Embed Size (px)
Citation preview
€ FILE COP
(.0 Report No. AMXTH-TE-CR-88026o Final Report
U.S. Army Toxic and Hazardous Materials Agency
PROPELLANT REUSE/RECOVERYTECHNOLOGY
(TASK ORDER NO. 7)
August 1988Contract No. DAAK11-85-D-0008
"g' J;' 2 9 9 ; i;
Prepared by:
Arthur D. Little, Inc.Acorn Park
Cambridge, Massachusetss 02140-2390
Prepared for: .... cr P...iic
U.S. Army Toxic andHazardous Materials Agency
Process Development BranchAberdeen Proving Ground, MD 21010-5401
'I:
DISTRIBUTION UNLIMITED
lSA"
The views, opinions, and/or findings contained in this report should not be construedas an official Department of the Army position, policy, or decision, unless so desig-nated by other documentation.
The use of trade i iames in this report does not constitute an official endorsement orapproval of the use of such commercial products. This report may not be cited forpurposes of advertisement.
I Final Report toUnited States ArmyToxic and HazardousMaterials AgencyAugust 1988
IPropellant Reuse/RecoveryTechnology
I (Task Order Number 7)
I Final Report
I A.A. Balasco* Program Manager
C.A. Jake - Task Leader, Hercules (RAAP)I R.C. Bowen
M.L. Hundley (RAAP)L.H. McDaniel (RAAP)L.L. Smith (RAAP)
I Principal Investigators
I.
I Distribution Unlimited
AL Arthur D. Little, Inc.Cor,,tct No. DAAKi 1-85-D-0008Reference 54147USA THAMA Reference AMXTH- TE-CR-88026
I89 3 ., .
UNCLASSIFIED2a SECURITY CLASSIFICATION AUTHORITY 3 DISTRIBUTION/AVAILABILITY OF REPORT
Unlimited2b DECLASSIFICATION /RDOWNGRADING SCHEDULE
4 PERFORMING ORGANIZATION REPORT NUMBER(S) l. MONITORING ORGANIZATION REPORT NUMBER(S)
Reference : 54147 AMTH-TE-CR-88026
6a NAME OF PERFORMING ORGANIZATION 6b OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION
Arthur D. Little, Inc. (if applicable) U.S. Army Toxic and Hazardous Materials Agenc
6c ADDRESS (City, State, and ZIP Code) 7b. ADDRESS (City, State, and ZIP Code)
Acorn Park Attn: CETH-TE-DCambridge, Massachusetts 02140-2390 Aberdeen Proving Ground, Maryland 21010-5401
8d. NAME OF FUNDING/SPONSORING 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBERORGANIZATION U.S. Army Toxic & (If applicable) Contract No. DAAKll-85-D-0008
8 Hazardous Materials Agency CETH-TE-D Task Ordei No. 78c. ADDRESS (City, State, and ZIP Code) 10. SOURCE OF FUNDING NUMBERS
Attn: CETH-TE-D PROGRAM PROJECT TASK WORK UNIT
Aberdeen Proving Ground, Maryland 21010-5401 ELEMENT NO. NO. NO. 7 ACCESSION NO.
I1 TITLE (Include Security Classification)
Propellant Reuse/Recovery Technology
12 PERSONAL AUTHOR(S) A. A. Balasco, R.C. Bowen, C.A. Jake, L.H. McDaniel, M.L. Hundley andL.L. Smith "
13a. TYPE OF REPORT 3b. TIME COVERED 14. DATE OF REPORT (Year, Month, Day) 5. PAGE COUNT
Final FROM _,I TO 31 August 1988 23416 SUPPLEMENTARY NOTATION
17 COSATI CODES 18. SUBX'CT TERMS (Continue on reverse if necessary and identify by block number)FIELD GROUP SUB-GROUP 0 Obsolete Propellants) * Solvent Extraction -).
I_ _ __ SPropellant Recovery/Reuse) 9 Propellant Resolvation .>sI Propellant Reclamation) (Continued)19 ABSTRACT (Continue on reverse if necessary and identify by block numbb Due primarily to changes in weaponsystems, the military has stocks of chemically acceptable propellants which are obsolete.Past disposal practices have been to incinerate or open-burn these stocks; however, extensivresearch has been conducted in the past at the Radford Army Ammunition Plant (RAAP) for thereclamation/reuse of solvent-based obsolete propellants. A literature review indicatedthat resolvation of propellant and recovery of selected ingredients from propellant viasolvent extraction were the optimal reclamation technologies. This study was conductedin order to demonstrate the feasibility of these technologies. The propellant resolvationI. ddid consisted of both laboratory and bench-scale evaluations; furthermore, appropriatehazards analyses of the pror-dures and equipment used in the evaluations were performed bythe RAAP Hazards Analysis Department. Several opi'ratine parameters were assessed includingresolvation time, propellant/solvent ratio, solvent/solvent ratio, and propellant particlesize.---
20 DiSTRIBUTION/AVAILABILITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATIONE]UNCLASSIFIEDUNLIMITED 0 SAME AS RPT. [ DTIC USERS UNCLASSIFIED
22a NAME OF RESPONSIBLE INDIVIDUAL 22b. TELEPHONE (Include Area Code) 22c. OFFICE SYMBOLPatricia A. Spaine (301) 671-2054 ICETH-TE-D
DD FORM 1473,84 MAR 83 APR edition may be used until exhausted. SECURITY CLASSIFICATION OF THIS PAGEAll other editions are obsolete. UNCLASSIFIED
I I''/
UNCLAsSIFIED
SECURITY CLASSIFICATION OF-THIS PAGE
18. Subject Terms (Continued)* Single/Double/Triple-Base Propellants" NC/NG/NQ" Military Specs
19. Abstract (continued)
The laboratory-scale studies indicated that obsolete propellant can be successfully resol-vated wihin 90 minutes using closely controlled operating parameters. More specifically,the percent nitrogen and viscosity of the nitrocellulose (NC) determine the solvent/solventand solvent/propellant ratios required to properly resolvate the propellant. It was alsodemonstrated that single-base propellants resolvate more readily with increased solvent/propellant ratios, i.e., more solvent than is used in the RAAP standard production processesOn the other hand, most multi-base propellants resolvate using production-established orslightly increased solvent/propellant ratios.
Bench-scale demonstrations were performed with M1 single-base propellant to optimize theoperating parameters. The propellant was ground, dewatered using a Sweco® Vibro-Energyseparator, dried in a forced air dry facility, and mixed in a series of thirty iterativetrials; the resolvated propellant from the last nine trials was extruded through a 4-in.vertical press. The extrudate was cut to length and processed in the standard RAAP produc-tion operations used to manufacture single-base propellant. The finished propellant, whichwas subjected to the applicable ballistic, chemical, and physical analyses, eitber 1net orexceeded military specification requirements.
Propellant ingredient reclamation via solvent extraction was also conducted as a part of thistudy. Following a preliminary hazards analysis, appropriate solvents were selected basedon solubility and distribution coefficient determinations. Laboratory-scale solventextraction procedures were then developed for single-, double-, and triple-base propellants.Three principal ingredients, i.e., NC, nitroglycerin (NG), and nitroguanidine (NQ), weresuccessfully extracted from single-, double-, and triple-base propellants. NC recoveryranged from 96-100% for single-base, 100% for double-base, and 88% for triple-basepropellants. NG recovery from double-base averaged 80% and 100% from triple-base propellantsNQ recovery from triple-base propellant averaged 82%.
Based on the results of the evaluations, design criteria information was developed for bothpropellant resolvation and solvent extraction of selected propellant ingredients. Inaddition to the operating parameters defined by these studies, safety and quality wereaddressed in both designs. Safety-related considerations requiring additional evaluationinclude remote materials handling, equipment and facility clean-up and containment ofpotential spillage. Pertinent quality assurance considerations were also addressed in orderto ensure the production of specification-grade propellant. For example, the design criteriinformation generated from these studies provide baseline data which can be used to developan appropriate quality assurance plan for any follow-on studies.
Using the design criteria information, pilot-scale resolvation studies should be conductedfor single-, double-, and triple-base propellant. Additional grinding/screening studiesshould be performed to obtain propellant particles passing a 12-mesh screen to ensureadequate resolvation. Alternate methods of dcwatering and drying of the ground propellant(other than forced air dry) for resolvation should be investigated. Bench-scale solventextraction studies to optimize the extraction of single-, double-, and triple-base propellantingredients should be performed. Final users' specification requirements should be delineateJto permit the use of resolvated propellant in current military weapon systems.
UNCLASSIFIED
SECURITY CLASSIFICATION OF THIS P5A3F I
I
IIIIII
(Radford Army Ammunition Plant) for Arthur D. Little,Inc. in fulfillment of a requirement for Task Order
Number 7 under Contract DAAKIl-85-D-0008.
IIIIIIII
II
I CONTENTS
H Page
1 1.0 Introduction 1
2.0 Laboratory-Scale Propellant Resolvation Studies 2
1 2.1 Propellant Selection 22.2 Solvation of Propellants 22.3 Resolvation Time 62.4 Operating Parameter Evaluations 62.5 Resolvation Testing Procedure 262.6 Results 372.7 Discussion 73
3.0 Bench-Scale Propellant Resolvation Study 75
I 3.1 Preliminary Extrusion Studies 763.2 Sample Preparation 793.3 Preliminary Resolvation Study 833.4 Processing 913.5 Sample Analyses 93
4.0 Solvent Extraction of Selected Propellant Ingredients 94
4.1 Testing Procedures 974.2 Solvent Extraction Flow Charts 1224.3 Solvent Extraction Results for Single-Base Propellants 136
4.3-1 Ml Propellant 1364.3.2 M6 Propellant 1404.3.3 MlO Propellant 140
4.4 Solvent Extraction Results for Double-Base Propellants 140
4.4.1 M2 Propellant 1404.4.2 M7 Propellant 1434.4.3 M9 Propellant 143
4.5 Solvent Extraction Results for Triple-Base Propellants 1444.5.1 M30 Propellant 1444.5.2 M30Al Propellant 1444.5.3 M31Al Propellant 148
I 5.0 Design Criteria Information 148
5.1 Particle Size Reduction 1495.2 Propellant Resolvation 1515.3 Solvent Extraction of Selected Ingredients 153
IIII
I Page
6.0 Conclusions 155
7.0 Recommendations 156
8.0 References 157UAppendixes
A Total Systems Hazards Analysis on Propellant 158Reuse-Recovery Technology
B Hazards Analysis of Equipment, Procedures, and 171Operations Planned for a Reclamation Process forthe Recovery of Obsolete Cannon PropellantsI
II
IIII11II ii
I
ILIST OF TABLES
I Page
I Compendium of terms describing various staLes 7
of propellant resolvation
3 2 Single-base propellant ingredients and functions 14
3 Multi-base propellant ingredients and functions 15
1 4 Chemical analyses of Ml single-base propellant 17
5 Chemical analyses of M6 single-base propellant 18
6 Chemical analyses of M1O single-base propellant 19
7 Chemical analyses of M2 double-base propellant 20
8 Chemical analyses of M7 double-base propellant 21
I 9 Chemical analyses of M9 double-base propellant 22
10 Chemical analyses of M30 triple-base propellant 23
11 Chemical analyses of M3OA] triple-base propellant 24
i2 Chemical analyses of M31AI triple-base propellant 25
13 Results of laboratory resolvation study for various 38I propellant particle sizes
14 Effects of major propellant ingredients on 72propellant resolvation
15 Laboratory propellant resolvation results using 74production solvent/solvent systems
16 Data collected from preliminary extrusion studies 77
3 17 Results of iterative trials for bench-scale resolvation studies 88
18 Results of chemical, phvsical, and ballistic analyses 92I for resolvated Ml propellant
19 Solubility data for propellant ingredients 95
20 Separation of propellant ingredients into groups 96
21 Distribution coefficients for propellant ingredients 98
II iii
I
Page
22 Liquid chromatographic standards for single-, 99
double-, and triple-base propellants
23 Detection limits for propellant ingredients in 120single-, double-, and triple-base propellants
3 24 Summary of Student's t-test for single-, double-, 121and triple-base propellants
25 Single-base propellant ingredient recovery 137
26 Double-base propellant ingredient recovery 141
27 Triple-base propellant ingredient recovery 145
1111a111I11l iv
I
LIST OF FIGURES
1 Nitrogen content and solubility in ether/ethanol 5I solvent system (2/1 ratio)
2 Whole grain M6 single-base propellant 10
1 3 Crushed M6 single-base propellant (passed through 11torn screen)
4 Coarsely ground M6 single-base propellant (retained 12on 20-mesh screen)
5 Finely ground M6 single-base propellant (passed through 1320-mesh screen and retained on a 50-mesh screen)
6 Laboratory mixing equipment used in resolvating 27propellant samples
7 Partially resolvated whole grain M6 propellant 28
(1/1 solvent/ propellant ratio and 65/35 ether/ethanolratio after 90 min)
8 Partially resolvated crushed M6 propellant 29(1/1 solvent/ propellant ratio and 65/35 ether/ethanolratio after 90 min)
9 Resolvated coarsely ground M6 propellant 30(1/1 solvent/propellant ratio and 65/35 ether/ethanolratio after 90 min)
10 Resolvated coarsely ground M6 propellant 31(1/1 solvent/propellant ratio and 65/35 ether/ethanolratio after 90 min) showing desired plasticity
11 Resolvated finely ground M6 propellant 32(1/1 solvent/propellant ratio and 65/35 ether/ethanolratio after 90 min)
12 Resolvated finely ground M6 propellant 33(1/1 solvent/oropellant ratio and 65/35 ether/ethanolratio after 90 min) showing desired plasticity
13 Plastic extruder for determining density and flow 34
characteristics of resolvated propellant
3 14 Flow characteristics of Ml single-base propellant 35
15 Flow characteristics of M6 single-base propellant 36
I 16 Single- and multi-perforated whole grain Ml 43single-base propellant
I
I
I17 Degree of solvation vs time for Ml multi-perforated 44
propellant (1/1 solvent/propellant ratio and 65/35ether/ethanol ratio)
18 Degree of solvation vs time for Ml single-perforated 45propellant (1/1 solvent/propellant ratio and 65/35ether/ethanol ratio)
I 19 Degree of solvation vs time for Ml single-perforated 46propellant (1.2b/1 solvent/propellant ratio and 65/35ether/ethanol ratio)
3 20 Degree of solvation vs time for Ml single-perforated 47propellant (1.5/1 solvent/propellant ratio and 65/35ether/ethanol ratio)
21 Degree of solvation vs time for M6 propellant 48(1/1 solvent/ propellant ratio and 65/35 ether/ethanol
I ratio)
22 Degree of solvation vs time for M6 coarsely ground 49propellant (1/1 solvent/propellant ratio and 65/35ether/ethanol ratio)
23 Degree of solvation vs time for M6 propellant 50(1.25/1 solvent/propellant ratio and 65/35ether/ethanol ratio)
24 Degree of solvation vs time for M6 propellant 51(1.5/1 solvent/propellant ratio and 65/35ether/ethanol ratio)
I 25 MlO single-base flake propellant 53
26 Degree of solvation vs time for MlO propellant 54(1/1 solvent/propellant ratio and 70/30ether/ethanol ratio)
27 Degree of solvation vs time for M10 propellant 55(1.25/I solvent/propellant ratio and 70/30ether/ethanol ratio)
3 28 Degree of solva'ion vs time for M1O propellant 56(1.5/1 solvent/propellant ratio and 70/301 ether/ethanol ratio)
29 Degree of solvation vs time for M2 whole grain 57propellant (0.4/1 solvent/propellant ratio and56/44 acetone/etnanol ratio)
Ivi
PageI30 Degree of solvation vs time for M2 coarsely ground 58
propellant (0.4/1 solvent/propellant ratio and56/44 acetone/ethanol ratio)
31 Degree of solvation vs time for M2 finely ground 59propellant (0.4/1 solvent/propellant ratio and56/44 acetone/ethanol ratio)
32 Degree of solvation vs time for M7 coarsely ground 60propellant (0.4/1 solvent/propellant ratio and49/51 acetone/ethanol ratio)
33 Degree of solvation vs time for M7 finely ground 61propellant (0.4/1 solvent/propellant ratio and49/51 acetone/ethanol ratio)
I 34 Degree of solvation vs time for M9 flake propellant 62(various solvent/propellant ratios and 44/56
If acetone/ethanol ratio)
35 Degree of solvation vs time for M30 whole grain 63propellant (various solvent/propellant ratios and40/60 acetone/ethanol ratio)
36 Degree of solvation vs time for M30 coarsely ground 64propellant (various solvent/propellant ratios and40/60 acetone/ethanol ratio)
37 Degree of solvation vs time for M30 finely ground 65propellant (various solvent/propellant ratios and40/60 acetone/ethanol ratio)
I 38 Degree of solvation vs time for M3OAl whole grain 67propellant (various solvent/propellant ratios and40/60 acetone/ethanol ratio)
39 Degree of solvation vs time for M3OAI coarsely 68ground propellant (0.4/1 solvent/propellant ratio3 and 40/60 acetone/ethanol ratio)
40 Degree of solvation vs time for M3OAl finely ground 69propellant (0.4/1 solvent/propellant ratio and40/60 acetone/ethanol ratio)
41 Degree of sol' .tion vs time for M31AI coarsely ground 70propellant (va ious srlvent/propellant ratios and45/55 acetone/ethanol ratio)
42 Degree of solvation vs time for M31Al finely ground 71propellant (various solvent/propellant ratios and45/55 acetone/ethanol ratio)
3 vii
I _
Page
43 Summary of pressure vs ram-rate ratios of 12-in. 80horizontal and 4-in. vertical presses
44 Sulfonated whole grain MI multi-perforated propellant 81
3 45 Mitts and Merrill knife grinder 82
46 Sweco ® Vibro-Energy separator 84
47 Particle size distribution of ground, sulfonated Ml 85multi- perforated propellant (wt% retained on standardTyler screens)
48 Remotely controlled 2.5-gal. Baker Perkins 86sigma-blade mixer
I 49 Single-base standard calibration data of area counts 100vs concentration for dinitrotoluene (DNT)
5 50 Single-base standard calibration data of area counts 101vs concentration for N-nitrodiphenylamine (N-NDPA)
51 Single-base standard calibration data of area counts 102vs concentration for diphenylamine (DPA)
52 Single-base standard calibration data of area counts 103vs concentration for 2-nitrosodiphenylamine (2-NDPA)
53 Single-base standard calibration data of area counts 104vs concentration for dibutylphthalate (DBP)
54 Representative HPLC chromatogram for single-base 105propellant ingredients
55 Double-base standard calibration data of area counts 106I vs concentration for 1,2-dinitroglycerin (l,2-DNG)
56 Double-base standard calibration data of area counts 1073 vs concentration for 1,3-dinitroglycerin (1,3-DNG)
57 Double-base standard calibration data of area counts 108I vs concentration for nitroglycerin (NG)
58 Double-base standard calibration data of area counts 109vs concentration for ethyl centralite (EC)
59 Representative HPLC chromatogram for double-base 110propellant ingredients
1 60 Triole-base standard calibration data of area counts 112vs concentration for nitroguanidine (NQ)
3viii
I
I _
PageI61 Triple-base standard calibration data of area counts 113
vs concentration for 1,3-dinitroglycerin (l,3-DNG)
62 Triple-base standard calibration data of area counts 114vs concentration for 1,2-dinitroglycerin (1,2-DNG)
63 Triple-base standard calibration data of area counts 115vs concentration for nitroglycerin (NG)
64 Triple-base standard calibration data of area counts 116vs concentration for 2-nitrosodiphenylamine (2-NDPA)
65 Triple-base standard calibration data of area counts 117vs concentration for ethyl centralite (EC)
66 Triple-base standard calibration data of area counts 118vs concentration for dibutylphthalate (DBP)
67 Representative HPLC chromatogram for triple-base 119propellant ingredients
68 Calibration curve and operating parameters for 123analysis of potassium (K+ ) cation via atomicabsorption spectrometry
69 Calibration curve and operating parameters for 124analysis of lead (Pb+ ) cation via atomicabsorption spectrometry
70 Calibration curve and operating parameters for 125analysis of barium (Ba+ ) cation via atomicaabsorption spectrometry
71 Solvpnt extraction procedure for Ml single-base 126propellant
72 Solvent extraction procedure for M6 single-base 127propellant
1 73 Solvent extraction procedure for M1O single-base 128propellant
74 Solvent extraction procedure for M2 double-base 130propellant
75 Solvent extraction procedure for M7 double-base 131propellant
76 Solvent extraction procedure for M9 double-base 132propellant
3 ix
I
Page
77 Solvent extraction procedure for M30 triple-base 133propellant
78 Solvent extraction procedure for M3OA1 triple-base 134propellant
79 Solvent extraction procedure for M31A1 triple-base 135propellant
80 Nitroguanidine (NQ) crystallization resulting from 147hot water extraction
81 Flow diagram of pilot-scale propellant resolvation 152process
GLOSSARY
TERM IDENTIFICATION
Ba+2 Barium cationBa(N0 3 )2 Barium nitrateCASBL Continuous Automated Single-Base LineDBP DibutylphthalateDNG Dinitroglycerin1,2-DNG 1,2-dinitroglycerin1,3-ONG 1,3-dinitroglycerinONT DinitrotolueneDPA DiphenylamineEC Ethyl centraliteFAD Forced Air DryH20 WaterHPLC High-performance liquid chromatographyKC1O 4 Potassium perchlorateKd Distribution coefficientK+ Potassium cationKNO 3 Potassium nitrateK2S04 Potassium sulfateLCL Lower control limitN NitrogenNC NitrocelluloseNG NitroglycerinN-NDPA N-nitrosodiphenylamine2-NDPA 2-nitrodiphenylamineNQ NitroguanidinePb+ 2 Lead cationPbCO 3 Lead carbonateRAAP Radford Army Ammunition PlantRQ Relative quicknessRF Relative forceRS Residual solventsSpG Specific gravityTV Total volatilesUCL Upper control limit
Al
EXECUTIVE SUMMARY
Due primarily to changes in weapon systems, the military has stocksof chemically acceptable propellants which are obsolete. Past disposalpractices have been to incinerate or open-burn these stocks; however,extensive research has been conducted in the past at the Radford ArmyAmmunition Plant (RAAP) for the reclamation/reuse of solvent-basedobsolete propellants. A literature review indicated that resolvation ofpropellant and recovery of selected ingredients from propellant viasolvent extraction were the optimal reclamation technologies. This studywas conducted in order to demonstrate the feasibility of thesetechnologies. The propellant resolvation studies consisted of bothlaboratory and bench-scale evaluations; furthermore, appropriate hazardsanalyses of the procedures and equipment used in the evaluations wereperformed by the RAAP Hazards Analysis Department. Several operatingparameters were assessed including resolvation time, propellant/solventratio, solvent/solvent ratio, and propellant particle size.
The laboratory-scale studies indicated that obsolete propellant canbe successfully resolvated within 90 minutes using closely controlledoperating parameters. More specifically, the percent nitrogen andviscosity of the nitrocellulose (NC) determine the solvent/solvent andsolvent/propellant ratios required to properly resolvate the propellant.It was also demonstrated that single-base propellants resolvate morereadily with increased solvent/propellant ratios, i.e., more solvent thanis used in the RAAP standard production processes. On the other hand,most multi-base propellants resolvate using production-established orslightly increased solvent/propellant ratios.
Bench-scale demonstrations were performed with Ml single-basepropellant to optimize the operating parameters. The propellant wasground, dewatered using a Sweco® Vibro-Energy separator, dried in aforced air dry facility, and mixed in a series of thirty iterativetrials; the resolvated propellant from the last nine trials was extrudedthrough a 4-in. vertical press. The extrudate was cut to length andprocessed in the standard RAAP production operations used to manufacturesingle-base propellant. The finished propellant, which was subjected tothe applicable ballistic, chemical, and physical analyses, either met orexceeded military specification requirements.
Propellant ingredient reclamation via solvent extraction was alsoconducted as a part of this study. Following a preliminary hazardsanalysis, appropriate solvents were selected based on solubility anddistribution coefficient determinations. Laboratory-scale solventextraction procedures were then developed for single-, double-, andtriple-base propellants. Three principal ingredients, i.e., NC,nitroglycerin (NG), and nitroguanidine (NQ), were successfully extractedfrom single-, double-, and triple-base propellants. NC recovery rangedfrom 96-100% for single-base, 100% for double-base, and 88% fortriple-base propellants. NG recovery from double-base averaged 80% and100% from triple-base propellants; NQ recovery from triple-basepropellant averaged 82%.
xii
Based on the results of the evaluations, design criteria informationwas developed for both propellant resolvation and solvent extraction ofselected propellant ingredients. In addition to the operating parametersdefined by these studies, safety and quality were addressed in bothdesigns. Safety-related considerations requiring additional evaluationinclude remote materials handling, equipment and facility clean-up, andcontainment of potential spillage. Pertinent quality assuranceconsiderations were also addressed in order to ensure the production ofspecification-grade propellant. For example, the design criteriainformation generated from these studies provide baseline data which canbe used to develop an appropriate quality assurance plan for anyfollow-on studies.
Using the design criteria information, pilot-scale resolvationstudies should be conducted for single-, double-, and triple-basepropellant. Additional grinding/screening studies should be performed toobtain propellant particles passing a 12-mesh screen to ensure adequateresolvation. Alternate methods of dewatering and drying of the groundpropellant (other than forced air dry) for resolvation should beinvestigated. Bench-scale solvent extraction studies to optimize theextraction of single-, double-, and triple-base propellant ingredientsshould be performed. Final users' specification requirements should bedelineated to permit the use of resolvated propellant in current militaryweapon systems.
xiii
1.0 INTRODUCTION
The military currently has stocks of acceptable propellants which areobsolete due to changes in the weapon systems for which the propellantswere originally produced. Additional quantities of waste propellant,i.e., propellant that does not conform to ballistic, chemical, orphysical specifications, are generated during normal propellantmanufacture. According to the Environmental Conference proceedings ofthe "Hazardous Waste Minimization Interactive Workshop" sponsored by ArmyMaterial Command (AMC) in November 1987, 158,000 metric tons of obsoleteconventional munitions are in the demilitarization inventory; 249,0003 metric tons are projected by the year 1993.
Past disposal practices have been to incinerate or open-burn obsoleteor out-of-specification propellants or explosives. For example, at theRadford Army Ammunition Plant (RAAP) alone, 88 metric tons ofsolvent-based propellants (single-, double-, and triple-base) are slowlybeing disposed by incineration or open burning. Extensive research hasbeen conducted in the past at RAAP for the reclamation/reuse of thesepropellants. An engineering evaluation/selection of recoveryalternatives was previously conducted in PE-796, "Propellant ReuseTechnology Assessment," l to evaluate the existing technologies forreprocessing waste propellants and to develop improvements that wouldminimize environmental discharge and conserve strategic materials. Theresults of this study indicated that resolvation of waste propellant andrecovery of selected ingredients from waste propellant via solventextraction were the optimal reclamation technologies.
I Based on the .engineering evaluation/selection of recoveryalternatives previously conducted in PE-796, laboratory-scale propellantresolvation studies were conducted on selected solvent-basedpropellants. The purpose of these studies was to define the optimumresolvation times necessary to achieve an acceptable colloid for solidpropellants. Several operating parameters were also evaluated during thecourse of these evaluations, including propellant particle size,propellant/solvent ratio, solvent/solvent ratio, ingredient addition, andremixing. The colloided propellant doughs were also evaluated inlaboratory-scale mixing and extruding equipment. Using the datagenerated during these evaluations, bench-scale demonstrations wereperformed with selected propellant. During the bench-scale evaluations,the operating parameters were optimized. The propellant was ground,mixed, extruded, cut, and dried; the finished propellant was thenanalyzed for ballistic, chemical, and physical conformance tospecification.
As determined by the engineering evaluation in PE-796, a number ofprocesses have been demonstrated for the recovery of propellantingredients.1 As a part of this project, certain propellant ingredientswere recovered from single-, double-, and triple-base propellants viasolvent extraction on a laboratory-scale basis. Following thepreliminary hazards analysis, appropriate solvents were selected based onsolubility and distribution coefficient determinations. Solvent
extraction procedures were developed for the three types of propellant.Two testing procedures were prepared for the solvent extraction studies:high-performance liquid chromatography (HPLC) and atomic absorptionspectroscopy. A statistical study was conducted to verify that the HPLCmethods developed for these evaluations were comparable to the analytical3methods delineated in MIL-STD-286B.
Based on the results of the evaluations, pilot plant design criteriainformation was developed for propellant resolvation and bench-scaledesign criteria information was developed for solvent extraction ofselected propellant ingredients. Three parameters were addressed in theboth designs: operation, safety, and quality.
2.0 LABORATORY-SCALE PROPELLANT RESOLVATION STUDIESILaboratory-scale propellant resolvation studies were conducted on
single-, double-, and triple-base propellants, e.g., Ml, M7, and M30. Inorder to select the optimum resolvation technology, several testingparameters were chosen based on the current production methods used atRAAP for the manufacture of solvent-based propellants. These parametersincluded colloiding the propellant, defining the various states ofsolvation to attain the desired colloidal system, and determiningresolvation times. Criteria to permit introduction of the resolvatedpropellants into standard manufacturing processes were established byoptimizing the testing parameters via laboratory-scale studies.
I 2.1 Propellant Selection
3Propellant selection was based on three criteria: base ingredients,production-established solvent systems, and grade of nitrocellulose(NC). Selected propellants (chosen to represent the bulk of the3propellants available for reclamation) contain at least one of the threebase ingredients: NC, nitroglycerin (NG), and/or nitroguanidine (NQ).Single-base propellant contains NC; double-base propellant contains NCand NG; and triple-base propellant contains NC, NG, and NQ. The solventsystems were chosen to be compatible with existing production solventsystems at RAAP. The grade of NC, i.e., nitrogen (N) content, determinesthe resolvation capability of the propellant; for example, triple-baseM30 (12.6%N) propellant more easily resolvates than single-base Mlpropellant (13.15%N) which consists of a blend of 12.5%N and 13.4%N NC.
2.2 Solvation of Propellants
The ingredient that mainly affects solvation capability ofpropellants is NC, a binder yielding gaseous decomposition products andenergy during the ballistic cycle. The %N and viscosity (a measure ofchain length or molecular weight) of the NC determine the solvent/solventand solvent/propellant ratios required to properly solvate the propellant
£2II
I ingredients in a mix. Proper ratios are necessary to ensure optimalprocessing of the propellant in subsequent manufacturing operations,e.g., blocking, extruding, cutting, and solvent removal. Single-basepropellant solvent removal is accomplished in the solvent recovery, waterdry, and air dry operations with the exception of M1O flake propellant;the solvents in M10, as well as all multi-base propellants, are removedin a forced air dry (FAD) facility.
Various solvent systems can be used in propellant manufacturing. Forinstance, an ether/ethanol system is used at RAAP for the production ofsingle-base and certain double-base propellants. An acetone/ethanol
system is used for other double-base and triple-base propellants. Inorder to ensure compatibility with standard production processes at RAAP,e.g., solvent removal, only those solvents used in the originalmanufacture of the propellants were considered in this study. Theproduction solvent systems in use at RAAP were selected on the basis ofthe NC blends (including %N and viscosity parameters) required to yieldspecific physical characteristics of the individual propellants.
5When certain solvent systems are used, the %N of the NC determinesthe amount of NC that is soluble. For example, NC having 10 to 12.6%N issoluble in a 2/1 ether/ethanol solvent system, whereas NC having >13%N or<10%N is not soluble. However, acetone dissolves NC having >1O%N. 2
Furthermore, as the %N decreases, solubility increases causing thepropellant to burn slower, thereby affecting the burning rate of the3 propellant.
Dilute solutions of NC (low viscosity) exhibit Newtonian behavior inthat the rate of flow is proportional to the applied stress or pressure.On the other hand, non-dilute solutions of NC (high viscosity) exhibitnon-Newtonian behavior. In non-dilute solutions of NC, the NC micelles(i.e., an ordered collection of submicroscopic fibrils) are more aligned,or parallel, than those in dilute solutions; this alignment is due to therestricted physical space available for the micelles to migrate andpossibly become misaligned. The relationship of viscosity to tensile
strength and micelle elongation is proportional, i.e., as viscosityincreases, the tenacity of the micelle directional alignment alsoincreases. During extension (stretching) of an NC film, which results inelongation of the micelles, additional alignment of the micelles alsooccurs, e.g., the more amorphous the initial state of the NC film, thegreater the tensile strength which is developed by extension because themicelles are aligned in a parallel fashion. NC films prepared withether/ethanol are more amorphous than NC films prepared with acetone.When acetone is used, the extension is reduced and less opportunity isprovided for the micelles to align in a parallel manner, resulting inlower tensile strength; when rupture occurs, a greater portion of themicelles are perpendicular to the axis of stress. As evidence by theabove discussion, the strength of NC films can be adjusted by the choice3 of solvent system. 2
In the manufacture of propellant at RAAP, NC is blended to specified%N and viscosity to yield desired physical characteristics of thepropellants during and following processing. Solvent systems areselected to aid in obtaining these desired characteristics. It must be
* 3I
Inoted that much of the biological structure of the original cellulose isretained in the NC. Cellulose fibers are comprised of layered structuresof fibrils, which are in turn composed of ordered layers of molecules.For example, single-base propellants utilize 13.15%N in the NC blends and
an ether/ethanol solvent system of approximately 2/1 to dissolve(gelatinize) the NC. As shown in figure 1, not all of the NC isgelatinized, a desirable characteristic in that a certain quantity ofundissolved, intact NC fibers enhance subsequent propellant processing byminimizing mechanical disintegration of the colloid. 2 Since NC is theprimary ingredient of single-base propellant, the ether/ethanol solventsystem is desirable for ease of solvent removal.
3 Acetone, on the other hand, has been shown to be the most effectivesolvent for NC. In the early years of propellant development, acetonewas rare and expensive whereas the ether/ethanol system was manufacturedfrom ethanol feedstock. Process developments for acetone productionsubsequently accelerated the development of multi-base propellants. Theinteraction between acetone and the NC is not restricted to the externalsurface (outer layers) of the NC; even when the amount of acetone isrelatively small, the acetone is absorbed in the interior and penetratesbetween the molecular chains, increasing the spacing between them. Thisphenomenon, termed swelling, takes place at random and is limited by thereplacement of the hydroxyl groups by other groups that endow themolecule with solubility. If the physical conditions of the replacementreaction are uneven, the distribution of substituent groups may be soirregular that one section of a chain may be soluble (particularly asection in a disorganized region of a chain) while another section(probably in an organized region of a chain where penetration of thesolvent has not been as effective) may not have undergone enoughreplacements to enable it to dissolve; here the residual hydroxyl groupsmay be numerous enough to prevent the chains from separating. With agreater amount of acetone, the molecular array becomes confused, thechain alignment is lost, and a gelatinous mass of no regular structureresults. Finally, when an excess of solvent has been added, themolecules are completely separated and a true solution is formed.
2
In the manufacture of double- and triple-base propellants, NG alsoserves as a plasticizing agent, i.e., solvent. Since varying the contentof NG has profound effects on propellant physical properties, e.g.,burning rate, brittleness, and tensile strength, F. S. Baker conducteddielectric studies to determine the manner in which the NG is dispersedin the NC matrix. 3 Baker, assuming that the cellulose structure ispreserved in the manufacture of NC, further supposed that there existed alimited number of sites possessing high interaction energies. As NG isadded to the NC, site occupancy is increased towards monolayer coverage,with monolayer coverage anticipated at an NG concentration of -25%.Addition of NG in excess of 27% leads to multilayer adsorption; above 30%the available sites are completely filled, indicating that NG isrelatively inefficient as a plasticizer. In the production ofpropellants, therefore, acetone is used to swell the NC, thereby exposingmore sites for NG adsorption. Furthermore, since NQ is not solubilizedin any of the solvents used at RAAP, the additional swelling provided bythe acetone permits the NQ to be interdispersed in the NC/NG triple-basepropellant matrix.
*4
I
I
I NOW i.-
S.•. .I ..o -" • * . 1'- OO- -O
00 0j
0 -
*1 I.*0101 0u 4
/'-f .1 .
L -m.-r -l
20 101-31Nk• % )o
"I'
-0P 00,
0 I I 12 I 14Nft gu(%)
II
I
Figure 1. Nitrogen content and solubility in ether/ethanol3n solvent system (2/1 ratio)
5
IThe solvent/propellant ratios are also determined by the NC content
of the propellant, i.e., the greater the NC content, the greater theamount of solvent required to dissolve the NC. The solvent/propellantratios must be closely controlled because of their effect on solventremoval which in turn affects propellant physical characteristics andballistics. Ethanol is used as a wetting agent to remove water from theNC and to prevent immediate plasticization of the micelles' outersurfaces on the addition of ether or acetone. For single-basepropellants, -10 lb of ethanol is added per 38 lb NC, 4 which dictates thesolvent/propellant ratio. For multi-base propellants, the %N in the NCand the NG content determines the solvent/propellant ratio. Formulti-base propellants produced at RAAP, the average solvent/propellantratio is 0.2/1.
A compendium of terms describing the degrees of propellantresolvation is shown in table 1. These terms describe the various statesof solvation to attain the desired colloidal system which is the intimatemixture of two substances, one of which, called the dispersed phase orcolloid (propellant), is uniformly distributed in a finely divided statethroughout the second substance, the dispersion medium (solvents). Theterms listed describe the degree of propellant solvation in the order asit occurs in the mixing process. Several terms are combined defining thevarious stages of solvation. One set of terms occurs twice since thepropellant will undergo these stages before and/or after the desiredcondition of plasticity is attained due to the production mix cyclerequiring an over-solvation step and a drydown step (solvent removal byvaporization).
12.3 Resolvation Time
3A 90-min time frame for resolvation was considered adequate to permitIntroduction of the resolvated propellants into the standardmanufacturing processes. The average production mix cycle time forsingle-base propellant is 15 min whereas the average mix cycle formulti-base propellant is 180 min. Mix cycle times of 90 min, i.e., themean of the single- and multi-base propellant cycle times, were chosen.Furthermore, during normal production, remix of single-base propellant,e.g., press heels and rework from subsequent cutting operations, requiresadditional mixing time (>15 min) due to the 13.15%N NC and the 87%+ NCcontent in the propellant whereas multi-base rework usually does notrequire additional mixing time.
2.4 Operating Parameter Evaluations
Four particle sizes, based on the test plan, 5 were evaluated toassess the effect of particle size on propellant resolvation: wholegrain, crushed, coarsely ground, and finely ground. The various particlesizes are defined as follows: whole grain as propellant requiring nopreparation; crushed as propellant grain slivers passed through a tornscreen on a Wiley mill; coarsely ground as propellant retained on a
3 6I
Table 1. Compendium of terms describing variousstates of propellant resolvation
Degree of solvationin order as it occursin the mixing process Term Definition
Excess solvent Solvent not absorbed inpropellant.
2 Grainy Unsolvated propellant pieceshaving a texture of fineparticles determined by tactile
or quality (for flake, crushed, orground propellants).
Grainy centers Unsolvated centers ofpropellant grains having atexture of a small hardparticle determined by tactilequality (for whole grainpropellants).
3 Poor consistency Solvent not distributed overpropellant evenly, therebyproducing various anomaliessuch as grainy pockets in thepropellant sample.
4 Softening depth Percentage of the grain thatcontains dispersed solventswhich Is measured by visual andtactile qualities and bypropellant grain length todiameter (OD) at the end of thetest (for whole grainpropellants).
5 Grainy-plasticity A combination of grainy andplasticity qualities areobserved in the propellant
or sample (for flake, crushed, orground propellants).
Grainy centers- A combination of grainy centersplasticity and plasticity qualities are
observed In the propellantsample (for whole grainpropellants).
6 Poor consistency- A combination of poorplasticity consistency and plasticity
qualities are observed in thepropellant sample.
7
ITable 1. (cont)
Degree of solvatlonin order as it occursin the mixing process Term Definition
7 Doughy Propellant is a dough free ofunsolvated propellant particlesbut is not pliable or workabledue to insufficient solvent in
or mixing resulting in the doughnot having the desired plasticqual1ty.
I Spongy Swelling of the propellant fromthe absorption of solventresembling elastic, porous, andabsorbent characteristics dueto the propellant beingover-solvated.
Plasticity Propellant is pasty (softmixture capable of being moldedor modeled of uniformcomposition) or elastic
8 0-75 % (capable of being flexible bybeing pliable when molded or
or modeled) forming a colloidalsystem that resembles a
9 75-100% pliable, workable dough free ofunsolvated propellantparticles. This condition isthe desired end product ofresolvating propellants.
10 Doughy See degree of solvationor number 7.Spongy
38
I
20-mesh screen; and finely ground as propellant passed through a 20-meshscreen and retained on a 50-mesh screen. Figures 2 through 5 are
I representative photographs of the four particle sizes evaluated in thisstudy.
The evaluations of production-established solvent/propellant ratiosshowed that increasing the amount of solvent enhanced propellantresolvation. The original intent was to vary the established RAAPproduction ratios in incremental steps of + 5%; however, single-basepropellants resolvated more easily with increased (i.e., greater thanproduction-established ratios) solvents whereas most of the multi-basepropellants resolvated using production-established or slightly increasedsolvent/propellant ratios. Established production solvent/solventratios, i.e., ether/ethanol and acetone/ethanol, are used at RAAP for themanufacture of single- and multi-base propellants, respectively; theseratios, of course, are dependent on the specific formulation ofpropellant to be produced. As in the solvent/propellant ratioevaluations, the original intent was to vary the production-establishedratio (i.e., 70/30 ether/ethanol) in incremental steps of + 5%. However,preliminary testing of the worst-case propellant, M1O, which has thegreatest NC content, indicated that deviation from theproduction-established ratio was not necessary since a solvent/propellantratio of 1.25/1 successfully resolvated the propellant. Asolvent/propellant ratio of 1.5/1, i.e., additional solvent, was requiredfor resolvation of the propellant using a solvent/solvent ratio of 65/35ether/ethanol. Based on these results, all propellants were resolvatedusing production-established solvent/solvent ratios.
Prior to resolvation, all propellants were analyzed for applicableformulation-specific chemical ingredients; lists of single- andmulti-base propellant ingredients and their functions are presented intables 2 and 3, respectively. The chemical analyses of Ml, M6, and M1Osingle-base propellants are presented in tables 4, 5, and 6,respectively. The only propellant ingredient out of specificationaccording to MIL-STD-652D is diphenylamine (DPA) in two of thesingle-base propellants: MI multi-perforated propellant for the 155-mmgun system (designated M3Al, lot number 60710) and M6 multi-perforatedpropellant for the 155-mm gun system (designated M119, lot number69877). Since these propellants are old lots and the OPA level is notbelow 0.2%, these are acceptable propellants in that the amount of DPAloss is well within the limits established in the storage specificationsdelineated in the applicable US Army/Hercules Incorporated contractualagreement.
The chemical analyses of M2, M7, and M9 double-base propellants arepresented in tables 7, 8, and 9, respectively. NG content is 0.05% lowfor the M2 propellant lot, which is not a problem due to storagespecification requirements, i.e., a certain percentage of NG loss ispermissible during storage. The ethyl centralite (EC) content for the M7propellant lot was low; however, since this propellant is a currentproduction item, this lot was blended to meet military specifications.
The chemical analyses of M30, M3OAI, and M31AI triple-basepropellants are presented in tables 10, 11, and 12, respectively. Asshown, none of the propellant ingredients were out of specification.
9
II3III _
__ 4~)
-IE'EU _ --
F-a)0~0
S.-I a)'U
a)I p -
S zC
5-.I*6 In - *1 a)
) F-
0IA
a)S.-I
IIII* 10
I
IIU)
00
OD z kD
OD,
:3a
CDC
:3
400S.-
0. CODiG)
(D a0)
IUU
3 V"
I ° C
,C
III a
!C
I0
* Lo
-
- L
-D -0r
= S--l 0. u
--(-
cc LL N
(It.
13
II Table 2. Single-base propellant ingredients and functions
Ingredient Function
Nitrocellulose (NC) A base ingredient that is a binder.Yields gaseous decomposition productsand energy.
Dibutylphthalate (DBP) Plasticizer. Peptizes binders such asNC so that fibers form plastics such aspropellant. Improves mechanicalproperties such as promoting Increasedelongation. Decreases energy.Decreases hygroscopicity.
Dinitrotoluene (DNT) Like DBP, it acts as a high boilingplasticizer-solvent, which aids inconferring upon the propellant itsproperties of non-hygroscopicity andflashlessness.
Diphenylamine (DPA) Stabilizer. Acquires decompositionproducts to inhibit decomposition anddecreases energy.
Potassium sulfate (K2S04) Flash and smoke reducers to inhibitcompletion of combustion and reduceflash (associated with radardetection). Particle size is3 important. Provides some energy.
Graphite Acts as a lubricant, thereby Increasingloading density. Also acts as aconductor for static electricity.
Ethanol (ethyl alcohol) Used in propellant manufacturing togelatinize 12.6% N NC. Ether isrequired with ethanol for NC having ahigher nitrogen content.
3 Ether (diethyl ether) Used in propellant manufacturingrequired with ethanol to gelatinize NCso that other ingredients can be boundinto It. Ether alone will not dissolveNC with any nitrogen content.
Water (H20) Used in propellant manufacturing tokeep NC wet and to purify. Keeps NCfibers from becoming tightly knit.Aids in cross linking NC so that3processing is facilitated.
I 14!
Table 3. Multi-base propellant ingredients and functions
Ingredient Function
Nitrocellulose (NC) A base ingredient that is a binder.
Yields gaseous decomposition products
and energy.
Nitroglycerin (NG) A base ingredient that yields gaseousdecomposition products and energy.
Nitroguanidine (NQ) A base ingredient that yields gaseousdecomposition products and energy.Gases are cool and much less gun barrelerosion is obtained than with otherpropellant bases.
Dibutylphthalate (DBP) Plasticizer. Peptizes binders such asNC so that fibers form plastics such aspropellant. Improves mechanicalproperties such as promoting increasedelongation. Decreases energy.Decreases hygroscopicity.
2-nitrodiphenylamine (2-NDPA) Stabilizer. Acquires decompositionproducts to inhibit decomposition anddecreases energy. (Also acts as ratemodifier.)
Ethyl centralite (EC) Stabilizer. Acquires decompositionproducts to inhibit decomposition anddecreases energy.
Potassium perchlorate (KClO 4) A burning rate modifier that promoteshigh rate for rockets. Alsocontributes energy.
Potassium sulfate (K2SO4 ) and Flash and smoke reducers to Inhibitbarium nitrate [Ba(N0 3)2 ] completion of combustion and reduce
flash (associated with radardetection). Particle size isImportant. Provides some energy.
Cryolite Flash reducer, Insoluble in water.Therefore, cryolite is good for slurrymix operations.
Graphite (glaze) Acts as a lubricant, thereby increasingloading density. Also acts as aconductor for static electricity.
15
ITable 3. (cont)
Carbon black Increases rate of burning. Opacifiesand prevents subsurface burning.
Ethanol (ethyl alcohol) Used in propellant manufacturing togelatinize NC.
Acetone Gelatinizes (peptizes) NC so that otheringredients can be bound into it.
Water (H20) Used in propellant manufacturing tokeep NC wet and to purify. Keeps NCfibers from becoming tightly knit.Aids in cross linking NC so thatprocessing is facilitated.
I
I
I
| 26
-00
LOArE%DO 1 *&-t a 4 4 - 4
.E nLo r- qr " ( 0C\Jc 0
LA r- WTr- 0)00 ko OLn ) -LA 0
-T (: - 000 CD. C L00
4- --
-02
04 1- r4 1.4 Z4 r2 00kcO c Ln - .0 w w
m LA*LA A00; 0 LA LA
r_ 0- - LA
4-
'4-0 .0
2A aI0 + t-*- E 0.
2 S-- 0
w mI+S-s I++II I 4-C> 05 C.C
V)~ 0 0 00 00 0Q E -LA
uL L - 006 00 aC3 00- LA.-
- 0-cc
uL Z a .. CL0)
I;3 ~O O-g LMLA, oa
IV
CL0
I-
II4- 4.
0c
LA A
04 0
0CCDa0 re
x U
CL S )0 0 0 00 a.O 0 .- -'
0 0 0 . Lac
0 > -t
3 I.4- ~ 4'0(d ro~a 4-> (2) cN- a)0
c) cu.c 4-ui -j 4a$ a
I~ z~o. IV A
4-I4j 7; Cr_ W 0= mC- C6(o18
edI 3= -
ICIlIrI
0
C>00 00
(4-
34 - GC)~ ~ ~ ~ 0)( Ee
CE E -
4- Z4~~~*~t E 00, - O(OO C)7 - -E
0 .) 0 .l 0
o 0000 0.'0 U
z e LA >4"- a Q, a
.00
>1.J >N- e0G 0
aIQ
c cuLAJlj* -7 O -0
.L L C~j 4A-- -- 1o
S. r- -w -
CL 0cu
C~C
+1 + III +1+1+I1 OL 0 0Q
C7
cu Lw
k.) (A3 ) C" 0
z 4- - (D
(n*4-'4- 0 m~ >.,
n~~-,C)w a~ 4-' (
rU1~~~I (.. *0 EV7U' U. -
4-.Zt n4L >
Uo 20 n a C 0 0: 'I j >4
I QUA 3 4 1 4 r d .
a;E 0
CLJ
a)0) 4-'0
0i- 0)
-4- . 0 .
0~ O -0 0 0= C-0xs
cu~~C C) c ooo(
00c
*00 0 00
(A (
V)Clj0 (
(* I0) -j tj 4-j
I~~~ ~ ,~ 1;U.~> 0)N- 0
I3. =I~-~ > 0 0 -- 0=00 0-- -00- f 0 4-'
I 21
a) C -o
0 C-'4-1
a Lo. > > C 0 0 - -0 Ow u
0- (D~ EE.C0 4 Ln (-"a)- :3-
C: z- z.
0 2n CC LO
Q. -. -o r-00-
w~j 0 0 0 O
cu >A M , ) \ 3 a
>I. rd 0 -
.- i.) - U - r
-- a41d 4 r oa) C
=~~~ =-a 4-ra- 'a
dI. D O('-0 Li ) >*- *- 0.
= Om- u
3 22
0
C* 4-4' a a
et3 o.-!'4-m 0. 0
* l c'*j IC di- 0 ~ 4-
-0 0~ 0
'4-0C
E
ro S. '-
oo - r- 3
4-,J4-
CU - U C ~ 0 ~
4.J V) 0 >I ~~ 0. -ZL( = I- gm
*OC 23_ -
I2' " e
0SIEq~
I
C W
4- (i (1
- - -2-4 a - M- 0 - C) CWC 0 aJ-c "- ur-- 0) 0 O
0. : 04-J 4- (cn 0 0 O OCk.
54-00
Cj I E
-c WrW43V4oS$.-. LmC
L~- 0 n L OC-fl- I I 1 0 a)
.0 c'ir- C;0C 0-U cNA C\JOU 4-
7E CD 0 00 0(A
-4-
S.(4A4 A(n CU (ACN - -Q c4
0JU .2 S- 0 1- l')
CUUC E *& -2 l=;2 1 ' C)* ~ - U0 .4-' *-.
- -U'0 C" 0 i 4 in n -L - 0
0 4- 4- C 4- r 0) 4-Z roQ-- -41 0 - >I 0 4-j
* 24
ICC-t -or AS. o v
1. 1-9 49- ' 97 C7 7 11 -11 -o1 M C 0 Q4.C -J 4- -'
a, 0 0. 0~
EEI1 04-) to
- ~ 0 o-n o Lc - 0 0 OCJ
ILA C 0a-~o O
(4-
* ~~ I0C 1.-OI 4-J
I ~ ~ ~ -.. >0 ~
LLO
*n U
Q- > 41 o
S.- .-Z M0 NC. f 2 0
0I' 4-J CP 4AXO> L)u
S. 0- >1 0 -JI 0. Z ZZ Cl. 7 Lm
* 25
2.5 Resolvation Testing Procedure
I Prior to the propellant resolvation tests, hazards analysisevaluations were performed to assure proper handling and safetytechniques of propellants. Applicable operating procedures were reviewedby the RAAP Hazards Analysis Department to ensure that specific safetyprecautions and controls were applied to each step of the laboratory
resolvation studies. The detailed hazards analysis is included asappendix A.
Resolvation tests were conducted on each propellant by weighing a 50-gsample and placing the sample into a transparent polyethylene sampleI bag. The propellant sample bags (6-in. x 6-in. x 6-mil) have vapor sealsthat fold down for taping to prevent solvent losses. Preweighed solventsfor testing solvent/propellant ratios in the specified solvent/solventratios of ether/ethanol or acetone/ethanol were then added to the sampleand the bag sealed. The sample bag was then placed in a bottle on aviscosity mixer that continuously tumbled the sample at a rate of 3.5 rpmto assure coating of the propellant by the solvents; the laboratorymixing equipment used for these studies is shown in figure 6. Thepropellant/solvent sample was inspected at 15-min intervals to record thetime and degree of resolvation.
The degree of resolvation was recorded per the terminology describedin table 1. If necessary, grainy pockets in the propellant/solventsample were removed by manually massaging the sample bag to assure aneven distribution of -solvents throughout the sample since no internalmixing, i.e., shearing action (work) imparted to the propellant via mixerblades, could be accomplished without solvent losses. Sample resolvationwas discontinued when the sample attained the desired degree ofresolvation or after 90 min of resolvation. Photographs of partiallyresolvated and acceptably resolvated M6 propellant are presented infigures 7 through 12; resolvation parameters are listed on the individualfigures. If the sample resolvated, the density and flow characteristics(extrusion pressure and flow) of the sample were determined.
I The density and flow characteristics of the resolvated propellantsample were determined using a plastic extruder (syringe), shown infigure 13, using the following known parameters: (1) volume, (2) weight,(3) total volatiles (TV), and (4) diameter as area of nozzle. Thedensity of the sample was obtained by blocking the sample at 20 psig,which was the measured pressure applied to the air cylinder. Flowcharacteristics were determined by measuring the time required to extrudea strand and the length of the strand over predetermined pressureranges. Propellant flow curves were generated and compared to actualproduction mixes (baseline data) to aid in determining if the sample wasover- or under-solvated. Examples of these curves are shown in figures14 and 15 for Ml and M6 propellants, respectively.
II 26
I
0)
-=I a) (.-
E
04
tot*_I LL.
27)
I CL0I - 00.
.4-)(a C)
CL Q0
-b 4
Ik 0
0- 4-)I ~40
S.-Q )
a) LO
-U>1
L- E
28u
0
I-
CL
* -W
C -
4-
CL
00IL 4-)ea
13 0
-)
to W
- LO
a) Lf)
S- 4-
0
*-0
L
I 29
(0
- -DC
E-4
C 4-J
C-0
0 oPC
-0eoc =C= 4-)
S.-
o,)
S.- Lo
4-) C
L
30.
40m
+-0
a) (o a
040
LS-
'-0
04-)S- (L)1 )
040
0.1
IV CL
(-0
0 S- 4o LA to
0n) v a
0-4
-
* U.
1 31
CL'U
0
C a
1 4W -)'Ct
p- rj
- U-.
a
CT)
32m
CL
OE
0
C-
- s-
(1)4- j
4- C :(> 0 S
4LLa-
330
S4- )Ix
LA-
I3
Uc
4-
41 0.''
0I
S.-
-
'Co CU-
35+
CJ I
-
*Ito
-r-4
I~ "4
36U
I2.6 Results
Propellant resolvation tests were conducted with single-, double-,and triple-base propellants. The resolvation tests for single-basepropellants were conducted using production-established solvent/solventratios of ether/ethanol designated for each individual propellant. Theresolvation tests for the double- and triple-base propellants wereconducted using formulation-specific, production-establishedsolvent/solvent ratios of acetone/ethanol. The complete laboratoryresolvation results for the four propellant particle sizes of each of thepropellants have been summarized in table 13 to facilitate the followingdiscussions. For each of the propellants discussed below, the degree ofsolvation attained is presented graphically in an accompanying figure; adescription of each degree of solvation is presented in table 1. Thedegrees of solvation described in this table were delineated in order toaccurately describe the stages of solvation which normally occur duringmixing of propellants in the standard manufacturing processes. Thedesired degree of solvation, i.e., the plasticity necessary to introducethe resolvated propellant into the standard manufacturing processes,shown in each figure is 9.
The single-base Ml propellant (both single- and multi-perforated)whole grains were not crushed because the whole grain itself isrepresentative of the crushed sample size (fig. 16). The resolvation ofMl multi-perforated propellant was conducted at solvent/propellant ratiosof 1/1, 1.25/1, and 1.5/1. The coarsely and finely ground samplesresolvated in 90 min at the production-established solvent/solvent ratioof 65/35 (ether/ethanol) for solvent/propellant ratios of 1/1 and1.25/1. As evidenced by the degree of solvation represented by they-axis in figure 17, only 60% plasticity or resolvation of the wholegrain/crushed MI multi-perforated sample occurred in 90 min. As shown infigure 18, the resolvation of Ml single-perforated propellant occurredfor the whole grain/crushed and ground samples within 90 min for thesolvent/propellant ratio of 1/1 at the production-establishedsolvent/solvent (ether/ethanol) ratio of 65/35 except for one groundsample. For the whole grain/crushed and ground Ml single-perforatedsamples at a solvent/propellant ratio of 1.25/1 and whole grain/crushedsample at a solvent/propellant ratio of 1.5/1, resolvation did not occurwithin 90 min due to inadequate mixing (figs. 19 and 20); therefore,these tests were repeated. The whole grain/crushed samples did notresolvate whereas the ground samples resolvated within the 90-min timeframe for solvent/propellant ratios of 1/1 (table 13).
The resolvation of M6 single-base propellant occurred for the coarseand finely ground samples within the allotted 90-min time frame for allthree solvent/propellant ratios of 1/1, 1.25/1, and 1.5/1 (figs. 21
through 24) for the production-established solvent/solvent ratio of 65/35(ether/ethanol). Only one each of the coarse and finely ground samplesdid not vesolvate in 90 min for the solvent/propellant ratio of 1/1.These tests were repeated since inadequate mixing occurred; the samplesresolvated within the 90-min time frame. The whole grain and crushed M6procellant samples did not completely resolvate in 90 min at varioussolvent/propellant ratios (table 13).
I 37
I
Iew(
N 0CL)
CLU
a.a
*0 4-4 4
0
A IU, 0 0
338
IA vit I(-0 ( ^t A ivai Wi
Vii L
U U
CL CL d c CL
;A Ln L LL-C 19a : P L. 0 . 0
LnL L L I
U, Ul Ul Cs
w3 %0O'I
,3j3
439
--
CL CL
CL CL
z) LOV! U)r-
cli 1
U) U)1 (01 C
cm CC Lcm
w) w aw u w Qu a
N ZI U~ ,
40)U)U
* C)0a0 Z) f 0aoCcu0
CL
01
CLC
-. c0-
OW U, n L Tco C ' T ) L n -c
0.0
al U,
I NI
'-U~ a)" cu Q)1 (U'J cuC" 4)1 a)
00. >Q.N
5 5vi (A(
'UjLU, 1~cu
- -- j -- U - 0
00: *0 0 0 C ' 0 0 C
5L 0 0. -- 41
0
-. 0 0.0- .0L
1AJ
co CD4 C 0 CDC
I- 0U (ACL
vi QJ
m 0O 0 C0
W 0 0
ot
00
> E Qa
c ra c
c ~ ~~ ~ ~ ~ o4 ' AC IL- Q .- .0 -0* ' ;
U I 0 I u IQ q) cuC a Sa0 a0 If I' I 0C a I c
CL >C (UC0> 0. w .
QJ A J
.0~ ~~ >0. 0 ~ .
CJ L00 0
L a
4 42
0 . 42....
IC
-o
S.-
LnC
* 40
I0 L
0.IY 4-0N 'E
N a)
a> CL
Ln C
3- Ln
97 k-Q 0
l CD* coi C) al -
44~~-
IL
~~-ca X--
C" LO
CO E
>N 06
4.W4
fICf
Iz~m&o So soil
-. 45
o -
P-4~
I I S4- (Z
II0
0 b c t 0 N 0O - CW4 e
46 0
-- 4-
C
*.4 C
LOW.'
(U
cc -0 CQ
uo.42Io j oto
47-
4-)c
0
V)
-4
a) c
4-PC
* 2 >
0~ c
a)
LLJ~j. 04-
x 448
I0
cI-IM -Li
I-
-0.0
E Co
41 0' a)
L.I0I.
5414
49-
4I ca-
I~~ Lnr~.
I0 I
* C\La.
- ~ 0)
I CICIA
0I CD c o i V Q w
50'
IEI1
.~- - b
~ -" -4-
Io I
0 0 cc 0
K, I I C51
I MID is a single-base flake in which a sample represents whole grain,crushed, and ground particle sizes (fig. 25). As shown in figures 26through 28, resolvation of the M1O propellant occurred for thesolvent/propellant ratios of 1.25/1 and 1.5/1 for the solvent/solventratio of 70/30 (ether/ethanol).
I Whole grain and ground samples of M2 double-base propellant did notresolvate within the allotted 90-min time frame when thesolvent/propellant ratio was 0.4/1 (figs. 29 through 31). The wholegrain sample resulted in "grainy centers" of unsolvated propellantgrains. The ground samples, both coarse and fine, resulted in "poorconsistency" due to the solvent not being evenly distributed over thepropellant, thus forming grainy pockets in the samples. Greatersolvent/propellant ratios of 0.6/1 and 0.8/1 resolvated the finely andcoarsely ground samples, respectively. However, increasedsolvent/propellant ratios did resolvate the whole grain and crushedsamples (table 13).
M7 double-base propellant resolvated within the 90-min time frame ata solvent/propellant ratio of 0.4/1. The crushed particle sizes containcase-hardened propellant pieces which require additional solvent (0.6/1)for resolvation; case hardening of the propellant granules occurs duringthe drying operations (primarily the water dry process). Ground samplesof M7 propellant resolvated within the allotted 90-min time frame at thesolvent/propellant ratio of 0.4/1, resulting in the same resolvationcurve for both coarsely and finely ground samples (figs. 32 and 33).
M9 is a double-base flake propellant in which a sample representswhole grain, crushed, and ground particle sizes. The M9 propellantresolvated within the. allotted 90-min time frame for solvent/propellantratios of 0.4/1 and 0.6/1. The greater ratio (0.6/1) resolvated thepropel.lant in 30 min whereas the lesser ratio (0.4/1) resolvated thepropellant in 90 min (fig. 34).
The resolvation of whole grain and ground samples of M30 triple-basepropellant also showed that particle size reduction of the propellant isa necessary requirement to obtain resolvation within the allotted 90-mintime frame (figs. 35 through 37). Resolvation testing over a range ofsolvent/propellant ratios (0.4/1 to 1/I) for the whole grain samplesresulted in all samples having "excess solvent" not being absorbed intothe propellant. The whole grain sample having a solvent/propellant ratioof 1/1 started to resolvate after 90 min, resulting in a sample havingpoor consistency." The coarsely and finely ground samples overlysolvated, resulting in a "spongy" mixture above the desired degree ofplasticity. Decreasing the solvent/propellant ratio from 0.4/1 to0.2/1-0.3/1 resolvated the ground samples to a more acceptable degree ofplasticity (figs. 36 and 37) since these propellant samples attainedborderline plasticity within 30 min. The crushed samples would notresolvate due to the greater amount of exposed case-hardened surface areawhich precludes solvent penetration in the NC matrix.
The resolvation of whole grain and ground samples of M3OAItriple-base propellant also showed that particle size reduction of thepropellant is a necessary requirement to obtain resolvation within the
I 52
I
UC
* 4)
533
1 CL0
c~I 0
IC 0-4
El 4- 1
-- S
0
I( cEr
a)
XI 0 4-)
C... 4 I4-)
4I -0
ci 0 C
L -
0 0 do - to to 4 CO) CU 4
I %Xo;1u.zs jo sejisoa
I 54
I 4-)
C)
I CL4-)ai
I I C)
(31
3: 0
C
C0 Q
.,- (31
fuC)
~C) C
0 aVQo 0
-S 15.-
0 cc to In . C§ cu -. a -
-U
I 55
II
I
4-)
C:
I I
(-)
O-
00
s.- .
II0-
Lr
a 0
ua;~uA~ram Jo*-r.co t-to.,.O No V 0 CQ 0
I !5
-1
II
w 0
0 Q
*- UCL (a
co
IV.4e
0
0 o
! o5
F 1 r r " " "r - - I--- -r 0
i C) I ,,lO 4ooo-zo
I
-- 4-,
04-
~> c
0.>
I --
Ib Io L v .
%zoT,41@.&tW ;a 0os
S8
Q)
00
Ci
10~~ rAc
CO4- C
CO) >
4- I 00cu.
I --
C\2) (1 C
0o 0I D t. 0 CC
I4
I I59
I1"
LC
C U
600
00
amC
10 Uc
Z: 0
(1)'.C
00
0
C/
4-1
i~~ ~ T (U--+
a~~ 0*t
UU.v&O j 010
I o -
0
.. ~ 0LO
0
0I) 4
-4-J 0
0 0.
4-j.
()0)
IT-- -r-
1 62
-4a
- C)
0l)
'4
oo
ci 0
I.4-1
0 0) 0 o
IZT4w&O Uooae
C") L~ L-
63C
00
X: 0
o
4-4(II0)
I 6a
C o
m~ C
-C
0. C
8. 0
0C)
I0 CL
fu a
IV
>0
30.'s&O 04. 00.se
o~6b
Iallotted 90-min time frame (figs. 38 through 40). Resolvation testingover a range of solvent/propellant ratios (0.4/1 to 1/1) for the wholegrain samples resulted in samples having "grainy" or unsolvatedpropellant for solvent/propellant ratios of 0.4/1 to 0.8/1. The wholegrain sample having a solvent/propellant ratio of 1/1 slarted toresolvate 75 min; however, at 90 min the sample was not a colloid. Thecoarsely and finely ground samples both resolvated in 30 min at asolvent/propellant ratio of 0.4/1, resulting in the same resolvationcurve. The crushed samples showed borderline resolvation at asolvent/propellant ratio of 0.4/1; increasing the solvent/propellantratio to 0.8/1 resulted in excess solvent over the propellant.
I Ground samples of M31AI propellant resolvated within the allotted90-min time frame when the solvent/propellant ratio was 0.2/1 (figs. 41and 42). At the greater solvent/propellant ratio of 0.4/1, thepropellant overly solvated, resulting in a "spongy" mixture. Thecoarsely ground sample was below the plasticity requirement forresolvation whereas the finely ground sample was above this requirement.The whole grain and crushed samples would not resolvate due to thegreater amount of surface area exposed, not allowing solvent penetrationinto the NG matrices.
I Table 14 shows the effects of the base ingredients (NC, NG, and NQ)on propellant resolvation of ground triple-base (M31AI, M3OAl, and M30),double-base (M2, M7, and M9), and single-base (Ml, M6, and MIO)propellants. The effects of NG and NQ on propellant resolvation appearto be minimal. As the percent of NC increases from triple-base tosingle-base propellants, the solvent/propellant ratio increases forpropellant resolvation. The multi-base (triple- and double-base)propellant solvent systems utilize acetone and ethanol. NC is also verysoluble in acetone, thereby reducing the solvent/propellant ratiorequired for resolvation of these propellants. NC is marginally solublein ether as well as ethanol, both of which are used in the single-basepropellant manufacturing solvent systems; however, NC will gelatinize ina 2:1 (ether:ethanol) combination of the two solvents. The increasedsolvent/propellant ratios for the single-base propellants are alsoinfluenced by the increasing NC content of these propellants. NQapparently does not influence propellant resolvation since th triple-and double-base propellants require the same solvent/propellant ratio forpropellant resolvation and NQ is insoluble in the individual solvents(acetone and ethanol) used in the manufacture of multi-base propellants.The presence of NG in the multi-base propellants probably influencespropellant resolvation since it is very soluble in acetone, which is usedin the multi-base propellant manufacturing solvent systems; NG also actsas a plasticizer that aids in resolvation. It is difficult to determineif the reduced solvent/propellant ratio required to resolvate multi-basepropellants is directly influenced by the NG or NC content of the
propellants.
However, as demonstrated by the resolvation of multi-basepropellants, NC content apparently affects propellant resolvation. Forthe ground samples having the lower NC content (M31AI propellant), thesolvent/propellant ratio of 0.2/1 proved adequate for resolvation; with asolvent/propellant ratio of 0.4/1, borderline resolvation approached the
1 66I
-S
a)0L a
0
a- )
CDL
0 CD
co
0r
E c
"0
0
I L.
00
02 0-4
0) -
0O-v&O j- + 0)>eU
67~ 1
-4-
U -0o
CD-0
Ca
0-
0 C:
*> 0
-aLO~( 0(2
C)
*znilwj ex~
56
U1I0
0.0
CD
ct
)c
4--
-0
~~0' 6
we a, CD
Jcu
-+o
a 0 ca
uolllv.&,Io4- 0oS.S~
q-4~ 69
w 0
IC C
UL -4)
L3 g 0 m
+ L4-jI>4-LMI'
£3 >.
U~j,4,w,&jo0 0.6.2o
9 00
QIr
- 0,
wra
-a C
Co
o w0~
Io >
CU 0
0~~~ ~ ~ ~ to 9 D .* M ~
I0148&OOj 026
e4 ~ ~ 71
CL
0.r~0S.- C '- I,:0l
C- S- -> cC
U): --.-- -3
" qr 4-
(Ji 0 0>0 0~00 ()
Id) *C) . .a) C rd
2) oor- 0 0~ L))- CL
a) r-2 >1 - 3:
c0--CIL rn 40. C 0 C
o Cr * *Ln S
V) 0-I
Ln 0.I a - ) - 0-
+ +1 +1)0- -C 4 1E C: 0 ( 0-0a)0= 4--
C) <D C) ) C C) 7
0 = 000.T L) 000 EEC 0-U a -V
U)16 CL a)A.-U 0.
0A 0A 0~ 00iS.-.-' f. ~ C .
Ei~~~~ Ca. E0 .04' UC ~C C" -
-~CC)- - C - 0~CCL3 DI C
4-7)2 S C
CL>
Udesired degree of plasticity. For the ground samples having a higher NCcontent than M31AI propellant (i.e., M3OAI and M30), a solvent/propellantratio of 0.4/1 yielded the desired degree of plasticity for M3OAl withborderline results for M30. The ground samples having the highest NC
content (M7) required a solvent/propellant ratio of 0.4/1.
2.7 Discussion
In summarizing these resolvation tests, the criterion of importanceis obtaining colloided propellant within a 90-min time frame to permitI introduction of the resolvated propellants into the standardmanufacturing processes. The parameters varied for the tests were (1)propellant particle size, (2) solvent/propellant ratios, (3)pvoduction-established solvent/solvent ratios, (4) ingredient addition,and (5) percentage of remix. The two parameters that greatly affectedpropellant resolvation were particle size and solvent/propellant ratios.Smaller particle sizes reduced the case-hardened area in the propellant,allowing greater solvent penetration for softening the NC matrix. This,in turn, reduced the solvent/propellant ratios for resolvation.
The production-established solvent systems used in propellantmanufacturing at RAAP are ether/ethanol and acetone/ethanol systems forthe single- and multi-base propellants studied, respectively. Thesolvent/solvent ratios of these solvent systems are determined by the %Nand viscosity of the NC used in the specific propellant products. A + 5or + 10% variation of these production-established solvent/solvent ratiosgreatly affects the mechanical disintegration of the colloid insubsequent propellant processing and final product physicalcharacteristics (ballistics). Therefore, since production-establishedsolvent/solvent ratios resolvated the propellant samples satisfactorily,variation of these ratios was not necessary.
The complete chemical analyses of the propellant samples (tables 4through 12) indicated propellant ingredient addition for these tests wasnot necessary. Based on these results, the propellant was utilized as
100% remix; therefore, partial remix would more closely approximate the3virgin mix blends in actual production processes.Not all of the whole grain samples resolvated in the 90-min time
frame as designated by the under-solvated propellant in table 15.Under-solvated propellant results from one of two conditions: (1) notenough solvents were added to the propellant sample to achieveresolvation or (2) case-hardened particles in the propellant sampleprecluded resolvation regardless of the amount of solvent added,resulting in excess solvents covering the dried propellant particles.The solvent/propellant ratio range tested with no propellant resolvationoccurring in the 90-min time frame for each propellant type is shown inparentheses in table 15. Those that did resolvate were the smallergrains of Ml single-base (single-perforated), M9 double-base flake, andM10 single-base flake. Both the Ml single-perforated and M1O flakerequire a 1/1 to 1.25/1 solvent/propellant ratio for resolvation; this isin close agreement with virgin single-base propellant where 10 lb of
1 73I
I"C:)
4, CA-
U-~ 0
cu 0
00
oo - C)C -4CLo -.
CL
o ~0- - 0
00
u 0
I 4I
0 _
4- 0 - 0- 0o 4-- a C C
a)~V V. -- I -I VI
04- 0L
0 )
0-0
0 >CUL
o -_ 0 0 WCS- - 0 - 4-i 4-' 4 4 ~a
-r 0~ 4- 0 I.
IAC rr CA L )C
CL u sa)
V1 Lm fl vi (ILI A10 I
C- -a)
GI 4-- ILV) >A 0 L
0 0c ~ -~ 03a .4 CZ~ -E -0m r- CLM 0nr
74
ethanol is required to wet 38 lb NC, resulting in a production-established solvent/propellant ratio for virgin material of 0.9/1 for M10flake and 0.62/1 for Ml single-perforated propellant. In the resolvationstudies, additional solvent is required to penetrate the case-hardenedpropellant for softening the NC matrix. The double-base M9 flakepropellant requires a 0.4/1 to 0.6/1 solvent/propellant ratio forresolvation. Again, the additional solvent was required to penetrate thecase-hardened propellant.
Not all of the crushed samples resolvated in the 90-min time frame.Those that did resolvate were multi-base propellants having less NCcontent than the single-base propellants. M7 double-base propellantresolvated using a solvent/propellant ratio of 0.4/1 whereas M2double-base propellant did not resolvate due to its greater NC content.M3OAI and M31AI triple-base propellant showed borderline resolvation atsolvent/propellant ratios of 0.4/1 and 0.7/1, respectively. M30triple-base propellant is a large grain propellant and did not resolvatedue to the crushed samples having a greater amount of case-hardenedsurface area exposed, not allowing solvent penetration into its NC matrix.
The coarsely ground propellant samples are those retained on a20-mesh screen. Nith one exception, the single-base propellantsresolvated using solvent/propellant ratios of 1/1 and 1.25/1, i.e., Mlsingle-base (single-perforated) propellant required a solvent/propellantratio ranging from 1/1 to 1.25/1. The multi-base propellants resolvatewhen the solvent/propellant ratio is 0.2/1 to 0.4/1. One double-basepropellant, M2, required a solvent/propellant ratio of 0.8/1 due to itsgreater NC content. One triple-base propellant, M30, requires asolvent/propellant ratio between 0.2/1 to 0.3/1 for resolvation. Theexact solvent/propellant ratio for M30 was not established since internalmixing, i.e., shearing action (work) imparted to the propellant via mixerblades, is required to ach' ve resolvation.
The finely ground propellant samples are those having particlespassing through a 20-mesh screen and retained on a 50-mesh screen. Allof the single-base propellants resolvated using a solvent/propellantratio of 1/1. The multi-base propellants resolvate when thesolvent/propellant ratio is 0.2/1 to 0.4/1. The double-base propellant,M2, required a solvent/propellant ratio of 0.6/1 due to its greater NCcontent; the finely ground particle size permitted 75% of solvent usageas compared to the coarsely ground particle size. The triple-basepropellants either overly solvated at a solvent/propellant ratio of0.4/1, did not resolvate at a solvent/propellant ratio of 0.2/1, orshowed borderline resolvation characteristics; again, internal mixing isrequired to establish the exact solvent/propellant ratios (especially forM30).
3.0 BENCH-SCALE PROPELLANT RESOLVATION STUDY
The two main parameters for the bench-scale study were that (I) thestudy be based on the results of the laboratory-scale study and (2) thepropellant resolvated during the mixing evalua.ons of the bench-scale
75
studies be properly resolvated to permit its introduction into thestandard RAAP manufacturing processes. Hazards analyses were conductedbefore and during the bench-scale evaluations to ensure conformance tosafety guidelines. Due to safety considerations of the availablebench-scale mixer, e.g., the presence of NG, only single-base propellantswere considered for the bench-scale study. More detailed informationconcerning the hazards analyses are included as appendix B.
Of the single-base propellants evaluated in the laboratory-scalestudy, both Ml and M6 were being produced on the RAAP production lines.M1 propellant was selected rather than M6 propellant because largerquantities of MI were available for rpsolvation; furthermore, the ease ofusing the MI die in the 4-in. press without modifications was a factor inselecting Ml propellant for these studies. Also, M6 propellant was notbeing dried in the standard drying operations but was being dried at thecontinuous automated single-base line (CASBL) where retaining thepropellant identity would be prohibited.
Due to production constraints and availability of propellant forresolvation, sulfonated Ml multi-perforated propellant for the i55-mm gunsystem (designated M4A2) was produced during the bench-scale study. Thispropellant does not utilize the same die as the currently manufacturednon-sulfonated Ml multi-perforated propellant for the 105-mm gun system(designated M7241 Furthermore, ,ne minimal sulfate content (1.0%) ofthe sulfonated yropellant (bench-scale quantity of -30 lb) would bediluted in the production quantities of non-sulfonated propellant (-3,000lb) being processed and not affect final propellant specifications. (Itshould also be noted that the processing water itself used in normalproduction operations contains -0.2% sulfate.)
Five tasks were delineated for investigation during the bench-scalestudy: (1) preliminary extrusion study using th 4-in. press, (2) samplepreparation by grinding sulfonated Ml multi-perforated propellant at theincinerator facilities, (3) preliminary resolvation study by resolvatingthe ground Ml propellant in a 2-1/2 gal. Baker-Perkins mixer, (4)in-process operations of processing the resolvated, ground MI propellantin the 4-in. press and standard propellant manufacturing operations, and(5) sample analyses and data reduction to determine final productanalyses of chemical, physical, and ballistic uniformity of thepropellant.
3.1 Preliminary Extrusion Studies
In order to delineate the optimum operating parameters for extrudingthe resolvated sulfonated propellant in the 4-in. vertical press for thebench-scale study, data was collected from both the 4-in. and 12-in.horizontal presses utilizing the ]05-mm die. The data, summarized intable 16, are separated into three parts: (1) mix data, (2) 4-in. pressdata, and (3) 12-in. press data. These data include recidual solvents(RS), total volatiles (TV), strand length, extrusion time, density,specific gravity, pressure, and temperature. These data were collectedfrom a 12-in. press (normally used in production) during the extrusion of
76
0 ITXL l fl1N '0 C-C) * C)
m (, - - c C m - C
0-~~~~~ .C I 0 * 'r **00
o" - 0" L
N ~~~I I~O~' ~A C'Jf c 00
0 n CII ".0 0
zX 0- ) rn PI, -fl :2i r- C). - ) N
-o No I- N 0 a) C U a %0"0 U~I I0' r'0 C c0 n ED -j
-I -Q C - 0 N -
z 'IT Lf') 00rm 'IT 0- 'T O D LI r
oN *c** NN *Lfl' 00 UC' - OM ofCDm: x * * I In r- C), rl
I ~~ ~ z CD' ~ N'~~ - 0
I *(I N
0 0N L
* 0 c 0n :3 rN~ 0 - 3 L %0N .Nr- 0 *r L l. c oITIr ~ ~cn r,-S - 0'-N-0
.- M . . .- .: C' . C j 0 N0t 0 0
GJ z '00 NN 'N 000 fOLn L~ 00
x *1 *o %0 *:3 *N *n I n cN0C" 10 -
X:0 E - r- C D D(
0 C14 0"0 liL - nL0Z r" 0.'0 'IT rO r- Ln
InIN ~ ~ C I--0 r -u - ' -" ~ C9 r * nr
CL ON "'C 0 .N'- -I 0 C'4 C.' *Z
- C.~l0" en'O r) ~ 1,0 * - C, C -
z LO o IT r" c' CP2I 00 \,n Tc cz0
0 .- C I o L
C .I E0 ' c '0:0 c,, cc ~ -" --'CQ (
M M- 0 - - -nU~ m - . nCl q - E0 .C= .-
-0 CD r - 4 q m - ,T wJ * j c, -
000. 1 0 . C0 0 2wli cn -0 0 CD--~- " r
77 0
I oG T ri L n Z
LA Ln
NO nL CNJI CL (DN LAOn Iq OnL)( r =r- LA
z~ *-N LO LO- IT~ * *
-o) -a' co n o aa'dU) I*- A q0~ *0oL
* I I 1 -InA r
z 0 0 n n n 0, a
C14 LA
(Z. C N c Ln LA .0'n~~O L r- Ln ID C% r, m m
ul)~ ~ ~ I -'N.11 1:oC
M~~ 0- LAT ''.0 CDA CD .- , C
o LA
cL I , , c~LA -~ N Ao -m LAO *LA cm C('10 2m C
z lr lz 00a' LA U- A A CD'0 0 LO'N~~a *-N LACZ - ~ * -
z c'1'. m'0 'N. m CiIq C z C' 0 * 1 * LA In 1 -L
-: mL q0 * 0'm r- O O a'N-r-- 7
z 4) ' O O LA m m 0C%1 'N. o co CDA 'N '. J
x Zo *r T -LA 0, -o ,r 00-C' w~' C, .L(M1 O '
co -
00 NT C,
m0~ 'n. LAI c'1 00
~LAC'10C' 'IT'LA *q0O0 x
M m V %D 0 * N ( 1 4 LO *- n *r CD0 m *-
oM c-' cj
z cLO *O- r'0 * 0 mC~ * CI. itI =$
7 *- I v w 17 'JO CL'0
On - -
-i o 0N0 (O NJ'N-( 1
X: CD 00~
0n T. En C0L CJ
E.: -5 C -. 0 5 -
*0m -D -0 a S -' 0- '
X:~S'1~3 U03~ ~*S.' ~ '..~ 00 ..- ScIr
78I n (% nL
'Inon-sulfonated M1 multi-perforated propellant. To obtain data on the4-in. press (not normally used in the production of single-basepropellants), 10 lb of the non-sulfonated Ml propellant was removed fromthe production mixes and extruded through the 4-in. press using the samedie and screen size as the 12-in. press.
I The mix data shown in table 16 represents propellant samples from theproduction line mixes at the completion of the mix cycle. The RS and TVdata were collected to obtain initial conditions prior to extrusion inthe 4-in. vertical press and the 12-in. horizontal press. The 4-in.press data includes mix RS and TV after extrusion, test press number,temperatures (outside and bay), volume displacement of the 3xtrudedstrand (for flow rate calculations), length of the strand over a timedcycle, specific gravity (SpG), weight and density of the strand, andpressure. The ram rate (X4) was calculated for the 4-in. press using the
flow rates from the 4- and 12-in. presses, the cross sectional area andnumber of dies, and the measured ram rate (X12) of the 12-in. press.Similar data was collected for the 12-in. press.
I The critical information required to extrude bench-scale mixes in the4-in. vertical press is pressure vs ram-rate ratio (XI2/X4 ) at known TV.This data shown in figure 43 would be comparable to the 155-mm die sincethe pressure must remain in the 12-in. press control limits to ensureproper processing of the propellant throughout the remaining operationsof cutting and drying (solvent recovery, water dry, and air dry). Theupper control limit (UCL) is 2,600 psig whereas the lower control limit(LCL) is 1,800 psig. The 4-in. press tracked extremely well with the12-In. press, i.e., only 19% maximal variation (30% variation isallowable). The range in ram-rate ratio of 7.96 to 17.52 is due to theTV of the extruded propellant. The lesser ratio of 7.96 had a maximum TVdifference between the extruded strands (3.59%) whereas the greater ratioof 17.52 had a minimum TV difference (0.31%) for the same test presses.In order to extrude bench-scale mixes in the 4-in. press, a TV range of28 to 40% is allowable with an ether/ethanol ratio of 1.9 for both dies(105-mm and 155-mm).
3.2 Sample PreparationUThe granular MI propellant selected for the bench-scale study is
cylindrical with the following dimensions: 0.18-in. diameter and 0.40-in.length (fig. 44). The size of the propellant granules was reduced withthe Mitts and Merrill grinder (Model Number 14-CSF) located at the RAAPincinerator facilities. A schematic diagram of this process unit isshown in figure 45. This unit is a knife grinder (or granulator) with acutting chamber containing rotating and stationary knives. A 30 x 40-in.screen containing nine thousand 3/16-in. diameter holes in the bottom ofthe cutting chamber limits output particle size through the bottomdischarge. The propellant is introduced into the grinder as a scparatefeed along with a water feed to eliminate the potential of fire. Thereduced propellant is then discharged as a slurry of ground propellantand water.
I 79I
00~I',i
04 '4x
CI)LoW CO
'4 C\JI '4 +'4 0o
3n 73 (L4
a.)~~V M.c. ~~
a40
Ai~~~~~' *i c ;_ 4 c i d c
I 80
L.
0.
C4C
81II
U WATER
I PROPELLANT
KNVE3 THROAT BAR
ISCREEN
SWM
* *- LUR DISCHARGE
Figure 45. Mitts and Merrill knife grinder
33
I The slurry was transferred to the separation facility to be dewateredby gravity filtering on a Sweco ® Vibro-Energy separator (fig. 46).Separation was accomplished by continually feeding the slurry to thecenter of the 48-in. screen (150 mesh, 0.0041-in. wire diameter,0.0026-in. wire opening) in which the dewatered solid propellant moved byvibration across the screen cloth to a discharge spout. The water passedthrough the screen to a lower discharge spout. The propellant wascollected from the separator in 20-lb (wet weight) increments and placedon trays which were loaded into drying cabinets. A total of seventeentrays were filled, with the top tray remaining empty so that thepropellant fines on the lower trays would not contaminate the FAD bayduring drying. Six cabinets were filled, resulting in 2,040 lb (wet
* weight) of ground propellant.
The propellant was dried at 140°F for 24 h in a FAD to reducemoisture content to <1%. Propellant coniaining >3% moisture alil' notprocess properly. Chemical analyses were conducted according LOMIL-STD-286B on the propellant prior to grinding and after drying; noingredient loss occurred during sample preparation according to thesecnemical analyses (refer to table 4 for the M1 propellantspecifications). However, the MIL-STD-286B analysis for total stabilizeruses DPA as the reference measure; HPLC analysis revealed that twoby-products of DPA (N-NDPA and 2-NDPA) were present in the propellant,thus distorting the actual DPA content. Because of these by-products,0.14% of DPA was added to the solvent mixture for the resolvation studiesto compensate for the actual OPA loss. The TV analysis indicated 0.15%water, 0% ethanol, and 0.02% ether. Photographs of the variouspropellant particle sizes, together with particle size distribution, areshown in figure 47. The propellant was then remotely dumped and packedout in ten drums, each weighing 135 lb.
13.3 Preliminary Resolvation Study
A remotely controlled 2.5-gal. Baker Perkins mixer was used toperform the bench-scale resolvation studies. The mixer, fabricated ofstainless steel, is jacketed for water cooling or steam heating. Thesigma-configured mixer blades (fig. 48) are rotated in a front-to-backspeed ratio of approximately 1.88 to 1.0. Nominal clearances for thismixer are summarized as follows: 0.035 in. between blade and bottom ofbowl, 0.125 in. between blade hub and bowl ends, and 0.180 in. betweenblade tips and bowl ends. A Reeves Vari-Speed motor drive unit providesfor manual variation from 57.5 to 230 rpm; this unit was originallydriven by a 2-hp motor with explosive classification of Class I, Group D,and Class II, Groups E, F, and G.
The Hazards Analysis conducted prior to the bench-scale studiesindicated that this motor did not meet proper electrical classification,i.e., Division 1, Class I, Groups C and D and Division I, Class II,Groups E and G. Therefore, a modified air-purge to the motor meetingthis classification was installed and safety approved. Mixer evaluationswere initiated using the ground, dried Ml propellant. Blade speeds of 15and 30 rpm were used to approximate the blade speeds of the standard
I 83I
I SCREEN SURFACE4"1 FROM FEED TO SCREEN
CETRTEDWISIML
DICAG
Fiue4.Iec~boEeg eaao
I8
IL
0
U) UJIU 0 100
co~ m
00
LU .- cu
U, -oI - c -
85=
UUII
Q)I0)
'U
.0I ItOS0~* U')
*1~
S..
* C)0.
2 -~'UI .. 2~I ax &0
CM
-e* w C)
0S.-
4-,I 0
C)4-,I 0S
C)I*UIU 86
I
production sigma-blade mixers. A total of thirty iterative mixing trialsusing 4 lb of ground propellant per trial were conducted in order tooptimize the operating parameters, witn the last nine trials beingprocessed in the standard production operations of extruding, cutting,and drying for sample analyses. The results of the iterative trials for5 the bench-scale resolvation studies are summarized in table 17.
Based on the results of the lboratory-scale resolvation :tudies, thefollowing operating parameters were established for Trial 1: asolvent/propellant ratio of 1/1, a sorption cycle of 15 min (required toallow the solvent ample time to permeate the dry propellant), and a mixcycle of 20 min. The mixing operation was stopped at 5-min intervals forexamination. The data (TV and approximation of the degree of solvationattained) indicated that maximum solvation was attained at I5-min mixtime; however, the mix was over-solvated and a 24-min arying cycle wasrequired to remove the excess solvent. A plasticity degree -f -90% wasattained.
To reduce the drying time, a solvent/propellant ratio of 0.8/1 wasevaluated in Trial 2 in order to more closely approximate standardproduction ratios. A 25-min mix cycle in conjunction with a 15-minsorption cycle was selected in order to compensate for the loweredsolvent/propellant ratio; however, propellant resolvation was attained in15 min, thus negating further evaluations of mix cycle times. Only a
5-min drying cycle was necessary to attain the required TV level forextrusion. As in Trial 1, -90% plasticity was achieved.
In Trial 3, using the same conditions for Trial 2, the procedure forcharging the mixer was reversed. In the first two trials, the solventwas added to the propellant in the mixer (to simulate actual productionprocedures); apparently, when the solvent is added to the propellant inthe mixer, a gelatinous layer forms, thus limiting solvent sorptionthroughout the contents of the mixer. In this trial the solvent wasplaced in the mixer and the propellant evenly distributed in it. Thisprocedure did not improve final plasticization (-70%) since the mix was
I over-solvated and a 15-min drying cycle was required to remove the excesssolvents.
In order to further enhance plasticization, Trials 4 and 5 wereperformed using a solvent/propellant ratio of 0.8/1 and a 30-min sorptioncycle. The solvents were added to the propellant to simulate actualproduction procedures. The additional time for the sorption cycle didnot change the drying cycle (table 17); however, plasticization greatlyimproved, i.e., -95% and -80-85% for Trials A and 5, respectively.
Trial 6 was also conducted with a solvent/propellant ratio of 0.8/1.Since indications of a gelatinous layer were observed in Trials 4 and 5,the solvent was placed in the mixer and the propellant evenly distributedin it. The sorption cycle was increased to 45 min, foilowed by a 15-minmix cycle and a 4-min drying cycle to achieve -98% plasticity.
In Trial 7, an actual standard production solvent/propellant ratio of0.7/1 wa: evaluated with the solvent added to the propellant (standardproduction procedure) using a 15-min mix cycle. Visual observation of
I 87I
41 c + o . . . .4. .~ . . . . . . . . . . . + + . . 4.
3' j 0 C,~- 0 3 a 0N C7, (71 (1 31 (71 a, al C a a C, al a'1 -0 a, a' '
410
-~~~r r-cl---J
In
C .40J 0=n1iLna oCN m C ' o 1.
'Cr C Cn 1= C: CCE)II I I I I
-j I
- - - - - -~ - - - - -
0. a~
*1 -z-1 1 0 10 1)
ImL-~ w 'D ."I ? m 'a n 73 m 03 -1 T: "o .D -A .0 7D m 'a
wN~ ' fl N 0 NNV1N(>3Ja 1 (D I I) l i i l - ) a
a a c88
I a .C- ---
Ithe mix following the 15-min mix cycle indicated that all propellantparticulates had not been plasticized, i.e., -70% plasticity. However,the under-solvated particles were evenly distributed throughout thecoalesced over-solvated propellant, indicating that additional mixing forfurther distribution of the under-solvated particles would not result inadditional plasticization. An 11-min drying cycle was required to removethe excess solvent from the over-solvated propellant.
To improve plasticization using solvent/propellant ratiosapproximating standard production ratios, the procedure for charging themixer was again -eversed in Trial 8, i.e., the solvent was placed in themixer and the propellant evenly distributed in it; furthermore, thesorp+'on cycle was increased to 30 min to achieve greater plasticity. Aplasticity of -80-85% was attained with a 3-min drying cycle.
Resolvation of the propellant for Trial 9 was conducted using thefollowing parameters which were optimized in Trials 1 through 8: (1)solvent/propellant ratio of 0.8/1, (2) 45-min sorption cycle, (3)propellant addition to the solvent, and (4) 15-min mix cycle. Toeliminate the drying cycle and determine that a 15-min mix cycle willresolvate the propellart, an additional 1 h of mixing was performed toverify that the 15-min cycle was adequate. A 95% degree of plasticitywas attained. Even though the TV level was low (19.48%), this mix wastested in the 4-in. vertical press. Excessive extrusion pressure (3,000psig) was observed due to the mix fouling the 16- and 40-mesh press
I creens. The particles on the screen, which were solvated on theoutside, contained dry pieces of propellant on the inside. This
indicated that the method of determining plasticity did not consider theincomplete resolvation of the larger particles
Trial 10 utilized propellant particles that passed an 8-mesh screento assure more complete solvation of the propellant to reduce fouling ofthe 16- and 40-mesh press screens. After a 45-min sorption cycle inwhich the propellant was added to the solvent, followed by a 30-min mixcycle to assure thorough mixing, a 15-min drying cycle was required toremove excess solvent from the plasticized (95+%) mix. Extrusion of thismix also resulted in high pressure (3,000 psig) and fouling of the pressscreens.
In Trials 11 and 12, propellant particles that passed a 10-meshscreen were evaluated using solvent/propellant ratios of 0.8/I and 0.7/1,respectively. The propellant was added to the solvent for a 45-minsorption cycle, followed by a 15-min mix cycle. No drying cycle wasconducted in either trial in order to ensure solvent-wet propellant (95-.%plasticity) for extrusion. Both mixes, which were extruded in the 4-in.press, resulted in high extrusion pressures (-3,000 psig) with fouling ofthe press screens; however, propellant strands were obtained. Visualinspection of the strand from trial 11 (0.8/1 solvent/propellant ratio)indicated that the strand was too soft for further processing, i.e.,cutting, whereas the strand from trial 12 (0.7/1 solvent/propellantratio) was processible. The high pressures and the fouling of the pressscreens prohibited complete extrusion of both mixes. Since propellantstrands were obtained from both of these trials, the drying cycle
I89I
following mixing was unnecessary; therefore, the drying cycle waseliminated in subsequent trials.
Trials 13 and 14 were performed similarly to trials 11 and 12,respectively, with an additional 5-min mixing time in trial 14 to assureeven distribution of the solvents. The same degree of plasticity (95+%)was obtained for both mixes. The extrusion of trial 13 resulted in lowpressure (-1,000 psig) with the resulting strand being too soft to cut.In trial 14, high pressure (-3,000 psig) during the extrusion resultedfrom fouling of the press screens; however, a short strand of propellantwas extruded before the extrusion was discontinued. Visual inspectionindicated that this strand was processable for cutting.
To determine if additional sorption time would have improved theextrusion in trial 14, i.e., reduce fouling of the press screens, trial15 was conducted using a 60-min sorption cycle and a 20-min mix cycle.The results were similar to those of trial 14; however, the additionalsorption time produced an improved quality strand for cutting, i.e., thesurfaces of the strands were smoother than in the previous trials.
Trials 16 through 20 were conducted similarly to trial 15 with oneexception. A 0.75/1 solvent/propellant ratio was chosen for two reasons:
(1) The 0.8/1 solvent/propellant ratio (trial 13) resulted in lowextrusion pressure (-1,000 psig) and a propellant strand thatwas too soft to cut whereas a 0.7/1 solvent/propellant ratio(trials 14 and 15) resulted in fouling of the press screens(prohibiting complete extrusion of the mix) even though thestrands were processable for cutting.
(2) Consideration was also given to further processing with respectto granule shrinkage that occurs during solvent removal in thedrying operations.
The propellant mix in Trial 16 was not extruded due to the unavailabilityof the 4-in. press. The addition of solvents to prevent the propellantmix from drying until it could be extruded would have negated the results
of the trial, i.e., an unknown solvent/propellant ratio. The extrusionpressures during Trials 17 through 20 were high (-3,000 psig), due toslight fouling of the press screens. The majority of the propellant mixin each trial was extruded before fouling of the press screens occurred,yielding acceptable strand lengths with varying surface qualities rangingfrom rough to smooth. These factors, i.e., high pressure, screenfouling, and varying surface qualities of the strands, indicated that theram rate of the press was inconsistent and the press required additionalhydcaulic fluid to increase the ram rate.
Trial 21 was conducted similarly to trials 16 through 20. When thismix was extruded, additional hydraulic fluid was added to oress head toincrease the ram rate, resulting in complete extrusion of acceptablepropellant 'strands and clean press screens. The addition of excesshydraulic fluid to the press head is not a normal operating procedure forthe production of Ml propellant in the 12-in. horizontal press.
90
Trials 22 through 30 were conducted as trial 21. The propellantstrands were processed through the cutting and drying operations asdescribed below. Chemical, physical, and ballistic analyses wereperformed on the finished MI product for comparison to the specificationsin MIL-STD-286B (table 18).I3.4 ProcessingI
As noted earlier, the propellant from the last nine of the thirtyiterative mixing trials was further processed in the standard productionoperations, i.e., extruding, cutting, and drying, to produce a finishedproduct. The extrudate from the 4-in. press was collected on cones for
transport to the cutting operation. During this operation, the strandsare fed into the cutting machine through holes in a feed bar. KnurleGrollers grip the strands and draw them inward through a middle and rearfeed plate, or cutting die, to the blades. The speed of the rollers issynchronized with the speed of the blades so that a length of strandequal to the desired length of grain will be drawn into the machine eachtime a blade passes the cutting die. A reservoir, positioned on top ofthe cutting machine, contains a coolant solution (water) which drops onthe die to keep the surface moist. This solution helps prevent excessivefriction and keeps particles of propellant from sticking to the ruttingbl ades .
bdThe grains of propellant are severed from the incoming strand andfall in stainless steel containers. After cutting, the propellantcontains -30-33% solvent which must be removed. The propellant grainswere placed in sausage bags (muslin bags equipped with drawstrings) toretain propellant identity throughout the remaining manufacturingprocesses, i.e., solvent recovery, water dry, and air dry. The sausagebags of propellant were placed in heated solvent recovery tanks which a-purged with inert gas to lower the oxygen content of the tank to a saioperating limit (-0.8%). Heated inert gas is forced through th,propellant to vaporize solvent which is then condensed and sent tcactivated carbon tanks for purification; the solvent remaining in thepropellant (-7.0%) is then removed to an acceptable level via the waterdry operation. The resolvated propellant remained in the solventrecovery tank for 36 h.
The water dry operation is accomplished in a large wooden tank inwhich heated water is circulated throughout the propellant. The retainedsolvent in the propellant has a high affinity for water; surrounding thegrains with water causes a condensation and diffusion of solvent at ornear the surface of the propellant into the water. The resultant voidsare filled by solvent from the interior of the grain. The water dryprocess is continued until the RS level meets specification requirements,i.e., 0.70% maximum. Following 90 h in the water dry tank, theresolvated propellant was sampled for RS analysis, resulting in 0.69%RS. The propellant remained in the water dry tank for an additional 27 h
* prior to transfer to the air dry tank.
I 9i
I
Table 18. Results of chemical, physical, and ballistic analyses
for resolvated Ml propellantIChemical composition
Propellant ingredient/ Specification M1 f/155mm M4Al
characteristic requirement Composite
Nitrocellulose (NC), % 85.00 + 2.00 85.05
Dinitrotoluene (DNT), % 10.00 + 2.00 9.75
Dibutylphthalate (DBP), % 5.00 + 1.00 5.20
Diphenylamine (OPA), % 1.00 + 0.20, -0.10 1.08
Potassium sulfate (K2S04), % 1.00 + 0.30 0.65
Total volatiles (IV), % 1.13
Moisture (H20), % 0.60 _+ 0.20 0.62
3 Residual solvent (RS), % 0.80 maximum 0.51
StabilityColor change No color change in >60 min
40 min (minimum)
Explosivity Shall not explode >5 hin <5 h
Physical dimensionsStd dev, 9 of mean
Specification Die Finished Spec (max) Actual
Length (L), in. 0.447 0.4161 6.25 1.22
Diameter (D), in. 0.266 0.1882 6.25 1.39
Perf dia (d), in. 0.022 0.0160
Web diff., s
% of web avg 15 max -5.36
L:D 2.10-2.50 2.21D:d 5.0-15 11.8
Web avg, in. 0.0354
3 Closed bomb (200 cm3 . +90F)
Relative quickness (RQ) 99.44
Relative force (RF) 99.74
9I
I 92
I
IThe air dry operation removes surface moisture from the propellant
grains by forcing heated air through the propellant grains and exhaustingit to the atmosphere. In the air dry process, the propellant is loadedinto tanks through which the circulating air is maintained at-130-150 0 F. After drying for 6 h (based on production operatingprocedures for Ml propellant), the propellant was sampled for moistureanalysis. The results of this analysis indicated 3.1% moisture, wellabove the defined specification requirements of 0.6 + 0.2%. However,additional drying of 18 h resulted in 0.62% moisture; this additionaltime was required since the propellant was contained in sausage bagswhich may have hampered drying. During packout of the final propellantproduct, samples were taken for various physical, chemical, andballistics testing. The results of these analyses are presented intable 18.
3.5 Sample Analyses
Chemical, physical, and ballistic testing was performed to determinecompliance of the reclaimed Ml propellant to the applicable propellantspecifications. The military specifications pertaining to eachingredient in Ml single-base propellant are summarized as follows:
Nitrocellulose (NC) MIL-STD-244Dinitrotoluene (DNT) MIL-STD-204, Rev. ADiphenylamine (DPA) MIL-D-98
Potassium sulfate (K2SO4 ) MIL-P-193Dibutylphthalate (OBP) MIL-D-218
Chemical analysis is defined as the determination of the percentagesof all ingredients present. The chemical analyses were performed incompliance with the test methods delineated in the item specificationsfor final lot acceptance testing or in MIL-STD-652D, "Propellants, Solid,for Cannons, Requirements and Packing." The standard procedures forconducting these tests are contained in MIL-STD-286B, "Propellant Solid,Sampling, Examination and Testing."
*Physical dimensions play an important role in propellant performancein that ballistic effects are controlled to some degree by the physicalform of the propellant, e.g., grain diameter and web distance. Physicaltesting consists of visual measurement, using a toolmaker's microscope,of propellant grain configuration and physical characteristics. Thefollowing parameters were monitored: grain diameter, grain length,perforation diameter, and web thickness.
Ballistic testing was conducted in the RAAP closed bomb, aheavy-walled cylinder capable of withstanding pressures up to 100,000psi. The closed bomb is equipped with a piezo-electric gage whichresponds to pressure changes; the testing was conducted under twoconstants: a volume of 200 + 10 cc and a temperature of 90 + 2°F. Thepurpose of the test is to determine the quickness voltage, i.e., the timerate of pressure rise (dp/dt) for a sample using a firing sequence ofthree shots of standard (Ml production-grade propellant) and three shotsof sample (reclaimed MI propellant) alternately. The results of the
93
I
Itesting, i.e., relative force (RF) and relative quickness (RQ), are usedto determine sample compliance to specification. RF is the ratio of themaximum pressure of the sample to the maximum pressure of the standard;RQ Is the average rate of change in pressure with respect to time (psi/s).
3 The propellant met all the specification requirements for physical,chemical, and ballistic testing except the K2SO4 content, which wasslightly low. This was due to excess leaching during the drying cycle;the sausage bags also retained excess moisture which continued to leachthe K2 SO4 following removal from the water dry tank.
1 4.0 SOLVENT EXTRACTION OF SELECTED PROPELLANT INGREDIENTS
Laboratory-scale solvent extraction studies were performed todetermine extraction efficiencies and affinities of selected solvents forspecific propellant ingredients in single-, double-, and triple-basepropellants. A preliminary hazards analysis was performed to review thedevelopment, sampling, storage, and process procedures to be used in thisstudy (appendix A). Appropriate solvents were selected based onsolubility and distribution coefficient determinations and ingredientextraction procedures were developed for the three types of propellant.Two testing procedures were prepared for the solvent extraction studies:high-performance liquid chromatography (HPLC) and atomic absorptionspectrometry. A statistical study was conducted to verify that the HPLCmethods developed for these evaluations were comparable to the analyticalmethods delineated in MIL-STD-286B.
The solvents to be evaluated were chosen on the basis of safety andtoxicity hazards as well as a literature review conducted during thepreparation of the test plan for this project. 5 These solvents includedwater (acidic, basic, and neutral), alcohols, and ketones, among others.Solubility determinations for each solvent on sixteen propellantIngredients were conducted. Distribution coefficient testing wasperformed on the non-miscible solvent pairs to determine separationcharacteristics of these solvents for ingredient reclamation. Thosesolvents that optimally concentrated propellant ingredients were then3selected for use in the extraction procedures.
Solubility testing for the solvents listed in table 19 wasperformed. The solubility data for propellant ingredients have beendivided into three general categories: very soluble (>0.100 g/mL),slightly soluble (0.010 to 0.100 g/mL), and insoluble (<0.010 g/mL).Solubility determinations for the solvents listed in the test plan werecompleted, with the exception of chloroform. Since it is expected thatno significant difference would be noted between the use of chloroformand methylene chloride, and because minimal use of chloroform is mandatedIn the RAAP laboratory policy, chloroform was not tested. It must benoted that these results represent relatively crude determinations ofsolubility and may or may not agree with literature values; however,because all testing was conducted in the same manner, the results arevalid. The propellant ingredients were then separated into four groups:hydrophilic, organophilic, insoluble, and others as shown in table 20.
I94I
I 6
as I cm
40
41 4
410
C6
595
Table 20. Separation of propellant ingredients into groups
Hydrophi1ic Organophilic
Potassium nitrate DinitrotolueneBarium nitrate DibutylphthalatePotassium perchlorate DiphenylamineLead carbonate 2-nitrodiphenylaminePotassium sulfate Ethyl centralite
Insolubles Others
Graphite Nitrocellulose (12.6%N)Carbon black Nitrocellulose (13.15%N)Cryolite Nitroguanidine
96
Based on the solubility data, distribution coefficient (Kd) testingof propellant ingredients in four non-miscible solvent pairs wasperformed: methylene chloride/water, hexane/water, toluene/water, andether/water. Only neutral water 'distilled tap water, pH -5.9) was usedfor this testing. The Kd results are presented in table 21; theseresults follow the trends that could be predicted from the solubilitydata. Determination of Kd was not performed on any insolubles, NC, or NQ.
As evidenced in table 21, there are no appreciable differences amongthe four non-miscible solvent pairs which were evaluated. Therefore, theether/water solvent pair was chosen since ether is an establishedproduction solvent at RAAP. However, during testing on M1O single-basepropellant, the consistency of the propellant matrix appeared to changeafter the addition of the ether, hampering completion of the extraction.All subsequent extractions were performed using methylene chloride toyield concise phases for separation; because of safety and toxicityconsiderations, hexane and toluene were not evaluated.
4.1 Testing Procedures
3The analytical standards for use in the liquid chromatographicanalysis of the extracts generated in the solvent extractions are shownin table 22. A stock solution, prepared in HPLC-grade methanol, wasvolumetrically diluteJ yielding the standards for the single-basepropellants and double-base propellants; a stock solution, prepared in75/25 acetonitrile/water (HPLC grades), was volumetrically diluted toyield the standards for triple-base propellants. Because of the limitedsolubility of NQ, it was necessary to modify the amount of NQ included inthe triple-base standard. A chromatographic standard containing the 1,3and 1,? isomers of dinitroglycerin (DNG) was prepared in order to analyzefor the possible breakdown products of NG in double- and triple-basepropellants. A stock solution, prepared in HPLC-grade water, wasvolumetrically diluted to give the standards also listed in table 22.These standards were injected separately.
Testing was performed to determine the linearity of the calibrationcurves resulting from the injection of propellant ingredients anddegradation products in the single-base standard. For the series ofcurves (figs. 49 through 53), correlation coefficients are >0.999 for allcomponents except N-nitrosodlphenylamine (N-NDPA), which has acoefficient of 0.998. A representative chromatogram for a single-basestandard is shown in figure 54 which also lists the chromatographic3conditions.
The linearity of the generated calibration curves for double-basepropellants was investigated; the results are presented graphically infigures 55 through 58. As with the single-base propellant calibrationcurves, correlation coefficients were very high, ranging from >0.999 for1,2-DNG to 0.998 for NG. A representative chromatogram for a double-basestandard is shown in figure 59; the chromatographic conditions are listedon the figure.
1 97I
ITable 21. Distribution coefficients for propellant ingredients
Organic phasePropellant Ingredient Methylene chloride Hexane Toluene Ether
Potassium nitrate <1 <1 <1 <1
Barium nitrate <1 <1 <1 <1Potassium perchlorate <1 <1 <1 <1
Lead carbonate <1 <1 <1 <1
Potassium sulfate <1 <1 <1 <I
Graphite ...Carbon blackCryolite ....Dinltrotoluene 983 1217 1590 1260Dibutylphthalate 614 890 700 1083
Diphenylamine 1290 971 1031 1173
2-nitrodiphenylamine 584 674 512 685Ethyl centralite 810 915 714 815
12.6%N nitrocellulose -
13.15%N nitrocellulose3 Nitroguanidine
Kd was not performed on the insolubles, cellulose nitrate, or
£ nitroguanidine:
[X] organic phaseKd =I
[XI aqueous phase
IIII
I| 98
I
Table 22. Liquid chromatographic standards for single-,double-, and triple-base propellants
Std 1 Std 2 Std 3 Std 4Ingredient, ppm (2x)a (x) (3/4x) (1/2x)
Single-base:
Dlnitrotoluene 2088 1044 783 522N-nltrosodiphenylamine 132 66 50 33Diphenylamine 200 100 75 502-nltrodiphenylamine 200 100 75 50Dlbutylphthalate 1172 586 440 293
Std l Std 2 Std 3 Std 4(2x) (x) (4/5x) (1/2x)
Double-base:
I Nitroglycerin 2062 1031 825 516Ethyl centralite 162 81 65 40
Std I Std 2 Std 3 Std 4(2x) (x) (4/5x) (1/2x)
Triple-base:
Nitroguanidineb 2039 2039 1631 1020Ethyl centralite 254 127 102 642-nitrodiphenylamine 142 71 57 36Dibutylphthalate 430 215 172 108Nitroglycerin 3600 1800 1440 900
Std I Std 2 Std 3 Std 4Isomers:
1,3-dinltroglycerin 634 317 254 1591,2-dinltroglycerin 160 80 64 40
a Std 1 Is the stock solution.b NQ Is in excess in Std 1.
1 .99
I~ __
UI,,
IM 0)
Il C*C 4 -
1000
Iu
*o _ C
- C-
a I
P-4. 41~
CDC
1-01
CA
* >
1OL-
3m CL
CA
P-1-
In in N 01
102
cc >
a 5
II
_ _ _ _ _ _ __ _ _ _ _ _
I I I-
KM 0
(ep~xuanIfc.L
-44
- 40-
0 CO
4-) '
K:( 4.-)
04-
O~LO
-
1044
3 integration & online plot from OPU memorySignal A B Annotation
I S wibw 265,20 214,20Range 100 100
Zero "10 10
M= 0
- It
1~I I
ESTD
I NAME GR SI TIME TYPE REF AMOUNT WIDTH dTIME GUOTIENT[min] cm] (min] m)n] C amountI
3 2.40NT A 4.936 VB 562.901 0.1744 NNDPA A 6.731 BV 227.525 0.231S CPA A 7.589 V8 369.1S6 0.259
6 2NDPA A 12.502 88 243.119 0.401
7 1BP B 16.773 Be 445.703 0.510
Ii Chromatographic conditions:
25 cm Resolvex C18 4.6 ID 5 um particlesFlow rate = 1.0 mL/min
I Mobile phase = 75/25 methanol/waterTemperature = 40*CInjection volume = 10 uL
i Diode Array detection at 265 and 214 nm
Figure 54. Representative HPLC chromatogram for single-base propellant
Ingredients
3 105
U.
W4 0-
4-J -
cD
CQ~Ir 0
0 0
106~
I>
II
I~~ I0 -
c107
*~ a1
CisC
000
Id I -
dC
CO)C
C at .e108
I>
IW
=0
4-1-
I,4-
109-
Integrafion & online plot from DPU memorySignal A B C AnnotationS wlbw 214,20 254,20 260,80Range 100 t00 100Zero % to 10 to
-4 It
21
-
4-
-- I7HiiI
-jA ZERO
131 iI I
14!
ESTO
I NAME GR SI TIME TYPE REF AMOUNT WIDTH dTIME CUOTIENTCmin] I ppm] Cmin] min] [amount]
2 I,3DNG A 2.080 BV 379.950 0.1303 I,20N6 A 2.448 VB 98.9820 0.150
NG A 5.50 8B S41.747 0.268S EC A 12.679 88 39.4348 0.2S6
Chromatographic Conditions:
20 cm Hewlett Packard RP-8 4.6 mm ID 5 tim particlesFlow rate: 0 to 6.0 min = 2.5 mL/min
6.0 to 8.5 min = linear ramp from 2.5 to 3.0 mL/min8.5 to 12 min = 3.0 mL/min
Mobile Phase: 0 to 5.4 min = linear gradient from 25 to 35% methanol/water5.4 to 8.6 min linear gradient from 35 to 55% methanol/water8.6 to 12.0 min = linear gradient from 55 to 65% methanol/water
Temperature = 40'CInjection volume - 20 wLOrode Array detection at 214, 254, and 260 nm
I Figure 59. Representative HPLC chromatogram for double-base propellantingredients
110
The linearity of calibration curves generated by the analysis oftriple-base propellant ingredients (and final degradation products) wasestablished. Figures 60 through 66 represent the calibration curves forNQ, 1,3-DNG, I,2-DNG, NG, 2-nitrodiphenylamine (2-NDPA), EC, and DBP,respectively. Correlation coefficients ranged from 0.9998 for DBP to0.9212 for 1,2-DNG. The calibration curves for NQ and the isomers of DNGshow a slight curvature at the upper end of the concentration range.This area of the curve should not be used due to the predictable(specification value or less) nature of the propellant samples. Togetherwith the acceptable correlation coefficients, use of the calibrationcurves is justified. Figure 67 is a representative chromatogram of atriple-base standard.
Detection limits were determined for the various propellantingredients analyzed by the single-base propellant method. Determinationof the detection limit was performed by injecting lesser amounts of thecompound onto the chromatographic column. The lowest amount which wasrecognized as a peak by the integrator at its most sensitive setting wasconsidered the detection limit. It must be noted that at this level apeak was clearly visible, and the signal-to-noise ratio was much higherthan 3. The detection limits for single-, double-, and triple-basepropellants are listed in table 23. Detection limits of double-basepropellant are considerably higher (100 ng injected) than forsingle-base propellants; this can be attributed to the much lowersignal-to-noise ratio present in the 214-nm region of the ultraviolet
3 (UV) detector.
Detection limits for the triple-base propellant ingredients anddegradation products were determined (table 23). Nitrate esters show ahigher detection limit due to the decreased signal-to-noise ratio in thewavelength band of interest. The detection limit presented for NQ ishigh because of the wavelength (340 nm) from which it is taken, ThoughNQ can be detected more easily at 214 nm, it was determined that withsuch a small capacity factor (k' value of 1.27), a selectivedetermination of NQ would be accomplished at 340 nm. This detectionlimit represents a minimum NQ concentration of 6.5 ppm.
A statistical study was initiated to determine if the HPLC methodswere comparable to the analytical methods delineated in MIL-STD-286B.Ten samples of each for a representative single-, double-, andtriple-base propellant were analyzed by both methods. The means werecompared by a Student's t-test to assure that both methods werecomparable (table 24).
Results for DNT and DPA in single-base propellants indicate nodifference at the 95% confidence level as determined by the Student's ttest to establish the difference between means. The t values of DNT andOPA were 1.3701 and 0.7032, respectively. The critical t value is 2.110fo" seven degrees of freedom. Results for DBP also indicate nodifference at the 95% confidence level as determined by the Student's ttest to determine the difference between means. The t value for DBP was1.397, with the critical t value of 2.110 for seventeen degrees offreedomi. Results for EC in double-base propellant indicate no differenceat the 95% level while NG shows no difference at the 90% level for either
I
I>
ruC r
I~ aO4-4 4-
0
I z +I -,
.ozP-44
C4C
CCto CV) Co
0-4%XLOO 1 0 J0
112
*CV)
0 ro
4-j
IIci C - (10
o wf m v C 0c C weUj 0i A 4 . 6 C i0 ,- jnxix
~113
II
I "Lo
"
0
L-
4-44
II 0°'-
I C\t4- .
o -ow
r-4
I
I 114
I.Im '/) m m m I I
IIII
CA
(A
* NC0C-)I1..
0I z 0
4-) -r~
I C0 C
a. -~~L..
0~* -I-)o3
C COC)
4, 4-)a 00
U) ~(A -4-'I
.- 0)0. ~)
..1-0~- C-)
Q* N
-4
U)
1-ILL
I I I I I I -~
-4 -4 .4'
3 (ep~zueno~1.)s~urLoo u..zy
II 115
I
eooT
z O
1- 000 V)
000
-o
C-
1164
*v >
00
o
0 to 04-J
0 -
-o 4
4-
ICII
co m
00
W4- w.- Wq 0 0 0 0 0
1170
00
.4--
0Cl
0-
S.-
1 03
IL 44
0 030a
IS.- 0
L0-
Lii
6-4%Zr~o laAC
118
Iegration & cnire:1TV I AL I :-
A 9 C .ntt:
S w!bw 214 .,' 254,20 ZT4' -.3Car.;e 100 I
Zero 10 0
Ki-J
A ZERO
7
8 A ZERO
-
-! .- -7 1. - N A ZERO0 -- - -- - - .?
1 1, -CN A -----9 ------ -- -- - - - - 0 11
I: i S I II I iI,A lI I
,t-- 4 i I I i_ i I I
ESTO
* NAME T!T!4E TY0E O5F A IOI.NT JIOTW dTIME Q'rTrElf-
Flwrt:0t i ierrm rm .5 to 3.0] mL/min-] £mou
6 to16m inO3 0.492I ,2-ONG A 4.B SP - .130
5 N4 A .75 vvm lia ra0m f5 0.4t0 A/B 1.042B 2-NOPA A !1.30S PP "-' 0.155 0. 00 A/B 0.995
10 SC A ':.-' B - : " 0.162 0.OOO g.S 1.003
I C BP ,- 14.51 9P - A 4 0,= .0C AiB 0,560
Chromatoqraphic Condi tions:
20 cm Hewlett Packard RP-8 4.5 mmn ID 5 umiz particlesFlow rate: 0 to 6 min linear ramp from 2.5 to 3.0 mL/min
6 to 16 min = 3.0 mL/min1bile phase: 0 to 5.4 min = linear ramp from 0 to 45% methanol/water
5.4 to 3.6 min = linear ramp from 45 to 55% methanol/water
8.6 to 12 min = linear ramp from 55 to 65% methanol/waterTemperature = 400CInjection volume: 20 pLDiode Array detection at 214,254,340 nm
Figure 67. Representative HPLC chromatogram for triple-base propellantingredients
119
Table 23. Detection limits for propellant ingredients insingle-, double-, and triple-base propellants
Propellant ingredient Detection limit, ng*
Single-base:
Dinltrotoluene 33N-nitrosodiphenylamine 41Diphenylamine 162-nitrodiphenylamine 16Dibutylphthalate 55
Double-base:
1,3-dinitroglycerin 791,2-dinitroglycerin 80Nitroglycerin 120Ethyl centralite 40
Triple-base:
Nitroguanidine 16311,3-dinitroglycerin 1591,2-dinitroglycerin 120Nitroglycerin 1442-nitrodiphenylamine 17Ethyl centralite 30Dibutylphthalate 52
The detection limit represents the amount injected onto the column and
analyzed by the method listed in the test plan.
I3 120
ITable 24. Summary of Student's t-test for single-, double-, and
triple-base propellants
3 Single-base propellants:
DNT OPA DBPMIL-STD-286B HPLC MIL-STD-286B HPLC MIL-STD 286B HPLC
X 9.98 10.12 1.00 1.06 4.99 5.01s 0.14 0.29 0.03 0.06 0.01 0.04n 10 9 10 9 10 9t 1.367 0.703 1.397df 17 17 17I
Double-base propellants:
I NG ECMIL-STD 286B HPLC MIL-STD 286B HPLC
19.76 20.11 0.43 0.45s 0.45 0.51 0.02 0.03n 10 10 10 10t 1.713 1.985df 18 18
3 Triple-base propellants:
NG 2-NDPA DBPMIL-STD 2868 HPLC MIL-STD 286B HPLC MIL-STD 2868 HPLC
19.86 19.40 1.32 1.40 4.54 4.53s 0.39 0.36 0.03 0.08 0.08 0.13n 10 10 10 10 10 10t 1.740 0.993 1.495df 18 18 18
i - means . standard deviationn . number of determinationst - Student's t value3 df . degrees of freedom: n1 + n2 - 2
II
I 121
I
the double- or the triple-base propellant. Results for DBP and 2-NDPA Intriple-base propellants indicate no statistical difference betweenanalytical methods at the 95% confidence level as established by theStudent's t-test to determine the difference between means.
i NQ extract samples were not analyzed by the Student's t test due tosolubility problems of NQ in the HPLC solvents. Therefore, NQ extractswere prepared in a very large volume of water (500 mL) to circumventthese inherent solubility problems; the results are discussed later inthe solvent extraction evaluations for triple-base propellant.
Analysis of water-soluble salts in the propellant extracts wasaccomplished by atomic absorption spectrometry. A Varian Model SpectraAA-20 was used for all atomic absorption spectrometry determinations.Standards were prepared by volumetric dilution of purchased 1,000-ppmstandards (Fisher Scientific, Fairlawn, NJ). Potassium (K+ ) cationstandards were prepared to be 0 0.5, 1.0, 1.5, 2.0, and 2.5 ppm; boththe lead (Pb+2 ) and barium (Ba+2) cation standards were prepared to be 0,1.0, 2.0, 4.0 10.0, and 20.0 ppm. Figures 68 through 70 represent thecalibration curves and operating parameters for K+ , Pb+2 , and Ba+2 cationstandards, respectively. The correlation coefficients ranged from 0.999for the K+ cation standards to 0.997 for the Pb+2 cation standards.
4.2 Solvent Extraction Flow Charts
Solvent extraction procedures were devised for each of thepropellants based on the solvent evaluations. Extraction procedures forsingle-base propellant.s are shown in figures 71 through 73 for Ml, M6,and M10, respectively. The procedures developed for the Ml and M6propellants contain inherent drawbacks, e.g., high energy consumption inevaporating solvents which can lead to further degradation of stabilizerand separation of three ingredients (DPA, DNT, and DBP) may not befeasible due to solubility differences.
* In order to assess the effects of the chosen parameters of interest,e.g., solvent selection and distribution of propellant ingredients in thesolvent, testing was conducted with a mixture of propellant ingredients,not the actual propellant itself. In this way, the solvent systems wereevaluated without the influence of the propellant matrix in theevaluation. The ingredients contained in MIO single-base propellant,i.e., NC, OPA, and K2SO4 , were weighed and mixed together as thoroughlyas possible. The appropriate extraction scheme was followed, but theresults of three extractions indicated that inherent problems existed inthe method. For example, DPA analysis, i.e., extraction of the DPA fromthe NC with ether, revealed the procedure to be irreproducible due tosampling problems encountered as a result of the inhomogeneity of themixture. The large raw material size, inefficient mixing, and absence ofNC plasticization caused an inability to reproducibly sample the mixture,precluding evaluation of the extraction method for individual propellantingredients.
3 122
I
II K
1 n SAMPLE I REPLICATE I MANUALH 0 . ) l hn t, ASR C I CiO CEn I KM I BC OFF
PRINTING .. .. ...-- I~~cLI)aIru £),"*PU '
A
8.0008.8 CONCEITRATIOH 2.8
SOLUTION GRAPHICSTYPE CURSOR
I Operating Parameters: Air/acetylene flame5 mA lamp current766.5 nm wavelength
1.0 nm slit widthsample aspiration
IIIII
I Figure 68. Calibration curve and operating parameters for analysisof potassium (K+) cation via atomic absorption spectrometry
Im 123
mI
PbSAMPLE I REPLICATE I MANUAL
0 33 ABSORBANCE CONCENTRATION BC OFFSPRINHt INTEGRATION 5.8 (sec) AIR-ACETYLENE
CALlBRATION GR PH
AI BS
I8.0 (OHCEHNTATION 22.
SOLUTION GRAPHICSTYPE CURSOR
I Operating Parameters: Air/acetylene flame5 mA lamp current217.0 nm wavelenath1.0 nm slit width
sample aspiration
IIIIII
Figure 69. Calibrati n curve and operating parameters for analysis of3 lead (Pb+ ) cation via atomic absorption spectrometry
124
I- .u e m uum mmlin llln
n
I Ba0 009 BLA REPLICATE I MANUALABSORBANCE CONCENTRATION BC OFFINTEGRATION 5.8 (sec) N20-ACETYLENE
CALIBRATION GRAPH
0.057 "
I AII5
8.8 CONCEWTIAIION 11.9m o .6 oi~trn A i I .
W IE CUSOR
I Operating Parameters: Nitrous oxide/acetylene flame20 mA lamp current553.6 nm wavelength0.5 nm slit widthsample aspiration
I
I
I Figure 70. Calibration curve and operating parameters for analysisof barium (Ba+ 2) cation via atomic absorption spectrometry
II 125
I
I-
> --
I. 44-1EE
m et3a)
S.0.oe --4-ed 41
c~ c.4
0 0 D
00
- - 4--u 0c
x >1 0a4
=0. CI C _SI. 4-1 -
E
C~C41
0 0
4-
4 (WLI4I e
LII. cCIl 4
L126
I;L
.4-J
I= 00a
S- 0oC a
(AJ
4, 00
0 n
I- C0 4-0 u0fo7
0* x
-4-I
* 4J.j 4U
LU.C L
.4..)127
CL
4--o
00ciI~ ~ a)4~)4-1- 0 U
i~ CA0
,a Li .a0 EL 1S. 4-J4
-~ 0
CA -
-Q128
ITo circumvent this problem and simultaneously evaluate the influence
of the propellant matrix, mixture homogeneity was ensured by using groundMlO propellant (particles passing 20-mesh screen and retained on 50-meshscreen). Prior to grinding, the whole grains of propellant were washedto remove as much of the graphite coating as possible. As shown in theflowchart kfig. 73), the ground propellant was then extracted withmethylene chloride to remove DPA, resulting in a solid containing K2 SO4and NC. This solid was extracted with water to remove K2SO4 and leave NCas the solid. This evaluation indicated that the propellant matrix doesnot adversely affect ingredient extraction.
Double-base propellant extraction procedures are presented in figures74 through 76 for M2, M7, and M9, respectively. Pretreatment of the M2and M9 propellant includes washing with water to remove as much excessgraphite as possible prior to grinding (particles passing 20-mesh screenand retained on 50-mesh screen). Pretreatment of M7 propellant islimited to grinding. Furthermore, the NG and EC in all double-basepropellants are left together in a liquid matrix for safety reasons.
I Solvent extraction procedures were developed for the triple-basepropellants being investigated. Figure 77 represents the scheme for M30propellant, with pretreatment of the propellant limited to grinding(particles passing 20-mesh screen and retained on 50-mesh screen);furthermore, the NG and EC are left together in a liquid matrix forsafety reasons. Figures 78 and 79 represent the schemes for M3OAl andM31Al propellants, respectively. For these two propellants, pretreatmentconsisted of a water-washing step to remove as much excess graphite aspossible followed by a grinding step. The NG and EC were left together
* in a liquid mat,*Ix for safety reasons in the M3OAI; the DBP and 2-NDPAwere also left together with NG in the M3lAl for safety reasons.
The extractions were begun with solvent/propellant ratios of roughly10 to 1 and 5 to I (weight-to-weight). A weighed amount of the groundpropellant was placed in a screw-top vial and the extraction solventadded to the vial. The top was replaced and the sample was agitated onan orbital shaker for -24 hours. At this time, the sample was allowed tosettle and the supernatant liquid drawn off. This procedure was reneated
three times with all of the extracts being combined. By the thirdextraction, the samples with a solvent/propellant ratio of 10 to I had asupernatant that was clear, while the samples with a solvent/propellantratio of 5 to 1 ratio had a supernatant that was slight discolored. (Theoriginal color of the extract varied depending on which type ofstabilizer was added to each propellant.) Based on these results, asolvent/propellant ratio of 10 to 1 was used for all subsequentevaluations. After removing the methylene chloride extract from thevial, the samples were dried under a gentle stream of air to remove anyresidual solvent. At this point distilled water was added to the vial to
begin the extraction of any water solubles. Again, three 24-hextractions were employed with the extracts from each trial beingcombined.
1I 129
I
00
CI* 70
(cli
0
~ 4-'
£ 0
3 ..
0 C-,u (SI - - t
0 -ts
(3) (/)
cm -m
f130
Iu
I 04-J 0
o o- In
4-)4-
CL. C-o03 u
4 a)
00C:I 0
4-
I uj
131,
CL
4-J
LL, 4 WI I0 V
0
00
0 4 WIL a)C)
a))
UULA0
I 132
-.-I CLa)
I -
C)
-oIr 0).4-ID L) 0
CL,
0)
4-J.x0
0L -4LJ0
1 o 0
z Ca)
C..) 0.
4-1-aI X
.C.)
u0
13
C
0
CS.,
-u Z
= 4-1
C-) -~-.1C0
4-0 en
-w 3
a) a) -"Jod,
0'4 C ~ 0 a
00
C-) 0
0 V) .- 4-'
V- 0 ca --
3 0'
4-'
0' Li C
~00ccO
0cAiLL
134
CL
0.
S.L)L
a)
o .4-J
00
0 -o
0A C> 4-
OL a-, u-3
70 in S-oau rC - 0-4.- 4-S.
X 4-0I -0I x4'~
S.- 2= ~ L Z -
= 0 -4D.4-j
It, CC4A
E n 4-1
4-J 4
4-J L) LA
LAU
0 i:
CAi
135
4.3 Solvent Extraction Results for Single-Base Propellants
I Solvent extractions using the procedures described earlier wereperformed on three single-base propellants: Ml, M6, and M10 (table 25).As described in the following sections, analyses of the extracts revealedthe viability of several of the solvent extraction procedures. Theresults of the analyses of the extracts indicate that NC recovery fromsingle-base propellants appears feasible. The recovered NC (13.15%N)could be used in any propellant requiring this grade of NC.
4.3.1 Ml Propellant
The first extract, i.e., methylene chloride, of M1 single-basepropellant removed the DNT, DPA, and DBP along with any decompositionproducts of DPA. The methylene chloride was allowed to evaporate andmethanol was used to selectively solvate the DPA, DBP, an d DPAdegradation prod'.cts. The extracts were filtered and the cr'/stallinematerial (ONT) was dissolved in acetonitrile. Both extracts were dilutedto 100.0 mL with HPLC-grade solvents and subjected to HPLC analysis. Theextracts for DNT ranged from colorless to light yellow; the color in thissample was attributed to DPA decomposition products not being thoroughlyseparated from the DNT. Since the DNT also has a finite solubility inmethanol, the longer the extracts were washed with methanol, the greaterthe chance of obtaining ONT in the DPA/DBP sample. DNT recovery in theDPA/DBP extract ranged from 54 to 82% of the specification value.Recovery in the DNT extract ranged from 0.1 to 35.8%, though theseextracts normally contained some DPA and DBP also. DPA and DBPrecoveries in the DPA/DBP extracts were fairly consistent. DPA wasnormally recovered at an average of 75% of the specification value whileDBP was recovered around the 90% level. However, due to the inability toeffectively separate the individual components of the first methylenechloride extract, this approach does not appear promising for DNT or DPA
*recovery.
The first water extract removed approximately 32% of the K2SO4in the sample (based on specification value), as determined by atomicabsorption spectrometry. Analysis of the extracts by HPLC methods didnot reveal the presence of any DNT, DPA, or DBP. The second waterextract (acidic water) to remove lead carbonate (PbCO3) from the samplewas also analyzed by atomic absorption spectrometry. Since certain Mlpropellants contain 1% PbCO3 , the extraction procedure was developed toremove and analyze any residual lead following processing. Resultsindicated that only a very small amount of the PbCO 3 was detected,representing 0.04 ppm (mg/L); this amount of PbCO3 apparently wasabsorbed during the water dry process and remained in the solid materialleft behind. Again, HPLC analysis of the extracts did not reveal thepresence of any DPA, DNT, or DBP. The solid material left behind shouldcontain NC and any of the above-mentioned ingredients that were onlypartially removed. Calculations assuming only NC present indicate about98% recovery; even accounting for all the K2SO4 and PbCO 3 remainingbehind, recoveries are still around 96%.
I136
I
Table 25. Single-base propellant ingredient recovery
Wt % Wt %Propellant (calculated) (specification) . recovery
NC Recovery
Ml (NC) (NC) (NC)84.2 85.0 99.183.6 85.0 98.43 82.4 85.0 96.9
M6 (NC) (NC) (NC)
3 86.1 87.0 99.185.6 87.0 98.485.7 87.0 98.5
I M1O (NC + graphite) (NC + graphite) (NC + graphite)100.5 98.2 102.498.9 98.2 100.683.7 98.2 85.2
3 DPA Recovery
Ml 0.74 1.00 74.0
0.66 1.00 66.0? 79.03 (0.13) 13.
0.77 1.00 77.0
I (_)
M6 0.71 1.00 71- 31 72.9(0.02) 1.6
0.75 1.00 75.4(-)
0.60 1.00 59.87 75.7
(0.16) 15.9j
I MIO 0.88 1.00 87.8
0.89 1.00 89.2
0.90 1.00 89.5
II 137
I
Table 25. (cont)
DNT RecoveryWt % Wt %
i Propellant (calculated) (specification) % recovery
Ml 6.11 10.00 6]. f 87.6(2.65) 26.5j
1 5.49 10.00 54. 90.7(3.58) 35.8
8.29 10.00 82.?; 83.0(0.66) 0.1
3 M6 8.01 10.00 80.1 85.1(0.50) 5.01
(8.74 10.00 87.4 87.8
6.63 10.00 66.3 87.4(2.11) 2l .l_
3DBP Recovery
M1 4.48 5.00 89.6I (_)
3.95 5.00 79.0I (_)
4.67 5.00 93.4
I (-)
M6 2.60 3.00 86.5 88.1(0.05) 1.61
2.75 3.00 91.6(-)
152.08 3.00 69.5 87.90.55 18.4
II
1138
I
Table 25. (cont)
Wt % Wt %Propellant (calculated) (specification) . recovery
K2S04 Recovery
M1 0.34 1.00 33.50.32 1.00 31.50.23 1.00 22.5
M6 0.46 1.00 45.70.50 1.00 49.80.52 1.00 52.2
MlO 0.10 1.00 10.40.11 1.00 10.60.05 1.00 5.4
m Residual PbCO3
Ml 0 .03 ......0.03 ---0 .0 5 . . .. . .
m
3 The number in parentheses represents the amount in the DNT acetonitrile
extraction.
IIIImmI 139
II
4.3.2 M6 Propellant
* The M6 propellant evaluations were performed much the same asthe Ml propellant and yielded very similar results. The methylenechloride extract removed the DNT, DBP, DPA and its degradation products.As in the Ml propellant evaluations, the attempt to separate the DNT fromthe DPA/DBP extract was unsuccessful.
3 The first water extract removed -49% of the K2SO4 in thespecification value, representing a sizable increase from the Mlpropellant. HPLC analysis showed no carryover of DNT, DPA, or DBP.Recovery of NC was again around the 99% level, when considering materialremaining as NC. Taking the entire amount of K2SO4 into account, therecoveries could be as low as 97%.
I 4.3.3 MIO Propellant
The MlO propellant was evaluated similarly to the othersingle-base propellants. The methylene chloride extract was analyzed andresults revealed that -90% of the DPA specification value was removed andrecovered. The water extract removed only -10% of the specified value ofK2SO4 . The variability in these levels of removal is uncertain at3 present. HPLC analysis of the water extract revealed no DPA present.
Recovery if NC was calculated to be -101%; the recovered NC wascontaminated with graphite, accounting for the >100% recovery. Sinceonly 10% of the K2S04 was removed, recoveries are at the 100% level(taking the entire K2SO4 amount into account).
4.4 Solvent Extraction Results for Double-Base Propellants
U Solvent extractions using the procedures described earlier wereperformed on three double-base propellants: M2, M7, and M9. The resultsof the analyses (table 26) of the extracts indicate that NC recovery fromdouble-base propellants appears feasible; however, the recovered NC(12.6%N) should only be used in the production of double-base propellant
I since the NC has previously been contaminated with insolubles and NG.
4.4.1 M2 Propellant
The first methylene chloride extraction was used to remove NGand EC, leaving NC, graphite, barium nitrate [Ba(N0 3 )2], and potassiumnitrate (KNO 3 ) behind. Analysis of this extract revealed NG to berecovered at about the 95% level, based on the specification value. ECwas recovered at approximately 98% of the specification value. Becauseof inherent safety problems, no further separation of this extractedmaterial was performed; furthermore, utilization of this extractedmaterial is not recommended.
140U
ITable 26. Double-base propellant Ingredient recovery
I Wt % Wt %Propellant (calculated) (specification) % recovery
NC Recovery
M2 (NC + graphite) (NC + graphite) (NC + graphite)
79.1 77.85 101.578.6 77.85 100.978.9 77.85 101.3
M7 (NC + Carbon (NC + Carbon (NC + Carbon3 black) black) black)
60.1 55.80 107.760.1 55.80 107.7
I60. 55.80 107.960.2 55.80 107.9
3 M9 (NC + graphite) (NC + graphite) (NC + graphite)
61.0 58.15 104.960.4 58.15 103.960.0 58.15 103.2
3 NG Recovery
M2 18.62 19.50 95.518.85 19.50 96.718.28 19.50 93.7
M7 30.23 35.50 85.230.47 35.50 85.830.47 35.50 85.8
M9 33.11 40.00 82.833.22 40.00 83.132.07 40,00 80.2
I EC Recovery
M2 0.60 0.60 100.00.58 0.60 96.70.59 0.60 98.3
5 M7 0.74 0.80 92.50.75 0.80 93.80.75 0.80 93.81
I 141
I
ITable 26. (cont)
IWt % Wt %
Propellant (calculated) (specification) % recovery
EC Recovery (cont)
M9 0.47 0.75 62.70.48 0.75 64.00.47 0.75 62.7
Potassium Salt Recovery
(KNO3) (KNO 3 ) (KNO 3 )
M2 0.37 0.75 48.80.32 0.75 43.20.37 0.75 49.9
I (KCI0 4) (KC104) (KCI04)
M7 3.45 8.05 42.94.23 8.05 52.53.40 8.05 42.2
5(KNO 3) (KNO3) (KNO 3 )
M9 0.19 1.50 12.50.97 1.50 64.40.82 1.50 54.4
Ba(N0 3)2 Recovery
M2 1.22 1.40 87.31.25 1.40 89.51.23 1.40 88.1
IIII
~142
I
IThe water extraction step removed the Ba(N03 )2 and KNO 3 , leaving
only the NC and graphite. Analysis of the water extracts by atomicabsorption spectrometry revealed the KNO 3 to be recovered at around the49% level, and the Ba(N0 3)2 to be recovered around 83%. Analysis of thewater extracts used for atomic absorption spectrometry analysis by HPLC3 showed no indication of either NG or EC in the water.
The NC recovered in this extraction procedure is contaminatedwith graphite, as well as any other propellant ingredient not completelyremoved. NC removal was around the 101% level; as noted earlier, the NCwas contaminated with graphite . Since most of the other ingredientswere not recovered at extremely high levels, it can be assumed that theNC/graphite mixture would be contaminated with low levels of some ofthese ingredients. Since there is a possibility that NG may be presentalong with the NC/graphite, this recovered NC/graphite should only beused for the manufacture of double- or triple-base propellants, i.e., thepresence of graphite in the reused NC may affect propellant ballistics.
I 4.4.2 M7 Propellant
3 The methylene chloride extract removed the NG and EC. NG wasrecovered at about 85% of specification value while the EC was recoveredat about the 93% level. Analysis of the water extract again revealedthat no NG or EC was removed in the second step. The second step waterextraction did remove the potassium perchlorate (KClO 4 ) at around the 43%level.
I The NC and carbon black that should have been left behind wererecovered at the 108% level. Again, since all ingredient recoveries werelow, the recovered NC is likely contaminated with other propellantingredients.
4.4.3 M9 Propellant
This propellant is the only one of the double-base propellantsthat did not contain an insoluble ingredient (refer to table 20) with theNC. The first methylene chloride extract removed the NG around 82% ofspecification value, while only about 63% of the specification value ofEC was removed. Apparently, the higher NG level in this propellantsomehow hampers the extraction of the EC in that there is an inverserelationship present between the amount of NG in the double-basepropellant and the amount of EC recovered.
The first extract removed roughly 58% of the KNO 3 in thepropellant as determined by atomic absorption spectrometry. Analysis ofthe water extracts by HPLC methods revealed no NG or EC was extracted inthe water step. The solid material (NC) left behind calculated to be104% of specification value, apparently accounting for the EC or KNO 3
that was not extracted.
Examinatiot) of all double-base propellant solvent extractionresults indicate that NC recovery may be feasible. However, since the NCin most cases contains other insolubles or contaminants, the recovered NC
3143
Ishould only be used in the manufacture of multi-base propellants. SinceNG removal is lower than 100%, the recovered NC should not be used in anysingle-base propellant due to possible NG contamination; furthermore,this blend of NC (12.6%N) is only used in multi-base propellants.
I 4.5 Solvent Extraction Results for Triple-Base Propellants
i Solvent extractions using the procedures described earlier wereperformed on three triple-base propellants: M30, M3OAI, and M3lAl(table 27). The results of the analyses of the extracts indicate thatrecovery of NQ is very feasible; the recovered NC (12.6%N)/graphitemixture could be reused but only in the production of triple-basepropellant, assuming no ballistic effects from the graphite.
4.5.1 M30 Propellant
The first methylene chloride extract removed the NG and EC. TheNG was recovered at around 96% of the specification value. However, theEC was recovered at a relatively consistent 127% of the specificationvalue. Since this analysis is based on separation, i.e., HPLC analysis.determination of the cause of this discrepancy is difficult; thisbehavior was also present in the M3OAI propellant evaluations.
Hot water extraction was used to remove NQ. The extraction wasconducted by placing the extraction vial on a hot plate (-95 °C) andloosely replacing the cap. The NQ, which was extracted into the hotwater, tended to recrystallize at the top level of the liquid as seen infigure 80. The liquid containing the NQ was saturated and precipitatedon the pipettes and flasks used for collection. Removal of thispteclpitate was easily accomplished by washing with copious amounts ofwater. The crystalline material was not removed from the vial; rather,it was allowed to redissolve in the next aliquot of water added forextraction. Analysis of these extracts by HPLC revealed the NQ to berecovered at around 88% of the specification value; analysis of the NQextracts for NG and EC revealed none to be present.
The solid material remaining behind should contain NC, cryolite,and graphite. Calculations indicated that from 97 to 111% of thespecification value for these three components was recovered. Furtherseparation of these components by solvent extraction is not possible.
4.5.2 M3OAI Propellant
I This propellant, as well as the M31AI discussed in the nextsection, required an additional extraction step to remove ambient watersolubles. The methylene chloride extraction removed NG at around 98% ofthe specification value and EC at about 165% of the specification value.Again, as with the M30 propellant, the reason for this unreasonably large
I 144
I
mTable 27. Triple-base propellant ingredient recovery
3 NC Recovery
Wt % Wt%Propellant (calculated) (specification) % recovery
M30 (NC+Cryolite+ (NC+Cryolite+ (NC+Cryolite+graphite) graphite) graphite)
31.7 28.50 111.228.2 28.50 98.9
3 27.8 28.50 97.6
M3OAI (NC + graphite) (NC + graphite) (NC + graphite)
m 29.9 28.15 106.229.3 28.15 104.137.7 28.15 133.9
I M31AI (NC + graphite) (NC + graphite) (NC + graphite)
18.8 20.15 93.16.0 20.15 79.419.0 20.15 94.3
I NQ Recovery
M30 42.59 47.70 89.341.60 47.70 87.242.79 47.70 89.7
M3OAl 39,09 47.00 83.2(1.96)* 4.2
38.59 47.00 82.1(1.98) 4.2
31.97 47.00 68.0* (1.83) 3.9
M31AI 39.90 54.00 73.9m (2.09) 3.9
37.96 54.00 70.33 (2.04) 2.0
42.47 54.00 78.6(2.29) 4.2
3 *The number In parentheses represents the amount in the ambient water wash.
I 145
I
ITable 27. (cont)
I NG Recovery
Wt % Wt%Propellant (calculated) (specification) % recovery
M30 22.35 22.50 99.321.83 22.50 97.020.92 22.50 93.0
M3OA] 22.36 22.50 99.422.29 22.50 99.121.80 22.50 96.9
M31AI 19.93 19.00 104.9
20.04 19.00 105.518.46 19.00 97.2
3 EC Recovery
M30 1.90 1.50 126.71.91 1.50 127.31.92 1.50 128.0
M30A1 2.58 1.50 172.02.47 1.50 164.72.34 1.50 156.0
I I2-NDPA Recovery
M31AI 1.37 1.50 91.31.42 1.50 94.71.64 1.50 109.3
I DBP Recovery
M3lAl 4.40 4.50 97.84.39 4.50 97.64.01 4.50 89.1
I _.2 S4 Recovery
M3OAI 0.78 1.00 77.50.89 1.00 88.60.94 1.00 93.5
M31AI 0.88 1.00 88.4S0.89 1.00 88.8
0.82 1.00 82.5
I~146
I
IK
is-.s- '' -
4- -
0
IL-
I4
Ivalue is unknown. There is no other propellant ingredient in either ofthese propellants that would co-elute with the EC.
The second extraction step was an ambient water wash. By atomicabsorption spectrometry analysis, between 75 to 79% of the K2SO4 wasremoved. Analysis of this water wash by HPLC did not indicate thepresence of any NG or EC but did indicate that roughly 4% of the NQpresent in the sample was extracted.
I The hot water extract, which showed formation of crystals,removed NQ around 68 to 83% of the specification value. The lowest levelwas due to incomplete extraction, as evidenced by the NQ crystals whichwere visible in the solid left behind. The NQ extracts were analyzed byHPLC and no NG or EC was present, nor was any K2SO4 indicated by atomicabsorption spectrometry.
The solid material left after all extractions contains NC andgraphite, as well as any component not completely removed. Taking the NCand graphite into account, recovery was around the 105% level. Onesample indicated a 134% recovery; however, residual NQ (as evidenced bythe lowest, i.e., 70.3%, NQ recovery shown in table 27) was present,
i thereby distorting the actual recoveries of NC and graphite.
4.5.3 M31A1 Propellant
The methylene chloride extract removed NG, 2-NDPA, and DBP. TheNG recovery was from 97 to 105% of the specification value, while the2-NDPA ranged from 91-109% and the DBP ranged from 89-98% recovery. Theambient water extraction removed about 88% of the K2SO4 , and, as in theM3OAI propellant evaluations, about 4% of the specification value forNQ. No NG, 2-NDPA, or DBP was detected in these extractions whenanalyzed by HPLC.
The hot water extract removed the NQ from 70 to 79% of thespecification value. HPLC analysis of this extract showed no NG, 2-NDPA,or DBP; atomic absorption spectroscopy analysis did not detect K2 SO4.The NC/grapnite mixture that remained after all extractions representedfrom 79 to 947. of the specified value.
Examination of all triple-base results indicate that recovery ofNQ is very feasible and that the NC/graphite mixture could be reused, butonly in a triple-base propellant. However, the NQ particle size andgraphite may affect ballistics.
5.0 Design Criteria Information
Preliminary design criteria were established for pilot plantpropellant resolvation studies and bench-scale solvent extraction ofselected ingredients. The common requirement for both technologies isparticle size reduction of the propellant with minimal ingredient loss
I 148
I
Iduring size reduction operations. After size reduction is accomplished,preparation of the propellant to be processed via either technologyconsists of reducing the moisture and/or ingredient addition orseparation for resolvation or solvent extraction, respectively. Based on
the results of the earlier testing, preliminary operating plans have beenformulated for both the propellant resolvation and ingredient extractiontechnologies.
* 5.1 Particle Size Reduction
* Particle size reduction consists of grinding the propellant in a safemanner for reuse. Previous studies at RAAP surveyed various sizereduction equipment to obtain a uniform distribution of particles fromthe grinding operations. 6 The review of size reduction equipment such ashammer mills, dicing machines, attrition mills, and knife grinders(granulators) showed knife grinders to be superior for reducing thevarious propellant sizes manufactured at RAAP. The Mitts and Merrillgrinder (fig. 45) has been very successful in size reduction ofpropellants for feed to the incinerators6, in previous resolvationstudies conducted at RAAP 7 , and for the bench-scale study of thisproject. Based on RAAP production experience and the abbreviatedgrinding time requirements for incinerator feeds, the optimal screen sizefor the grinder has been established to be 3/16-in. diameter holes.Fouling and blinding of screens less than this size occurs to due to thegrinding time requirements; screens of larger size result in largeparticles of propellant being discharged in the slurry, hinderingsubsequent pumping operations.
The optimal propellant particle size requirement was established forMl single-base propellant during the bench-scale resolvation study. Thisrequirement is that propellant particles passing a 12-mesh screen willresolvate in the 2-1/2 gal. Baker-Perkins mixer (sigma bladeconfiguration). Particles of M7 propellant greater than those retainedon a 12-mesh screen have successfully been resolvated in aproduction-size sigma blade mixer. 7 Therefore, additional studies arerequired on the various propellants (single-, double-, and triple-base)to determine optimum particle size requirements. If propellant particlesthat pass a 12-mesh screen are required for the resolvation of thevarious propellants, either the grinding time can be increased to obtainthe smaller particles and reduce fouling of the grinder screen orscreening operations following grinding can be optimized to obtaincorrect propellant particle sizes.
The optimal propellant particle size requirement was not establishedfor solvent extraction of selected propellant ingredients. The solventextraction studies utilized ground propellants that passed a 20-mesh
I screen and were retained on a 50-mesh screen.
The major drawback of the Mitts and Merrill grinder is the safety3 requirement of using water as the coolant for grinding 8 and the resultantexcess surface moisture on the ground propellant. However, previouslaboratory studies of various coolant media for grinding (100% water, 50%
149
I
water/50% ethanol, and 100% ethanol) at various temperatures (hot andcold) showed minimal ingredient losses with water as the coolant.1
Ethanol was evaluated as a coolant because it is a common solvent forprocessing both single- and multi-base propellants and will notplasticize the propellants as the ether/ethanol and acetone systems;plasticization would preclude pumping of the slurry to the dewateringoperations. These studies showed that 100% water as a coolant mediaresulted in losses, i.e., leaching, of salts and NQ. In 50% water/50%ethanol, greater salt losses along with stabilizer and NG lossesoccurred; however, the amount of NQ loss decreased. The least NQ lossoccurred in 100% ethanol; however, the greatest losses of salts,stabilizers, and NG resulted.
Therefore, water was selected as the grinding coolant media due tominimal leaching of propellant ingredients, e.g., salts and NQ. In orderfor the propellant to meet final specifications, these ingredients can beadded directly to a production mix whereas NG needs to be added as premixfor safety reasons. The stabilizers can be added to a mix by solvatingthem in the ether or acetone to be added to the mix for resolvation.Water was also selected as the coolant media based on previous hazardsanalysis studies 6 ,8 and the hazards analysis performed for this project(appendix B). With water as the selected coolant media, excess surfacemoisture accumulates on the propellant particles due to the groundpropellant being discharged from the grinder as a slurry. Excessivesurface moisture interferes with subsequent processing; therefore,surface moisture must be reduced to <3% via dewatering and drying.
In previous resolvation studies conducted at RAAP7 , dewatering waseffected by collecting the slurry in muslin bags to permit excess waterdrainage from the ground propellant. These bags were then stored in thewater-wet condition in plastic drums until the propellant could be driedfor resolvation studies. During storage, significant ingredient lossesoccurred as a result of being stored in the water-wet condition. Severalevaluations were conducted to dry the propellant, i.e., remove thesurface moisture; the most acceptable method was drying the groundpropellant in a FAD for 96 h at 140 0F. After cooling, the propellant wasmanually transferred from the bags into grounded conductive plastic-linedtubs. Ethanol was then added to the tubs to reduce dust duringpropellant transfer from the tubs into the mixer. Additional ingredientloss, e.g., approximately 3% NG, was incurred since the propellant wasstored throughout the testing period in the ethanol-wet condition.
IDuring the bench-scale evaluations of the current study, the groundpropellant was dewatered on a Sweco ® Vibro-Energy separator in order toreduce the time required for drying in the FAD. This dewatering approachreduced the total drying time to 24 h. Drying ground propellant in a FADto reduce surface moisture is advantageous in that flake propellants (MlO
and M9) are currently dried in FADs, remote dumping of trays is anestablished operation, and no solvent vapors are present For vaporignition. However, use of the FADs to dry ground propellant has twoinherent disadvantages. The operation is labor intensive, resulting inincreased costs. Furthermore, personnel would be exposed to the dust
I150I
generated from transferring the dried propellant from the remotely filledi drums into the mixer.
Even though potential static discharge from propellant dust has beenshown to be within acceptable safety limits, 9 minimal dust generation isdesirable to alleviate personnel exposure. As demonstrated in previousstudies, 7 transfer of ethanol-wet propellant reduces personnel exposureto dust during handling, i.e., transfer from grounded conductiveplastic-lined tubs into the mixer. The use of ethanol-wet propellantcould be considered for propellant resolvation in pilot-scale studies;however, ethanol should be added immediately before the ground propellantis added to the mixer and not during storage. On the other hand,ethanol-wet propellant is undesirable for solvent extraction ofpropellant ingredients since the initial extraction step utilizesmethylene chloride; the interaction of the two solvents would adverselyaffect solubility separation of the selected ingredients.
1 5.2 Propellant Resolvation
Pertinent operational aspects, including safety and qualityconsiderations, were addressed in the propellant resolvation studies.The results of the bench-scale study showed that in order to obtain aprocessible single-base propellant mix, a 60-min cycle to permit solventsorption is required prior to a 15-min mix cycle. Production-establishedsolvent systems can be utilized with solvent/propellant ratios as low as
0.75/I for single-base propellants; for multi-base propellants, theinitial solvent/propellant ratios can be those optimized in thelaboratory-scale study since a production line mix of ground1 dried M7propellant was successfully processed in a previous study) using asolvent/propellant ratio of 0.4/1.
Two significant points of departure from standard production mixesshould to be considered. One is the method of contacting the groundpropellant with the solvent mixture, i.e., the solvent mixture should beadded to the mixer followed by an even distribution of the groundpropellant. If the propellant is added to the mixer first followed bythe solvents, a gelatinous layer forms on the propellant limiting solventsorption throughout the mix. The second point of departure is that thebench-scale mixes were made with a sigma blade mixer rather than theBeken mixer which is normally used in single-base propellant production.
Preliminary safety design criteria for follow-on pilot plantoperations were developed in the bench-scale studies; the completehazards analysis report is presented in appendix B. Figure 81 depicts aflow diagram of the pilot-scale propellant resolvation process.Pertinent quality assurance considerations, e.g., ingredient addition atthe mixer, were also addressed during the bench-scale studies in order toensure the production of specification-grade propellant. The followingdesign criteria information for a pilot-scale propellant resolvationprocess are based on the parameter constraints established from theresults of the laboratory and bench-scale resolvation studies:
I151
I
a"DTE'ap LLRHTi1 is RUS
EMERCEN Cy CONSERVATION(REL IEF VENT VENT
ZVS CALMALYaleS
RAATER
WATERYU (ACITATED SOLVENT $MTN)
ALCOHOL-WETPROPELLANT
D
ETOCTES
CHEMAICAL U
:OLVNT ACOHO
FIRI TRAGE RCOVERYCAUSTI
Figure~~~ ~~ 81 lwdarmo iltsaepoelnreslvaio procSes
EMERCENCY CONdSMEATIO4
AE10O"VATION .....LLTNTCACITATEC) ~X RARSTADARDCSE
JAAGITATED)
EROPE L LAN
MATME
CHLCMICALEAER O..
ANALYSISINR
CAB
RECOV ERY SISYSTEI
MOTE::
,. CAUSTIC USED FOR NITROCLYCERIS OEITS'JCYIOA.
Z. :EMODTE CSNKI.C. SECOND0 DEMATERI14C.MAND RESOLVATIONW OPERATIONS
QSUIPPED RITAFN S SIOH YTTV
)f pilot-scale propellant-ocess
5h2 .
II. Single-base propellant particles passing an 8-mesh screen were
determined to be the optimum size for resolvation in thebench-scale studies; however, additional testing should beconducted to determine if this particle size is the optimal3 particle size requirement for multi-base propellants.
2. The production-established solvent/solvent ratios adequatelyresolvated both single- and multi-base propellants.
3. A solvent/propellant ratio ranging between 0.70/1 to 0.75/1proved optimal for single-base propellants; thesolvent/propellant ratios established in the laboratory-scaleresolvation studies should be used as a point of departure todefine the optimal ratios for multi-base p.opellants.
4. A 1-h sorption cycle is required for single-base propellants;the length of the sorption cycle for multi-base propellants mustbe established.
5. A 15-min mixing cycle is required for single-base propellants;the length of the mixing cycle for multi-base propellants must
* be established.
6. The necessity and required length of the drying cycle times(required to remove the excess solvent from over-solvatedpropellant following mixing) must be established for theproduction of specification-grade products.
7. Following resolvation, standard production operationsestablished for the formulation-specific propellant should beutilized to produce a finished product.
I 8. The existing hazards analyses (appendixes A and B) must beupgraded for pilot-scale evaluations. All equipment andoperating procedures must also be reviewed by the RAAP HazardsAnalysis Department for each propellant to be evaluated.Safety-related considerations requiring additional evaluationinclude remote materials handling, equipment and facilityI clean-up, and containment of potential spillage.
9. Resolvated propellant should be utilized in selected propellantformulations; oroper chemical, physical, and ballistic testingshould be conducted on propellants manufactured to assurespecification compliance.
I5.3 Solvent Extraction of Selected IngredientsPreliminary bench-scale criteria information for solvent extraction
of selected ingredients from ground propellant was based on the resultsof the laboratory investigations. The following design criteria3 Information for a bench-scale solvent extraction process of prooellant
153
I
Iingredients are based on the parameter constraints established from theresults of the laboratory-scale studies:
I. Particles passing a 20-mesh screen and retained on a 50-meshscreen were utilized in the laboratory studies. Additionaltesting should be conducted to determine if this particle sizeis optimal.
2. The solvents, methylene chloride and water, used independentlyas described in the laboratory studies, can selectively extractthe ingredients using 10 to I (weight-to-weight) extraction
l ratios (solvent/propellant) for each solvent.
3. The solvent/propellant ratio used for the extractions should be10 to 1 (weight-to-weight).
4. Three 24-h extractions followed by combining the extracts arenecessary for optimal ingredient recovery.
5. Orbital shakers were used in the laboratory studies; additionaltesting is required to determine bench-scale equipment.
I 6. The main ingredient for recovery is NC from both single- andmulti-base propellants followed by NQ recovery from triple-basepropellants. Quality parameters of the extracted ingredientsmust be established due to contamination constraints:
a. The 13.25%N NC from single-base propellant is contaminatedwith graphite in certain cases, e.g., MIO single-basepropellant.
b. The 12.6%N NC from double-base propellant is contaminatedwith insolubles and NG.
c. The 12.6%N NC from triple-base propellant is contaminatedwith insolubles, NG, and NQ; furthermore, the NQ crystalscould be contaminated with other triple-base ingredients.
7. Extracted ingredients should be utilized in selected propellantformulations [i.e., NC (13.25%N) is only used in single-basepropellantsj and proper chemical, physical, and ballistic
testing should be conducted on propellants manufactured with the*reused ingredients to assure specification compliance.
8. Solvent (methylene chloride) use consideralions such asconducting extractions below the lower explosive level or abovethe upper explosive level, equipment compatibility, personnelexposure, solvent reuse or disposal, and solvent handling shouldbe assessed by the RAAP Hazards Analysis Department.
II 154
I
I6.0 CONCLUSIONSI
1. The percent nitrogen (%N) and viscosity of the nitrocellulose(NC) detprmine the solvent/solvent and solvent/propellant ratios requiredto properly resolvate propellant.
2. Testing of M1O propellant, which has the greatest NC content,indicate that deviation from production-established solvent/solventratios was not necessary.
3. Single-base propellants resolvate more readily with increasedsolvent/propellant ratios (i.e., greater than production-establishedratios).
4. Most multi-base propellants resolvate using production-Pstablished or slightly increased solvent/propellant ratios.
5. Smaller particle sizes, which reduce the case-hardened area inthe propellant, allow greater solvent penetration to soften the NCmatrix, thus reducing the solvent/propellant ratios required forresolvation.
6. Obsolete or out-of-specification propellant can be successfullyresolvated.
7. The results of the statistical study verified that thehigh-performance liquid chromatographic (HPLC) methods developed forthese evaluations were equivalent, with regard to precision and accuracy,to the analytical methods delineated in MIL-STD-286B.
8. NC recovery from single-base propellants ranged from 96 to 100%;the recovered NC (13.15%N) could be used in any propellant requiring NChaving this nitrogen content.
9. The maximum recovery of dinitrotoluene (DNT) in single-basepropellants was -35%; however, complete recovery of the remainingingredients represents only -5% of the formulation-specificationingredients.
10. NC recovery from double-base propellants averaged -100%,however, the recovered NC (12.6%N) should only be used in the productionof multi-base propellant since the NC has previously been contaminatedwith nitroglycerin (NG) and insolubles, e.g., carbon black and graphite.
11. NG recovery of -80% is attainable in double-base propellants;however, only 2 to 10% of the remaining formulation-specificationingredients can be recovered.
12. Separate recoveries of NC and nitroguanidine (NQ) fromtriple-basc propellants averaged -88 and 82%, respectively; however, therecovered NC (12.6%N) should only be used in the production oftriple-base propeliant since the NC has previously been contaminated withNG and insolubles. Furthermore, the recovered NC should not be used in
155
I
Ithe production of double-base propellant due to possible contamination byI NQ.
13. Virtually 100% recovery of NG is attainable in triple-basepropellants; however, only 2 to 7% of the remaining formulation-specification ingredients can be recovered.
14. Safety in handling represents the major concern in the reuse ofNG recovered from the multi-base propellants.
7.0 RECOMMENDATIONS
1. Pilot-scale resolvation studies should be conducted for single-,double-, and triple-base propellant.
2. Additional grinding/screening studies should be performed toobtain propellant particles passing a 12-mesh screen to ensure adequateresolvation.
3. Alternate methods of drying of the ground propellant (other thana forced air dry (FAD) facility) for resolvation should be investigated.
4. Bench-scale solvent extraction studies to optimize theextraction of single-, double-, and triple-base propellant ingredientsshould be performed.
5. The HPLC methods developed under this project should be utilizedto determine both ingredient addition in pilot-scale propellantresolvation stuoies and ingredient recovery in bench-scale solvent
i extraction studies.
6. Final users' specification requirements should be delineated topermit the use of obsolete or out-of-specification resolvated propellant3 in current military weapon systems.
INIII
I 156
I
8.0 REFERENCES
1 F. W. Nester and L. L. Smith, Propellant Reuse Technology Assessment,PE-796, Contractor Report No. AMXTH-TE-CR-86076, U.S. Army Toxic andHazardous Materials Agency, Aberdeen Proving Ground 21010-5401, 1986.
2 F. rD Miles, Cellulose Nitrate, Interscience Publishers Inc., NY,1955.
3 F. S. Baker, et al, "Dielectric Studies of NitrocelluloseNitroglycerin Mixtures," Royal Ordnance Factories, ExplosivesDivision, Waltham Abbey, Essex, UK, May 1983.
4 E. C. Worden, Nitrocellulose Industry, Vol. II, D. Van NostrandCompany, NY, 1911, p 906.
5 Test Plan for Arthur D. Little, Inc., Task Order Number 7, Reuse andRecovery Technology for Energetic Material, Contract No.DAAKll-85-D-0008, Hercules Incorporated, Radford Army AmmunitionPlant, Radford, VA, Sept. 5, 1 86.
6 D. E. Rolison and R. L. Dickenson, The Production Engineering of anAutomated Incinerator for the Disposal of Propellant and ExplosiveWaste and Evaluation of a Prototype Waste Propellant Incinerator,PE-209 and PE-263, Contractor Report No. RAD 100.10, HerculesIncorporated, Radford Army Ammunition Plant, Radford, VA, 1975.
7 J. H. Agosti, Process Design for Disposal of Scrap Propellant,PE-425, Contractor Report No. RAD 100.10, Hercules Incorporated,Radford Army Ammunition Plant, Radford, VA, 1976.
8 T. W. Ewing, Preliminary Hazards Analysis of the Use of the Mitts andMerrill Hog to Grind Alcohol-Wet Benite Propellant, PE-425, HerculesIncorporated, Radford Army Ammunition Plant, Radford, VA, 1973.
9 T. W. Ewing, Electrostatic Hazards Evaluation for the Handling ofFinely Divided Ml Propellant, Report No. HA-75-R-4, HerculesIncorporated, Radford Army Ammunition Plant, Radford, VA, 1975.
II
I 157
I
I~t
* APPENDIX A
Total Systems Hazards Analysis on Propellant Reuse-Recovery Technologjy
I
II
I
3 158
II
F
.S] Uy. Is Pr of rew job.
I Attachment MemorandumRADF0O ARMY AMMUMITION PtANT
April 1, 1987
II
I
HI-87-M-44
Total Systems Hazards Aralysis or,Propellant Reuse - Recovery Technology
Report No. 1
3 DIGEST
Objective
3 The objective of this study is to conduct a Total Systems Hazards Analysis(TSHA) in propellant reuse-recovery technology from laboratory-scaleinvestigations to the development of design criteria for pilot plant designand testing. This specific report identifies and evaluates the potentialhazards to personnel and facility during (1) laboratory-scale solubilitydeterminations of selected essential materials in different solvents and (2)extraction of selected e:sential materials from single- and multi-basedpropel lants.
SI-rary and Conclusions
This interim safety review documents the Preliminary Hazards Analysis (PHA) oflaboratory propellant resolvation studies for extraction of propellants and
j essential materials using selective solvents.
7-10
U A-173 NIV 1.14
HI-87-M-44 2 April 1, IQ3?
The planned propellant resolvation/extraction studies can be safely performedat Radford because personnel are well versed in both safety and operatingprocedures and small quantities (100 grams) of energetic materials areinvolved. Table 1 lists typical operating procedures and sppcific nafetyprecautions for controlling potential hazards to personnel. These controlsare applied to each step in the laboratory studies. The personnel andoperational safety controls listed in Table 1 evolved from years of laboratoryand propellant experience.
This hazard study is being done to fulfill partial requirements of a contractbetween Hercules Aerospace Company and Arthur D. Little, Inc. (ADL) who is theprime contractor for U. S. Army Toxic and Hazardous Materials Agency(USATHAMA). The total contract is identified as TASK ORDER NUMBER 7.
Recommendations
No recommendations are given for the laboratory phase of the reuse/reclamationprogram at this time.
Future Work
This is the initial report on the reuse and recovery technology. Theremaining hazards analysis studies are outlined in Table 2.
INTRODUCTION
The military has stocks of chemically acceptable propellants which areobsolete for one reason or another. For example, the gun system may havechanged and a safe and efficient method is needed to reprocess the propellantinto a new configuration that can be used for a different weapon.
Also, propellants that are nonconforming to chemical and ballisticspecifications are available for reclaiming/reprocessing into chemically andballistically acceptfble products.
Past practice has been to burn or detonate unneeded or unacceptablepropel-ant. This technique requires a waiver to destroy any quantity ofpropellant and unnecessarily destroys reuseable material.
This hazards analysis study consists of multiple phase safety assessments.This initial effort is devoted to performing the initial PHA of laboratory-scale studies that will: (1) determine solubility of selected propellants andessential materials, (2) determine extraction capability of selected essentialmaterials from propellants, and (3) determine resolvating and extrusionparameters. This safety review is being done as concepts are presented; nospecific equipment or process design are yet available.
7-11
160
IthHI-87-M-44 3 April 1, 1987
D DISCUSSIONJ
Methodol oiv
IThe TSHA of the Propellant Reuse/Recovery Technology is being conducted usingthe Hercules Evaluation and Risk Control (HERCT) technique. The HERCtechnique is a practical method of hazards isolation, evaluation, elimination,and control; it is a formal engineering approach to quantitatively evaluateprocessing hazards. The basic objectives of this safety analysis program are:(1) identify hazards, (2) eliminate or control the hazard, and (3) providesystem design and operating criteria. The procedures for performing thequantitative risk analysis is described in references 1 and 2 and the HERCtechnique fulfills the requirements for system safety specified in MPBMA OSM3385-1.Inherent Safety Features
Radford personnel have extensive experience in the handling and testing of
potentially hazardous explosives and solvents. Over the years, operatingprocedures have been developed to safely sample, store, process propellantsand explosives for testing and dispose of small quantities of materials usedto manufacture these products. The sample, preparation, and analyticallaboratories used for these tests have specific operating procedures andsafety features that preclude discharge to the atmosphere. Numerous examplesof operating procedure controls and/or safety features are listed in Table 1.These safety controls and precautions are self-explanatory.
*Laboratory Extrusion Studies
Procedures, equipment, and instrumentation to measure extrusion pressure andrate does not exist within the laboratory. It is essential that measurabledata be obtained during the resolvation study for comparison with full-scalepress results. The simple device shown in Figure I was used for an initialeffort at measuring extrusion pressure and extrusion flow.
A sample of resolvated propellant was placed in the plastic extruder(syringe). The following parameters were known: (1) volume, (2) density,5(3) weight, (4) total volatiles, and (5) diameter or area of nozzle.
The pressure applied to the air cylinder was measured. The time to extrudethe strand and the strand length were measured. The overall pressure on thesample was calculated. This simple device allowed measurable initial attemptsfor quantification of extrusion results.
This type test can be safely done within the laboratory because only smallvolumes (-20 mL) of propellants are used and the plastic walls of theextrusion vessel will rupture prior to sufficient pressure being applied forinitiation. To further improve safety, this operation is performed under ahocd to reduce solvent exposure. A shield is also used for personnelprotection.
37-121161
I-T-M-44 4April 1, 1987
5 AP~& -4ND DISCLAIMER
Within the scope of work, Hercules warrants that it has exercised its bestefforts in performing the hazards analysis reported herein, but specificallydisclaims any warranty, expressed or implied, that hazards or accidents willbe com.pletely eliminated or that any particular standard or criterion ofhazard or accident elimination has been achieved.
MAH:bpw
IAttachments
I
l
162
HI-87-M-44 5 April 1, 1987
REFERENCES
!HERC' Engineering Analysis Manual, Edition 1, Hercules Incorporated, AlleganyBallistics Laboratory, HERC No. 73-116, December 1973, Hercules Proprietary.
2HERCS Risk Analysis Manual, Edition 1, Hercules Incorporated, HERC No. 75-79,October 1975, Hercules Proprietary.
7-14163
I
0
0 - 4 -
1 ~~~L
Q H
I -~ a- (L
0,
7 --1a16
0 cc0K -
0- 0 -5 ac 110
CM
00 4; L. 0 0
C~ L .
1.0 L1 . C0 w u.t 0.0.= -
o1 0
* C.CP 0
0-1 0; 0, X w 0)
m IC CL S- A-. - C L ' 4C C3- , to J 4' C 01 1"
U. -. 01 'V 0 71
*~ C C i 4'CL1 64 .0 L- '*0- 2 Z5 Luq *'A 45..6."4 0a0 L.. L. 2~ aV z z
. 4.0. VI C- 4 11.i L A -3 L. .0 - l
Ll CCL 'A0V -O .0 .0'41 : . )0 0 4.-C~~~~S &4 C L.4 ;Q LIE t
0'rT 0- 1- OL 4Q. 1 al L. 0 .)L
00 4C 0J Cc .
o C L C
0 6V
C t
W, 0
41 C 1.-aL.
Q. C'
3 7-16
165
-~~C 0.n- -
- K -K -K -C
' -e
00 m 0 - -
a IllZ CC 4z40 7z V cu
40 * ~W 0.WV C 4 c . 1-c
IV le0 0 6
aD = 0.L W - 0 c1 . 1w L4- -, 06 L. Uu 4 u
c4 -A .4 .49 0' Ls (4 4j 4j 0M Li..4 4445U a ~ .0 0 a L.V C'V w 0 ~
C . *WL G 10.C =4 @ L. D Z C- 4
%. CLD S. O L -L 6. .4-E V .V L45 '
40 C
o 01 05m ZI x1 t U 'A. 0 k^ .0
It ~ C0. C E 0
-L -~CL
0:' 1.06 S.~ ~u
GD US7-17
7-67
Q6C
cc- -t
* CO a). CC)
-tr -w
4o. Cr t.) - a C
w tC 0.CL~ . w a 0 ~C
rm 4.0 - E. LEL 1
cL L.- 0 6 'V. - V C L 1. _ C - .>. .- L. a6 cCC *- . 0 0 43 cu c, '
I- -~ cL I.. J. L. a CL L. Go a6C C '
= C - i t
C 6 ~ 0. ~ 6. 'AQ 9 L. m9 ~'V- >. 'A- 41 In z~ J .9 ~ 4 0
c . A - 0 /C 3O1a -- /
fo
a.C cr
a >,,4 C C
wl m u ,CL. - L -
7-18 C
-~ I- 4.167
CC
a C~ 0
-,C 0' i!AC
4j c
a. a'V %- "zV ' tr
S. M. - .
ml~ a Q = 0 0. v -a m V
w M ~ A -. -. cm %. 0 (
o -i.- 'A. .x a
I. W .
4) o .5C
.5 06 CL. E L. a. %
C- ' C ~-
~~~~*~~ a~' D L .~~~
.10C. I.
*~7-19
168CC
0~~0 0 C0 CC CC- 7 LN N -NN3 - - - - - C -. r- ~ -
CD c =CC) oC.. Ll -0
I" c
C. ol 46 7Z
0 W 9d M -U lU -. . m1,~4'~1 w
IL c0 L . - A I-. W ~ C~
3 641 1. 4'CL 0. 06 C
0 10.0. - C - c . .1 --
.2 L. <W CL IV t.- C1 4wJ~ 1- J 6 4 60 4l 0- C 1. Q~ .M
6r4 41 22 ~ .210.2 06 .? CL 4.- ~ 1 ul. 1 q0l 4I ..
c4 "Iu cc=2L - 1 1 4J L . Z- 060
L.U4 L. 041 0 41 L.J. . U0 Ef. U
41-1
C 06
C~ 0. -
~ 44,
* 7-20
3 169
I3 Table 7
Schedule for Hazards Analys~s Events
I Time 108'
3 Event April May June July Aug Sept Oct Nov Dec
1. Completion of PHA onLaboratory Scale Reuse/Reclamation Technology
2. Repair of 2.5 GallonI Mixer (by others)
3. Risk Assessment of 2.5Gallon Mixer for Single-Base Use
4. Procedure Review for 72.5 Gallon Mixer
5. Submit Safety/DesignCriteria for a PilotPlant to Reclaim/ReuseSingle-Base Propellant
I 6. Final Report
III
I _=Start
ACompletionI
3 7-21
170
I
i
IIII
APPENDIX B
Nazards Analysis of Equipment, Procedures, and OperationsPlanned for a Reclamation Process for the Recovery of3Oasolete Cannon Propellant
IIII
IIII3 171
I
Safety Is pW of )eI job.
fi| HE RCUL.E.S Memorand-RADPO2D ARMY AMMUNITION PLANT At
I
Hazards Analysis of Equipment, Procedures, andI Operations Planned for a Reclamation Process for the
Recovery, of Obsolete Cannon Propellants
IDIGEST
i ObJective
The objective of this study is to perform a Total Systems Hazards Analysis(TSHA) of equipment, procedures and operations planned for a reclamationprocess for the recovery of obsolete cannon propellants. This study was
conducted to ensure safety to personnel and facility and to determine
compliance with safety risk requirements of Army safety document MPBMA OSM385-1.
Surmary and Conclusions
This final report documents the TSHA performed of equipment, nrocedures, andoperations planned for reclamation of obsolete cannon propellants. Plannedoperations from storage throtugh finished propellant operations were subjectedto an engineering risk assessment. TLse assessments were extracted fromprevious reports where applicable and updated if required. Appropriateanalyses were performed on equipment and/or procedure changes made in theseparticular operations.
172
RA-173 REV. 8-84
IHI-87-M-125 2 October 28, 1987
Upon implementation of the study recommendations in Table 1, the plannedpropellant grinding (size reduction), resolvating, and extrusion of reused Mlpropellant can be safely performed. The overall risk to personnel andfacility will then conform to Army safety requirement MPBMA OSM 385-1 as shownin Table 2.
Safety design criteria are provided for the design of a pilot plant to reclaimcannon propellants.
Any changes to the system design or operations as planned to date will3 invalidate the findings of this study and require reassessment.
Recommendations
I Recommendations to eliminate and/or control real or potential hazards toacceptable levels are listed in Table 1.
3 Future Work
Table 3 outlines the initial Hazards Analysis effort for the single-basepropellant reclamation project. Procedure reviews and completion of anOperating Hazards Analysis (OHA) of dry-run reclamation operations have beendone.
INTRODUCTION
The military has stocks of acceptable propellants which are obsolete for onereason or another. For example, the gun system may have changed and a safeand efficient method is needed to recover and reprocess the propellant for adifferent weapon. Also, propellants that are nonconforming to ballisticspecifications are available for reclaiming into chemically and ballisticallyacceptable products.
Past practice has been to burn or detonate unneeded or nonconforming3 propellant. This technique unnecessarily destroys reuseable material.
This hazards analysis study consists of multiple-phase safety assessments.The initial phase was a Preliminary Hazards Analysis (PHA) of laboratory scalestudies to determine (1) the solubility of selected propellants,(2) extraction capability, and (3) resolvating and extrusion parameters. Thisreport documents the quantitative assessments of equipment and operationsplanned for reclamation process for recovery of obsolete cannon propellantsfrom initial storage through size reduction, resolvating, extrusion, cutting,drying, and storage.
I DISCUSSION
The following sections discuss the methodology and material response data usedto risk assess the equipment, procedures and operations planned for reclaimingobsolete cannon propellants.
I 173
I
IHI-87-M-125 3 October 28, 1987
Methodology
This safety assessment used the Hercules-developed Hazards Evaluation and RiskControl (HERCT) safety techniquel, 2 to quantitatively assess the riskpotential for injury to personnel or facility damage to equipment, procedures,and operations planned for a reclamation project that will recover obsoletecannon propellants. This study will use existing plant equipment andprocedures to transport, grind (size reduction) and cut the finishedpropellant. The HERY technique permits identification and quantitativeassessment of in-process energies for starting fires or explosions whenapplied to propellant during handling and processing. For this study,in-process energies are compared to Ml propellant material initiation data toestablish safety margins and determine the probability (risk) for fire,explosion, or personnel injury during the recovery of obsolete cannonpropellants. Raquiremen s for operational sifety risk levels are defined inArmy safety document MPBMA OSM 385-1 for four hazard levels in terms of riskto personnel and facility.3 Unacceptable risks are eliminated or controlledto acceptable levels by engineering changes to equipment and/or facilities orby procedural changes as feasible.
The Hazards Analysis techniques used during this study to identify, eliminate,or control hazards are the PHA and engineering risk assessment. An OperatingHazards Analysis (OHA) was performed during the dry-run reclamation processwhen equipment, procedures and operating conditions were safety assessed.Table 3 indicates when each task is done.
Preliminary Hazards Analysis
The PHA qualitatively identifies potential hazards during concept and designI stages when it is most economical to make changes. References 4-7 areexamples of PHA's performed previously on equipment planned for propellant3 reclamation.
Engineering Risk Assessment
The engineering analyses quantitatively assesses the probability of
potentially hazardous events identified and assigns an accident severity levelas shown in Appendix A. This assessment is used to quantify potential hazardsidentified during the PHA and the OHA and assures that an acceptable risk isachieved in accordance with MPBMA OSM 385-1. The process flow in thereclamation program is briefly described in Appendix B and is depicted inFigure 1. The engineering risk assessment is followed by the OHA. The OHA isexplained after an explanation of what frequencies, events, and materialpresent columns in Table 4 represent.
i 'Registered trademark of Hercules Incorporated.
_174
I HI-87-M-125 4 October 28, 1987
|Value assignment and calculations of frequencies, event, and material present3 probabilities in support of the line items of Table 4 are shown in Appendix C.
Operating Hazards Analysis
This safety technique assesses humans as a potential contributor of initiationinto the system analysis; it combines operating procedures and equipment todetermine where human error can occur and attempts to evaluate the
* consequences of such error.
Material Response
5 Background
An important aspect nf any hazards investigation is to define the initiationand explosive characteristics of the propellant physical and chemical statespresent. This testing has been done for the various process mixtures of Mlpropellant as shown in Table 5. As can be seen by reference to Table 5, Ml
m propellant is not unduly sensitive to mechanical impact and friction stimuli.
Fine particles of Ml are sensitive to Electrostatic Discharge (ESO), see Table5, but the reclamation project is designed to minimize the generation of fines.
Ml propellant does not react readily to flame initiation (Critical Height) andshock (Critical Diameter) but a brief explanation of each test is needed.Critical Height to Explosion (Che) tests are performed to define thesusceptibility of a material to transit from a burning to an explosionreaction once initiation has occurred. The Critical Diameter Test (CD) forexplosive propagation determines a material's susceptibility to propagate anexplosive reaction. Again, a double-base propellant is provided forcomparison of the two propellants. The data in Table 5 indicates that the Mlshould not transit from a burning to an explosive reaction in the plannedoperations of this reclamation project. The data indicates that the Ml canpropagate an explosive reaction, if.confined. Confinement necessary for thepropagation was not identified in the equipment to be used.
3 Hazards Analysis
I. Removal from Storage
I The hazards analysis was performed starting with movement of thepropellant from the New River Storage Area, see Figure 1, and is completedupon chemical, physical, and ballistic testing after the Ml propellant ha:been processed through the Finishing Area. Obsolete propellant that willbe used for this study must be removed from storage, transported toRadford, and stored until it can be passed through the size reductionoperation just as if the propellant were to be incinerated.
175
Imu nm m m NIN~TI
I HI-87-M-125 5 October 28, 1987
I2. Grinding and Dewatering
I Radford has a long and safe history of grinding waste propellant forincineration. References 8-10 are Hazards Analysis reports that discusswhy this operation is safe. However, briefly stated, this operation issafe because the propellant is ground in the presence of copious volumesof water and water flow failure automatically deactivates the grinder.Minor fires have occurred during propellant grinding without consequenceor facility damage. Pump casings and impellers are lined withenergy-absorbing rubber to reduce propellant initiation by mechanicalstimuli during movement.
3 The propellant recovery process makes use of the SWECO system to dewaterground propellant. This system can be safely used, see Table 4, but isnot the updated SWECO normally associated with propellant screening. TheSWECO at the Incinerator has a metal tie down in the center of thescreen. The tie down provides unnecessary metal-to-metal friction andimpact points. This deficiency has been controlled by "potting" theentire area around the tie down. Periodic reexaminations are needed toassure that the "potting" has not deteriorated. Also, piece marks shouldbe aligned on the SWECO frame, where they can be easily seen, in order todetect slippage which can lead to excessive friction. The volume of waterpresent in this equipment will be sufficient to inhibit a sustainedpropellant burning reaction.
The SWECO separator equipment design and operations are not ready for use;pumps and parts need replacement. Current procedures and training are notup-to-date. These items must receive attention prior to starting
i propellant grinding and screening.
If a SWECO is to be used in the pilot operation, the hollow ring holdingthe screen should be replaced with a solid ring. The hollow ring couldfill with propellant fines during extended use and present a propellantconfinement hazard.
Capability exists to spill contaminated water and propellant on the scalesand floor around the SWECO. Spillage provides two potential hazards:(1) personnel exposure during cleanup and (2) capability to contaminate
area outside the building since the water can flow outside. Provisionsmust be made, before operations, to contain any spillage within thebuilding.
i. Prooellant Traying and Drying
Normal practice after propellant grinding is incineration. However, inthe reclamation project, the propellant will be triyed or bagged after ithas been dewatered by the SWECO. The ground propellant will have excesssurface moisture ('1% moisture) removed in a Forced Air Dry (FAD) building.
I3176
I
IHI-87-M-125 6 October 28, 1987
Risk assessment of potential friction, impact, electrostatics and thermalenergies associated with both manual and equipment operations, as relatedto planned drying were assessed. Examples of operations/equipment
* subjected to risk analysis were:
1. Operation of the SWECO Separator.
2. Manual traying or bagging of ground propellant.
3. Manual handling and equipment movement in FAD operations such as:
I a. Opening FAD doors--impact and friction.
* b. Temperature controller operation.
c. Buggies passing over spilled propellant.
3 d. ESD.
These operations were assessed to be acceptably safe, see Table 2 for asunmary of the data and refer to Table 4 for complete details on the riskassessment.
4. Resolvating and Remixing
Small samples of ground, dried Ml propellant will be brought to Building3677, C-9 Mix House, for resolvating and mixing in the 2.5-Gallon Mixer.Approximately three pounds of the ground Ml will be placed into the mixerand approximately four pounds of alcohol and ether added for resolvation.This operation has several potential hazards that merit discussion andthese individual events are discussed in the following paragraphs.
5. Baker-Perkins 2.5-Gallon Mixer
This equipment has been subjected to several hazard assessments.4-7 Eachof these safety reviews outlined work required to allow the processing ofmulti-base or high-energy propellants. In Reference 7 are all therecommendations that have been made that will allow the mixer to be usedfor processing multi-base and high-energy propellants. However, thisstudy is directed toward using the mixer for Ml single-base propellant
only. Those recommendations that are applicable to single-base have beenextracted from Reference 7 and are included in Table 1 with a note onstatus.
3 a. Propellant Contamination
The Baker-Perkins 2.5-Gallon Mixer was not designed as a production;ropelldnt mixer. It has open gears for spilled propellant to fallinto, dry, and to be easily ignited by the rotating gears. The mixerwas built prior to Occupational Safety and Health Administration
I 177
I
HI-87-M-125 r October 28, 1987
(OSHA) regulations; an open rotating shaft serves as a point forpotentially igniting spilled propellant and :s a hand hazard. Ifoperated as designed, the system has an unacceptable probability of2.5 x l- 4/h for an incident. The gears and shaft are now protectedfrom propellant contamination by covers. Once these covers are inplace and the operating procedure requires effective housekeeping(current procedure is deficient in this area) the probability for anaccident falls to an acceptable 2.5 x 10-10 /h, see numbers 71 and 72of Table 4.
Another potential thermal initiation hazard exists from propellantgetting into the mixer glands. The glands are not typical mixerglands in use at Radford. Maintenance records do not identify thepacking or the last time the glands were refurbished.11 Calculations
I were made, see Appendix D, that show the ability of the glands, iftightened excessively, to ignite propellant. Therefore, the glandswere inspected and replaced before the reclamation project began and awork order (700139) accomplished this task. When this repair work hasbeen completed there is an acceptable probability of 3 x 10-8/h for anincident, see Table 4 number 62.
b. Volatile Vapor Ignition
The electric motor that drives the 2.5-Gallon Mixer is rated for ClassI, Group D vapors. Both diethyl ether and alcohol are required duringresolvation and .the motor is not rated for ether. This need notdisqualify the motor for use, however. Currently, the motor has amodified purge which does not work effectively, see Figure 2. Changeswere made to the system as follows:
The sheet metal pipe was capped at the cut shown on Figure 2; airfrom the plant system is applied directly against the motor end andforced out at the other end. Measurements and calculations show aminimum of 94 cubic feet per minute of air passing over the motor.The bay volume is in excess of 4,680 cubic feet, see Appendix E.The bay fumes are removed by an air driven eductor at approximately1,560 cubic feet per minute or a complete air change about everyfive minutes.
The flammability limit for the solvent system is 1 2 1.9 to 36% byvolume.
I Calculations show a volume of 0.064 cubic feet per minute, seeAppendix E, of solvents being released. Thus, it can be seen thatthe evaporated solvents released in the bay are considerably belowthe f',ammability limits.
The air purge for the electric motor is not a complete Type X purge asrequired by the National Fire Code. Power is cut off on purge loss.However, no time delay exists to electronically assure the required
I 178
UHI-87-M-125 8 October 28, 1987I
volume of purge air passes past the electric motor. However, theamount of purge air passed is sufficient because procedures requirethe purge to be started during initial equipment checkout andsufficient air volume will have passed the motor before the current is
I applied;
The probability of this system malfunctioning and causing an incident
is an acceptable I x 10l- per operating hour, see Table 4, number 59.
6. Extrusion and Cutting Resolvated Propellant
a. General
By reference to Figure 1, it can be seen that small quantities of MImade in the 2.5-Gallon Mixer will be extruded using the four-inchpress; normal practice is to use the 12-inch press for extrudingsingle-base propellants. Another difference is the absence of thepreblock step. Both actions are being taken because the quantity ofmaterial produced in the 2.5-Gallon Mixer is small, approximately fourpounds of propellant per mix.
Personnel and facility are protected during the extrusion ofpropellant from the press. Operation is remote, blow-out panelsrelieve bay pressure, and fire protection exists. Inert gas is usedto inhibit the potential for adiabatic initiation hazard (during raminsertion and withdrawal). Several abnormal events could lead to anincident with the press and are worthy of discussion.
b. Frictional Heating
It has been demonstrated that grit in propellant (foreign material)can cause ignition at low velocity by friction between steel andDelrin.13 The probability for an incident with the four-inch press isan unacceptable 5 x 10-/h, if foreign material is present, see Table4, number 80. Without the foreign material present, the probabilityfor an incident is an acceptable <1 x 10-10 /h, see Table 4, number79. Thus, it is imperative that the mix be free of foreign material.The Ml is passed through a metal detector prior to grinding and ishand loaded into and out of the mixer. Since the resolvatedpropellant is also hand loaded into the four-inch press, any foreignmaterial should be detected and removed before extrusion.
c. Comoressional Heating
Compressional heating of air bubbles during propellant extrusion hasbeen suspected is a potential initiation source. Using the computermodel for a 12-inch press ' resulted in a three-inch diameter bubbleraising the temperature to about 1490C,14 which is near the Mlinitiation point. However, a three-inch diameter bubble is notexpected since the material will be hand loaded into the four-inchpress.
I179I
iHI-87-M-125 9 October 28, 1987
d. Impact Initiation
The potential for an impact initiation hazard due to an out-of-alignment ram head impacting propellant on the basket top could easilycause initiation. Such an event provided an unacceptable risk of5 x 10-4/h, see Table 4, number 83.
Procedures require observation to assure that alignment does exist andit will be easy to wipe off any excess propellant after loading. Withthese events controlled, an acceptable probability of 5 x 10-9/h3 exists, see Table 4, number 82.
7. Cutting, Solvent Recovery, and Water Dry
3 These are standard operations; the only difference is the smaller quantityof Ml propellant involved. These events have been assessed previously andassessed to be safe.
15
I Safety Design Criteria for Design and Operation of a Pilot Plantfor Obsolete Cannon Propellant Recovery
5 One of the safety objectives of this small-scale study is to provide safetydesign criteria for a pilot plant capable of reclaiming several thousandpounds of propellant per day. Although this specific study was directedtoward single-base propellant, the pilot plant would need to processmulti-base to be economically feasible.
Shown in Figure 3 is a simplified flow sheet for the reclamation ofpropellant.16 Based on this concept, the guidelines in Table 6 wereformulated.
IMAH:bpw
I Attachments
IIIII180
I
HI-87-M-125 10 October 28, 1987
REFERENCE3
*HERC3 Engineering Analysis Manual, Edition 1, Hercules Incorporated,Allegany Ballistics Laboratory, HERC No. 73-116, December 1973, HerculesProprietary.
2HERC5 Risk Analysis Manual, Edition 1, Hercules Incorporated, HERC No.75-79, October 1975, Hercules Proprietary.
3Safety Program for Modernization and Expansion System projects, 12 May 1980.MPBMA OSM 385-1.
I 4j. R. Murray, "Hazards Analysis of the 2.5-Gallon Baker-Perkins Mixer(Building 3677)," HI-75-M-l, March 6, 1975.
5J. R. Murray, "Hazards Analysis of the Baker-Perkins 2.5-Gallon Mixer,"HI-75-M-2, June 12, 1975.
6M. A. Hundley, "Safety Review of the Baker-Perkins 2.5-Gallon MixerFacility," HA-80-M-50, October 3, 1980.
7T. W. Ewing, "Updated Hazards Analysis Study of the 2.5-Gallon Baker-PerkinsMixer," HI-86-M-103, November 6, 1986.
8M. L. Griffith, "Hazards Analysis Evaluation of the Proposed Methods ofIncinerating Liquid Explosive Waste," November 7, 1978, HA-78-27-M.
9H. W. Carter, "Hazards Analysis Study of the Automated Incinerator Facility5 - Report No. l," February 24, 1972, HI-72-M-l.
10M. L. Griffith, "Risk Analysis of Grinding and Dewatering of SomePropellants and TNT Based Explosives," HA-81-R-8, June 12, 1981.
lIReview of Records by Maintenance Personnel, July 1987.
12Fire Protection Guide on Hazardous Materials, Fourth Edition, 1972.
131nvestigation of a Press Fire During Extrusion of M31AI Triple-Base3 Propellant, October 2, 1985.
l4 . A. Insley, Calculation of Bubble Temperature, July 30, 1987.
3 15R. L. Asbury, J. L. Evans, "Production Engineering Project PE-406," July1977.
16M. A. Hundley, "Hazards Analysis Review of Propellant Reuse Technology,"HI-86-M-24, February 28, 1986.
181
- L-
oco
V) S- k.
4.9 Q C6-
C0 FI V w
w E zj, = n
V SC6C..-
aI -Z aiC4
WU %. - 4 .. ,
14 c w~ = L. L m = " o- 0 04- C. :1: - . 4 - -3
'U C6 j. S- A a-
- Z I m - C- - or 4- 4 - -
I ~11 ,44 WZ M= vi. CZ. Q CL' ~ 0.
H oA~~ co~ 0 X. 0 0C- 4.Cii ~2 ~u-....~ a~o.u*. z.-e..d
0t
c0
0 0 C a, a, - . iaj C
N182
U 12
4).0MI
I- CL
cu 4 0- 10 4
-, -i I) v C)V 4 V 4 ) V IV
3 ~ ~ L S.f i . vi' S_( . (A C A C (A %_
Z vi W 0. 4 a-a
mVC, C- ~ 0'
-C C~0 L.
=- 'a L.;) k CC
'a4d wzoa )C 4 A
- 0 v.' NCc C
X0' CT M Q u0 a . C C
CLC
- CC )L -
S_ M0w Q V L. 4'MV 4)L .0(
Qj u L . 4 . -'A ) O cV
c V- 'a 4)0
CL a - ' - ~ C 6 .1-V
CL aU
dL,
'a
183
ICa 013
c U. c c.0
0vi L- ", -
.etn a-
C. C0~
GJCC ZJ~'Cu~~~> to Q~~~~~
- Lc a.La4jL W
CC0 c
CL. CM'a i c
VV C
2C 4u C:m cmv = -c
w -i -1 .M'UL CE. C'. In
t c- 4j CJ MJC
-CE -3 -LJ 'A 0
0 00
184.
* 14
oc I Io9
4J
4J t
a'
C
0 I C. I I
C10v
' ~ ~~ . I . ~
4-J I
4.
4-
IUw U
S- m cc0
4-1 LO
:; Cj C)-
S- X 00 ~ CC -*-i"
185
15
Table 3
Schedule for Hazards Analysis Events for Reuse-Recovery of M1 Propellant
Time 1987
Event April May June July Aug Sept Oct Nov Dec
m 1. Completion of PHA on A Completed April 1, 1987Laboratory Scale Reuse/
m Reclamation Technology
2. Prepare 2.5-Gallon Mixer _
for Single-Base Use (byothers)
3. Risk Assessment of 2.5- //_ A This reportGallon Mixer for Single- completesBase Use this item
4. Procedure Review for2.5-Gallon Mixer
5. Submit Safety/DesignCriteria for a PilotPlant to Reclaim/ReuseSingle-Base Prope 1lant
m 6. Final Report
III
* A = Start
A= CompletionII
~186
I
P.~ 16
C--
-IL
I-- L
j ~-:*O cnL
* ~ ~ . ~ - -- -
en (ne
LLJ W,
2 l to-
*0 I I SI* 0 U.S
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _126_
- ~9 a:S I'A .StJ .
t. I- Q CL
bS U.1 1-SU L. J
C-L
e4 3 (30 '0
c 0
Is c
06 0
isS 1- c z- v c I i
N187
- 17 -
6. - .
cN~ ILr -. 0
ca a. n,La'm 1
-w
U I
IIII In
I Ins. 4f-c
z4 so- w. ms- 41 coL
cL~ u .A
0 c C
1.I iv (% 6 ; U-~ ~~~~J =~J c Li61 -!0,CI-s.I J m o u L . "
.92
a Q5 a 0D a 0 m 0
UU
0*-0
r*t.- - r- 04 -l LO 0U
C Lj 0.
zl a
V9 0
C64. C'
a% 4) V~ 41' j (a . cu cAl - 2:;Ia- eL; 2 0 0.!
NL ON -M
j Lo mmc . U 0 . 0 m 0 . w c1j ~ ~~ 2 .
'A z z _ _w_ _wS_ _ _ _ __coL_ _._
a. m4 kj Q~ Q 'a22 -.-- CI1.T
.0L-t m .lj~iI -. _ _ _ _ _ _ _ _ _ _ _ _
189* -
I19
00c- %
m -- fn
w 0 0 40 0, 0
-~ 4* CJ3 oe - IzjL
9 asc. c
g 0 00 0 0 0 0
-~ ua0 0 040 fa- -- - S~~ 2: ~ ~ A.I Ia a' U
vi tU 1' w.. C2 ) -vi
.- w m 0q 0.
OI Q -m6
0 .UM
I00 C a >. 0 4
rU cu- ) L
4-aaI 2Afn e" Q C
a a a 190
20
__ ___
S.A 4IIC
U24
.. I VV -V V -v
CA1
w X - - -
z InUw jj LL - -
ciJ I - .I"ca- - -
-- 0 eni ~ ~ 0
A A
.. -i . . *
U , r I'l -W cc= CA eW W WUi
Q b en
-- LIM C, ~ - - C5 -N C C
CA
0~ - Ci
"'Ohl
'A A I
SC, .- LA a w 2Q Z )iCL 0 .0 '1 CL c
= (o - 01-0"WU 49gu
0~~ 'm0uO ,L 0 U6-U1 S. 0 'A41L
c ~.J ~ .LA 4j ~ O ~ 4 L4J In6L0 L. t2 W~ 0 MS W ~ 5- =u'~ Z
C; 0I C ~ O ~ U C~V 4-'U cc - U AO S4 ~ -A
ma. ~
191
21
coJ
- a .
o o o o e o CD
C -0. 0U 0 0 0 0
4 - - en- - -
-. Z- - -
4 -a I I I I II I
La w. . L.A
L.. II . -I - ut 7"
. =CCU.)
'A IM.
'II CL AA . IA - V z: 1
'4- c. 2- 1" j Q 0, A Q
L.A -t s- %Jh 0
L.J M L6 6 w w - - -
Z. X. - - A. - 4n
I - -- 4
L~ .L4 L 41'1924
I 22
a a
alll
3 0 .. SZ 0 00 0 ~l
C- VI VV
~~C-4~ i~ ~Sn1
CO mn
cm, u. U
w ao a *
'So - - c
=, =0
00f
S- ri 12 1- 0j(AU M=A A
,-If CL Is aa aj
Is L. -a-~ ~3 a., j M
L-in~-193
- 23 ~
,...
-1° - °- - - ° -
,.,. -, ,, ,., - ,-,
... "..,"
LLW
Cd .4 In
o U
-- S .,--< =
Da 0. ~ aaa
- -,
U,
a. -
cli
"S - -, -
199
an In CYJL~
CL -l G i '
N NSn%..N
~C 22. -u 0CL: V I M.. Q 1 0 . w i m 1c
q- 10'A2 c mc .- : - 'A I I
ci -K. &C -a , ICC 2 .Z
C- '; V -j _- s- Z
Ig I M 0 - ca CU C -J -0 C6
-a CL 0 CL
a,. P-0.a
-00
194
24
Z 00
00 0 0 0
0 d. 0.1
C6
-6 1" -- -
144
cv IS) Co AS - IWAF o- I I- It A~~~ -A M..d WU L6 L. .
cn I o I I- I -
ad
-~4 -~ en - -
Irv. 0
CA 9) q.'a
Z;J 04 7" . Z .CL =0 Z0 j 'A Z.0 (I~C IS - -
U,
- 'o
U. a. ma.
W) = A 0 .. (
QA l= c -6 c
op 'A C. w.' C
3. 4.1 7;- o-- -= ZCc W.I 8'. 2 -.'
195
U -*25
-h 1 -
L6 Lm
41
41 mn4 en
0L- U- U- a l *
hl- In .
w cuwl--
Z 4-w A
*..4I 0.I
V V3 - )
'I~~ VI
aa c wI %- M - a)
cc wl -, kA 41r-
cu .2 E
QW - -( = ; Ia
'Al cI-- I
0I a) L .VI - U U
X4I -0 4. 4 . 4tn~41VI..cc
z - UC * C *.196C
I
i Table 5
Sensitivity Initiation Characteristics for Different Forms of Ml Single-Base Propellant with M26 Double-Base for Comparison
Initiation TestMethod Units Condition Comoosition Temperature Test Value
Exolosive
Impact a ft-lb/in. 2 Steel-Steel M1 fines Ambient 6.7
Ml Extruded Ambient 20
Strand12-25% TV
MI Granule Ambient 24.7After Solvent
Recovery 191 TV
Ml Granules Ambient 8.0Dry
Slidinga psi 9 fps Steel-Steel Ml Dry Flakes Ambient 60,000 08
Friction M Green Mix Ambient 176,900 @ 8
Solvent Wet 30'. TV
Ml Extruded Ambient 70,000 @ 8Strands
Ml After Ambient 77,285 0 8SolventRecovery
191 TV
M Finished Ambient 69,1.55 0 8Dry Granules
Electrostatic a Joules N/A Ml fines Ambient 0.0013Discharge
Ml Cut Ambient 75.0GranulesSolvent Wet20-29% TV
Ml Finished Ambient 75.0Granule Ory
Diethyl Ether Ambient 0.0013
Vapor AirMixture
Ethyl Alcohol Ambient 0.075Vapor AirMixture
aThe Threshold Initiation Level (TIL) is described as the level above which initiation can occur as establishedby 20 failures at the indicated level, with at least one initiation at the next test level.
lII
197
I
Table 5 (cont)
Initiation TestMethod Units Condition Comoosltion Temperature Test Value
Thermal
Autoignitiona OF- N/A Ml N/A 329
Exoosure to Flame and Shock
Flameb InchesCritical HeightHeight to AtExplosion
Diame.er
1-inch Steel pipe Mi Dry Ambient >7diameter Granulespipe
Ml Granules Ambient >35Water Wet-Soakedin H20 for 4 1/2
minutes priorto test
M1 Macerated Ambient >48Propellant 35% TV
Criticalc Inch Steel pipe M1 Dry Ambient <0.25Diameter Granulesfor ApproximatelyExplosive 0.5-0.61 Ambient >2.0Propagation moisture (unconfined)
Ml Paste Ambient l.Lumps
Volatile Materials
Diethyl Etner Vapor: Flarmiable limits by volume 1.9 to 3o percent
Ethyl Alcohol Vapor: Flammable limits ty volume 3.3 to 19 percent
aAutoignition Temperature or temperature where propellant automatically ignites.
bCritical height is defined as the confined material height above which an explosion can occur when subjected
to bottom flame initiation produced by a 12-gram bag igniter (50/50 mixture of 2056 casting powder and Class 6black powder).
cCritical Diameter is defined as the confined material dimension above which an explosi reaction can be
rooaaated when subjected to a shock impulse produced by a Composition C-4 donor (L:D ratio of 3:1 plug 1 inch1-r iasting c~p,.
198
28
Table 6
Preliminary Safety Design Criteriafor the Design of a Pilot Plant to Reuse Propellant
1. The Propellant Reuse Operation should be one continuous operation in onelocation and-extensive manual material handling steps should be eliminatedthereby reducing personnel exposure. For example, propellant should beunpacked by mechanical methods to reduce exposure. Slurry movement orsimilar material handling steps should be employed.
2. Each reuse operation run should be type propellant specific. For example,the line should be cleaned prior to the Ml run and cleaned upon completionof the Ml run.
3. Propellant size reduction should be done using equipment similar to thatused in the Waste Propellant Incinerator operation. The grinder has beendamaged by tramp metal in propellants but sustained burning reactions havenot resulted during grinding due to the large volumes of water present.
4. The only coolant used in the size reduction operation should be water andan excess of water must always be present.
5. Initial dewatering of propellant is to be done in approved dewateringsystems such as a SWECO separator with modifications necessary forpropellant operations.
6. All equipment must be approved type for explosive service.
7. Entire plant layout should be in a sump to facilitate clean-up and torecapture any spills.
199
(ttasr t fo Ka rd rtmishnS Area(nhinamav) "o
Storage at . l i. ver "( 3)( ) ) e )"
(1 -)(3) (W'"ITransporc co Scorafe
Score or ci rerator Cnpack and lepack
I .'vc.Ave CConveyor
SLur:'.
Conveyor co Slurry TAn"
Grinder j CLr ITransporc to
FAD FAD a 1coFAD . ______I____ Tray Sve
(5)
Storae y.
F ry iber Transport (I.) t ransport R4ouse
I{
I SoventRecvery tran- Cccin '..oc5 ChesicaL. Ad4A1ionI T ranspOrt ~
Trns Traysa 16.Lc Ori in and
I1So vn R c v r po ts A sco oL v t n
coraeLco( ) ( (3) (.
I Cc icr
I: SCC otC,: ?(3) PhyeLea di"lOsLO-9
"Ls/cum antd volaciles
(6) . ,eivit - t a-c -a rinc l. - a 200
ITesc znlv once
30
II
Sheet Metal Building - ?3 - Pipe Wall/
i Cut Hereand Capped
1 47 Plant Air
Flow Switch
Electric3 Sheet Metal DrivenMo tor
I
l|Metal Band
IFigure 2. Schematic of Modified Purge System
of the Electric Motor
III
201I
32
APPENDIX A
EXPLANATION OF HAZARD EVALUATION/PROBABILISTIC ANALYSIS TABLES
I Column No. Title Description
1 Operation States the operation, specific task, andwhether normal or abnormal operation is beingassessed.
2 Units of States the appropriate energy units forAnalysis Process Potential (Col. 5) and Material
Response (Col. 6).
3 Materials of The materials of construction associatedConstruction with the potential hazard are specified.
4 Combustible The combustible (Ml propellant, diethylether, and alcohol vapors) that is presentwhere a potential hazard is named.
5 Process The process stimuli or energy that can bePotential (PP) generated by the potential hazard. This is
determined by direct measurement, laboratory3 simulation, or calculation.
6 Material The threshold initiation level (highest testResponse (MR) level at which no initiation is evidenced in
a fixed number of trials, usually 20)established from initiation tests for a givencombustible.
7 Safety Margin Equal to the materlal response (MR) divided(SM) by the process potential (PP) less one
I Sx = -
8 Frequency (f) Frequency is 1 where continuous process isinvolved, or the frequency per hour if anintermittent operation.
9 Probability of Ep is the probability of the hazardous eventEvent (Ep) occurring and is numerically equal to one for
normally occurring events and is establishedfrom the appropriate equipment or humanfailure rate for abnormal events.
3 202
iI
* 33
13APPENDIX A (cont)
Column No. Title Description
10 Probability of C p is the probability of combustible materialMaterial Present being present where and when the potential(Cp) hazard occurs. The sequence of events
necessary for the combustible to be present(for example, whether normally present or asthe result of an accidental condition orprocedural error) is considered in
establishing the probability.
1 11 Probability of IP is determined statistically comparingInitiation (Ip) material response and process potential.
Safety margins and probit plots are used forthis determination.
12 Probability of SP is the probability of transition fromSustained initiation to burning. Where the potentialBurning (Sp) hazard is in the presence of quantities of
combustible the most severe condition istaken; that is, So = 1. Where thecombustible is present in smaller amounts, asthe result of migor spills, Sp is either1 x 10- or lx 10- .
13 Probability of Fp is the product of f x Ep x Cp x Ip x Sp.Fire (FP)
14 Probability of T is the probability of transition from
Transition (Tp) sustained burning to an explosion and iseither one or zero, depending on whether thecritical height to explosion is exceeded(in-process material height >Che) or not (seeTable III).
15 Probability of Xp is the product of Fp x Tp.Explosion (Xp)
16 Hazard Category The potential hazard is classified inaccordance with MPBMA OSM 385-1 to reflecthazard level (see below). The severity ofthe hazard alone, and not its probability of5occurrence, determines hazard category.
II 203
I
34
IAPPENDIX A (cont)
3 Hazard Severity
Hazard Severity Categories/Accident Categories Application:
Hazard severity categories are classified by MIL-STD-882B into fourcategories, based upon the most severe result of personnel error, proceduraldeficiencies, environment, design characteristic, or subsystem or componentfailure or malfunction. When the necessary conditions exist and the necessarysequence of events occur, then a hazard severity category becomes thecorresponding category accident. The probability values given in hazardanalysis indicate the probability of the transition occurring from hazard toaccident. The hazard severity categories are defined as follows:
(1) Category Ia (Catastrophic) - Conditions such that the failure modeoccurrence will cause system loss or large-scale environmental damage.
(2) Category 1a (Catastrophic) - Conditions such that the failure modeoccurrence will cause death or permanent total disability to one or morepersons.
(3) Category IIa (Critical) - Conditions such that the failure modeoccurrence will cause critical system damage or some environmental damage.
(4) Category IIS (Critical) - Conditions such that the failure modem occurrence will cause permanent partial disability to one or more persons.
(5) Category IIIa (Marginal) - Conditions such that the failure modeoccurrence will cause minor system damage or some environmental damage.
(6) Category III (Marginal) - Conditions such that the failure modeoccurrence will cause temporary total disability or lost time injury not3 covered by category I or II .
(7) Category IV (Negligible) - Conditions such that the failure modeoccurrence will not result in injury, occupational illness, or system damage.
The relationships between the accident categories and effects on the3 system are further explained in Exhibit 1.
2III 204
I
I 35
IAPPENDIX A (cont)
3 Hazard Severity
Exhibit 1. Accident Categories and Effects on System.
EFFECTS ON SYSTEM
3 ACCIDENT
CATEGORY EQUIPMENT PERSONNEL
3 Ia SL or LSED ---
Is D or PTD
3 IIa CSD or SED ---
I IS PPD
I I IL MSD or SED ---
IIS --- TTD or lost timeinjury not coveredby category Is orlIB
IV No Damage No Injury
SL = System LossCSD = Critical System DamageMSD = Minor System DamageLSED = Large Scale Environmental DamageSED = Some Environmental DamageD = DeathPTD = Permanent Total DisabilityPPD = Permanent Partial DisabilityTTD = Temporary Total Disability
IIII1 205
I
1 36
I APPENDIX B
PROCESS DESCRIPTION FOR REUSE-RECOVERY OFMl SINGLE-BASE PROPELLANT USING THE
2.5-GALLON MIXER
IA small quantity (:3500 pounds) of MIMP propellant (Lot 413) isstored in the New River Magazine. No further need exists for thispropellant; it will be returned to Radford and reclaimed inaccordance with the operations shown in Figure 1. As can be seen byreference to Figure 1, the reuse-recovery project is labor intensiveand operations will need to be streamlined for a pilot plant toreduce personnel exposure and simultaneously effect processingeconomics.
IIiIiIIiIII3 206
I
37
41C 4) 034 CUC3 L. C)4 43-0 4
en s- cCC3 o. w 3 3. ci m3 M CL V m U3
CL *.w 4- * 1 4 3 I4%A.04u 00 0. 01. 0 00
43~~ -u u-0 Iu0 W. . - . 40
L. I" CL -1- 4 1-4 1-) 1-3 143 1 43 -433 1-3 143 0.. .0 0. *0. 00 43 = .0.30m
0 -1 3 - 1 -- -1
10
S- Cl w0.0 1. w3 0.4mcu m
43 %- 'a 0. 1- r-
j I4 w3 Q3 L- j .- S- w w
o3 S. 43 a . 0 . 4.1- 0 - 0 -4u-4
CL 43 C - - - - 5S. L. A;Qd 4) J9 c
w430 C - 0 0j
0" C, 7 IC = = C3 4 3.. 4.. a t0 u 1- 1. r4 .0 1 04
C- m43
Ii 1' etI S: Ij i: E a a4 mc0 0.
>1~1 00 00aIj 14.
0 00 co 10 ca moz
- CO 434 CL 43 43 a -4
61.
II014
Ul :2a
207
38
s- 411.0~~~4 o04IC ~ -
-o-C . .14I 4- - 4-u
C~~~~l C.c1UI-C0..
mC; en C. 4 C L w
I- -. uC "oC 44 -L. 'U ' 44 413 1o0 W U LC 4
CC~~z u-..1..L C L61 .0 1
4
(10
I-
cL1. V1 L 3m n'6- ' I c z & ,"c~C Is M " 1 c4 A C %'A L m L - oM L
3t .Cj -s.ICj sm05,-> ILC 0' n% .'
m IC. Le z IC 41c Im IIC
-0 C. (m C 1.C - -C -
,A C
LI~ cto L.4l1.. I
= L; ca. C
1.32
2080
39
en -. m m) S. m
31 CL C- - CLnV 0
4) 0.. 0.. I I -1. nIn 4) 4) 4.1 4- ~ 0 4)41 4
L6 W0 a. CL1- r- 6WS410 I-
a-C C- *.0
C C 1 - 0) 0)14 -
4a -i 4)4
CD .. . I-
CD~ Ci C C.> IV
'A 'U CL- 0 'U a'41 a- Q1
1- 1- W- Li 0 0. a. 0. 0.
0.4 ) n
VI - j 0 -.
00
c C0w
Irc )C. cc l
.20
40
CU~ en -1 CL-
4, - v I -aQ.4 L w, 41 CLI
a~ C0 0U 41
I/I 41. a S_ "J CL 41
Q. 0m =C4 - t I- W Q.4.
ci - C ,- . 1 ci o L410 41= . 1 * 4 1 1C .0
Z 13. tu 0 00 CM0. _
IIW.. CL 43CL
m -0 LWc
L i -i U'QQj(
-L. 1.. M. CL
en m3 41
L! U2.
W ze vi 4141414
Im
C L
0., 0% 10 .- C - U
04) C M0. -C u9. 0w 21 'A. 4 0.3 40 C3=~ 00 S. 41. 43 a %_ 43 00 0.i = Q
-C Ed.41 .14 C C C4 4A1 00.CL 01.00. )4 M - 4
C CCc
4)a C= 41. U.
414
I 6 C 03 1
210
41
c. L. Us
-- o
GJ-' uL -
ena. ena cu (n m Q
s- - i. O
4jL L '
L.L
Q 71-jcLr
IJ 0,
w Iz C)'U a.c
Ij -
ca 4) cL o- I- aC z Q
M QwL. c 0 L 'U s-x
O-LL o Q rL 0G
I- r
-s 13C $ - = Lmfl0 =C M, L6IA Q
M' UCC
ft -Z I M
3c L 4 4
Z~J 211C
42
Ij 1 41 1 In en 0 a f C
m' rLu lo 0.0 Q... -C C o .0 '0. . .
*6 CL CL.
1.. a 00 a040 L O1. . 0
L. d)O 0. .il 0~ 0 0 0 0c0 o 0
4'tm.. 0.i 00 00 W.4 0.0 1.' 1.'
0 cu0 0~ al 0 d L L. U. . j 5-0 10
CL .C 0 0' w
0
0 .0 0 L 4
UMO .-1j 0. 0 ' 4
0 0 C 0 £ LIM4 a $ "C u C.
ei .4. M.41-s-.m
0 0- C0 LO c4 .0
cm - C0.- .CL
0 --
4'~~S c *~00
1- 1.. 1-C 1-C .. 'A .C1C -
W, ci a ir CL S- u ~ S- CL74- 1 s
41 ej a- Q- =. Q- 1- v 1 u = w
.I (A '7'4 '.0 '0 a A' 0 4I
V; A *1- 4A 0. 1- CC970 L.i0 C40 40 40
0.00 0.00 0.00 0.00 0. 0
.21
43
41 4 4 413.A3
c3 43 01 01c- c CU 2
40 13 C. 43. 34 U4
CL.W 3U CL 4343d -Q- C -o4-I I -6- M- C. -M- 2 34 3
0..a 0..C 4) 1C A0 C m CL CL-~~S 4341- Q33- E g
- = .C. a4. V . .iV a1 Qj a.4 (W. o.0 30 L0U 4 V0.U 4 43w1a 421. 41 4.
43~~~~r C 3 33 :~3-~ .4-a 1. 14.~0 43 L.3..V c
Q; 0
-Z cm-
C -.
-43 S- U-3 43 43 >.- 433 44 44
S0 s- 0. CCCC I-CL C
00 cu Q OC QC Q0 0
A. 0..3'.QA- L. S- S- L .413 ) c u 4 U ) c j WzQ~.3C ~ 3C et3
1. E.
4341
2133
U 44
cu C5 4c, 4
CLC
CL .4 (W o. CL m.L0. C.L 0.0
0j oCc
Q.4 4.
a4~-
Lo -o
~II-
44).
.0 W4) W4
4214
45IAPPENDIX 0
IiCALCULATED TEMPERATURE RISE IN A TIGHT GLAND
1H = 359000 P MsAT where
1 HP = 33,000 ft-lb per minute
J = 778 ft-lb per Btu
P = horsepower loss
I M = mass of material in gland - propellant and packing say 0.5 lb
S = specific heat of propellant 0.36 Btu/IbOF
T = temperature, OF
3You can rewrite thisAT = 33,000 ;(236P) OF per minute
t -Ms
If we assume a 100% loss of hp due to packing
AT = (236)(2) = 4720F per minute temperature riset
However, a more realistic loss would be 25-50% loss for the mixer to turn.Thus, (236)(0.5) = 118OF per minute rise in temperature.
For Ml the autoignition temperature is 3290F. Consider normal ambient3temperature to be 750F. Thus, AT is 329-75 = 2540F.
aT = 254 = 2.2 minutes for the conditions just described to start a fire.at M
This calculation has assumed that no heat loss occurs from the glands and thatabsolute ignition results. Actual operating results using full size mixers donot verify these calculations. Solvent-wet propellant is difficult to igniteand even more difficult to sustain burning at atmospheric conditions.
I1
3 215
I
3 46
APPENDIX E
IGENERATION AND REMOVAL OF FUMES FROM THE MIXER BAY
3Mixer Bay DimensionsL - 19'7" say 19 1/2'W - 15'll 1/2" say 16'H - >15' -
Bay Volume >4680 ft3
Bay Openings.
Door: 6' Wide
Air Forced into Bay:
Air Purgi for Electric Motor3 94 ft/ min without motor fan running (measured value)
Volume of Solvents Evaporated:
0.064 ft3/mina Ether and Alcohol Mixture
Volume of Air Removed Each Minute by Eductor
SQ = AV where
Q = volume removed, ft3/min
A = cross sectional area of eductor, ft2
V = velocity of fluid, ft/min
A = r 2
IA = [3/12]2
A = 0.196 ft2
V = 8,000 fpm - measured value
Q = (8000)(0.196) = 1568 ft3/min removed
Room Volume: >4680 ft3
>4680TW = total air change in room about every 3-4 minutes. Say 5 minutes to
account for the rafter volume.
aCalculations by W. M. Walasinski.
I216
I
I
I DISTRIBUTION LIST
Defense Technical Information Center 12Cameron StationAlexandria, Virginia 22314
Defense Logistics Studies Information Exchange 2U.S. Army Logistics Management CenterFort Lee, Virginia 23801
I CommanderU.S. Army Toxic and Hazardous Materials AgencyAttn: AXXTH-CO-PAberdeen Proving Ground, Maryland 21010-5401
CommanderU.S. Army Toxic and Hazardous Materials Agency 14Attn: A.XTH-TE-DAberdeen Proving Ground, Maryland 21010-5401
iIIIi
iII
i /t Arthur D. Little, Inc.
