AIAA-2001-1129-974

(c)2001 American Institute of Aeronautics & Astronautics or Published with Permission of Author(s) and/or Author(s)' Sponsoring Organization.

A01-16908VIIVI A

AIAA2001-1129Detonation Initiation Studies and PerformanceResults for Pulsed Detonation EngineApplicationsFred Schauer, Jeff Stutrud and Royce Bradley*Air Force Research Laboratory, Propulsion DirectorateWright-Patterson AFB, OH 45433Innovative Scientific Solutions, Inc.Dayton, OH 45440

39th A1AA Aerospace SciencesMeeting & Exhibit8-11 January 2001

Reno, NVFor permission to copy or to republish, contact the American Institute of Aeronautics and Astronautics,

1801 Alexander Bell Drive, Suite 500, Reston, VA, 20191-4344.


AIAA 2001-1129

DETONATION INITIATION STUDIES AND PERFORMANCE RESULTSFOR PULSED DETONATION ENGINE APPLICATIONS

Fred Schauer and Jeff Stutrud

Air Force Research Laboratory, Propulsion Directorate

Wright-Patterson AFB, OH 45433

Royce Bradley

Innovative Scientific Solutions, Inc.

Dayton, OH 45440

Abstract

An in-house computational and experimental program to investigate and develop an air breathing pulsedetonation engine (PDE) that uses a practical fuel (kerosene based, fleet-wide use, "JP" type) is currently underwayat the Combustion Sciences Branch of the Turbine Engine Division of the Air Force Research Laboratory(AFRL/PRTS). PDE's have the potential of high thrust, low weight, low cost, high scalability, and wide operatingrange, but several technological hurdles must be overcome before a practical engine can be designed. This researcheffort involves investigating such critical issues as: detonation initiation and propagation; valving, timing andcontrol; instrumentation and diagnostics; purging, heat transfer, and repetition rate; noise and multi-tube effects;detonation and deflagration to detonation transition modeling; and performance prediction and analysis. Aninnovative, four-detonation-tube engine design is currently in test and evaluation. Preliminary data are obtainedwith premixed hydrogen/air as the fuel/oxidizer to demonstrate proof of concept and verify models. Techniques forinitiating detonations in hydrogen/air mixtures are developed without the use of oxygen enriched air. An overviewof the AFRL/PRTS PDE development research program and hydrogen/air results are presented.

Introduction

Recent renewed interest in pulsed detonationpropulsion concepts has prompted a concerted effortbeing made by the U.S. Air Force (AFRL), U.S. Navy(NRL, ONR, and the Naval Post Graduate School),NASA, and several research contractors (AdroitSystems Inc., Advanced Projects Research, Inc.,Pennsylvania State University, Enigmatics, and majorengine manufacturers), to develop a low-cost, practical-fueled, pulse detonation engine. Conceptually, a pulsedetonation engine (PDE) offers few moving parts, highefficiency, high thrust, low weight, low cost, and easeof scaling. These make the PDE an attractive alternatveto jet turbine engines for small disposable engines. Adrawing that illustrates the simplicity of the PDE cycleis provided in Figure I. The near constant volume heataddition process, along with the lack of a compressioncycle, lend to the high efficiency and specific impulse,simplicity, and low-cost of pulse detonation engines.Pulse detonation engines have the potential for

operation at speeds ranging from static to hypersonic,with competitive efficiencies, enabling supersonicoperation beyond conventional gas turbine enginetechnology. Currently, no single cycle engine existswhich has such a broad range of operability.

Pulsed detonation propulsion research has beenfunded by AFRL since the early 1990's, but most of theefforts have been performed out-of-house bycontractors. In 1997, an AFRL/PRTS (the CombustionSciences Branch of the Turbine Engine Division,Propulsion Directorate at Wright-Patterson AFB, Ohio)in-house PDE research and development program wascreated. Principal interests lie in the air-breathingarena; a similar pulsed detonation rocket engine(PDRE) program is being conducted by AFRL/PRSA atEdwards Air Force Base, California.

The in-house PDE program was established inorder to make AFRL's unique resources available forthe development of this technology. Traditionally, we

1American Institute of Aeronautics and Astronautics


Fill TubeFuel

Air

DetonateMixture Exhaust

Refill Tube, RepeatFigure 1. Conceptual pulsed detonation engine cycle.

have used AFRL's advanced computational modeling,diagnostic measurement techniques, and test facilitiesto work with government agencies and their contractorson the development of advanced combustor concepts.Much of this research was focused on deflagrationphenomena while trying to avoid detonations. In orderto work with pulsed detonation phenomena, AFRL hasset out to develop the facilities, diagnostics, modelingtools, and experience necessary to contribute andprovide unique resources for the maturation of pulsedetonation technology.

The second motivation of the in-house programwas to produce publishable PDE data from which codesand performance predictions can be anchored and/orvalidated. Currently, there is a great deal of dissensionon PDE performance within the community.1

Detonation physics and detonation engine blow-downare highly sensitive to initial conditions, boundaryconditions, and multi-dimensional geometry effects.Most of the available data and models are proprietaryand not shared across the community, making itdifficult to assess the current status of PDEperformance and capability.

For the Air Force, a practical-fueled PDE meansJP/air detonation. This requirement creates severaltechnological hurdles that must be overcome in order tofield such a PDE. Complex hydrocarbon fuels, andparticularly liquid hydrocarbons, are difficult todetonate in air, typically requiring hundreds of kilo-joules to directly initiate a detonation.2 For this reason,a practically fueled PDE becomes a deflagration todetonation transition (DDT) minimization process sincethe fuel burned during detonation initiation does notproduce thrust efficiently while it is burning at lowpressures. Furthermore, since thrust is generated witheach detonation cycle, Figure 1, it would be beneficialto raise the operating frequency in order to producemore thrust. Higher operating frequencies also havebenefits from an unsteady inlet, nozzle, and noisegeneration perspective, but create complications in

other areas including valving, mixing, shortenedresidence time requirements, and increased heat loads.

In cooperation with other governmentorganizations performing PDE research and as thedeveloper of one of two government in-house PDEresearch engines, AFRL/PRTS has established a nichefor itself in tackling the above issues. The ONR fundedresearch engine at the Naval Post Graduate School inMonterey, California is directed towards liquid fuelinjection, atomization, and mixing3 and AFRL's engineresearch is focused on detonation initiation andrepetition. While there is crossover in the twoprograms, we are confident that if we can develop apremixed vapor-fueled/air PDE, the results of theNavy's research will provide the basis for making itwork on liquid fuels.

Approach

AFRL's unique resources4 have been used todevelop three areas in which AFRL can contribute tothe development of PDE technology. In broad terms,these areas are modeling, facilities and instrumentation,and research hardware development and testing.

AFRL's detonation modeling work is described inmore detail elsewhere.5 Recently this work has beenextended to three dimensional calculations and studiesof detonation initiation schemes such as the Shelkinspiral calculation in Figure 2. Both calculations employweak initiation of hydrogen/air mixtures, but the upperframe is a straight channel and the lower contains a twodimensional representation of a Shelkin spiral. Thebrighter areas in Shelkin spiral calculation indicate thatthe "hot spots" critical for DDT events are moreprevalent with the extra geometry. In the interest ofspace, further discussion of detonation modeling willnot be addressed within this paper.

The Pulsed Combustor/Detonation EngineResearch Facility (D-Bay) is capable of supporting upto 60,000 Ibf thrust experiments, with integrated remote

American Institute of Aeronautics and Astronautics


Figure 2. CFD pressure map of DDT event with 2Dtube and Shelkin spiral.

control and instrumentation systems. Pulsed thrustmeasurements from 3 to 1,000+ Ibf are accurately madewith a damped thrust stand mounted on the existingengine thrust stand. Up to 6 Ibm/sec (3 kg/sec) of 100psi (680 kPa) air is available and high-capacity inletand exhaust stacks are useful for self-aspirating designsand atmospheric exhaust. A direct connection to aliquid fuel farm via a high-pressure/high-capacity fuelpump retains the facilities ability to feed large-scale60,000 Ibf thrust engines. The facility test stand,damped thrust stand, with an installed research PDE areshown in Figure 3. The damped thrust stand itself sitsupon the large capacity static thrust stand and the roll-up door to the exhaust tunnel is visible on the right.

A hardened remote-control room is adjacent to the

Figure 3. Pulsed combustor/detonation engine teststand, damped thrust stand, with installed researchPDE.

750,000+ ft3 test cell. A minimum of 2 feet ofreinforced concrete is situated between the test cell andpersonnel during testing. Such precautions arenecessary when dealing with the high noise levelsassociated with PDE operation. Control of all pulsedcombustor/detonation engine operations and dataacquisition is done via a LabVIEW based interface withduplicate manual emergency shutdown and safetysystem controls.

Choked flow measurements are employed toaccurately regulate and measure oxidizer and fuel flowto pulsed engine experiments. These choke pointsisolate the measurements from the downstream pressureoscillations of pulsed valves. Each flow systemcontains a pressure controller, a choked orifice plate orcritical flow nozzle, and a surge tank to set and hold arequired flow rate even with unsteady combustor valveflows. As long as the flow is choked, flow rate can bevaried by changing the pressure and choked area.

In addition to conventional (low Hz and kHzfrequency) data acquisition and control systems whichinclude intake, fuel, and purge system instrumentation,the facility is equipped with up to 16 channels of high-frequency data acquisition at up to 5MHz. These maybe used for high-frequency pressure transducers,thermocouples, photodiodes, or advanced laserdiagnostics. A IMhz framing rate digital camera is alsoavailable for advanced laser diagnostics and imagingtechniques.6 High frequency pressure transducers andphotodiodes are currently installed with plans for digitalSchlieren experiments to begin at a later date.

Due to the nature of this facility, testing is notlimited to small-scale PDE experiments. Conventionalfull-scale turbine engine tests are possible makinghybrid turbo-PDE's a future research possibility in thisfacility. It is envisioned that several smaller scale(< 1,000 pound thrust) experiments could take placeacross the test deck or a single large-scale (10,000+pound thrust) engine test could be performed. As withmost of PRTS's test facilities, easy swap-out of testhardware is expected and accounted for in the initialtest-facility design.

Due to the critical timing issues in pulseddetonation engine operations, the high frequencyvalving tends to be both expensive and highlyconstrained. During the design of a research PDE,many options were considered that were either tooexpensive, had severe limitations in operating range, orboth. The research engine design selected is basedupon valving found in a General Motors Quad 4, DualOverhead Cam (DOHC) cylinder head commonly usedin the Pontiac Grand Am automobile. This PDE design



Figure 4. Second generation "Quad 4" based PDE during assembly (left) and as installed.

has an extremely broad operating range andconfiguration, with up to four detonation tubesoperating at up to 100 Hz each. The engine has provenreliability and durability, although the loss of the firstgeneration engine design did occur due to fatigue in thetube mount area after approximately 200 hours of hottime.7 The head and tube mount systems have beenredesigned to permit higher frequency operation, quickvalve system and detonator tube configuration change-outs, and eliminate the areas where fatigue became aproblem. The second generation engine is shown inFigure 4 during assembly and as installed in thedamped thrust stand.

The operating conditions of PDE's are verysimilar to internal combustion engines and many of thecomponents can be shared. By driving the overheadcams with an electric motor, the four valves in each ofthe four cylinders can be made to operate at between0.5 and 50 Hz. With minor modifications, thefrequency limit can be increased to 100 Hz for anaggregate maximum frequency of 400 Hz. Currently,several different detonator tube configurations areavailable including single 2" (50 mm) diameter by 3'(900 mm) tube, single 3.5" (90 mm) diameter by 3'(910 mm) tube, and multiple tube versions of each ofthe previous configurations. Provisions for lubrication,cooling, ignition, and fuel delivery are integral to thecylinder head/intake manifold assembly. The electricvalve-train drive motor, which is grossly oversized buta readily available component, is clearly visible on theleft side of the frame in Figure 3, along with the valvetrain drive parts.

The two intake valves in each cylinder, visible inFigure 5, are used to feed premixed air and fuel intodetonation tubes, which are attached to an adapter platesecured by the head bolts. In the current configuration,

the head and detonation tubes are installed horizontally,and the intake valves are the upper pair. Cold air flowsthrough the exhaust valves in reverse as a purge gas tobuffer hot products from igniting the next incomingcharge and to convectively cool the inside of thedetonation tube walls. The extra exhaust valve orvalves in this four-valve-per-cylinder design could alsobe used for an axial predetonator or additionalcombustion air if necessary.

Somewhat uniquely, this PDE is operatedpremixed, minimizing mixing and stratification issues.The large pop-off valves and check valves visible inFigure 3 are some of the precautions used to preventcatastrophic failure in the event of an engine backfirethrough the premixed intake section. Up to fourdetonation tubes can be run at 90 degrees out of phase,with various diameters ranging up to -3.5 inches(85mm). The main combustion air and purge air linescontain ball valves for each detonation tube feed systemso that the engine can be run with one tube, two tubes180 degrees out of phase, or all four tubes. A rotary

Figure 5. Research 'Quad 4' PDE detonation tubeadapter plate with visible intake (upper) valve pair, purgeair (lower) valve pair, and conventional spark igniter.



position sensor is adapted to the intake camshaft toprovide both an index of the valve timing sequence andthe relative position of the valves. This signal serves asthe master timing signal for the ignition and dataacquisition systems.

An eight-channel igniter/fuel injection control boxis triggered off the rotary position sensor. Separatecontrol of each detonation tubes igniter and/or fuelinjector can be accomplished with this system, althoughcurrently vapor fuels are premixed with the combustionair via a separate critical flow nozzle and flow controlsystem. Do to the high noise levels associated withPDE testing, all controls and data acquisition areperformed remotely from an isolated control room. Allof the control systems and data acquisition systems areLabVIEW based and integrated into one 'virtualinstrument' with back-up manual shutdown and safetysystems. This virtual control panel is extremelyflexible and can control all aspects of the PDE'soperation including: lubrication, operating valve drivemotor speed, fuel flow, main combustion air flow,purge air flow, timing, ignition delays, and automaticshutdown in the event of a critical system failure. Bychanging the position of a few manual ball valves andpushing a few switches in the virtual control system, theengine configuration can be switched from one tubeoperating to four tubes in a matter of minutes.

The engine is to be used for performanceprediction validations and serve as a test-bed forresearch of detonation initiation and DOTminimization, heat transfer, noise levels, pulsed ejectorconcepts, and multi-tube interactions. Initial testingand proof-of-concept is being done with hydrogen asthe fuel due to the increased detonability versuspractical liquid hydrocarbon fuels. A vapor propanefuel system has also been constructed in order to workwith a complex-hydrocarbon that detonates much likekerosene based JP type fuels. This will eliminate theatomization and mixing of liquid fuel complicationsthat increase the difficulty of practical PDE design andallow us to focus on detonation initiation and highfrequency operation. As mentioned previously, ONRfunded research is tackling the difficult problems ofliquid fuel atomization and mixing for PDEapplications.3'8 Recently, active cooling has beenimplemented along with expanded fuel systems so thatindefinite run times are possible. Further details on theresearch facility and engine are available elsewhere.9

Results and Discussion

Cold flow testing of the systems began in early1999, with the first hot firing on 9 September. This

Figure 6. In-house research 'Quad-4' PDE, success onfirst attempt, 9 September, 1999: 91 second operationat 8Hz, single 2" diameter by 3' long tube,stoichiometric H2/air, conventional ignition.

initial test was done at low fuel flow conditions tominimize the amount of hydrogen in the test cell inevent of a failure. The first test was done with fullyinstrumented intake and purge systems, detonation tubesurface temperature thermocouples, four high frequencypressure transducers along the length of the tube,damped thrust, and two black & white video cameras.An image extracted from one of the video cameras isshown in Figure 6.

The initial testing produced very good qualitativeresults, with four runs of up to 91 seconds duration.These runs, which were un-cooled, were cut shortbecause the tube surface thermocouples were epoxymounted and the epoxy melted. The thermocoupleshave since been re-affixed more robustly. The sharp'CRACK' sound and flash of the exhaust werequalitative indicators of detonations which contrastedwith the softer 'wumpff sound and flame visible outthe back when the engine deflagrated due to off-stoichiometry conditions.

Although data has now been obtained withpropane/air, the results presented herein will focus onhydrogen/air operation. Results presented wereobtained with a single aluminum 2.0" (50.8mm) IDtube that was 36" (915mm) long. Conventional weakinitiation was employed at the head end (via the sparkplugs visible in Figures 4 and 5) with a 3.5 msecignition delay. The fuel/oxidizer mixture wasstoichiometric and premixed hydrogen/air with a 50%clean air purge fill ratio. The above operatingparameters and an operating frequency of 16 Hz appliesto all data herein unless otherwise stated.



Figure 7. High frequency pressure traces from in-house PDE engine. Measurement locations at 3, 15,21, 33" axial distances from head, -200 msec durationshown.

Initially it was found that the detonation did nottransition from deflagration until near the end of thedetonation tubes. The addition of a Shelkin or shockingspiral produced much faster transitions and higherthrust levels. A 3/16" wire diameter spiral with a -1.8"pitch was placed in the first 12" of the detonation tubes.This spiral produced overdriven detonations by the 9"axial location. High frequency pressure transducermeasurements, as seen in Figure 7, indicate measuredwave speeds of 1959 m/sec. Further experimentalverification of detonation wave speeds was provided byphotodiode measurements shown in Figure 8 with aderived wave speed of 1959 m/sec. These results are inexcellent agreement with the stoichiometrichydrogen/air wave speed of 1968 m/sec publishedelsewhere.10

C :..._-O) »- I

Since the initial single tube tests, the engine hasbeen run in multi-tube mode, demonstrating both two-tube operation 180° out of phase and four tubeoperation 90° out of phase. A wide variety offrequencies have also been demonstrated along withoperation of the tubes using partial fills.

£i

20

15

10

°0 10 20

frequency (Hz)

30 40

Figure. 9. Thrust versus frequency.

PDE's are highly scalable, as demonstrated inFigure 9. The thrust is observed to increase linearlywith frequency, with the engine making no thrust whennot operating as expected. This data also demonstratesthe accuracy of the thrust measurements, as the errorbars are +/- 0.5 Ibf (+/- 2.2 N). Such thrustmeasurements have been demonstrated with the currentsystem down to 3 Ibf (13 N) but the accuracy and thrustrange can be varied with configuration changes.

The impact of ignition delay can be assessed with

^^

h-

7.5

7

6.5

6

5.5

5

4.5

4

_ , , , j , . . I | : , . . , , , , , i

Af B«*5ni» ..•,

V •»• *f» <!***: ; ; : :— - • - - - - - • - - - • - - • - - • ; - • • - • • - -

_ . . . . . ; . . . . . . . . . . : . . . . . . ; . . . . . . : . .

L.....J ....... I....... . . . . . L

, . . | ' • , : ; I

• 16 Hz

c 12 Hz

: :

'̂ : ~

3.5-5 0

Figure 8. Photodiode results at 27 and 33" locations. Figure.

5 10 15 20 25

Ignition Delay (msec)10. Thrust versus ignition delay.

30



6 ———

5 -

4 -

3 -

_c

8000

....*.

• 12Hz Thrust

• 1 GHz Thrust x 12/16

o0)

Q.c/)

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6

Tube Fill Fraction

Figure 11. Thrust (left) and fuel specificthe data in Figure 10. The plot of thrust versus ignitiondelay contains data for two different frequencies.Ignition delay here is defined as the time inmilliseconds between the intake valve closingcompletely and spark plug firing. Obviously, withpremixed operation, negative ignition delays are to beavoided as they can result in combustion before theintake valves are closed with consequent backfiringthrough the intake system. It was found that premixedoperation reduces the sensitivity of performance toignition delay, as some PDE systems have beenobserved to detonate only within a narrow ignitiondelay window of only a few milliseconds. However,certain trends are apparent at both frequenciespresented.

The ignition delay is observed to produce highand low spots. Because detonability is sensitive tochanges in pressure, the initial low spot is surmised tobe a result of attempting to initiate a detonation in theexpansion resulting from the closing of the intakevalve. The peaks at 3.5 msec ignition delay arebelieved to occur due to the presence of the subsequentcompression wave. These behaviors are then observedto repeat at periods corresponding to the acoustic lengthof the detonator tube.

PDE scalability is also accomplished via variationof the volume of the tube filled with detonable mixture.Via volumetric flow control, the tube fill fraction wasvaried as shown in Figure 11. Results are shown fortwo frequencies to cover a range of fill fractions whileremaining within the limits of a single choked-orificevolumetric flow control range. Thrust measurementson the left are scaled by frequency to 12 Hz. Due to thelinear relationship of thrust and frequency, Figure 9, thediffering frequencies collapse on one another when

7000

6000

5000

4000

3000

2000 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6Tube Fill Fraction

(right) impulse versus tube fill fraction.scaled by frequency. Further confirmation of thisphenomenon is evident in the fuel specific impulse ploton the right in Figure 11. Here frequency is accountedfor when dividing by the fuel flow and no otherfrequency scaling is required.

At a fill fraction of 1.0, the entire tube is filledwith fresh reactants for each cycle. At fill fractions lessthan 1.0, only part of the tube is filled with freshreactants with the remainder occupied by either a purgecycle or hot expanded products from the previous cycle.At fill fractions greater than 1.0, the entire tube is filledwith the excess detonable mixture presumably forminga free cloud at the tube exit. Thrust versus rube fillfraction is plotted on the left in Figure 11. The thrust isobserved to increase with tube fill fraction untilabruptly leveling out at a fill fraction of 1.0. Sincereactants detonating outside the tube are unconfined,they do not produce any thrust as shown for the fillfractions greater than 1.0. Note that even at tube fillscorresponding to only 30%, more than half the peakthrust is still obtained. This results in up to double theefficiency as shown by the increased fuel specificimpulse at fill fractions less than 1.0. This trend, whichhas been confirmed by Li, Kailasanath, and Patnaikusing CFD,11 is a result of longer blow down timesproduced by the increased acoustic relaxation length forpartial tube fills. Effectively, purge air or previouscycle products are pumped by the detonation, resultingin the same higher mass/lower delta-velocity efficiencygains found in modern high-bypass turbofans. Theremarkable efficiency gains are partly due to theincreased efficiency of shock coupling in the PDE ascompared to the viscous coupling in the turbofan.

In addition to the effects of frequency, ignitiondelay, and fill fraction, the impact of stoichiometry was



6.5

.Q

w 5.5

4.5

7000

6000

5000

0> 4000CA)

CO 3000

2000

1000

• Schauer, Stutrud, andBradley Experiment

——— Shepherd Calculation

•I-

0.5 1.5

PHI

2.5 3^ 1.5

PHI

2.5

Figure. 12. Thrust and fuel specific impulse versus stoichiometry.

also examined as shown in Figure 12. Thrust on the leftand fuel specific impulse on the right are plotted versusa wide range of stoichiometries. As expected fromdetonability data,2 a stable region is observed atstoichiometries near 1.0. At fuel rich conditions, boththe detonability and thrust are observed to fall offgradually with increasing fuel to air ratio.The detonability and thrust fall off more quickly on thefuel lean side of the stoichiometry curve. As withpartial tube fills above, more than half the thrust isobserved to occur even with only half thestoichiometric fuel to air ratio. This results in a higherefficiency for lean operation as confirmed on thespecific impulse plot.

The current experiments are graphed in Figure 12along with Joe Shepherd's analytical calculations12 withexcellent agreement across a wide range of equivalenceratios. The experimental error bar shown represents thevariation in thrust possible by changes in ignitiontiming alone, Figure 10, as the data was collected with aconstant ignition delay. Shepherd's results do notconsider deflagration to detonation processes. The falloff in specific impulse observed experimentally forequivalence ratios less than 0.75 can be attributed to therapid growth in DDT distance as the cell size gets muchlarger than the detonation tube size.

Summary and Conclusions

Pulse detonation engines are an extremelypromising alternative to small, disposable-jet turbineengines. The Air Force Research Laboratory hassupported PDE research for some time, and an in-houseprogram of the Combustion Sciences Branch of theTurbine Engine Division at Wright-Patterson AFB has

been established to produce shareable benchmarkperformance data. In addition, the in-house programhas been used to harness AFRL's unique resources inorder to contribute to the development of pulseddetonation propulsion technology in the form ofmodeling, facility, and research components.

It is expected that the deflagration to detonationtransition modeling can be used as a tool in thedevelopment and design of a practical-fueled detonationinitiator. The pulsed combustor/detonation engine testfacility has been developed as a cost-effective testresource that meets many of the unique needs requiredfor PDE testing. The remote controls and highfrequency data-acquisition systems have beenassembled to provide test support for researchersworking in collaboration with AFRL. The facility canhandle everything from bench scale experiments fromacademia to full-scale hybrid engine concepts fromengine manufacturers. Moreover, it is hoped thatresearchers will take advantage of this nationalresource.

A research PDE was successfully designed, built,and operated under the in-house program using aninnovative valve system based upon the "Quad-4", a 16valve, four cylinder automobile engine from GeneralMotors. The resulting engine is capable of a broadrange of frequencies and configurations with up to fourdetonation tubes. Data from the engine is beingpublished with the intent of providing non-proprietaryPDE data against which performance codes andpredictions can be benchmarked. The "Quad-4" PDEserves as a research tool and test-bed for detonationinitiation concepts, high frequency operation, heattransfer studies, multi-tube detonation engine operation,and pulsed ejector research. The engine, which was

8American Institute of Aeronautics and Astronautics


operated successfully for the first time in the fall of1999, demonstrates the affordability and ease ofscalability of PDE technology. The first generationengine operated for over 16 million cycles andapproximately 200 hours of detonations beforecomponents failed due to fatigue. The enginedemonstrated that PDE's can operate for extendeddurations even with low cost materials and designs. Asecond-generation engine design has been completed toreplace the failed engine with numerous designimprovements to durability and capability.

Hydrogen/air data have been presented on theeffects of frequency, ignition delay, fill fraction, andfuel/air equivilance ratio. The resultant findingsprovide insight for scaling thrust and improvingefficiency of PDE hardware. Data sets are available forcollaborative studies, including flow conditions andheat transfer data not published herein. Additional dataon propane/air detonations are available for qualifiedresearchers.

There is much work to be done in developingvalving, detonation initiators, noise suppressiontechniques, thermal protection systems, intake andexhaust nozzles, and control systems before a JP/airfueled PDE becomes practical. AFRL/PRTS wouldlike to invite the community to consider AFRLresources for further PDE research. With high qualitymodeling, research facilities, and an in-house PDEengine, AFRL can work with other organizations andcontractors, as done in the past with turbine enginetechnology, to maturate and transition PDE technologyto the field.

Acknowledgements

Special appreciation must be expressed to thetechnicians and support personnel, both in-housegovernment employees and on-site contractors whomade this work possible, particularly Dwight Fox(ISSI) who helped build much of the research PDEengine and Walt Balster (ISSI) who recently joined usas a facility technician. Dr. Vish Katta and Dr. L.P.Chin have performed their usual miracles on themodeling side and are gratefully acknowledged for theircontributions to this program. We also wish toacknowledge the technical leadership of Dr. MelRoquemore and Dr. Robert Hancock (AFRL/PRTS).

References

1. Kailasanath, K., Patnaik, G. and Li, C.,"Computational Studies of Pulse DetonationEngines: A Status Report," AIM 99-2634 (1999).

2. Kaneshige, M. and Shepherd, J. E., Detonationdatabase, Technical Report FM97-8, GALCIT,(1997).

3. Brophy, C., Netzer, D. and Forster, D., "DetonationStudies of JP-10 with Oxygen and Air for PulseDetonation Engine Development," AIAA 98-4003(1998).

4. Hancock, R. D., Gord, J. R., Shouse, D. T., Schauer,F. R., Belovich, V. M. and Roquemore, W. M.,"AFRL Combustion Branch (PRSC) Aero-propulsion Research and Development Activities,"Proceedings of the International Test andEvaluation Association (ITEA) Conference (1999).

5. Katta, V. R., Chin, L. P. and Schauer, F. R.,"Numerical Studies on Cellular Detonation WaveSubjected to Sudden Expansion," Proceedings ofthe 17th International Colloquium on the Dynamicsof Explosions and Reactive Systems. Heidelberg,Germany (1999).

6. Gord, J. R., Tyler, C., Grinstead, Jr., K. D.,Fiechtner, G. J., Cochran, M. J. and Frus, J. R.,"Imaging Strategies for the Study of Gas TurbineSpark Ignition," Presented at the SPIE's 44th AnnualMeeting & Exhibition, Conference 3783 on OpticalDiagnostics for Fluids/Heat/Combustion andPhotomechanics of Solids, Denver CO, 23 Jul 99.

7. Schauer, F., Stutrud, J. and Bradley, R., "PulseDetonation Engine In-House Research at AFRL,"invited paper at the 13th ONR Propulsion Meeting,Edited by G. Roy and P. Strykowski, Minneapolis,MM, 10-12 August (2000).

8. Gabriel D. Roy, "Pulsed Detonation Phenomena forAir Breathing Propulsion," JSABE 99-7127 (1999).

9. Schauer, F. R., Stutrud, J. S., Anthenien, R. A.,Bradley, R. P., Chin, L. P. and Katta, V. R.,"Development of a Research pulse detonationengine," invited paper at Joint meeting of theCS/APS/PSHS JANNAF Subcommittees, 18-22October, Cocoa Beach, FL (1999).

lO.Soloukhin, R. I., Shock Waves and Detonations inGases, Mono Book Corp, Baltimore (1963).

11.Li, C., Kailasanath, K. and Patnaik, G., "ANumerical Study Of Flow Field Evolution In APulse Detonation Engine," AIAA Paper 2000-0314(2000).

12. Shepherd, J., Personal Communication, CaliforniaInstitute of Technology, 24 August (2000).


Documents

AIAA-2001-1129-974