I JOHN H. CLARK1 LOUISE LETENDRE 1 MARC CONNALLY

NAVAL ORD STATIONflNDIAN 1 DALE R. HITEARMY BELVOIR R&D COMD/FORT HEAD I MKABELVOIR I DAVID CARPENTER 1 MKP

I L VILLAHERMOSA 3 CHARLES GALLAGHER I LUCIANO SEDILLO

I TECH L!BARMY BENET WPNS NAVAL POSTGRADUATE I DURWOOD I. THRASHERLAB/WAIERVLIET SCHOOLAMONTEREY 1 ROBERT VACEKI SMCAR-LCB-TL I CODE 1424 LIBS DIR 1 DAVID P. WEAVER

ARMY DEF AMMO CTR NAVAL RSCH LAB/WASHINGTON, AFAThI.SGLIN AFBSCH/SAVANNA DC I R. E. TAYLOR

1 A. H. BAUSMAN 2 CODE 2627

AI'WL/KIR'1l.AND AfTBARMY FGN SCI TECH NAVAL RSCH/ARLINGTON 1 SUL-TECH LIBCTMCHARLOTFLSVILLE I DR. RICHARD S. MILLER

I JAMES HEADDEN EIR/PAITRICK AFBNAVAL SEA SYS I S. CRESS

ARMY LABORATORY COMD/WASHINGTON, DCCOMD/WATERTOWN I SEA-09B312--TECH L.B FTD/WPAFIB

I LEWIS R. ARONIN I ARNOLD CROWDER1 M. M. MURPHY NAVAL SURFACE WARFARE I D 1IOJILER

CTR/DAHLCREN I R(',N MOREDOCKARMY MSL COMD/RFDSTONE I TECH SERVICES, CODE E231ARSENAL OGDEN ALC/HILL A-B

1 SUSAN L_ BURROUGHS NAVAL SURFACE WARFARE I JAMES ITO2 DOCUMENTS, AMSMI-RD-CS-R CnTISILVER SPRING1 FRANCIS E. HART 1 S. HAPPEL SACIOFFUIJT AFBI MR. J. MICHAEL LYON I NRI-STINFO

NAVAL UNDERWATER SYSARMY MS-/SPACE INTEL CMD/NEWPORT WRDC/WPAFBCTRIREDSTONE ARSEN,-L 1 TECH LIB 02 112 I WILLIAM G. BEECROFT

I AIAMS-DL I JAMES R MCCOYNAVAL WPN C'IR/CHINA LAKE - POPS

ARMY RSCHI OFC/RESEARCH 1 CODE 3131 1 , 'NNLT}'I. RONALDTRIANGLE PARK I JOHN BARKMAN 1 JAMES W. SUMMITI1 D. M. MANN I JOEiL BERNARD 1 T•CIINICAL RPTS SEC, ISL

I DAVID A- CIARAMITAROARMY STRATEGIC DEF i SUSAN C. DEMAYCOMD/HUN'ISVILLE I RICK F-.AfliERSION - NASA

1 A. B. DUMAS 1 C. E. FtRA.NKF'NBERC-EIRI D. C. SAYLES i JOHN M. HITNER NASA JOIINSON SPACF-1 MONT J. STESVENS CIR/ItIUUSION

WHITE SANDS MISSILE 1 REX DUfLVNTTIIARANGE/WHITE SANDS NAVY STRAT SYS PROJ

1 TECH LIB OFC-WASHINGTON. DC NAASA /lANt,I.fHYHAMPf ONI TECH LIB BR lID 1 S'IJ,".R I C JON3WS

I MS-18' TECII LIB

457

NASA LEWIS RSCH BOEING/SE.ATTLE LORALCTIR/CEVELAgD 1 SK-38 LWH PROC SIJPV AERONUJTROIIC/NEWPORT

3 L JARABEIC BEACHCALIFORNIA INST I TIECH INF SVC--DDC ACOS

NASA SCI TECH INF FAC/3WI TECHNOLOGY/PASADENA1 ACCESSIONING DEPT I LIB ACOSISTANDING ORDERS LORALIAKRON

NASA WALLOPS FLT CALSPAN/BUFFALOILBCTR/WALLO)PS ISLAND I N.A. HERDIC LTV MISSILES & ELCT GP/DALLASI PEJGGY CHESrER I LIB 3-58210, MAIL. STOP EM-0s

DIRECTEDNASAMARSI-ALL SPACE FLIGHT TECHNOLOGIES/ARLINGTON MARQUARDTNAN NUYSC'1R 1 NANCY CHESSER I LIIB

I AS24D1 TIEN-SEE WANG FAILURE ANAI.YSIS/'ALIO ALTO MARTIN MAR [ETTA-AERO &

1 ROBERT S. FRANKLE NAVAL SYS/BALTIMORENASA/WASHINGTON, DC 1 DEBORAH ISENHARTI W. R. FRAZIER, CODE 050 Fmfc CORP/MINNEAPOLIS

1 C. CLIFT MARTIN MARIETTAIDENVESR

-Otbar Govt. - GENERAL APPLIED IRC I S16

SCIENCE/RONKONKOMA MARTIN MARlE'rrA/ORLANDOCENTR~AL INTEL I SALLY KISS I MP-30 - TICAGCY/WASHINGTON, DC

I OCR/DSD/DL8 GENERAL DYNAMICS/FORT MCDONNELL DOUGLAS/ST. LOUISWORTH I C. IL. LAUER

DEF INTEL I MRS. V. G. HARRISAGENCY/WASHINGTON. DC MORTON'nl4OKOL4BRIGHAM~

1 P. MIJRAD, DT4A GENERAL DYNAMICS/POMONA CITY

D TCINOI MZ-4-20-CH- LIBI J MAR fl-IA WARD

C-MRALEXANDRIA GENERAL DYNAMICS/SAN DIEGO MORTON T-ilOKOij-UNThyI~l-E2 DTIC-FDA I R. E. AR NDAL 1 C. S. WECKWARTH

GENERAL ELECTRIC/PRINCETON PENNSYLVANIA STATE- Non-Go~'et. 1 PATRICA PLYMPTON UNIV/STAILE COLLEGE

I DANIEL KIELYAEROJET GENERAbJSACR.,QENTO GRUMMAN I K. K. KUOI IT-CH INF CIR AEROSPACF/BEIIPAGE

I HAROLD B. SMITH PRINCEi IoNAEROJETISACRAMENTO COMIUJS'IION/MoNMOUJll i1 R. MIRONENKO HERCULES/MAGNA JUJNCTION

ARSAELSAGLSI C. F. PARTRXIDGEj I N. A. MESSINA

I P. W. GREEN HERCULES/MCGREGOR RAND/SANTA MONICA1I RDU BLDG A-100 I D. S. HELFER

CHEMICALS/ALLENTOWN HERCULES/RAJ)FORD RAY]]IiFON/I3EDFORDI C. I- ANDERSON I B. JONES I J. DIJRBEN

ATEANlIC 1IERCLJLE.S ROCKEF CENIER RCCRFMNRE-SEARC")AJNESVII-A.E I M. L. WOODS IJAMES ('LIRLEIY

3 TECHNICAL INI- C`TRJIiJAYI. RO( KWIiIt..J('ANO(,A PARKAVCOfWILMINGTON I N. RUBlINS-ILN, 6.303 2 lThCII INF CT(-l

II D. HOWARDL.OCKHIEED. MSI. & SPACL/PAJJ R )CKWI ILI A)OWNI:Y

BATTELLE MEMORILA ALTOII (-)0-3-AOINSTRICH[AND I BLD)G zol

I DONALD R. KALKWARI- I MR. (2IIARIE', I.EON( ROCKW.I-I./SFiAL B IACI II AI l .ICi/AMl'I'ON

458

SANDIA NATIONAL[ABS/ALBUQUERQUE

1 DIV 25 IS, MARK ORIJBELUCHI LIB 3141, RHODA YAZZA

SOUTHWEST RESE,-JCHINSnTTmrEISAN ANTONIO

1 ROBERT D. ARMOR

STNEENGINEER[NOIWJNI3VMjE

1 D. OWETNS

SUN DSWÀNDIROCJCFOpD1 C. A. HULL

TALLEY/MESA1 SECURITY OFFCE

TEXTRON/EVERE1TI L T. NAZZARO

THIOKOLIEuICrON~I D. 2. MCDANIEL

TRW/REDONDO BEACHI1 THELMA4 FAIRFAX, SAN

BERNARIN~OI S/1930

UNIDYNAMIC~SIPHOENIXI C. RTElJHOUSE

LJNI[ED TECHNOLOGIES/EASTHARTFOR.D

I BETTY JOHNSON

UNITED TECHNOLOGIES/SAN JOSEI TECHNICAL L11BRARY

UNITED'D!CHNOLOGIESrWINDSOR LOCK-S

I MILTON MARCUS

UNI VERSALPROPULSION/PHOENIX

I I I. J. MCSPADDEN

UNIVERSITY CALIFORNIN/LoSALAMOS

I REPORT LIDI, P.164

WIES1NG14OUSr-ELECTRIC/SUNNYVAiE

I BElT-Y TYSON

WYLE LABORATORI1ES/FL,SEGUNDO

I PAUL M 'flJRKJIIj-IMER

459

Documents

… · CPIA PUBLICATION 550 VOLUME I OCTOBER 1990 Reproduction is not authorized except by specific permission. AD-A240 864,iIID11011,11, il I il …