Home

University of Illinois at Urbana

1. 113 2 AN 240 1195 19 92 9 94 39 0 2 AN 245 1265 21 08 11 62 N A 3 AN 250 1790 29 83 8 38 N A 4 AN 248 1764 29 40 8 44 N A 6 AP 335 1440 24 00 12 88 9 00 7 AP 336 1281 21 35 14 54 N A 8 AP 320 1686 28 10 10 40 2 35 9 AP 307 N A N A N A N A 10 AP 315 3282 54 70 5 76 N A The next series of tests included four AP propellant tests on S P ID s 6 9 as well as four Teflon tests to perform DIC calibration These tests also exhibited fewer problems that were carefully diagnosed and solved Some of the problems included issues with the solid state relay computer memory and EMF noise with the thermocouples It was during these tests that the solid state relay was abandoned The LabView program was converted to an optimized executable for each test and the HVAC system schedule was modified to coincide with test times The Picoscope was also added to record high frequency pressure data during the thermal runaway and cookoff phenomenon Small adjustments were also made to the other diagnostic systems Figure 3 8 and Figure 3 9 show the improvements made in the imaging diagnostic The image to the left is saturated with light from the resistance wires and blurry The image on the Figure 3 8 CCD Image of S P 2 Figure 3 9 CCD Image of S P 25 29 Shown in Figure 3 10 are the temperature and pressure data from the first AN propellant during the final stages of cookoff This experiment was the only AN propellant that di
2. 14 15 16 List of References Baer M R Hobbs M L Gross R J Schmitt R G Cookoff of Energetic Materials Tech Rep SAND 98 1946 Proc of Eleventh Symposium International on Detonation August 1998 CONF 980803 OSTI Albuquerque Sandia National Laboratories 1998 Baer M R Gross R J Gartling D K Hobbs M L Multidimensional Thermal Chemical Cookoff Modeling Tech Ref SAND 94 19090 Proc of JANNAF Propulsion Systems Hazards Subcommittee Meeting San Diego CA Aug 1 5 1994 CONF 9708123 2 Albuquerque Sandia National Laboratories 1994 Baer M R Kipp M E Schmitt R G Hobbs M L Towards Assessing the Violence of Reaction During Cookoff of Confined Energetic Materials Tech Ref SAND96 1376C Proc of JANNAF Combustion Subcommittee and Propulsion Systems Hazards Subcommittee Joint Meeting Naval Post Graduate School Monterey CA Nov 4 8 1996 CONF 961194 2 Albuquerque Sandia National Laboratories 1996 Blumm J Lindemann A Meyer M Strasser C Characterization of PTFE Using Advanced Thermal Analysis Techniques Bavaria Germany NETZSCH Sep 27 2007 Chew W M Jones D J Johns P Reed J M Subscale Screening of Solid Propellants to Support the Army Insensitive Munitions Program Army Aviation and Missile Research Development and Engineering Center Weapons Development and Integration Directorate Redstone Arsenal AL Department of Defense TB 70
3. Table 4 3 Temperature Percent Differences Average Percent Differences from Mean Minutes R 1 3 R 2 3 R 3 3 Avg 0 5 2 07 2 29 3 97 2 18 5 10 2 32 2 34 0 65 1 77 10 15 1 89 1 89 1 31 1 70 15 20 0 69 1 08 0 39 0 72 20 25 1 98 2 25 1 60 1 94 For the pressure data both DAQ and Picoscope values were normalized to the beginning atmospheric pressure A simple subtraction of the starting pressure provided the normalization The pressure analysis followed the manner of the temperature by calculating a mean between the tests and calculating the average standard deviation through five minute intervals The DAQ and Picoscope pressure measurements were analyzed separately Table 4 4 Pressure Standard Deviations in Psi DAQ Pressure Pico Pressure Minutes Avg Std Dev Seconds Avg Std Dev 0 5 0 09 0 5 0 2 00 5 10 0 11 0 0 5 0 85 10 15 0 28 0 5 1 0 73 15 20 0 55 l 1 5 0 88 20 25 0 46 1 5 2 0 82 2 2 5 0 78 2 5 3 0 73 3 3 5 0 68 3 5 4 0 63 4 4 5 0 61 32 For analyzing the spectroscopy data each percent composition measurement per scan was averaged to a mean and then used to calculate the standard deviation in percent composition In Table 4 5 left is the average standard deviation for all relevant atomic mass units for a full scan In contrast on the right is the averaged standard deviation for an individual atomic mass unit averaged across all scans It shoul
4. 05 1 0 15 20 25 3 0 35 Pico Technology www picotech com Figure F 22 S P 12 Pico Pressure JO Lust 3250 pe 3300 Figure F 23 S P 12 DIC Table F 6 S P 12 Spectroscopy E HE Ab O S O O JA Z Q mr ch eE D O SiS Sib min Numerous scans of same percent compositions as 1st row baseline 14 14 14 14 13 12 10 72 71 72 72 72 73 73 73 73 72 72 71 70 69 11 67 15 26 61 47 14 26 43 33 41 82 S P 13 Propellant AP200um AP60 130um HTPB IPDI DOA Test 14 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Good Comments This test was the second in the repeatability series The DIC results were the same as S P 12 S P 13 Temperature vs Time 350 300 N LD o N o o R 1 3 in DA R 2 3 in 100 A Temperature C LA un o R 3 3 in 50 0 o 200 400 600 800 1000 1200 1400 1600 1800 Time Sec Figure F 24 S P 13 Temperature 83 Pressure Psig 10 0 33 2 4 S P 13 DAQ Pressure vs Time 00 3320 3340 3360 3380 3400 3420 3440 3460 3480 00 PA SIN DARAS Time Sec Figure F 25 S P 13 DAQ Pressure 660 0 my DC 640 0 620 0 600 0 580 0 560 0 540 0 520 0 500 0 480 0 460 0 05 00 05 1 0 T 20 25 30 35 40 45 s Pico Technology www picotech com Figure F 2
5. Hydrochloric Acid Hydroxyl Terminated Polybutadiene binder Insensitive Munitions Isophorone Diisocyanate curing agent Insensitive Propellant Screening Test Protocol Material Density Quadrupole Mass Spectrometer Sandia National Laboratories Standard Operating Procedure Solid Propellant Super Small Scale Cookoff Bomb Temperature Time Variable Confinement Cookoff Test Virtual Instrument LabView 1 Introduction The earliest recorded use of propulsion is attributed to a Greek named Archytas in 400 BC where steam was used to propel a wooden bird in flight 7 Two hundred years later the concept of solid propellant originated when the Chinese developed black powder for entertainment and eventually warfare Their version of today s black powder was a simple mixture of charcoal sulfur and rock salt The Chinese started using the compound in religious and funeral activities by filling bamboo sticks sealing them and throwing them into a fire to produce explosions It is believed that some of these were improperly sealed allowing the first version of rockets to propel out of the fires Over the next thousand years the Chinese developed the concept into fire arrows that are first recorded as being used for weapons against Mongol invaders in the battle of Kai fung fu in 1232 A D 7 Over the next five hundred years the use of rockets in warfare spread from the Mongols to the Arabs French and later the British The applicat
6. Pressure ae BR e u 1600 1620 1640 1660 1680 1700 1720 1740 1760 1780 1800 5 Time Sec Figure F 17 S P 6 DAQ Pressure S P 6 Pico Pressure vs Time KA KA D Pressure Psig Dardu h uha a ddin aiibi 0 08 0 18 0 28 0 38 0 48 Time Sec Figure F 18 S P 6 Pico Pressure 78 Figure F 19 S P 6 DIC CO 44 Table F 5 S P 6 Spectroscopy 4 3 ES O 3 8 O a A Z Q TA e a x o SiS Sib H Numerous scans of same percent compositions as first row baseline 14 13 12 11 73 73 74 75 76 76 76 76 75 75 74 72 13 23 65 52 36 44 14 30 27 79 S P 12 Propellant AP200um AP60 130um HTPB IPDI DOA Test 13 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Good Comments This was the first repeatability test DIC results only exhibited the thermal expansion of the carbon fiber case S P 12 Temperature vs Time 350 300 250 200 R 1 3 in 150 R 2 3 in Temperature C 100 lt R 3 3 in 50 0 200 400 600 800 1000 1200 1400 Time Sec Figure F 20 S P 12 Temperature 80 Pressure Psig 10 6 B 2 0 4 2 2900 620 0 600 0 580 0 560 0 540 0 520 0 500 0 480 0 460 0 Eu S P 12 DAQ Pressure vs Time A JI ee 3000 3050 3100 3150 Time Sec Figure F 21 S P 12 DAQ Pressure
7. 15 68 60 70 S P 9 Propellant AP200um APS50um HTPB IPDI DOA Test 6 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Good Comments This test was nearly a complete failure exhibiting computer issues that lead to the executable being created as well as random failure of the solid state relay causing a sporadic heating rate S P 9 Temperature vs Time 300 N LD o N Oo o R 1 3 in Temperature C uw o R 2 3 in 100 R 3 3 in 0 500 1000 1500 2000 2500 3000 3500 Time Sec Figure F 11 S P 9 Temperature 71 Table F 2 S P 9 Spectroscopy H CH CH OH NH HO N CO CHO O Ar CO 2 14 16 17 18 28 29 32 40 44 Numerous scans of same percent compositions as first row baseline 0 5 1 0 1 74 1 15 1 0 0 5 1 0 1 75 1 14 1 0 1 5 1 0 1 75 1 14 1 0 1 5 1 0 1 76 1 13 1 0 1 5 1 0 1 77 1 12 1 0 1 5 1 0 1 77 1 11 1 1 1 5 1 0 1 78 1 10 1 1 1 5 1 0 1 78 1 9 1 1 1 5 1 0 1 79 1 8 1 2 1 5 1 0 1 78 1 7 1 2 1 5 1 0 1 77 1 7 1 3 1 5 1 0 1 77 1 6 1 4 1 5 1 1 1 76 1 6 1 5 1 5 2 1 1 75 1 6 1 5 1 5 2 1 1 75 1 6 1 5 1 5 2 1 1 75 1 6 1 5 1 5 2 1 1 74 1 6 1 5 1 5 2 1 1 73 1 6 1 6 12 S P 8 Propellant AP200um APS50um HTPB IPDI DOA Test 7 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Bad Comments Fluctuation in t
8. In slow cookoff the distributed thermal damage allows convective burning which produces accelerated combustion and a rapid flame spread Experiments focused on the presence of cracks and porosity in the thermally damaged material which increases the specific surface area It was found that cracks formed during the initial stages allow flame penetration deeper into the propellant and enhance the convective burning with higher burn rates at the crack tips Hypotheses disagree whether the cracks or the specific surface area is the main element in dictating thermally damaged energetic material and it is likely that neither can be neglected 23 In 1995 S Y Ho published a paper that contained a similar experimental design and research elements related to the scope of this report Ho performed both slow and fast cookoff experiments characterized as 6 C hr and 72 C hr respectively The test apparatus was a modified super small scale cookoff bomb SSCB defined by Naval Weapons Center NWC shown in Figure 1 4 13 f Som Wedi Le a e i log pale ES SE Prepare K i E e T zech Eu SZ LL if tg 4 MT en LL Iisegs esd l SS A Ka d LI TZ SE Ee 1 pr Sea ibero GE 1 Hee ng 3 L Papei verk Figure 1 4 Modified SSCB Taken Directly from Reference 13 Ho studied the effect of percent composition of plasticizer on the reaction violence in both fast and slow cookoff From his research Ho
9. it was found 25 that a higher magnification was needed in order to accurately measure the carbon fiber surface displacement A new achromatic doublet lens of focal length 100 mm was installed and the camera was moved farther away from the test apparatus in order to double the magnification The camera was mounted to a translation slide for proper focus adjustment To achieve smaller speckles an airbrush gun and small atomizing high temperature paint was utilized An inert sample of Teflon was used to validate the imaging diagnostic and digital image correlation code Images of Teflon at 25 C and at 250 C were processed with the DIC It should be noted that pre processing of the images centered the expansion and eliminated the bulk sample movement in the images that would have saturated the DIC results Careful cropping of the images then allowed a cleaner calibration of the DIC results which were checked with thermal expansion coefficient of Teflon between the stated temperatures To easily demonstrate the working DIC result in small images a single image of Teflon was manually expanded with a 5 expansion and ran through the DIC MATLAB code Figure 3 6 is the original image followed by the expanded image and vectors plotted over it The vectors represent the movement from the first to the second image Figure 3 6 Teflon 26 Figure 3 7 Expanded Teflon with DIC The spectrometer system experienced a few small difficulties During
10. 113 0 FORMAT 12 IDA IA 1 12 Yes PPI a ell d 101 E58 2E1 45 FORMAT I JQATTA G A Me a OE Ee oad E 100 F6 2 ES8 2E1 FORMAT A9 13 A1 FORMAT A4 100 A4 13 A1 A7 FORMAT A4 100 A5 13 A1 A8 FORMAT A6 1X A3 1X A4 FORMAT 2x A11 done Skip vver the Error Handling Section GOTO 999 ERROR HANDLING SECTION GW CONTINUE WRITE ERROR WRITE OPEN statement failed WRITE GOTO 999 E CONTINUE WRITE ERROR WRITE READ statement failed WRITE CLOSE inFileUnit ERR 930 GOTO 999 930 CONTINUE WRITE ERROR WRITE CLOSE statement failed WRITE GOTO 999 E CONTINUE RETURN END 98 DIC Matlab Code WD DD d On we Go Mr de fs de e e de de 0 0 Go Go Y wow Y Go NN NN N NN NN Hh PP PRP k RP k k k k an e WON Pop OD am Ms 0 Pic d OD DD o M OD DD 0 D M Do Ve Test code for the DIC of Slow Cookoff for Brad Horn Author Julio Barros Modified by Bradley Horn clear all cle To use correct slash for mac vs windows machine if ispc 1 Slash else slash end Choose directory and Read all the images pathDir uigetdir Choose the Folder in which you want to process files dir pathDir slash tif Make the DIC Results Folder mkdir pathDir slash DIC J ResultFol pathDir slash DIC Make th
11. 13 The low standard deviations shown in the results prove the repeatability of the setup for future tests and cookoff research Results have also proven that slow cookoff heating rates can be achieved regardless of which standard is chosen Due to the inherent difference in experimental setup the large pressures found in the Modified Super Small Scale Cookoff Bomb and other small cookoff experiments cannot be obtained In place of a steel pipe the carbon fiber case used to house the solid propellant and the top quartz window fail at containment at around 8 psig The containment fails before the pressure buildup can reach typical values of 190 MPa or 27 5 k psig 13 The spectroscopy results from the QMS spectrometer have validated the expected species from the CEA analysis Yet the experimental composition showed large amounts of H when the expected combustion products were predicted to be 0 3 hydrogen Due to the structure of the Quadrupole Mass Spectrometer there is a significant backflow of hydrogen that skews the results The flow through the spectrometer is split into 2 paths The primary path is for the bulk flow while the second path is sampled from the bulk flow in small increments The samples pass through the filament Quadrupole mass filter and the Faraday Cup detector Both paths are recombined before the diaphragm pump allowing low atomic mass molecules such as hydrogen to back flow into the sampling path See reference 16 for mor
12. 33561 0 27309 0 00257 0 27057 0 00002 0 10578 THERMODYNAMIC PROPERTIES FITTED TO 20000 K 1 1794 2 3069 1 2066 477 4 0 29531 2 3069 1 4762 0 4615 1 3351 0 5266 0 7487 0 00976 0 00127 0 33994 0 27309 0 00350 0 26663 0 00002 0 10578 1 1796 2 3076 1 2066 477 4 0 29536 2 3076 1 4766 0 4616 1 3352 0 5268 0 7486 0 00975 0 00127 0 33996 0 27309 0 00350 0 26662 0 00002 0 10578 1 1318 2 0419 1 2193 468 1 0 28357 2 0419 1 2854 0 4505 1 3181 0 4851 0 7704 0 01225 0 00059 0 33419 0 27309 0 00230 0 27178 0 00002 0 10578 WI FRACTION ENERCY TEMP SEE MOTE EJ FC MOL K 0 8800000 295770 000 623 150 0 0070800 74000 000 623 150 0 0928800 2970 000 623 150 0 0200400 311630 000 623 150 91 1 1660 2 2291 1 2101 477 5 0 29505 2 2291 1 4345 0 4585 1 3335 0 5214 0 7546 0 01049 0 00113 0 33817 0 27309 0 00314 0 26817 0 00003 0 10578 FileEditor Repeatability out 0 F 0 00000 AFUEL 0 000000 R EQ RATIO 1 123859 PHI EQ RATIO 0 000000 THERMODYNAMIC PROPERTIES F BAR 0 54675 e 627 95 RHO KG CUM 2 9015 1 E EJ EC 6953 00 U EJ EG 7141 44 C EJ EC 12268 6 8 EJ EC E 8 4650 M 1 m 27 707 dLV dLP t 1 00943 dLV dLT p 1 1661 Cp EJ EC K 2 2299 CAMMAS 1 2100 SON VEL M SEC 477 5 TRANSPORT PROPERTIES CASES ONLY CONDUCTIVITY IN UNITS OF MILLIWATTS CM X VISC MILLIFOISE 0 29509 WITH EQUILIBRIUM REACTIONS Cp KJ KC K
13. 7 5 amperes Heating element version III is uniform easily manufactured requires fewer electrical connections lower amperage and costs under 75 with roughly 3 hours of labor per experiment Most ceramic tubes are destroyed during cookoff eliminating cleaning efforts A table of time and expenses is located in Appendix E Figure 3 5 is an engineering schematic of the new heating element and basic test apparatus 23 Figure 3 5 Heating Element Version III Schematic Following repeatability tests modifications were made to the power control and heating in order to achieve slow cookoff heat rates Originally the Variac was set to 90 volts and turned on via the solid state relay for a constant heat source throughout the experiment The result was a heating rate near 12 5 C min which is an intermediary heating rate between slow and fast cookoff characteristics This constant heat source was used for simplicity and quick test times when validating the experimental setup In order to achieve a slow cookoff heating rate near 3 3 or 6 C hr a different solid state relay was used to pulse the power at varying frequencies Due to inherent circuitry issues the Variac was replaced by regular 120 125 wall outlet voltage as a power source A temperature controller and an additional thermocouple were installed to control the relay and pulse the power to achieve slow cookoff heating characteristics 3 3 Diagnostic Implementation Significant improveme
14. A N A Bad Comments The melted sample clogged the bottom gas port eliminating pressure and spectroscopy data The red T couple trace is hidden under the green trace below S P 2 Temperature vs Time 300 250 200 R 1 3in 150 Temperature C R 2 3 in 100 R 3 3 in 50 0 200 400 600 800 1000 1200 Time Sec Figure F 3 S P 2 Temperature 65 S P 3 Propellant AN HTPB IPDI Test 3 Mass Percents 70 27 86 2 14 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good N A N A Good Comments This test proved T couples cannot be recycled blue trace Melted propellant clogged the bottom gas port Acquired images were acceptable but still had room for improvement DIC Results exhibited only the small oscillation in the field of view S P 3 Temperature vs Time 350 300 N uw o N Q o R 1 3 in KA LD o R 2 3 in 100 R 3 3 in Temperature C 50 0 200 400 600 800 1000 1200 1400 1600 1800 Time Sec Figure F 4 S P 3 Temperature Figure F 5 S P 3 DIC 66 S P 4 Propellant AN HTPB IPDI Test 4 Mass Percents 70 27 86 2 14 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good N A N A Good Comments The green trace was sporadic due to being grounded instead of the ordered ungrounded The pressure chart displays when the bottom gas port clogged allowing the mass spectrometer to draw a vacuum S P 4 Temperature vs Time 320 27
15. OPEN fails go to the error handling section OPEN UNIT InFileUnit STATUS 0LD FILE InFileName ERR 910 OPEN UNIT 0utFileUnit STATUS UNKNOWN FILE 0utFileName ERR 910 Initialize data array DO j 1 ScanMax SumP 3 0 0 DO i 1 nmax Value j i 0 0 AMU 3 1 0 0 PercentAMU 3 i 0 0 ENDDO ENDDO Read in the first lines of titles READ UNIT InFileUnit FMT 70 ERR 920 a b c WRITE OutFileUnit a b c READ UNIT InFileUnit FMT ERR 920 d Point 1 M M 1 991 e WRITE OutFileUnit d Point 1 M M 1 991 e Read in all the Date Time Stamps and Values DO 10 N 1 NumScans READ infileunit A END 10 buff READ Unit buff FMT 30 END 10 ERR 920 Datemm N Datedd N Dateyyyy Ni Timehh N Timemm N Timess N Timep N READ buff 22 END 10 Value N m m 1 992 TotalP N Value N 992 CONTINUE Sum up the partial pressures to a single value per AMU per scan DO N 1 NumScans l 1 AMU N 1 Value N 1 Value N 2 Value N gt Value N 1 Value N 5 DO j 2 99 l 1410 AMU N j Value N 1 5 Value N 1 1 Value N l 3 4 Value N 1 2 Value N 1 1 Value N 1 Value N 1 1 Value N 1 2 Value N 1 2 Value N 1 4 ENDDO 1 1 10 AMU N 100 Value N 1 5 Value N 1 4 Value N 1 3 Value N 1 2 Value N 1 1 Value N 1 ENDDO Process data to see AMU s of interest AMUs that change DO i 1 100 Delta 0 0 96 Delta abs AMU 1 i TotalP 1 AMU NumScans i TotalP NumScans IF D
16. heating rate is defined as 3 3 0 5 C hr after heating the sample at 5 C min to 50 C and held constant for 8 hours for thermal conditioning 6 This standard for slow cookoff is also found in references 1 13 26 Another common standard in literature and industry is 6 C hr following a similar initial thermal ramp 13 The heating rate greatly determines the ignition temperature as well as the severity of the damage Contrary to intuition as the heating rate is increased the ignition temperature increases while the level of damage decreases As a result slow cookoff is more dangerous than fast cookoff and extensive research has been done to design composite casings that help in preventing slow cookoff conditions Additional research has been done on modifying the composition of insensitive munitions IM to be less susceptible to stimuli 27 as well as slow cookoff environments During the initial stages of heating the physics governing the behavior of the propellant is strongly dictated by basic heat transfer and quasi static mechanics The first stage of cookoff is an endothermic process represented as stage a in Figure 1 1 below As the temperature increases thermal expansion occurs and thermal coefficients change due to the material transformation Once the point of self sustained reaction or thermal runaway occurs stage b the rate of gas pressurization increases significantly The mechanical interaction of the propellant and
17. imported due to file size The processed data was then plotted in order for qualitative analysis All excel files are also intended to be presented to the computational group for validation of the computational code 53 D Solid Propellant Slow Cook off Standard Operating Procedure Solid Propellant Slow Cook off Standard Operating Procedure SP SCO SOP Latest Revision 10 3 11 SP SCO TEST ____ S P ID S P Composition by Mass Date S P Mass w case w out Heated to Temperature Ignition Yes No Propellant Preparation 5 Days prior to test 1 See Propellant Fabrication SOP for instructions 2 Transfer recorded propellant properties to this page and Diagnostic Record Test Preparation 1 Day Prior to Test 1 Attach lid to tank inside chamber ee Carry optics board into chamber and clamp down to lid L Wire the heating element and verify its operation with a variac Insert silicone seals between the steel top and bottom flanges 2 Insert propellant and attach all thermocouples with sealant typically4 With the heat shield removed check that propellant is properly seated in the top and bottom plates and plastic bearings Firmly bolt down top section in an alternating pattem Connect T couples to card x3 and T couple to extension for T controller 9 Thread cable bundle from outside into chamber through back top hole and connect the following a Power wires to designated power terminals b BN
18. pressurization within the confinement dictates the type of reaction upon ignition For AP HTPB propellants the cookoff becomes exothermic at about 247 C 22 Chemical reactions 1 Gas escaping FWIEGAS Dynamics b porous propellant KON g 4 r E ee a G Combuston i iTe Tasa j E 3 CN Tee f 2 i i K pa y a S BR Abisting surtnce Z A f S Y e a P r Heat Fiux g x a Gas Pressure KU z Ma k Prope tont A ef j 1 ef KR 7 Porous flow 5 COA Chemical reactions Solid Gas Thermo Mechanical Hot spot imtemal combustor EN ti i ma mn Figure 1 1 Stages of Cookoff 11 Figure 1 2 Thermal Runawav Phvsics 11 Once ignition is reached the sample is categorized based on the speed and severitv of the combustion process as deflagration explosion or detonation The extensive definitions and qualifications for each categorv are explained in detail in the TB 700 2 regulatorv document Department of Defense Ammunition and Explosives Hazard Classification Procedures 6 A deflagration is a rapid moving combustion wave with a flame velocity on the order of 1 10 mm s while an explosion also includes a pressure burst and fragmentation A detonation event is separated from explosion by the criteria of a supersonic propagation wave which results in a severe blast effect and fragmentation The critical diameter of AP HTPB propellants exceed 200 mm or 7 87 inches which is much greater than the si
19. suggested a possible criterion in categorizing slow versus fast cookoff being the reversal of the temperature gradient especially during thermal runaway In fast cookoff the hottest area is on the surface of the propellant which was 30 40 C higher than the center while slow cookoff exhibited nearly uniform temperature distribution within 30 minutes The point of ignition surface versus center correlates to fast versus slow heating as well as reaction violence The reaction violence was experimentally quantified by the peak pressure the impulse and the initial rise dP dt Ho also found that with slow heating the HTPB AP propellant became structurally tougher due to the oxidation of the HTPB binder and as a result exhibited higher thermal conductivity resulting in higher temperatures more uniform thermal distribution and more violent cookoff reaction It should be noted that Ho performed experiments measuring angular thermal gradients and as expected found the variance to be insignificant and less than 3 C throughout cookoff 13 1 4 Problem Statement Traditionally cookoff is modeled with a thermal chemical model that is decoupled from the mechanical behavior of the energetic material Ignoring the mechanical state of the propellant allows for a good approximation during fast cookoff but induces non negligible errors during slow cookoff It is now understood that the thermally damaged mechanical state of the propellant favors conditions
20. the initial tests with AN propellant nearly all tests failed to produce either pressure or spectrometer data due to clogging of the gas port Since AN propellant has a melting temperature lower than its decomposition ignition temperature the propellant melted and clogged the bottom sampling hole and eliminated any pressure or spectrometer data During a particular test molten propellant permanently clogged the capillary tube The gas connections were modified slightly and a 0 5 micron gas filter was added before the new capillary tube to eliminate the problem for future tests 3 4 LabView Programming LabView programming was a significant part of initial research LabView VI examples existed only for the temperature and spectrometer diagnostics Progress on the programming was slow until a course was taken in May 2010 During a weeklong course all aspects of LabView programming were covered at a rapid pace resulting in a CLAD certification Certified LabView Associate Developer With the acquired skills an extensive program was built The temperature VI example was modified for appearance and optimized to control the sampling rate and a few other visual displays A completely new pressure virtual instrument was designed using Measurement and Automation Explorer to create tasks that controlled the NI DAQ card for the pressure sampling The large spectrometer VI example was also modified to 27 include background subtraction a lockable graph
21. the top quartz window from an off center fiber optic cable The reflected light off the bottom polished surface would be collected via a second fiber optic cable The path length would be relatively long of 4 inches due to the propellant length By using fiber optics several different instruments could be used to analyze the data Available to us here are ten different spectrometers for spectroscopy on the core gases in the visible UV or near IR region with various detector combinations allowing either high 18 sensitivity 90 QE high signal to noise ratio spectra and or high speed data acquisition e g the FK CCD camera collects spectra every 1 us 11 The second method currently being explored is to investigate species in the mid infrared region of 3 5 microns Species like HCI display strong absorption peaks in this range and are not detected by the mass spectrometer as discussed later The proposed design involves inserting a sampling window into the gas lines between the pressure and capillary tube below the test apparatus By sampling below the apparatus instead of through the top window interference from the resistance wires can be avoided The proposed design uses a resistance wire heat source to collimate light through CaF lenses that serve as windows into the gas line The transmitted light would then travel into a Czerny Turner configuration for a monochromator and finally a camera detector For mid infrared spectroscopy a cool
22. with an operating range from slight vacuum to a max pressure of 285 psig The pressure is measured from the core region with a max temporal resolution of 1 ms and accuracy of 50 Pa on a range of up to 2 MPa The pressure rise in the final stage of cookoff requires a higher resolution than the initial stages and capabilities of said sensor Figure 2 9 Gems Pressure Sensor Figure 2 10 Gems Pressure Sensor amp Gas Filter 15 The Gems sensor signal is fed into a control box that connects the sensor to a small 24 V DC transformer for power and to a BNC output From there the signal is terminated with a resistance of 100 ohms in another control box and then connected via BNC cables to the DAQ card and Picoscope LabView converts the voltage into pressure using a pretest calibration and saves to file both the original and processed signals The pressure signal read into LabView is recorded at static levels of 1 Hz while the signal read into the Picoscope allows for a high frequency pressure measurement from auto ignition to the mechanical failure of the carbon fiber case Figure 2 11 NI USB 6008 DAQ Card 2 3 3 Imaging A fundamental check on the mechanical component of the modeling code is material deformation or displacement The inner annulus surface can be imaged for a rough internal displacement and core shrinkage after off axis distortion corrections A gridded surrogate sample would be used in order to experimentally solve for the nee
23. 0 220 R 1 3 in 170 gt R 2 3 in Temperature C 120 R 3 3 in 70 20 0 500 1000 1500 2000 Time Sec Figure F 6 S P 4 Temperature 67 Pressure Psig S P 4 DAQ Pressure vs Time 10 0 500 1000 1300 2000 2500 3000 Time Sec Figure F 7 S P 4 DAQ Pressure 3500 4000 Figure F 8 S P 4 DIC 68 4500 5000 S P 10 Propellant AP200um APS50um HTPB IPDI DOA Test 5 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Good Comments This was the first AP test and the thermocouples were grounded instead of the ordered ungrounded The voltage setting was too low and had to be adjusted mid test for the AP No pressure was observed during this test likely due to wiring in the pressure diagnostic S P 10 Temperature vs Time 350 300 N uw o N Q o R 1 3 in KA LD o R 2 3 in Temperature C 100 R 3 3 in 50 0 500 1000 1500 2000 2500 3000 3500 Time Sec Figure F 9 S P 10 Temperature Figure F 10 S P 10 DIC 69 CO 44 Table F 1 S P 10 Spectroscopy Sls ab O EIS e Ole al A Z S E a am e si El H Numerous scans of same percent compositions as first row baseline 14 13 13 12 12 11 11 74 74 74 74 75 75 75 10 75 75 76 76 76 77 76 76 76 76 76 75 75 75 74 71 10
24. 0 2 Department of Defense Ammunition and Explosives Hazard Classification Procedures Tech Rep NAVSEAINST 8020 8C June 17 2005 Dumoulin Jim A Brief History of Rocketry Kennedy Space Center Science and Technology Home Page NASA Spacelink System 25 Aug 2011 Accessed 02 May 2011 lt http science ksc nasa gov history rocket history txt gt Geisler An Integrated FCO Scaling Plan Draft Issue brief 2005 Geisler Fast Cook Off Overview Issue brief 2005 Geisler FCO Fragmentation Scenario Issue brief 2005 Glumac N Matous K IllinoisRocstar LLC Experimental and Computational Program for Slow and Fast Cookoff for Insensitive Munitions Testing Tech Rep A08 027 0721 2009 HITRAN Database Version 13 0 Center for Astrophysics Oct 28 2011 lt http www cfa harvard edu hitran gt Ho S Y Thermomechanical Properties of Rocket Propellants and Correlation with Cookoff Behavior Journal of Propellants Explosives Pyrotechnics vol 20 pp 206 14 1995 Hok Heng A Laser Scanning Technique for Measuring Solid Propellant Burning Rates MS Thesis Department of Mechanical Engineering University of Illinois at Urbana Champaign 1999 Isophorone Diisocyanate IPDI Partially Validated Method 2034 United States Department of Labor Occupational Safety amp Health Administration Apr 1988 lt http www osha gov dts sltc methods partial pv2034 2034 html gt Lemke Brad An
25. 0 44 Numerous scans of same percent compositions as first row baseline 1 9 2 1 2 72 1 14 1 0 1 5 2 1 2 72 1 13 1 0 1 5 2 1 2 72 1 13 1 0 2 5 2 1 2 71 1 12 1 1 2 5 1 1 2 73 1 11 1 1 2 5 1 1 2 73 1 10 1 1 2 5 1 1 2 73 1 9 1 2 2 5 1 1 2 71 1 9 1 3 2 5 1 1 2 71 1 8 1 3 2 5 2 1 2 71 1 8 1 5 2 5 2 1 2 70 1 8 1 5 2 5 2 1 2 70 1 7 1 6 2 5 2 1 2 69 1 T 1 7 2 5 2 1 3 65 1 6 1 11 3 4 3 1 2 57 1 6 1 17 3 3 5 1 2 43 1 4 1 28 17 2 4 1 2 37 1 4 1 22 19 3 3 1 2 44 1 6 1 14 88 G CEA Results Below is the input file followed by the output file for the CEA Chemical Equilibrium with Applications program written by McBride and Sanford found on the NASA website 31 The program was run under the repeatability series experiments to validate the mass spectrometry data Temperatures of 329 5 C 354 7 C and 354 8 C were used along with pressures of 6 88 psia 6 14 psia and 7 93 psia corresponding to the measured conditions for the S P 12 14 propellants The formulas for AP HTPB IPDI and DOA were entered into the input file along with the enthalpies of 295 77 2 97 74 311 630 respectively in units of kJ mol The problem was ran as a T P problem As can be seen from the results expected products are CH4 CO CO2 NH3 HCI H2 H20 and N gt Current File Repeatability inp problem oase Repeatability tp t 0 329 5 354 7 354 8 p psia 6 88 6 14 7 93 reaot name AP wt 88 t 0 350 h kj mol 295 77 N1H4CL104 name 1PDI wt 0 708 t 0 3
26. 143 1 2 5139 1 2 3507 1 2 2414 1 2 2409 1 3 0419 1 E KJ EG 7003 60 6948 68 6948 45 7000 69 6944 89 6944 66 7007 06 U EJ EC 7183 53 7137 34 7137 15 7180 78 7133 77 7133 57 7186 80 C KJ EG 12081 1 12294 5 12295 4 12101 6 12316 0 12316 9 12055 5 8 EJ EC K 8 4253 8 5145 8 5149 8 4641 8 5548 8 5551 8 3772 M 1 n 27 848 27 670 27 669 27 823 27 639 27 638 27 878 dLV dLP t 1 00767 1 00985 1 00986 1 00800 1 01018 1 01019 1 00727 90 0 54675 627 85 2 9020 1 6953 23 7141 63 12267 7 8 4646 27 708 1 00942 FileEditor Repeatability out dLV ALT p Cp EJ EC K GAMMAZ SON VEL M SEC 1 1391 2 0836 1 2169 467 9 1 1735 2 2725 1 2081 477 4 TRANSPORT PROPERTIES CASES ONLY CONDUCTIVITY IN UNITS OF MILLIWATTS CM E VISC MILLIFOISE 0 28369 0 29519 WITH EQUILIBRIUM REACTIONS Cp EJ EC K CONDUCTIVITY PRANDTL NUMBER 2 0836 1 3187 0 4482 Cp EJ EC K CONDUCTIVITY PRANDTL NUMBER 1 3188 0 4873 0 7677 0 01194 0 00063 0 33496 0 27309 0 00245 0 27112 0 00002 0 10578 2 2725 1 4589 0 4598 1 3344 0 5242 0 7514 0 01009 0 00120 0 33914 0 27309 0 00334 0 26733 0 00002 0 10578 1 1736 2 2732 1 2081 477 4 0 29524 2 2732 1 4593 0 4599 1 3344 0 5244 0 7513 0 01008 0 00121 0 33915 0 27309 0 00334 0 26732 0 00002 0 10578 1 1451 2 1179 1 2151 467 8 0 28378 2 1179 1 3449 0 4469 1 3194 0 4892 0 7654 0 01168 0 00067 0
27. 17 2004 Naval Air Warfare Center China Lake CA 2004 Rocco J A F F Lima J E S Frutuoso A G lonashiro K Iha M Matos J R Su rez Iha M E V Thermal Degradation of a Composite Solid Propellant Examined by DSC Tech Ref 1388 6150 Journal of Thermal Analysis and Calorimetry vol 75 2 pp 551 557 2004 Schmitt R G Baer T A Millisecond Burning of Confined Energetic Materials During Cookoff Tech Ref SAND 97 2668C Proc of JANNAF Propulsion Systems Hazards Subcommittee Meeting West Palm Beach FL Oct 27 31 1997 CONF 9710108 Albuquerque Sandia National Laboratories 1997 Stanford Research Systems QMS 100 Series Gas Analyzer User s Manual 2000 Unicomposite Carbon Fiber Tube Unicomposite Technology Co LTD 2010 Victor Andrew C Insensitive Munitions Technology for Tactical Rocket Motors Victor Technology San Rafael CA 1994 Published in part in Tactical Missile Propulsion by American Institute of Aeronautics and Astronautics Inc 1996 Wallace Ingvar and Duane Blue Insensitive Munitions Aluminized Propellant for Tactical Boosters Tech Ref N60530 91 C 0254 China Lake Naval Air Warfare Center Weapons Division NAWCWD Thiokol Propulsion N68936 97 C 0268 2000 45 Appendix A Typical Ingredients of Composite Solid Propellants Type Percent Acronym Typical Chemicals AP Ammonium perchlorate Oxidizers AN Ammonium nitrate crystalline 0 70 KP Potassium p
28. 2 2299 CONDUCTIVITY 1 4350 PRANDTL NUMBER 0 4586 WITH FROZEN REACTIONS Cp KJ KC K 1 3335 CONDUCTIVITY 0 5215 PRANDTL NUMBER 0 7546 cua 0 01048 co 0 00113 co2 0 33819 BCL 0 27309 H2 0 00315 E20 0 26816 NE3 0 00003 H2 0 10578 THERMODYNAMIC PROPERTIES FITTED TO 20000 K 92 H Data Processing Codes Temperature Data Smoothing FORTRAN Code C Usersihorn8 Desktop Processing Codes Tsmooth f Wednesday March 30 2011 3 41 PM CO oO OO AW AWAY HAO A et E ONA A St fr Si A A a El C C Status Filename Tsmooth f Written by Brad Horn Email horn8 illinois edu Dec 20 2010 Course Masters Thesis Program Tsmooth f Due date N A errors in compilation compiles but does not perform partially complete complete Description this program will take the temperature data and smooth it program Tsmooth implicit real s a h o z implicit integer 4 i n character 0 inputfile outputfile This code will read data from a file average every n points write out the averaged dataset write 6 10 format 2x Data Smoothing Routine 2x Enter the name of the input file read 5 inputfile write 5 20 format 2x Enter the name of the output file read 5 outputfile write 5 30 inputfile outputfile format 2x inputfile a60 2x outputfile a60 write 6 40 format 2x Enter the number of points to average read 5 n write
29. 5 33 23 23 23 23 32 31 23 23 31 22 23 22 22 22 22 21 30 29 24 25 10 10 10 10 28 27 26 26 28 20 26 29 19 18 18 18 18 18 17 15 25 32 36 38 37 36 36 36 37 29 23 22 23 31 24 25 26 28 11 38 DR a Y Lui 40 178 845 Images acquired during the slow cookoff were the first of a solid propellant sample with the 2X magnification installed following the repeatability series Despite the increased magnification the only DIC movement was bulk sample movement and the thermal expansion of the carbon fiber case Figure 4 5 and Figure 4 6 are examples of the images with the DIC vector plot overlay Figure 4 5 S P 25 at 25 C AS Rh Don Table 4 8 cont 0 bech bech p be bech bech bech 39 2 hh P nn P FA 27 28 93 64 65 65 67 68 p bech bech be bech bech bech bech Figure 4 6 S P 25 at 75 C 8 8 10 11 11 11 12 13 haha oo 15 CO OO On Ort 5 Conclusion and Recommendations 5 1 Conclusion The work performed here is not complete yet many key conclusions can be drawn from the documented results The primary conclusion is that the test apparatus has proven operational and repeatable The repeatability experiments were of an intermediary heating rate between slow and fast cookoff characteristics with ignition temperatures similar to those of fast cookoff
30. 50 h kj mol 74 C 12 H 18N 202 name ETFB wt 9 288 t 0 350 h kj mol 2 97 C 7 332 H 10 982 O 0 058 wt 2 004 t 0 350 h kj mol 311 630 C 22 HE 42 0 4 output massf short transport plot t p rho 89 FileEditor Repeatability out G s g NASA CLENN CHEMICAL EQUILIBRIUM PROCRAM CEA2 MAY 21 2004 BY BONNIE MCBRIDE AND SANFORD CORDON REFS MASA RP 1311 PART I 1994 AND MASA RF 1311 PART II 1996 problem oase Repeatability tp t 0 329 5 354 7 354 8 p psia 6 88 6 14 7 93 reaot name AP wt 88 t 0 350 h kj mol 295 77 N1H4CL104 name IPDI wt 0 708 t 0 350 h kj mol 74 C 12 H 18N 202 name ETFPB wt 3 288 t 0 350 h kj mol 2 97 C 7 332 E 10 982 O 0 058 name DOA wt 2 004 t 0 350 h kj mol 311 630 C 22 H 42 04 output massf short transport plot t p rho end THERMODYNAMIC EQUILIBRIUM PROPERTIES AT ASSICHED TEMPERATURE AND PRESSURE CASE Repeatability REACTANT WI FRACTION TEMP SEE MOTE EJ EC MOL Xx NAME AP 0 8800000 295770 000 623 150 NAME IPDI 0 0070800 74000 000 623 150 NAME SITES 0 0928800 2970 000 623 150 SAME DOA 0 0200400 311630 000 623 150 0 F 0 00000 AFUEL 0 000000 R EQ RATIO 1 123859 PHI EQ RATIO 0 000000 THERMODYNAMIC PROPERTIES F BAR 0 47436 0 47436 0 47436 0 42334 0 42334 0 42334 0 54675 T E 602 65 627 85 627 95 602 65 627 85 627 95 602 65 RHO EC CU M 2 6363 1 2 5
31. 6 50 n format Averaging i8 points open unit 20 file inputfile status old open unit 21 file outputfile status unknown Initialize the sums and counter to zero istop 0 icount 1 xsum 0 0 ysum 0 0 zsum 0 0 93 C Users horn8 DesktopiProcessing Codes Tsmooth f Wednesday March 30 2011 3 42 PM e Read file and sum After n values are read in compute the average and write it c Reset the sums and counters after everv n values e do while istop eq 0 read 20 end 101 x y Z xsum xsumtx ysum ysumty zsum zsumtz icount icountt1 if icount ge nti then xavg xsum icount 1 yavg ysum icount 1 zavg zsum icount 1 write 21 60 xavg yavg zavg B FORMAT 3 F8 2 5X xsum 0 0 ysum 0 0 zsum 0 0 icount 1 end if end do 101 continue c c If the number of data points is not evenly divisible by n just average the last remaining points ed if xsum ne 0 0 and ysum ne 0 0 and zsum ne 0 0 then xavg xsum icount 1 yavg ysum icount 1 zavg zsum icount 1 write 21 60 xavg yavg zavg end if write 6 70 format 2x All done end 2 94 Spectrometry 1 100 AMU Data Processing FORTRAN Code i ts a Beliebte a C Filename SpecDatal00 f C Written by Brad Horn C Email horn8 illinois edu C Date Dec 20 2010 C Course Masters Thesis C Program SpecDatal00 f C Due date N A c E Status errors in compilation E compiles but does not perform partially c
32. 6 S P 13 Pico Pressure Figure F 27 S P 13 DIC Table F 7 S P 13 Spectroscop Sy HE gs O 3 8 O O NI Z Q mir T AE D O Ss Sib H Numerous scans of same percent compositions as 1st row baseline 14 13 13 12 11 72 72 72 73 74 74 75 75 74 74 73 73 72 70 64 55 45 13 19 27 15 37 45 26 24 85 S P 14 Propellant AP200um AP60 130um HTPB IPDI DOA Test 15 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Good Comments This was the last of the repeatability series that exhibited the same results as the others S P 14 Temperature vs Time 300 250 200 R 1 3 in 150 R 2 3 in Temperature C 100 R 3 3 in 50 0 200 400 600 800 1000 1200 1400 1600 Time Sec Figure F 28 S P 14 Temperature 86 Pressure Psig S P 14 DAQ Pressure vs Time 10 On 00 ESF e 3100 3150 3200 3250 3300 3350 3400 3450 3500 3550 3600 B 2 Time 5 Sec Figure F 29 S P 14 DAQ Pressure 620 0 600 0 580 0 560 0 540 0 520 0 500 0 480 0 460 0 05 00 05 1 0 15 20 25 30 35 40 45 3 Pico Technology www picotech com Figure F 30 S P 14 Pico Pressure Figure F 31 S P 14 DIC Table F 8 S P 14 Spectroscopy H CH CH OH NH HO N CO CHO O Ar CO 2 14 16 17 18 28 29 32 4
33. AP propellants A new design was implemented to satisfy the design requirements Figure 3 3 A Figure 3 3 Heating Element Version II Figure 3 4 View of Setup Version II Version II used 0 5 stainless steel square blocks as vertical posts The posts were drilled to hold 5 ceramic arms held bv set screws to create a tree structure of 5 lavers each consisting of 1 resistance wire looped twice around the center for the inside and outside diameters Each layer was connected in parallel to create a circuit of resistance 4 2 ohms that was applied a power of 50 volts yielding 12 amperes This heating version allowed for more uniformity in layer spacing and greater structural integrity However major flaws of limited space to thread the resistance wire increased the time to clean and rewire the setup As shown in Figure 3 4 the heating element required a large number of electrical connections Version III returned to the concept of vertical ceramic posts but machined with diamond tipped drill bits as shown in Figure 2 3 and Figure 2 4 Nichrome 0 01 thick resistance wire threaded through the holes completed two circuits each of 4 loops and resistance of 37 5 ohms Both circuits connected in parallel formed a resistance of 18 75 ohms that when applied a voltage of 90 volts yields nearly 4 25 amperes A final modification uses 0 02 diameter resistance wire in a single segment of 8 loops yielding 16 6 ohms that when applied 125 volts yield
34. C camera amp pressure cables to respective instruments c Power cables for camera pressure sensor TCIC card and light 10 Insert steel gas line from inside chamber through hole and connect both sides 11 Start up RGA Mass Spectrometer according to Mass Spee SOP 12 Start LabView Cookoff Control and test communication to all diagnostics 13 Make fine adjustments to camera position and light source for picture quality 14 Check that the proper SPID is filled in create an executable 0000000000 15 Allow spectrometer and LabView to run for at least 4 hours prior to testing Pre Test Set up Day of Testing 16 Clear lab of all non assisting personnel 0000000000 17 Alert a supervising faculty of test at least 1 hour prior to testing 18 Start PicoScope 6 on laptop with Ch A 500ms div 2KS single trigger 540mV rising with 10 pre trigger graph at X10 0 scale and 28 offset 19 Verify that all diagnostics are being read by LabView 6 signals 20 Start up QuickCam on adjacent computer with settings 320 X 240 Ds Ow E tS SO 54 9132 OS e dd dd a A tata 21 Record initial values and settings on Diagnostic Record P 22 Turn on exhaust fan close and secure all doors 23 Attach Do Not Enter sign to outside of door to MEL 1304 es Test Procedure 24 Abide to all general laboratory safety procedures 25 Begin recording data by computer via LabView wait 5 minutes et gt 26 Start ramp program
35. Cell Champaign IL 61821 1903 E Juniper Dr 208 Elmwood Rd Mahomet IL 61853 Urbana IL 61801 55 Diagnostic Record Sheet S P ID SP SCO TEST ____ S P Composition by Mass Date S P Mass w case w out Heated to Temperature Ignition Yes _ Nu Propellant Properties Mass of carbon fiber case g Total Mass of propellant sample g Mass of propellant inside case g Initial Values and Settings Barometric pressure in room inHg Room ambient Temperature CR Initial Thermocouple Properties Channel Color PR Temperature C Channel Color____ R___ Temperature S Channel Color R___ Temperature C Channel Color R___ Temperature C Heating Final Start time of recording AM PM Start time of heating AM PM Time of Ignition AM PM Stop Time of recording AM PM Barometric pressure in room inHg Room ambient Temperature C Assessment of Process include comments on violence of cookoff 56 Propellant Fabrication Standard Operating Procedure 5 Days prior to test Case Preparation Cut carbon fiber casing ID 75 to 2 long sections E Drill 4 holes with small bit evenly spacing half of the circumference Weigh and record each carbon fiber mass Clean and assemble casting stand apparatus and insert cases _ Tape completely around outside of each carbon fiber case and label Gy ee ES Pre Mixing 1 Determine propellant composition and total mass of propellan
36. E EAR a A es 84 HigureF 26 5 P13 PICO Press ube iii das a Enas e aa AREE RTT a aS 84 patio TEE 85 Figur F 28 RI Temperature sssini etere e 86 Figure BEE KEEN 87 Figure E 30 SP E Pic PESE A A 87 Figure P31 SP lA KEE 88 VII List of Tables Table 3 1 Preliminary Tests Overview cescicronanicna ii darias EE 29 Table 4 1 Repeatability Test Overview EN 31 Table 4 2 Temperature Standard Deviations in Degrees Celsius oooconoccnnononoconoccnonacononcnononannnonnno 32 Table 4 3 Temperature Percent Differences 2 cic ssccsnveesjecctsvascassleyiadaasustae susngosata Eed deed 32 Table 4 4 Pressure Standard Deviations in Dei 32 Table 4 5 Spectroscopy Standard Deviations in Percent Composttons 33 Table 4 6 Cookoff Response of Modified SSCB AP HTPB 88 12 Propellant 13 34 Table 4 7 S P 25 Slow Cookoff Overview EE 36 Table 4 8 S P 25 Spectroscopy minicadena ii E i 37 Table E 1 Time Table amp Expense Summary i 61 Table Es2 Parts List ee 62 Table F 1 S P 10 Spectroscopy iii a a ea 70 Table P 25 P 9 PoCo COPY EE 72 Table EE EE 74 Jaeger 76 Table Ban e REENEN geen a GEN 79 Table F 6 Ee e ET 82 AS Spe itostop A E EA EE E EA 85 Table BSP TA Spectroscopy ni a ado 88 VII Al AN AP DIC DOA HCI HTPB IM IPDI IPSTP QMS SNL SOP S P SSCB VCCT VI List of Terms and Svmbols Aluminum Ammonium Nitrate oxidizer Ammonium Perchlorate oxidizer Digital Image Correlation Dioctyle Adipate plasticizer
37. Experimental and Computational Investigation of Methane Air Partial Oxidation Tech Rep UMI 3223647 PhD Dissertation Department of Mechanical Engineering University of Illinois at Urbana Champaign 2006 44 17 18 19 20 21 22 23 24 25 26 27 Lengelle G J Duterque and J F Trubert Combustion of Solid Propellants Tech Ref RTO EN 023 Proc of RTO VKI Special Course on Internal Aerodynamics in Solid Rocket Propulsion in Rhode Saint Genese Belgium May 27 31 2002 Chatillon Cedex France ONERA 2004 Mahanta Abhay K Monika Goyal Pathak Devendra D Rheokinetic Analysis of Hydroxy Terminated Polybutadiene Based Solid Propellant Slurry ISSN 0973 4945 E Journal of Chemistry vol 7 1 pp 171 179 2010 McBride Bonnie J Gordon Sanford Computer Program for Calculation of Complex Chemical Equilibrium Compositions and Applications II User s Manual and Program Description Tech Ref NASA RP 1311 June 1996 lt http www grc nasa gov W W W CEA Web RP 1311 P2 htm gt Optics of Spectroscopy Tutorial HORIBA HORIBA Ltd 1996 2011 lt http www horiba com us en scientific products optics tutorial gt Rattanapote M K Atwood A I Towne S E Johnson S S Buffum M C Wilson J E Brady V L A Fast Cookoff Assessment of U S Navy Legacy Weapons Proc of Insensitive Munitions amp Energetic Materials Symposium San Francisco Ca Nov 15
38. G WEE TE 5 154 PLOMO a A A A A 8 2 Experimental Equipment and Process ee Ee 10 Deb ER e EE 10 2 2 Heating Ple met ae 11 E E KEE 13 231 Ku cin dais 14 230239 PrESSUTE EE 15 RE A pea eae ease p ea sk BEO weve jat Ba PASSES a e a ana EnA 16 LIA Spectroscopie riras R AER e a A a AOU fa az re 17 24 Optical e PON A 18 O e o A O 20 3 Preliminary a ne aa 22 3 1 Origina Apparatus tisjir Ta a a 22 3 2 Heating Element Versions usina i e a a ioa 22 3 3 Diagnostic Implementation site 24 Say Mee Programming ele 27 3 5 Initial Test Results iii insistiet a se GA eda ug 28 4 AP Propellant Results and DISCUSSION cccsccessseceesseceeseeceeseeceeacecseceeceeceeceeeeeceeeeeseeeeeees 31 Bide Repeatability Resu E 31 4 2 Slow C okotf Results ai dance 35 5 Conclusion and Recommendations ccioiesisiescvcdscaseresnessavessaasnesdncosvadaconsegedasacenssdeensdaunnceonna 40 Dek e ET Ee DEE 40 E 42 List OP RETCRE A Et 44 ee EE 46 A Typical Ingredients of Composite Solid Propellants sms nen en nnnznn 46 B LabView Desi ii A A A setae 47 C Data Analysis PrO e as 49 D Solid Propellant Slow Cook off Standard Operating Procedure ooooooccccoccccnonccononccnnnnnnnnno 54 E Time EX penses it Parts Listivaninini iia dedicacion 61 F Eookort Data i sine hc ta i ia 63 G CEA SRESU Sl e ki fa ed 89 H Data Processing Codes urna e g a A 93 List of Figures Figure lol Stages pr Cookoff lis ts 4 Figure 1 2 Thermal Runaway Ph
39. IC iii ini ia ub idi 27 Figure 5 8 CCD Image of SP 2 a EE 29 fisure 3 9 CCD Image of Po a al cd ds 29 Figure 3 10 5 P 1 Data AN ia i ii ib a d a 30 Figure 4 1 Repeatability R 3 3 Temperature Plot iii Senne 34 Figure 4 2 Repeatability DAQ Pressure Plot lis 35 Figure 4 3 S P 25 Temperature afin eege e A Ei 36 Pisure 44S WEE 37 Figure 455 A A O 39 Proure ASIA 13 U e At A it 39 Figure B 1 Cookoff Control VI Front Panel accion de 47 Figure B 2 Cookoff Control VI Block Diagram sintio ii 48 Figure C 1 Amira S P Image Processing s s ba a Ni aaleaace eee vd le sa ena 50 fiSUre 12 A Mra EE 53 Fig re E a ita sel itti Soe e e A haha iaa 63 Figur P 2 S P 1 3 2DAQ Pressures ini illa 64 are De SP LEMA a a aeS 65 Eigure P 4 S P 3 Temperaturas 66 Figure E S SP DIO rias 66 fi ire F 6 5 P 4 Temperature it A e E 67 Figure Ra SPA DAQ Eege 68 figure E EECH 68 Figure FO SP EE ee 69 figure 10 SR 10 DIC reacia a di 69 Fig re ENEE 71 Figure RE E Terre iksi Sistina ei EEN 73 rer SES DAQ Press ane ee EE EEA 74 Figure EA ge 75 Figure F 15 5 P DEO tad 75 Fig re F 16 S P 6 ENEE 77 figure F 17 5 P 6 DAQ Pressure tt ane lee ate ee 78 Freure ELS 5 P O Pico Pressure Ee ee 78 Lisure EOS PO DIC ie Ai fn 79 Figure F 20 S P 12 Temperatura i sE Ei aiai 80 Figure FZS P12 DAQ Pressure td eet 81 fi ureF 22 Pico Presup A a tt 81 Figure 23 PAD i i b NEEN 82 figure F 24 SP13 Temp ratute cn ensiesitys soia 83 Figure E 25S P13 DAQ Press Ure xiv A E
40. RS QMS200 Mass Spectrometer Operating Precautions Due to the large quantities of water seen in the combustion environment it is important that upon finalizing testing a low pressure slightly greater than stp supply of nitrogen be purged into the sampling lines and to the inlet capillary to minimize the water in the sampling lines as well as the amount of water seen by the mass spectrometer pumping systems According to the Stanford Research Systems the time frame for dry gas exposure is somewhat variable but should be done to alleviate any large quantities of water vapor in the system exponential decay and thus the largest percentage the water is evacuated in a short time period Water does not detrimentally affect the ionizing filament However prolonged exposure of corrosive gases and water vapor to both the diaphragm and turbo molecular pumps can cause early failure The Mass Spectrometer is designed to operate continuously but the lifetime of the pumps can be increased if the ultrahigh vacuum UHV environment of the RGA is minimized during times of idle or limited use See proper idle and shutdown procedures It is important however to purge the svstem with a drv gas as outlined above Due to possible condensation in the sampling lines a blow out procedure should be done after each dav of testing This is done bv removing the sampling line closest to the mass spectrometer With this part detached the other end of the line can be detac
41. SOLID PROPELLANT SLOW COOKOFF EXPERIMENTS WITH ADVANCED DIAGNOSTICS BY BRADLEY J HORN THESIS Submitted in partial fulfillment of the requirements for the degree of Master of Science in Mechanical Engineering in the Graduate College of the University of Illinois at Urbana Champaign 2011 Urbana Illinois Advisor Professor Nick Glumac Abstract Slow cookoff of solid propellant was experimentally studied at the University of Illinois by examining the behavior throughout the cookoff process The general concept involved preparing axial symmetric samples of chosen compositions and casting them into carbon fiber cases The cured samples were then placed inside the test apparatus to slowly heat until auto ignition while continuously recording data used to help validate a complex thermal chemical mechanical code being built to model the cookoff behavior of energetic materials Preliminary work focused on developing the test apparatus and successful implementation of all diagnostics Experimental data taken for the scope of this research included three thermocouple readings of varying radius for spatial temperature gradients pressure and spectrometry measurements on decomposition and combustion gases and images of the surface of the carbon fiber case used in digital image correlation to yield mechanical displacement Single base ammonium nitrate propellant was the first propellant composition used in developing the test apparatus Follow
42. Swagelok gas filter Figure 2 10 followed by a stainless steel capillary tube manufactured by Stanford Research Systems Inc to reduce the pressure 3 decades down to below 5 mbar or 500 Pascals Stainless steel gas lines then connect through the back of the sound enclosure to the Stanford Research Systems 200 Quadrupole Mass Spectrometer A plastic exhaust line runs from the spectrometer back into the sound enclosure and partway into the exhaust vent in the ceiling of the chamber The spectrometer is capable of measuring from 0 to 200 amu with resolution of 100 ppm Since the capillary tube is near the bottom port temporal resolution is achieved without significant through flow from the sample A scan from 0 100 amu is completed in 30 seconds The residual gas analysis identifies the composition of the combustion products such as H2 OH 17 CH N2 O2 NH3 CO or CO2 However this analysis requires stable species as the time resolution is relatively poor when compared to absorption spectroscopy The SRS spectrometer 1s shown in Figure 2 14 Figure 2 14 SRS 200 Quadrupole Mass Spectrometer 2 4 Optical Absorption The optical absorption is a key diagnostic to either validate mass spectrometry results or explore species not documented with the mass spectrometer There are two possible methods of conducting optical absorption on the core gases as a validation of the mass spectroscopy As previously mentioned a source can apply light through
43. a fft2 al NFFI NFFI 32D FFT fftb fft2 b2 NFFI NFFI 32D FFT cci real ifft2 ffta fftb tCorralation Map Find the Peak peak max cci temp dv temp dx find cci peak f Find the displacement dx size ccl 2 2 temp dx 2X lens calibration dy temp dy size cci 1 2 2X lens calibration dxmm dx 0 0082682292 old 1X calibration 3dymm dy 0 0082662292 old 1X calibration dxmm dx 0 0084623847 dymm dy 0 0084623847 X j itw 2 1 Y Lritw 2 1 displacement ind X Y dx dy damm dymm ind ind 1 end end number sprintf 2305d k save filename displacement 3 Plot the Results figure Visibie off imshow A hold on Plot the vectors on top of the image quiver displacement 1 displacement 2 displacement 5 displacement 6 3 Command below will save a file as fig must change imagename to fig 3hgsave imagename print dtiff imagename 100
44. al degradation on the propellant performance Since the 1990 s Sandia National Laboratories SNL have been developing a complex model to simulate the cookoff environment First a complex code was built to model solid propellant for defined conditions during the final stages of cookoff 3 The code used shock physics to model the stages of cookoff between deflagration to detonation DTD in porous energetic materials Results indicated that the percent porosity greatly affected the characterization of the violence of cookoff as well as the time scale of the event The time scale from ignition to the end of the event ranged from 16 90 us depending on the percent porosity 5 The conclusions from the research were that a lot more understanding was needed on the state of thermally conditioned material prior to ignition in order to better understand the final stages of cookoff Later at SNL a finite element analysis code was developed as of August 1994 2 A combination of already existing codes COYOTE II CHEMEQ XCHEM 1D and CHAPARRAL were modified to implement heat conduction and decomposition multistep rate equations thermal chemistry and radiation computing respectively According to the paper the mechanical aspect of the problem other than simple material addition or deletion was not yet included due to limitations in finite element multidimensional solving methods At the time of this project computing limits were a large factor in all
45. ce is complete IMPORTANT to make any modifications to the mass spectrometer e g changing capillarv tubes remember to first deactivate the ionizer by pressing on the filament button to the off stop position Otherwise the filament could burnout Pumps Only Procedure State 2B This mode prepares the system but does not draw any sample gas The RGA will operate but only background will be shown Takes several minutes to start pumps lights will be bright green when finished Startup Sequence e Turn on only Mechanical Pump and Turbo Pump switches state 2B 58 Continuous Sampling State 4 from 2B Turn off filament before startup sequence and alternate method Start up sequence e Turn off the turbo pump Switch wait for light to extinguish e Turn on in order the Capillary Flow Valve Turbo Pump and Sample Inlet Valve Alternate method Turn on Capillary Flow Valve then Sample Inlet Valve switches This requires a slow pulse release action to eliminate trapped gas and is slightly slower than above A normal clicking sound will be heard This is not the preferred path Idling Procedure States 3 amp 2B When not sampling idling states minimize the load on the pumps and extend their life Idling allows the background to be pumped out more efficiently System is designed to run 24 hours Procedure e Short Periods State 3 o Close Sample Inlet Valve Important to eliminate water e Long Periods State 2B o Close both Valv
46. cess 3 1 Original Apparatus The original experimental setup was mounted to a steel table as pictured in Figure 3 1 and in Figure 3 2 The core components were slightly modified for improved thermocouple placement and attachment for the new heating element The setup was mounted to a small two foot square optics board to eliminate component vibrations and allow mounting of other diagnostics and shielding Proper shielding was implemented in order to protect the instruments and facilities Sheets of 1 8 thick steel mounted to L brackets and bolted down to the optics board serve as shields The shield protecting the camera was drilled for optical access and protected with 1 4 thick high temperature quartz glass and 1 2 thick clear plastic The adjacent shield was drilled for access to the pressure transducer CCU CAMERA LONG W 0 OBJECTIVE PRESSURE IRANSDUCER Figure 3 1 Original Experimental Setup A Figure 3 2 View Without Heat Shield 3 2 Heating Element Versions The original heating element was insufficient for reliable tests Nichrome 0 02 thick resistance wire wrapped around ceramic posts in series composed the original heating element version I The power source for beginning tests used a Variac and a constant voltage setting During heating the wires became flexible and thus provided heating that was inconsistent and 22 non uniform The initial version was also incapable of reaching temperatures required for
47. cles are pure white and the background is completely black 7 Select Particles e Right click remapped green file labeling label field e Check 2D amp 3D of Interior on top left e Check All slices in bottom left and select magic wand e Click on a particle e Adjust graphical selection until all particles are selected e Scroll through all slices to check particles on first and last slices may not be selected e Click the button 8 Cast label field e Right click green label field compute cast field 51 10 11 12 13 14 15 16 17 18 e Select label field click apply Distance Map e Right click green label field compute distance map e Set Chamfer Weights property to float click Apply Equalize e Right click file image filter equalize e Select 3D set contrast limit to 12 click apply Gaussian Smoothing e Right click file image filter Gaussian smoothing e Select 3D click apply Label field e Right click file labeling label field e Select all particles in same manner as before e Be sure not to select background while getting most particles Arithmetic e Right click file compute arithmetic e Enter A B e Click white box of arithmetic icon and set A to the label field and B to the smoothed data used to create the label field e Click apply Arithmetic e Right click result compute arithmetic e Enter A MaxValue into expression where A is the result and the max value is found by clicking on the r
48. d be noted that a scan takes 30 60 seconds to complete and depending on when during the scan cookoff occurs greatly affects which atomic mass units change the most during that scan and as a result the standard deviations Table 4 5 Spectroscopy Standard Deviations in Percent Compositions Scan Avg Std Dev Amu Avg Std Dev 1 0 067 2 0 918 2 0 039 12 0 038 3 0 051 14 0 118 4 0 051 15 0 044 5 0 054 16 0 087 6 0 054 17 0 032 7 0 054 18 0 120 8 0 076 27 0 044 9 0 081 28 1 066 10 0 111 29 0 050 11 0 204 32 0 617 12 0 193 40 0 014 13 0 194 41 0 023 14 0 181 44 0 547 15 0 223 50 0 023 16 0 194 17 0 198 18 0 206 19 0 188 20 0 168 21 0 227 22 0 418 23 0 270 Ignition 1 459 23 1 276 Validation of the spectrometer results was performed using the Gordon McBride Equilibrium Solver 19 The repeatability experiments were used to simulate the environment with the T P solver and with the correct composition See Appendix G for CEA input and output files The predicted products according to the solver are CH4 CO CO2 NH3 HCI H2 H20 and N2 Virtually all expected products are found in the mass spectrometer except HCl and the correct amount of H20 The reason is that these species condense through the gas lines 33 As discussed in Chapter 3 the surface imaging did not yield surface displacement results with the DIC code The displacement found was due to small oscillation in the field of view and the thermal
49. d not clog the bottom sampling port and prevent pressure and mass spectrometry data Temp C Pressure psi 600 500 400 300 200 S P 1 3 2 Temperature amp Pressure a D we e Pressure a Temp R 2 3 0064 1190 1200 1210 1220 1230 Time seconds Figure 3 10 S P 1 Data AN 30 4 AP Propellant Results and Discussion Following the preliminarv tests and successful implementation of the initial diagnostics the next step was establishing repeatabilitv of the diagnostics and experimental operation For this task a series of three tests were performed on AP propellant with identical composition The composition was AP200um AP60 130um HTPB IPDIDIDOA with mass percents of 54 56 33 44 9 29 0 71 2 00 respectively which follows the 88 AP and 12 Binder composition used in previous research 4 1 Repeatability Results For the test series each propellant sample was labeled with a number and weighed prior to testing All samples were heated by a constant heat source by having the Variac at 90 volts for twenty minutes and then raised to 95 volts until ignition The reason for the increase was due to a plateau effect in the heating rate of the first sample which required more power As a result the next two samples were heated identically In Table 4 1 each S P propellant is shown with its total propellant mass excluding the carbon fiber case Experimental data included is the ignition temperature cookoff time in seconds and minute
50. d to different annulus sizes and allows simultaneous casting of up to 9 samples The stand is placed inside a vacuum oven at 60 C and applied a slight vacuum for 4 days During this process the propellant slurry cures to a solid with the consistency similar to a small particle Rice Crispy Propellants from the same batches used in experiments were scanned and processed for quantitative analysis of the propellant structure prior to cookoff See Appendix C for more information The choice of propellants and compositions was ultimately dictated by the need of the research sponsors Initiallv AP HTPB propellant without metal additive was used because of the ease of manufacturing and propellants without metal additive are easier to model and thus served as the starting point for the experimental test matrix However it should be noted that most AP 20 modern propellants are commonly aluminized and thus future experiments should include Al powder into the composition Double base propellants were eliminated from the scope of the research due to manufacturing limitations and for future experiments will have to be contracted out for the propellant manufacturing 21 3 Preliminary Work The initial setup was designed for a proof of concept and needed improvements before reliable experiments were to be conducted A full year was dedicated to improving the test apparatus and verifying consistent performance The following chapter documents this pro
51. ded corrections However for this research the outer case surface is imaged for material deformation A hole is precut into the heat shield to allow a small optical access from the side to provide on axis measurements A CCD camera with micron level precision and frame rates up to 30 Hz is used to 16 acquire images However due to memory limitations and a long time scale images are acquired only every 2 minutes during slow cookoff Using digital image correlation DIC the sequential images are used to give material displacement as a function of time DIC is a complicated process that requires a speckle pattern in order to measure changes in speckle location as time progresses Commercial DIC codes are on the order of twelve thousand dollars but open source codes do exist For our project a small Matlab code was used See Appendix C for more information on the data analysis procedure The CCD camera and protective shielding are shown in Figure 2 12 The field of view of the CCD camera is 752 by 480 pixels and 6 36 by 4 06 millimeters A simple conversion yields 118 17 pixels mm for the DIC calibration Figure 2 12 CCD Camera and Shielding l Figure 2 13 CCD Camera 0 5 mm Markings 2 3 4 Spectroscopy Gas analysis provides a critical validation of the chemical portion of the computational modeling Mechanical sampling was implemented to perform residual gas analysis Connected via gas lines to the bottom port is a 0 5 micron 1 8
52. e heating rate can be set to any desirable rate and be within the same range of variance The slow cookoff test presented new and interesting features in the result The first new feature is that the slow cookoff ignition temperature was 359 17 C This ignition temperature was nearly identical to that of S P 13 and S P 14 in the repeatability series differing by 5 C This value corresponds closely to those of fast cookoff ignition temperatures presented in Table 4 6 13 The second new feature was the results observed by the mass spectrometer As shown in Chapter 4 2 significant decomposition of the propellant sample occurred in the temperatures of 225 231 C The decomposition was relatively slow producing negligible amounts of pressure according to the pressure diagnostic In the later stages of cookoff little change was observed in the core gases composition unlike the intermediary heating rate experiments During the repeatability series experiments the last two scans exhibited the largest change in composition These two features are likely linked together and a result of the main design difference between the experimental apparatus and those discussed in Chapter 1 3 By having mechanical gas sampling constantly draw gases from the annulus core the region is not allowed to slowly store decomposition gases and buildup pressure Increased pressure has been linked to faster ignition times and thus lower ignition temperatures As a result of this p
53. e DIC Plots Folder mkdir pathDir slash DICPlots PlotFol pathDir slash DICPlots Interrogation for window size itw 64 SGives the spacial resolution in pixels same size in X and Y directions ovl itw 2 Amount of overlap of the ITW s increases the spacial resolution NEFI itw 2 Size of the FFT Loop over all of the images Cl for k 1 1 length files 1 Depending on sampling rate change increments imageA pathDir slash files k name imageB pathDir slash files k 1 name change increment for correct correlation pathstr name ext versn fileparts imageB Filename ResultFol slash name MAT imagename PlotFol slash name tiff Open the images pair A imreadiimageA B imread imageB 3 Unmark this if images are RGB A rob2gray A B rgb2gray B displacement ind 1 Counter for i 1 ovl size A l itw for j 1 0v1l size A 2 itw extract the image in the interrogation spot al A i i itw 1 3 3 itw 1 bi B i ititw 1 j j itw 1 Change to double precision al double al bi double bi 99 48 43 50 51 52 53 54 55 56 57 58 53 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 EN 82 end Get rid of the mean for each window al al mean al bi bi mean2 b1 b2 bi end 1 1 end 1 1 SRotate the second complex conjugate Perform the cross correlation calculation fft
54. e discussion For the imaging diagnostic the setup has been proven operational to resolve spatial displacement and movement for inert Teflon samples that correspond to thermal expansion degradation The repeatability experiments didn t exhibit results and following a 2X magnification upgrade the DIC results still did not show movement greater than the thermal expansion of the carbon fiber case or bulk sample movement It is reasonably assumed that due to the annulus shape and rigid carbon fiber containment the solid propellant mechanically deforms and expands internally without affecting the containment walls except during the rapid 40 combustion stage of cookoff Without DIC results the mechanical aspect of the modeling code cannot be verified Single base propellants of varying composition and geometry can consistently and efficiently be manufactured in house at the University However the attempt to manufacture double base solid propellants failed and indicated that future work with more complex propellants will require outside manufactured propellants Current work is being done to negotiate the contracted samples Propellant can be scanned and characterized structurally by image processing in an effective manner helpful to the computational modeling group Following the repeatability tests the apparatus was modified to achieve slow cookoff characteristics The heating rate is well within the 0 5 degree range of 6 C hr In fact th
55. e sample in line with the center Four axially spaced bolts apply pressure to form seals Holes through the bottom plates allow access for the thermocouples Seated inside the top plate is a quartz window for optical access inside the annulus of the propellant sample Seated inside the bottom plate is a polished reflective stainless steel piece with a gas sampling port The reflective surface serves as a mirror to bounce off light from an off axis source that can be collected to perform absorption spectroscopy The sampling port connects to the pressure transducer and the capillary tube used for the gas analysis Figure 2 1 depicts the general schematic Detector Camera or Spectrograph Light Source Laser flashlamp IR Top Plate Lamps Bottom Plate b Sample for Reflective SS composition amp surface pres sure Thermocouple Array Figure 2 1 General Schematic 11 The bottom plate is mounted on optical post assemblies secured to an optics board The optics board provides a rigid surface to attach all shielding diagnostics and wiring shown in 10 Figure 2 2 The optics board is then placed inside a sound enclosure and strapped down to an existing structure present for unrelated experiments The sound enclosure is capable of handling up to 15 grams of explosives Attached to the top of the chamber is an exhaust fan for proper ventilation of combustion gases All connections extend outside the chamber to a w
56. ed photonic based camera system is required greatly increasing the diagnostic price Typical units are on the order of 75k The proposed design in Figure 2 15 is currently being pursued M1 Czerny Turner Monochromator M2 Diffraction Grating Camera Detector Pressure amp Mass Spec Heat Source Figure 2 15 IR Optical Absorption Diagram 19 2 5 Propellant Preparation Our research group has had extensive experience casting solid propellant in house For initial tests AN propellant was used with composition by mass of 70 AN 27 858 HTPB and 2 142 IPDI Once the apparatus was verified functional AP propellant used in previous research was duplicated with composition of 88 AP 62 200 micron 38 60 130 micron and 12 binder The binder was comprised of 77 4 HTPB 16 7 DOA and 5 9 IPDI The solid propellant was hand mixed according to the Propellant Fabrication SOP found in Appendix D Remote controlled automated mixing is expensive and sometimes can be more dangerous than by hand No significant difference in propellant structure exists between mixing methods Figure 2 16 depicts the propellant slurry after mixing in a plastic bag Figure 2 17 Propellant Casted in Stand Figure 2 16 Propellant Slurry Following careful mixing the propellant is quickly cast into carbon fiber cases in the manufactured propellant stand Figure 2 17 to form the annulus shape of the propellants The stand can be adjuste
57. ei Y VISAResourceName E A Partial Pressures Out ml Cm H Ni i Ge gi pu EEK f Partial Pressures Out lo ei ft Complete Scan Scan Figure B 2 Cookoff Control VI Block Diagram 48 C Data Analysis Procedure All data needed some form of post processing before it could be easily graphed or qualitatively analyzed Sampling rates combined with the test duration time made very large data files Although the original data is still useful to the computing group a smaller averaged data file was created in order to easily plot and analyze Simple FORTRAN files were created in order to smooth out and shorten the temperature and pressure data files with a moving average The codes take the name of the input file name of the output file and number of points to average as the user input The result is a smoothed lower sampling rate output file that can easily be plotted See Appendix H for the FORTRAN code used for the temperature files A similar code that differed in the number of columns read was used for the pressure data The pressure data from the Picoscope was also processed with the FORTRAN code However this data was only smoothed every ten data points in order to not distort the dynamic pressure data Another FORTRAN code was created to process the spectrometer data Originally the spectrometer took data from 1 to 100 amu at intervals of 0 1 amu per scan This created a large text f
58. elta ge 0 1 THEN WRITE OutFileUnit 50 Plot AMU i ENDIF ENDDO For each scan sum up the pressures of positive partial pressures DO i 1 NumScans DO j 1 100 IF AMU i j gt 0 0 THEN SumP 1 SumP i FAMU i j ENDIF ENDDO ENDDO Process data to get percentage of total pressure for each AMU If percentage of total pressure is negative disregard DO i 1 NumScans DO j 1 100 PercentAMU i j AMU i j SumP i 100 0 IF PercentAMU i 3 1t 0 0 THEN PercentAMU i j 0 0 ENDIF ENDDO ENDDO WRITE TO FILE SECTION Write data to output file WRITE OutFileUnit 60 Time AMU i i 1 100 Total Ei DO i 1 NumScans WRITE OutFileUnit 40 Datemm i Datedd i Datevvvv i Timehh i Timemm i Timess i Timep i AMU i j j 1 100 TotalP i ENDDO Write percentage of SumP data for all data WRITE OutFileUnit y WRITE OutFileUnit 65 Time SAMU i i 1 100 Sum of Ei DO i 1 NumScans WRITE OutFileUnit 45 Datemm i Datedd i Datevvvv i Timehh i Timemm i Timess i Timep i PercentAMU i j j71 100 SumP i ENDDO Close files if the CLOSE fails go to the error handling section CLOSE UNIT InFileUnit ERR 930 CLOSE UNIT OutFileUnit ERR 930 Tell user that program is finished WRITE 6 100 97 B FORMAT 2x inputfile A60 2x outputfile A60 il r2x Number of Scans 12 0 FORMAT 12 1 121 14 1X 12 1 12 1X 12 1X
59. erchlorate KN Potassium nitrate NP Nitronium perchlorate Metal Fuels also Al Aluminum act as a combustion 0 30 Be Beryllium stabilizer Zr Zirconium modifies burn rate HTPB Hydroxyl terminated polybutadiene Fuel Binder 5 18 CTPB Carboxvl terminated polybutadiene colybutadiene type PBAN Polybutadiene acrylonitrile acrylic acid PBAA Polybutadiene acrylic acid MAPO Methyl aziridinyl phosphine oxide l IPDI Isophorone diisocvanate eee TDI Toluen 2 4 diisocyanate OS 1 3 5 HMDI Hexamethyl diisocyanide react with polymer Ge binder DDI Dimeryl diisocyanate TMP Trimethylol propane BITA Trimesoyl I 2 ethyl azidrine Gab HMX Cyclotetramethylenetetranitramine Explosive fillers SR solid 0 40 RDX Cyclotrimethylenetrinitramine NO Nitroguanadine DOP Dioctyl phthalate Plasticizer Pot life DOA Dioctyl adipate control organic 0 7 DOS Dioctyle sebacate liquid DMP Dimethyl phthalate IDP Isodecyl pelargonate GAP glycidyl azide polymer l NG Nitroglycerine GE DER DEGDN Diethylene glycol dinitrate liquid BTTN Butanitriol trinitrate TEGDN Triethylene glycol dinitrate TMETN Trimethanolethane trinitrate Excerpt from Heng Hok s adaptation of Rocket Propulsion Elements 6 Ed 14 denotes experimental propellant only 46 B LabView Design It should be noted that the LabView virtual instrument was programmed in sections some of which started from supplied example versions A sample virtual instrument was supplied with
60. es Capillary and Sample Lowest load on the pumps Shutdown Procedure Do not store with diaphragm pump under vacuum for long periods See Manual for negative effects and correction procedures Always turn off filament before shutting down Procedure e Turn off switches in reverse order of startup o Sample Inlet Valve o Turbo Pump o Capillary Flow Valve o Mechanical Pump Short Periods less than 30 minutes stores a vacuum pressure e Turn off Main Power switch Long Periods more than 30 minutes State 1 Venting e SLOW While the turbo pump is still coasting to a stop open the capillary flow valve with the capillary still connected to the inlet Allows slow venting e FAST Wait for the turbo pump to coast to a stop Remove the capillary and open the capillary flow valve Useful when filling the system with a dry gas for storage e FASTEST Not recommended When the turbo pump is still coasting to a stop remove the capillary and open the capillary flow valve Do not vent the system to pressures above atmospheric If the capillary is connected to a gas at pressures above 1 bar monitor the pressure during venting Stop just as the pressure reaches atmospheric Overpressure Protection The RGA has a built in protection to turn off filament if pressure is too high however this should not be relied on to turn off the filament When shutting down the system turn off the QMS filament before the turbo pump is turned off 59 S
61. essing in Amira 1 Open Data e Click Open Data open the file e Itis the reconstruction txm file usually the largest and around 5GB e Read complete volume into memory will take a couple minutes 50 2 Display orthoslice e Right click on green file click orthoslice e In Properties box in bottom left note and record the voxel size pixels to microns 3 Crop image e With orthoslice shown click on green file in properties box click crop button e Drag green perimeter box down to selected region e Crop down to only solid propellant e Crop away all edges of propellant where distortion lines appear e Scroll through all slices in orthoslice repeat crop if needed e Change to xz view in orthoslice and crop in z direction for a square shape 4 Gaussian Smoothing e Right click green file image filter Gaussian smoothing xy plane apply 5 View Labelfield e Right click smoothed green file labeling label field e On left side it will show range of values in a graph e Observe and note the range usually around 8000 15000 e Delete the label field 6 Intensity Remapping threshold e Right click smoothed green file image filters intensity remapping e Set min and max to 0 and 65000 respectively e Set alpha to about 500 this is the value left and right of the beta that will be remapped e Set beta to about middle value of the range noted above e Click apply display orthoslice on result e Adjust values and redo intensity remapping until parti
62. esult and looking at the highest number in the properties box e Click apply Save e Save green result e Ifnext step crashes Amir repeatedly crop result before continuing Watershed e Right click result labeling watershed segmentation e Set Threshold to 3 98846 amp depth to 11 359 e Set output to Unsigned integer 32 bit click apply Cast Field e Right click watershed result compute cast field e Select 16 bit unsigned integer click apply Display Orthoslice e Right click result orthoslice e Select mapping type to color map 52 e Select edit options load color map and load the provided color map 19 Redo Watershed e Adjust watershed values and repeat until desired result all particles are presented in different colors Figure C 2 e Balance between particles being split and separate particles being joined see picture below e Reapply cast field 20 Shape Analysis e Right click green file measure shape analysis click apply 21 Measure Volume e Use 3D ruler in top menu and measure all 3 dimensions e Record dimensions 22 Save the data as a CSV file units are in microns e Send CSV file voxel size and scan dimensions to computational group A i Figure C 2 Amira Particles Lastly all raw and processed was imported to an excel file for each experiment The file contained a cover page with all relevant information about the solid propellant and testing parameters Only the results from the summary DIC code were
63. expansion of the carbon fiber case Figure 4 1 and Figure 4 2 show the temperature and pressure of the repeatability test series respectively The temperature plot shows the thermocouple placement at 3 3 of the total annulus radius which is near the inner wall As can be seen the temperature history is very similar except during the thermal runaway stage leading to ignition Temperature of R 3 3 In 320 jA 270 LA 220 gt S P 12 170 gt S P 13 120 Temperature Celsius S P 14 70 20 o 200 400 600 800 1000 1200 1400 1600 Time Seconds Figure 4 1 Repeatability R 3 3 Temperature Plot The repeatability experiments were of an intermediary heating rate between slow and fast cookoff characteristics As seen from Table 4 6 the ignition temperatures for the HTPB AP 12 88 2 0 DOA correspond to the composition used for the repeatability series 13 With a time to ignition of 25 minutes and ignition temperatures in the range of 325 355 C the test series corresponds closest to fast cookoff characteristics when compared to Table 4 6 Table 4 6 Cookoff Response of Modified SSCB AP HTPB 88 12 Propellant 13 DOA Heating Rate Temperature C Time min Response 2 Fast 336 10 25 Deflagration Slow 253 122 8 Deflagration Explosion 4 8 Fast 347 9 58 Deflagration Slow 254 374 7 Deflagration 6 5 Fast 332 10 25 Deflagration Slow 146 6 Explosion Table 4 6 cont 8 Fast 353 11 1 Deflagratio
64. extension wires FHS 2 Finned Heat Sink 1 2 deg C W 19 00 Omega Engineering Inc SSR mount OSTW CC E F Type E Female glass filled connector cable clamp 4 10 Omega Engineering Inc Female T couple connector OSTW CC E M Type E Male glass filled connector cable clamp 2 95 Omega Engineering Inc Male T couple connector OST CC Tool Assembiv tool holding fixture for OST series 50 00 Omega Engineering Inc Tool assemble T couple connectors CN7523 1 32 DIN Controller 97 00 Omega Engineering Inc Temperature controller CN amp 485 USB 1 Mini Node Com Signal Converter 99 00 Omega Engineering Inc Connector T controller OCW 3 Extended 3 vr warrantv 25 00 Omega Engineering Inc Warrantv T controller C106 0X Ammonium Nitrate 1 lb 4 60 Firefox Enterprises Inc AN C165B R45 M HTPB 1 gal 50 60 Firefox Enterprises Inc HTPB C150C ORM IPDI Isophorone Diisocyanate 1 qt 33 60 Firefox Enterprises Inc IPDI C108 OX 1 Ib Ammonium Perchlorate granular 200 micron 8 60 Firefox Enterprises Inc AP large particle C109B OX 1 Ib Ammonium Perchlorate xfine 60 130 micron 10 80 Firefox Enterprises Inc AP small particle C1461 1 qt Dioctyl Adipate DOA liquid Plasticizer 13 15 Firefox Enterprises Inc DOA 352278 UtiliTechPro 260 Lumen LED MR16 GUI10 38 flood 24 98 Lowe s Bulb for off axis light source 62 F Cookoff Data In the graphs below that all data was processed and smoothed down to a smaller data set for presentation purposes As a result actual ig
65. for T controller and record time on Diagnostic Record tea 27 At 200 C or 75 of ignition Temperature begin video recording via web cam 28 Observe until desired temperature or ignition never leave unattended 29 Turn off power to wires and stop ramp program immediatelv AN 30 Allow LabView to run and record for 5 minutes see gt 31 Stop web cam recording and LabView saving ss L Post test 32 Check for proper saving of all data and shut down LabView turn off filament 33 When saving the webcam video be sure to trim the video to 30 seconds 34 Shut down Mass Spectrometer according to Mass Spec SOP e 35 Wait for 10 min from end of test for gases to vent out of chamber r 36 Vacuum out chamber while waiting for setup to reach ambient temperature 37 Apply bottled Nitrogen to gas lines for 10 min to flush see Mass Spec SOP Ss 38 Disconnect T couples power cables gas line and all BNC connections 39 Remove testing optics plate from chamber 0000000000 40 All parts should be thoroughly inspected cleaned with ethyl alcohol and steel wool and replaced if nesded 1 A AJ In Case of Fire 1 Call 911 and activate fire alarm 2 Fire Extinguishers are located in the hallways 3 Contact Dr Glumac and or Krier NOTES Emergency Contacts Brad Horn Nick Glumac Herman Krier 309 752 3616 Cell 217 244 8333 Office 217 333 0529 Office 502 S Mattis Ave Apt D 217 586 6467 Cell 217 898 4700
66. for thermal runaway and increases shock sensitivity Previous literature 3 5 7 10 21 27 focuses on the effects of various environments on the level of violence in cookoff by assessing the fragment size of debris resulting from confinement failure In a paper from SNL it is noted These simulations strongly suggest that the determination of the thermal damage states prior to the onset of ignition is the key issue toward determining the degree of reaction violence associated with cookoff 1 The experimental work presented in this work is part of a multiphase research endeavor aimed at producing a complex chemo thermal mechanical model to accurately predict the behavior for various solid propellants The experimental tests conducted serve to validate the construction and performance of the complex coupled model Multiple diagnostics collect data to validate the temperature distribution decomposition gases pressure and surface displacement In our work an adaptation of the VCCT and SSCB experimental setups is used primarily to study the state of the energetic material from the start of heating up to the point of ignition 2 Experimental Equipment and Process 2 1 General Apparatus The test apparatus is designed to accommodate either cased or uncased cylindrical solid propellant of variable length from 1 2 and of 0 75 diameter The propellant sample is seated firmly inside two spacer plates top and bottom that hold th
67. going through the entire directory in increments of three images Then images 49 corresponding to chosen temperatures typically in increments of 25 C are copied into an additional directory This folder is then processed in single image increments to provide a more qualitative digital image correlation For the slow cookoff directories of images care should be taken to ensure the proper order of the files in regards to PM versus AM The code outputs a data array of x and y position dx and dy in pixels and dx and dy in mm s The code also maps the vector field on top of the processed image and saves the figure to disk Before the code could be ran each directory of images was screened for unusable images All images before the start of heating after cookoff and those of bad quality were deleted prior to processing to prevent the code from failing or producing misleading results Propellant analysis was also performed in accordance with sponsors wishes Propellant samples from batches used in experiments were removed from the carbon fiber case and scanned with the Xradia Micro XCT machine in Beckman Institute with the 4X lens The data files from the scan were processed with the program Amira on the Krasner computer in the Visualization Laboratory also in Beckman institute Pictured below is an image of the processing in Amira Figure C 1 Amira S P Image Processing A procedure outlined below gives instructions for the image proc
68. he light from the resistance wires caused poor imaging and prohibited DIC analysis This test led to adjustment of the light source A high lumen light source saturates past the light from the resistance wires eliminating the oscillation By lowering the gain on the camera and limiting the iris the problem is resolved S P 8 Temperature vs Time 350 300 N LD o N o o R 1 3 in KA u o R 2 3 in Temperature C 100 R 3 3 in 50 0 500 1000 1500 2000 Time Sec Figure F 12 S P 8 Temperature 73 Pressure Psig S P 8 DAQ Pressure vs Time 10 8 6 4 2 0 ek ann aa dnne ae a SEET f 1 0 500 1000 1500 2000 2500 2 Time Sec Figure F 13 S P 8 DAQ Pressure Table F 3 S P 8 Spectroscopy H CH CH OH NH HO N CO CHO O Ar CO 2 14 16 17 18 28 29 32 40 44 Numerous scans of same percent compositions as first row baseline 0 5 1 0 1 74 1 15 1 0 1 5 1 0 1 74 1 14 1 0 1 2 1 0 1 74 1 14 1 0 1 5 1 0 1 75 1 13 1 0 1 5 1 0 1 76 1 12 1 1 1 5 1 0 1 76 1 10 1 1 1 5 1 0 1 77 1 9 1 1 1 5 1 0 1 77 1 8 1 2 1 5 1 0 1 77 1 7 1 2 1 5 1 0 1 77 1 7 1 3 1 5 1 1 1 76 1 6 1 4 1 5 1 1 1 75 1 6 1 4 1 5 2 1 1 74 1 6 1 5 1 5 2 1 1 73 1 6 1 6 1 5 2 1 1 71 1 6 1 7 74 S P 7 Propellant AP200um APS50um HTPB IPDI DOA Test 11 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Good Comments This was the fir
69. he temperature controller The temperature controller establishes a starting temperature of 25 C and then ramps to 50 C in 5 minutes followed by thermal conditioning for an hour at constant temperature After which the propellant is heated at a rate of 3 3 or 6 C hr until ignition in accordance with literature 1 13 26 characteristics for slow cookoff Virtually any heating rate or combination thereof can be achieved The relay and temperature controller are pictured in Figure 2 5 and Figure 2 6 respectively Man CS NOREERONG Se S OLIO STATE RELAY amp y 1239 oum our wm IF ES loja Figure 2 6 Temperature Controller Figure 2 5 Solid State Relay 2 3 Diagnostics Continuous diagnostics record three temperature readings two pressure readings surface imaging and mechanical gas sampling The pressure and residual gas analysis are taken from the propellant core while the temperature measurements are spaced equally at thirds of the annulus radius The surface imaging is of the propellant case or the surface of the propellant if uncased Together there are six signals simultaneously read and recorded by LabView Nearly two months were spent designing and programming a custom LabView virtual instrument to effectively control and record all diagnostics The VI is fully documented in Appendix B In addition a webcam was also utilized to provide visual confirmation of cookoff ignition 13 2 3 1 Temperature A three
70. hed to blow through the detached section First make sure that the filament on the mass spectrometer is off Next the mass spectrometer sampling line connection should be carefullv removed This does not mean detach the capillarv from the mass spectrometer but from the connection to the reaction cvlinder This connection should not be capped since at this point it is assumed that the mass spectrometer pumps are still pumping All sampling lines should be blown out This can be done with house air or nitrogen It is recommended to use house air since it is cheaper than bottle nitrogen However if additional pressure is needed after trving house air bottled nitrogen can be used 60 E Time Expenses amp Parts List Table E 1 Time Table amp Expense Summary Time Minutes Cookoff Experiment Time Table Tasks starts from end of test Researcher Assistant Properly save data and unplug power sources 30 Allow for everything to cool down 120 While waiting to cool 10 min cut down and edit web cam video Vacuum up carbon fiber and residue in sound enclosure 10 Disconnect all connections amp remove optics board 10 Remove and discard T Couples 10 Remove all shields 15 Complete disassembly and ready for cleaning 30 Clean all parts and shields with ethyl alcohol 30 min to dry 45 While drying vacuum and clean optics board with et
71. hyl alcohol 10 Process data smoothing plotting etc 60 Back up transfer propellant data to external hard drive for processing 20 Fabricate a set of 8 ceramic tubes 180 Blow out gas line and make sure not clogged 30 Reassemble parts and ready for rewiring 30 Rewire whole setup and test functionality 90 Fabricate new heat shield 30 Reinstall shields 5 Set up optics board T couples gas line etc wait 1 hr for sealant 100 Finish last parts of setup and ready for test next day 30 Initiate test until cookoff 60 Total time spent per test 495 420 8 25 Y Cookoff Experiment Expense Summary Cost 8 ceramic tubes and bits to drill them 70 98 Top quartz window 27 77 4 Thermocouples 45 68 each 182 72 2 seals seals top and bottom of propellant 3 Heat shield T23 Random parts electric wire tape sealant etc 5 Total cost per test 296 72 61 min hrs Table E 2 Parts List Partit Description Unit Vendor Application 8746K17 High Temp Cermaic tube 1 4 OD 3 16 ID 12 long 13 94 McMaster Posts for heating element 4490A42 Diamond Pltd grinding bit 1 8 shank 1 16 head Dia 14 58 McMaster Drill bits for ceramic posts 8880K22 Nickel Chromium C wire 02 Diameter 1 4 Ib spool 23 61 McMaster Heating resistance wire 74515A34 10 302 tube Silicone sealant 24 43 McMaster RTV T Couple seals 9036K769 25 X4 X 012 blue finished 1095 steel coil 92 34 McMaster Heat shields 84815K41 Very High Temp Heat Resist Glass Cera
72. ic 1s discussed further in the conclusions 660 0 mv DC 640 0 620 0 600 0 580 0 560 0 540 0 520 0 500 0 480 0 460 0 0 5 0 0 0 5 1 0 1 5 20 25 3 0 3 5 40 45 Figure 4 4 S P 25 Pico Pressure The spectroscopv data file was large from a full scan being completed and saved roughly every minute In order to display the percent compositions below scans of roughly the same composition were deleted and the first column displays the number of scans that the row of percent compositions is repeated The astute reader will observe that the first and last rows of compositions were static values for long periods of time on the order of 14 hours The middle section of dynamically changing composition occurred between the temperatures of 225 231 C Table 4 8 S P 25 Spectroscopy of OH N Scans H CH CH NH H2O CO CHO O Ar CO 2 14 16 17 18 28 29 32 40 44 780 1 5 2 1 3 72 1 14 1 0 291 1 5 2 1 2 73 1 13 1 0 77 1 5 1 1 2 74 1 12 1 0 44 1 5 1 1 2 74 1 11 1 0 46 2 5 1 1 2 73 1 10 1 1 36 2 5 1 1 3 74 1 9 1 1 34 2 5 1 1 3 75 1 8 1 1 37 Table 4 8 cont 75 25 75 18 16 11 77 78 78 78 78 76 74 73 71 10 11 70 68 12 13 14 15 16 17 18 19 20 21 66 65 63 61 60 59 56 93 51 49 22 23 46 41 12 13 14 15 16 17 18 19 20 21 23 40 23 39 38 23 23 36 35 3
73. ick Lynch Jeffrey Mason Jennifer Peuker and John Rudolphi Thank you for all your wisdom and help It has been a pleasure working with all of you Cheers to you all Special thanks to my research assistant Mark Janowski who was involved in nearly all aspects of my research Another special thanks to Julio Barros for helping with the digital image correlation code Special thanks to Andy Tudor for the help and instruction in image processing of scanned propellants I would also like to thank my friend and fellow student Mark Shoemaker This journey was made easier by having a friend along the way My last acknowledgment is to my undergraduate physics professor and advisor Dr Christopher Fasano Words cannot express my sincerest gratitude for your diligent mentoring and guidance Everything that I have achieved through undergrad and graduate studies has been a direct result of your encouragement Thank you very much The work in this thesis covers the experimental portion of a collaboration between IllinoisRocstar Notre Dame and the University of Illinois supported by the United States Army project number W31P4Q 11 C0076 Experimental and Computational Program for Slow and Fast Cookoff for Insensitive Munitions Testing IV Table of Contents EE E VI List of Tables ati o ld o la do dd Vill List of Terms and SmbolS asin aci ni A ss ii ik pa IX ER ee a elo a 1 lla Solid Eenelter eendeitege dE 1 1 2 Slow Cookotf Fundamentals nina Dee 2 U
74. ile that was nearly useless The spectrometer processing code takes the name of the input file name of the output file and number of complete scans as user input Then the code sums up the partial pressures from 0 5 below the selected amu to 0 4 above and outputs a single partial pressure value per amu for all scans In the original file the last data entry for every scan is a total pressure measurement that was found to be inaccurate Therefore the code sums all of the positive partial pressures of a scan as the total pressure and ignores all negative pressures Each pressure measurement is then divided by the scans total pressure to give a percent composition per amu for all scans Another feature originally in the code pointed to dynamically changing amu See Appendix H for the spectrometer data processing FORTRAN code The shown code was for tests that scanned from 1 100 amu The code differs from the others in the length of variables and of the do loops It should also be noted that before the spectrometer FORTRAN could be ran on the initial log file a carriage return needed to be inserted at the end of the file and the result was saved as a Unicode UTF 8 format The original DIC code in MATLAB was written by Julio Barros and slightly modified to perform to desired specifications The code that is shown in Appendix H performs DIC between sequential images for a chosen increment DIC analysis Typically the directory of images is processed once
75. imentation during which the amplitude of noise is reduced to 0 02 C However for the slow cookoff experiments where the test duration is on the order of 50 70 hours the HVAC system is required Upon installing the temperature controller the noise returned which was diagnosed and reduced by adding a ground wire to the probe tip of the thermocouple used for the reference signal For the long test times most of the remaining noise is smoothed out in post processing The pressure system required modifications for communication Control boxes were built to wire the transducer to power and a BNC connection The BNC line was then terminated with a 100 ohm resistor to obtain the proper voltage range for the NI DAQ card Due to computer limitations a Picoscope was added to the signal to acquire higher frequency pressure data during the final stages of cookoff The DAQ pressure line serves primarily as a static pressure measurement The imaging system went through significant improvements Trial and error discovered a proper method for applying a usable speckle pattern for digital image correlation Program improvements allowed for live adjustment of brightness and contrast These controls combined with the manual gamma control allowed for proper brightness and contrast adjustment An external light source at 80 degrees off axis from the camera was installed in order to provide additional light for proper imaging After the repeatability tests were performed
76. ing scale and a few other optimizing changes To create the imaging virtual instrument a LabView programming language was purchased from EPIX the vendor that makes the frame grabbing card already installed A complete customizable program was created and adapted for the cookoff control The template was then sold back to EPIX for reimbursement of the programming language The LabView Front Panel and Block Diagram for the Cookoff Control Virtual Instrument are shown in Appendix B 3 5 Initial Test Results As discussed earlier initial tests were performed with AN propellant with a composition of AN HTPB IPDI and mass percentages of 70 27 86 2 14 respectively The AN propellant was easier to mix and avoided the hazard of HCI production during combustion unlike AP propellants The first two tests were performed with the heating element version II and were performed in order to test and fix diagnostic issues The first test revealed initial lighting issues in obtaining usable images of the propellant surface as well as a mistake in stopping LabView before it could record the last spectroscopy scan Adjustments to the VI were made to allow brightness and contrast adjustments and a five minute wait period was added to the standard operating procedure to allow proper time for saving of all data The second AN test exhibited the first occurrence of the grounding and noise issue with the thermocouples Image brightness was still an issue and the AN propella
77. ing successful implementation and testing of the diagnostics a series of tests were conducted to establish repeatability of all developed methods and procedures For these tests a common ammonium perchlorate composition previously used was duplicated to compare the data 14 After successful repeatability tests single base propellants were tested in accordance with slow cookoff standards The development of the test apparatus results of all tests and current status of all diagnostics are discussed in detail I Dedication This thesis is dedicated to my family and friends to my parents Guy and Barb Horn for always encouraging me throughout my education to do my best and for being proud of me no matter the outcome to my brothers Corey and Rob for always being there for me and serving as role models throughout my life to my entire family whom I love with all my heart without you I would be lost and without direction II Acknowledgements I wish to thank Prof Nick Glumac and Prof Herman Krier for all their guidance and wisdom throughout my studies Prof Krier said If we knew of what we were doing they wouldn t call it research I now appreciate the process of research and without your guidance I could not have completed that journey Thank you I would also like to thank my past and present research group including David Allen David Chonowski Michael Clemenson Drew Coverdill David Joyce Joe Kalman Lance Kingston Patr
78. ions of solid propellant and rockets went in waves varying in use between warfare and entertainment During these years small improvements were made including improved powder for sustained burn time tubular launch design and multistage rockets However little improvements were made to the accuracy of rockets whose effectiveness relied on sheer numbers The first presence of rockets in America was during the Battle of 1812 and led to the rockets red glare in the Star Spangled Banner By the early 1900 s scientists from Germany Russia and the United States were exploring the advancement of rockets and soon turned to the use of liquid propellant Goddard in the U S was the first to use liquid fuel and was soon followed by von Braun in Germany during WWII On October 4 1957 Sputnik was launched spurring the space race between the Soviet Union and the United States 7 1 11 Solid Propellant Solid propellant is comprised mainly of fuel and oxidizer that when ignited rapidly produces a large amount of hot gas By mixing an oxidizer into the composition the propellant is able to combust without the presence of external oxygen from the atmosphere If contained the rapid production of pressure can build until mechanical failure of the confinement yielding an explosion When allowed to vent through a nozzle the release of hot gases are directed to produce thrust powering a rocket into motion Once ignited solid propellant is extre
79. lausible correlation it can be hypothesized that the ignition temperatures of solid propellant are linked more to the ability of the confinement to store decomposition gases and buildup pressure and linked less to the actual heating rate of the solid propellant Obviously a fast heating rate can reduce the time for decomposition and thus the amount of gases stored in the confinement 41 5 2 Recommendations Several actions are recommended to improve the current diagnostics and continue research As discussed in the Preliminary Work Chapter significant steps were taken to reduce noise in the thermocouples from various sources The last source of noise was the installment of the temperature controller which was battled by installing a ground wire to the probe tip of the thermocouple used for the reference signal As a result there is a small difference between the temperature controller reading and the cold junction compensated temperatures read by the TCIC card A possible method to avoid this difference is to remove the thermocouple and communicate the temperature read by LabView to the temperature controller via the NI DAQ card Other methods could be explored to reduce the thermocouple noise including refurbishing of the TCIC card The heating rate of 6 C hr was chosen for smaller test durations According to the NATO STANAG 4382 test 6 the defined slow cookoff rate is 3 3 C hr Future experiments could be adjusted to this standard in
80. low is the mass of the propellant excluding the carbon fiber case ignition temperature duration of heating average heating rate and maximum observed pressures in psig The heating rate excludes the initial thermal ramp and conditioning Table 4 7 S P 25 Slow Cookoff Overview S P Mass Tign C Time min Time hr C hr Max Psi DAQ Max Psi Pico 25 15 858 359 17 3120 52 6 09 2 36 3 59 Figure 4 3 below is the temperature history Note that the traces overlay each other due to the slow heating rate which allows nearly complete thermal equilibrium within the sample Also displayed on the chart is a linear fit to the trace The slope of the line yields a heating rate of 6 042 C hr slightly better than the numerically computed average heating rate S P 25 Temperature vs Time H 250 T 4 150 R 2 3 in Ou al eg 100 R 3 3 In 50 p Linear R 3 3 in 0 0 500 1000 1500 2000 2500 3000 Time Min Figure 4 3 S P 25 Temperature The pressure data obtained during the cookoff event was less than the non slow cookoff experiments In relation the pressure violence observed during cookoff was minimal The top quartz window heat shield most of the ceramic tubes and even two thermocouples survived the cookoff event Video recording of the final stages of cookoff via a webcam proved visually the absence of the typical fireball explosion that occurred in previous AP propellant tests This top
81. mely difficult to extinguish without a rapid depressurization Solid propellant is categorized as composite or either a single base double base or triple base Double base propellant contains fuel and oxidizer chemically mixed to form a homogenous substance usually consisting of nitroglycerine and nitrocellulose Commonly the propellant is found in a powder grain or crystal form that can be combined with other additives and a plasticizer to form a heterogeneous composite mixture with desired attributes Single base propellants generally consist of a powder oxidizer mixed with a binder for fuel that forms a heterogeneous composite mixture with other additives to change the properties Common oxidizers used for rockets include ammonium perchlorate NHACIO AP ammonium nitrate NH4NO3 AN potassium perchlorate KCIO KP and potassium nitrate KNO KN 14 In single base compositions the rubber binder commonly hydroxyl terminated polybutadiene C 332H10 98200 058 HTPB is the main source of fuel and serves to hold the granular mixture together Plasticizers help in the elongation of the mixture during processing and dioctyle adipate C22H4204 DOA is a common choice A small percentage of mass can be additional curing agents like isophorone diisocyanate Ci2HisN20 IPDI to help solidify the binder 15 See Appendix A for a detailed list of common ingredients 14 Additional materials can be added to improve or modify material characteristics o
82. mic 2X2X 197 1 82 McMaster Camera Shielding 7532K14 plastic sleeve bearing 3 4 ID 13 8 0D 6 39 McMaster Mount S P 8632K42 1 16 thick 12X12 40A extreme temp silicone Rubber 9 01 McMaster Sheets top and bottom S P seals 1357132 2 Diameter X 1 4 thick quartz glass 27 77 McMaster Top middle glass in test apparatus 8546K26 PTFE Teflon rod 1 1 8 diameter sold by foot 21 67 McMaster Propellant stand machine on lathe 51135K212 High T Rubber Tubing 3 80D 1 41D Black per foot 1 00 McMaster nsulating tube mount T Couples 41254282 General Hard Carbon Steel Blade 5 4 5 5X 025 X18R 11 63 McMaster Band saw blade in 2308 shop 92414A457 2 252 1D 1 4 Screw Size Steel Spacer 5 94 McMaster Spacers heating element 71385K528 1 4 Dia ceramic tube fuse 20 Amp 250 VAC 2 35 McMaster Fuse for red Variac Transformer ACMI 2260 High Temp Resistant White Ink quart size 67 55 American Coding amp Marking Ink Co Ink air brush speckle pattern 0100SSC 760 Torr Stainless Steel Capillary tube for QMS200 150 00 Stanford Research Systems Inc Spectrometry diagnostic SS 2TF 05 5 micron 1 8 swagelok connection type T filter 95 20 Swagelok St Louis Spectrometry diagnostic SSRL240DC25 DC control signal 280 Vac line with 25A 26 00 Omega Engineering Inc Power management TJ120 CXSS 020U 2 5 SB Custom T c w Trans Joint quote H 009976690 45 68 Omega Engineering Inc Thermocouples EXTT E 24 SLE 25 Insulated T C wire 38 00 Omega Engineering Inc T Couple
83. n Slow 257 288 Deflagration Explosion 8 5 Fast 347 10 08 Burning Deflagration Slow 263 447 8 Deflagration The pressure history shown is of the DAQ pressure which shows greater variance than the Picoscope pressures due to inconsistent vacuum as a result from the mechanical gas sampling It was evident throughout the repeatability and additional tests that the seals around the propellant sample would inconsistently hold a vacuum for small periods of time For a complete presentation of the repeatability data see Appendix C DAQ Pressure 5 4 3 z 2 e d S P 12 3 E 0 9 P 13 si S P 14 2 3 A Time Seconds Figure 4 2 Repeatability DAQ Pressure Plot 4 2 Slow Cookoff Results Following the repeatability tests improvements were made to the camera setup as discussed in Chapter 3 Changes to the power control and heat source for the resistance wires allowed a slow cookoff test to be conducted with slow cookoff heating characteristics The temperature controller program established a starting temperature of 25 C and then ramped to 50 C in 5 minutes followed by thermal conditioning for one hour at constant temperature After which the propellant was heated at a rate of 6 C hr until ignition in accordance with literature guidelines 13 The higher standard for slow cookoff was chosen in order to reduce the time for a complete experiment 39 The sample used for the slow cookoff test is summarized below Shown be
84. nition temperatures are higher than graphically depicted Ignition temperatures are defined as the highest temperature recorded before thermal runaway or auto ignition Also a good status on the imaging diagnostic does not imply that DIC data processing yielding results Spectroscopy results when shown are scans with percentages of the main molecules in the composition The scans shown are those preceding ignition with multiple scans be omitted from the beginning which were of the same baseline composition S P 1 3 2 Propellant AN HTPB IPDI Test 1 Mass Percents 70 27 86 2 14 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Bad Bad Comments This was the first AN test with a single temperature trace and was the only AN propellant to yield pressure data The LabView instrument was stopped early before the spectroscopy scans could be saved to disk following ignition S P 1 3 2 Temperature vs Time R 2 3 350 300 250 200 150 Temperature C 100 50 0 200 400 600 800 1000 1200 Time Sec Figure F 1 S P 1 3 2 Temperature 63 Pressure Psig S P 1 3 2 DAQ Pressure vs Time 40 35 l SR si DSP A le O 0 52300 2320 2340 2360 2380 2400 2420 2440 2460 2480 2500 A IE 10 Time Sec Figure F 2 S P 1 3 2 DAQ Pressure S P 2 Propellant AN HTPB IPDI Test 2 Mass Percents 70 27 86 2 14 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good N
85. nt melted and clogged the bottom port eliminating any pressure or spectroscopy data The next two tests were performed with the heating element version 3 that used 6 loops of wire Both tests used the AN propellant composition and allowed small advancements in imaging and noise reduction in the temperature diagnostics After conferring with Omega Engineering new ungrounded thermocouples were ordered to help reduce noise During these tests the bottom port clogged again eliminating pressure and spectroscopy data Thus the decision was made to move to AP propellant compositions in order to effectively test the diagnostics For the first AP test the heating element was adjusted to 8 wires as noted in Chapter 2 2 due to the higher ignition temperatures for AP propellant The composition for the propellant was AP200um APSOum HTPB IPDI DOA with mass percents of 54 56 33 44 9 29 0 71 2 00 respectively During the first AP propellant test S P 10 the voltage setting was initially too low and needed adjustment A loose wire connection eliminated any pressure data Also the thermocouples were grounded instead of the ordered ungrounded All new thermocouples are 28 now tested upon receiving to confirm the correct specifications Table 3 1 summarizes the preliminary tests See Appendix F for a complete presentation of the preliminary tests Table 3 1 Preliminary Tests Overview S P ID Propellant Tign C Time Sec Time min C min MaxP psi
86. nts were made to the diagnostics prior to successful tests An extensive step by step diagnosis was used to make the thermocouple output reliable and free of interference The thermocouples were individually wrapped in rubber spiral cable wrap New 24 mounting holes in the bottom plate were drilled to allow for rubber tube inserts to electrically isolate the thermocouple from the test apparatus External ground wires were applied to the test apparatus and to a control box built to supply power to the pressure camera and thermocouple card A small section of rubber hose was inserted in the gas line to electrically isolate the spectrometer from the test apparatus After conferring with Omega Engineering representatives the original grounded thermocouples were replaced with less susceptible to noise ungrounded probes and all unused channels of the TCIC thermocouple card were shorted out by bridging the positive and negative posts The spacer plates securing the propellant were drilled out to allow plastic bearing inserts to electrically isolate the carbon fiber case from the rest of the test apparatus Figure 2 3 A portable power conditioner was installed to reduce noise in the power source The primary cause of noise was diagnosed as the HVAC system in an adjacent room During operation the noise amplitude was 5 C and fell to 1 C after all steps were taken For the beginning tests the HVAC schedule was modified to allow small windows of exper
87. ocating memory to finer mesh element calculations As a result larger elements that neglected small scale chemistry and element refinement methods were used The refinement methods used in the code at that time were a new research area in computational modeling A common experimental setup used for the study of cookoff is the variable confinement cookoff test VCCT used by the Naval Surface Warfare Center Figure 1 3 is the VCCT finite element mesh used by Sandia National Laboratories 23 Top flete ant Hanoi band Gres Recon A fi SE SLD LE 5 phe Daun Hale EA A Side view ol VCCT Cross section of VOCT Figure 1 3 VCCT Finite Element Mesh Taken Directly from Reference 23 The SNL finite element code briefly incorporated the first stage of cookoff up to ignition temperatures by modifying the energetic material into a thermally damaged state categorized by porosity specific surface area and crack density Later on SNL incorporated JAS into the model to better account for this mechanical behavior Variable confinement cookoff tests VCCT predicted that as much as 10 of energetic material decomposes prior to ignition indicating a strong dependence on mechanical physics in modeling the early stages of cookoff It was also noted that during slow cookoff the thermal damage is largely distributed throughout the sample when compared to the surface localized thermal damage resulting from fast cookoff
88. omplete 6 X complete C Description C This program will take the spectrometer data and process it Cs e s e s e es oe s esesee e es e e e essees se ee e s es e s se se s s e C Main Program PROGRAM SpecData IMPLICIT NONE INTEGER 4 InFileUnit OutFileUnit 3 nmax ScanMax amus i j k 1 n m NumScans PARAMETER nmax 2020 PARAMETER ScanMax 100 PARAMETER InFileUnit 25 PARAMETER OutFileUnit 20 INTEGER 4 Datedd ScanMax Datemm ScanMax Dateyyyy ScanMax 1 Timehh ScanMax Timemm ScanMax Timess ScanMax 2 Timep ScanMax REAL 10 Point 1 2001 Delta TotalP ScanMax 1 Value ScanMax nmax AMU ScanMax nmax 2 PercentAMU ScanMax nmax SumP ScanMax CHARACTER 60 InFileName OutFileName a b c d e CHARACTER len 20000 buff Gases USER Eh RT AAA ee e Prompt user for the name of the input file WRITE Enter name of the input file including extension READ InFileName WRITE Enter name of the output file including extension READ OutFileName WRITE Enter the number of complete scans READ NumScans WRITE 5 InFileName OutFileName NumScans Sa enn amp INITIALIZE SG LON See c Open Files if the OPEN fails go to the error handling section OPEN UNIT InFileUnit STATUS OLD FILE InFileName ERR 910 OPEN UNIT OutFileUnit STATUS UNKNOWN FILE2 OutFileName ERR 910 95 Open Files if the
89. order to obtain additional results to compare and validate with The software program can easily be adjusted for varying parameters of the ramp soak program Regardless of the heating characteristics a series of about 6 tests including varying compositions of double base propellant will likely comprise the test matrix for the cookoff project Due to the lack of results from the DIC for mechanical movement changes can be made to the imaging diagnostic Modifications could be made to eliminate the flexibility in the top and bottom seals which after long periods of heating seem to allow the sample to move in the vertical direction Another form of imaging can be obtained through the top quartz window of the inner annulus geometry By calibrating with a gridded surrogate sample the inner mechanical deformation could be resolved In a different approach it could be useful to investigate the mechanical structure of an unconfined solid propellant This approach would prevent pressure and mass spectrometry data but could make it easier for infrared absorption In conjunction with this test is the possibility for an open atmosphere experiment conducted by simple ignition from a resistance wire Such experiment would primarily investigate the dynamic absorption analysis and the material deformation of the final stages of cookoff and could be conducted in a large air blast chamber As previously mentioned the mass spectrometer data is flawed by the backflow of lo
90. ording to the industry standard TB 700 2 6 Typically large scale tests have been done with fast cookoff characteristics where the damage is measured qualitatively along with possible temperature and pressure measurements as a function of radius from the object In this manner a large amount of literature focuses on regions of damage or the level of violence rather than the state of the material during the stages preceding ignition 3 5 7 10 21 27 Recently studies have been moved to correlating between small and full scale test methods in the search to establish a reliable small scale methodology for testing and screening for materials that will not pass full scale tests The drive for this research is to provide a small scale alternative adoptable by the TB 700 2 regulations Full scale tests are expensive and can be dangerous Establishing a reliable correlation between scales will allow classification of energetic materials in a cheap and efficient manner Tests being developed are the Critical Diameter D the Burning To Violent Reaction Thermal Properties Characterization TPC and the Critical Impact Velocity CIV tests 5 The TPC test is the most relevant test designed with diagnostics to determine ignition temperatures exotherms endotherms weight loss volume increase and combustion products A separate test uses a small motor that is ignited after similar heating in order to measure pressures that reveal the affect of therm
91. orkstation controlled from an adjoining room for three levels of containment and safety precautions Figure 2 2 Entire Apparatus Inside Sound Enclosure 2 2 Heating Element The heating element is comprised of 0 02 thick nichrome Nickel Chromium C resistance wire threaded through ceramic posts When applied a voltage the nichrome resistance wire radiates to heat the propellant sample to any desired temperature up to 400 C at variable rates Eight ceramic posts are positioned in an octagonal pattern centered round the propellant with a distance of 2 375 from each post to the center This pattern creates a nearly angularly symmetric heat source Each ceramic post is machined with diamond tipped drill bits to allow 8 loops of nichrome resistance wire A single wire completes 8 loops through the ceramic tubes with a resistance of 16 6 ohms When applied a wall outlet voltage the circuit yields 7 5 or less amperes Figure 2 3 and Figure 2 4 show the heating element from a top and side view 11 Figure 2 4 Side View of Heating Element The complete circuit is connected to a solid state relay from Omega Engineering that has a 330 VAC with 25 ampere limit The solid state relay is controlled via a temperature controller 12 that uses a 4th thermocouple as a reference signal The thermocouple is placed at the same depth as the yellow thermocouple discussed in section 2 3 1 A software package from Omega is used to program and run t
92. r combustion performance Metal additives such as aluminum Al boron B or magnesium Mg are used to add additional energy to the combustion process However metal additives add instability and an increased danger when manufacturing and handling During the past century due to the military NASA and small rocket enthusiasts significant research has been conducted on various compositions and materials down to the effect of geometrical configuration of the grains 1 2 Slow Cookoff Fundamentals Cookoff refers to the process of heating an energetic material until it reaches self or auto ignition The ignition temperature is the point when the material enters a self sustained reaction that is an exothermic dynamic process that will continue without external heating Cookoff is a thermal chemical and mechanical complex process There are two main categories of cookoff determined by the heating rate which are slow and fast cookoff The division between the categories is not always well defined and depends on which source is referenced The general 2 range is that slow cookoff is on the order of 2 6 C hr where fast cookoff is loosely defined as a fuel fire on the order of 70 C hr with the distinction between the two being undefined In work by Geisler 9 three heating rates characterized as low medium and high with corresponding rates of 13 8 1200 and 90k C hr According to the slow heating NATO STANAG 4382 test the slow cookoff
93. running with filament on four hours prior to any data acquisition Checklist Before Startup Capillary tube attached to capillary inlet finger tight only no tools required Exhaust cap removed from exhaust port Connected to exhaust system if needed Power cord connected to QMS Serial cable connected from RGA on the QMS and COM port to the computer Continuous Sampling Procedure State 4 Continuous sampling requires that system be set to state 4 in the state diagram all switches on Preferred Startup Sequence e Set the four switches on the control panel to off and turn on the main power switch e Turn on switches in order o Mechanical Pump o Capillary Flow Valve o Turbo Pump o Sample Inlet Valve e Start the QMS program Choose the COM port that the QMS is connected to and then press the Connect button After a short initialization the QMS is ready To confirm communications under the Head menu choose Get Head Info A box will appear showing information about the QMS e Inthe software select the UtilitieslPressure Reduction menu item and enter the Pressure Reduction Factor for the capillary Check the box to enable the factor e Click the filament button on the toolbar to activate the ionizer Click the GO button on the tool bar and a scan will begin User may switch on all switches in correct order in a rapid fashion The spectrometer will initiate startup sequence in order and will be bright green when startup sequen
94. s overall average heating rate and max pressures measured from the DAQ and Picoscope Table 4 1 Repeatability Test Overview S P Mass Tign C Time Sec Time min C min Max Psi DAQ Max Psi Pico 12 1474g 329 5 1446 24 10 12 68 6 33 6 88 13 14 62g 354 7 1619 26 98 12 18 5 98 6 14 14 1484g 354 8 1573 20 22 12 48 7 96 7 93 For the temperature analysis each thermocouple location was averaged between the three tests to yield a mean This mean was used to calculate the standard deviation which was then averaged for time periods of five minute intervals The last interval is where thermal runaway occurred and contains the highest variance The Avg column is the average of the standard deviations for all three thermocouple placements for the given time period 31 Table 4 2 Temperature Standard Deviations in Degrees Celsius Average Standard Deviation Minutes R 1 3 R 2 3 R 3 3 Avg 0 5 3 34 3 23 3 46 3 34 5 10 4 72 4 67 1 27 3 55 10 15 4 59 4 69 3 43 4 24 15 20 1 88 2 95 1 02 1 95 20 25 6 31 6 98 4 66 5 98 A separate temperature calculation was performed for each thermocouple location to yield the percent differences from the previously calculated mean Again these measurements were grouped into five minute intervals for qualitative assessment Throughout the experiment it can be assumed that the temperature analysis is within 4 precision between repeated tests
95. st test to add the Picoscope pressure measurement However a short in the pressure wiring eliminated results above atmospheric reading The DIC processing showed negligible movement except the oscillation in the field of view S P 7 Temperature vs Time 350 300 250 200 R 1 3 in R 2 3 in R 3 3 in 150 Temperature C 100 50 o 200 400 600 800 1000 1200 1400 Time Sec Figure F 14 S P 7 Temperature Figure F 15 S P 7 DIC 75 CO 44 Table F 4 S P 7 Spectroscopy E ES O EIS o Q al N Z S Z D e FS El H Numerous scans of same percent compositions as first row baseline 14 13 12 11 73 73 74 74 75 76 75 74 74 72 68 10 14 21 62 52 33 40 11 40 34 29 46 76 S P 6 Propellant AP200um APS50um HTPB IPDI DOA Test 12 Mass Percents 54 56 33 44 9 29 0 71 2 00 Diagnostics Temperature Pressure Spectrometry Imaging Performance Good Good Good Good Comments This AP test was successful in all diagnostics DIC results still did not produce movement other than the oscillating field of view This was the last test before the repeatability series S P 6 Temperature vs Time 300 N LD o R 1 3 in 100 e R 3 3 in Temperature C 6 un o KH o 200 400 600 800 1000 1200 1400 1600 1800 Time Sec Figure F 16 S P 6 Temperature S P 6 DAQ Pressure vs Time 3 5
96. t needed for the batch plan for 15 grams more than needed 2 Clean lab workstation and locate all PPE goggles gloves lab coat 2000000000 SSC 3 Gather all chemicals to workstation from storage cabinets 4 Know safety precautions incase offre Sa 5 Locate emergency exits alarms and fire extinguishers Mixing 1 Weigh out and add oxidizer to mixture i e AN AP mix for m RES 2 Weigh out total masses of metal additive i e Aluminum HTPB binder DOA and additional additives and mix thoroughly in plastic bag for 15 min 3 Weigh out and add isocyanate such as IPDI to mixture mix for 15 min 4 Immediately cast final mixture into carbon cases tamping frequently Curing 1 Carefully transport propellant samples into vacuum oven KSE 2 Apply slight vacuum and a temperature of 60 C apply vacuum slowly 3 Cure for 4 days in vacuum oven monitoring frequently Les 4 Remove samples from casting stand and remove tape SE 5 Record each new mass of carbon fiber case and propellant 0000000000 _ 6 Label samples and store in flammables storage cabinet until use Emergency Procedure fire CAUTION Burning propellant will release hazardous gases In case of fire if possible contain fire and move to sink turn on water Tf fire is uncontrollable evacuate room and pull fire alarm e tes Notify officials 57 Mass Spec Standard Operating Procedure QMS 100 200 Series Mass Spectrometer System should be
97. the TCIC thermocouple DAQ card by Omega Engineering and another sample code was also provided by Stanford Research Systems to control the mass spectrometer top left and bottom left of Figure B 1 respectively Each sample code was modified significantly to achieve the below versions The section of code that controls the CCD camera was made possible through a sub vi software package supplied by EPIX Inc called XCLIB Lite The section of code was then sold to EPIX for open source distribution in return for the reimbursement of the XCLIB Lite software package The pressure code was written from scratch using Measurement amp Automation Explorer from National Instruments to create users tasks All of the code was implemented together with universal user controls and safety features The entire virtual instrument went through over 400 saved versions and over two months of development by Bradley Horn Certified LabView Associate Developer Mach MarTemp 5 08 Cookoff Status Savi Date Time Stamp Aio EAs Mas 41 11 18 11 04 05 35 725 PM y Total Pressure 9 04E 3 Linear Enable Background j Pressure PSI Mode Analog 75954 Units mTorr Scan Log Setup Start Scanning _stop scan Je M 1 Kik 6 17 57 PM Figure B 1 Cookoff Control VI Front Panel 47 General Control EEES XD A Background Temp SE enn wt gt E Ee B
98. thermocouple array comprises the temperature diagnostic The temperature distribution of the propellant sample is measured temporally in the radial direction Custom type E thermocouples from Omega Engineering have a thin probe lead of 0 02 diameter The predrilled carbon fiber case allows access to the solid propellant Figure 2 7 shows a schematic of thermocouple placement Note that the distance marked for each thermocouple is from the outer edge of the carbon fiber case which adds 1 32 to the depth 3 4 Diameter Figure 2 7 T Couple Placement Schematic The thermocouples have millimeter spatial precision with a max sampling rate of 5 Hz due to computer memory limitations Accuracy of 0 015 Kelvin is possible however the overall accuracy of the temperature diagnostic is 0 5 C due to uncontrollable noise Section 3 3 describes all steps taken to minimize the electrical noise in the thermocouples Data is collected by an 8 channel data acquisition system from Omega Engineering Figure 2 8 The TCIC card can read data at rates of 1 kHz far greater than the sampling rate of the thermocouples 14 Validation of the predicted temperature field provides a fundamental check on the thermal component of the computational model Figure 2 8 TCIC 8 Channel Thermocouple Card Image from Omega Engineering http www omega com pptst TCIC html 2 3 2 Pressure Attached to the bottom port is a Gems piezoresistive pressure transducer
99. w atomic mass molecules such as H2 A modification to the mass spectrometer would install an 42 external diaphragm pump duplicating the one already in the machine The two paths of flow in the spectrometer can then be separated to eliminate any backflow inherent to the QMS design and essentially convert the QMS into a pseudo UGA Universal Gas Analyzer The UGA is the next model provided by Stanford Research Systems The pump is fairly expensive and is currently being employed as an upgrade Calibration experiments will validate the mass spectrometer results in cooperation with the infrared absorption Subsequent experiments will likely include a matrix of some single base but primarily double base solid propellants It is possible that metal additives such as aluminum could also be added A heat flux sensor could be installed to gain exact knowledge of the heat flux at the propellant wall and possibly to help determine the thermal conductivity of the propellant samples A knife edge experiment could be conducted to measure the rate of diffusion of various gases through a small thin propellant disc More in depth analysis could also be performed on the tomography investigation Samples can be XCT scanned prior to thermal degradation and reimaged prior to ignition to obtain the mechanical deformation due to various thermal environments Lastly a large scale cookoff test could also be explored 43 3 8 9 10 11 12 13
100. ysics lluna deeNA ee Ee 4 Figure 1 3 VCCT Finite Element Mesh Taken Directly from Reference 23 6 Figure 1 4 Modified SSCB Taken Directly from Reference TI3lL eenenennnnnn 7 fisure2 l General Schematic bla a dat f ema i Sa 10 Figure 2 2 Entire Apparatus Inside Sound Enclosure sse nnnennennzznnnenanennnanznnnnnz 11 Figure 2 3 Top View of Heating Element iaia iba NEEN EAn 12 Figure 2 4 Side View of Heating Element i bii i KA A AL 12 Fi ure 2S Solid State Relay E 13 Figure 2 6 Temperature Controller uni ta 13 Figure 2 7 T Couple Placement Season ta a 14 Figure 2 8 TCIC 8 Channel Thermocouple Cardin d cs Qatar hadnt 15 are DEE 15 Figure 2 10 Gems Pressure Sensor amp Gas Filter ini 15 Figure 2 11 NI USB 6008 DAQ Card ss siti a i it ida ita 16 Figure 2 12 CCD Camera and Md ci 17 Figure 2 13 CCD Camera 0 5 mm Marks ii A ela ele caaand tienes 17 Figure 2 14 SRS 200 Quadrupole Mass Spectrometer c ooooccconoccnonoconononcnnnononnnononnncnonnnnncnnncncnonecinns 18 Figure 2 15 IR Optical Absorption Macri ideo 19 Fisure 2 16 Propellant STUY A E E EAE cee A R 20 Figure 2 17 Propellant Casted in Stade pisito diri oie 20 Figure 3 1 Original Experimental Setup iii EE 22 Figure 3 2 View Without Heat Shield uuu a 22 Figure 3 3 Heating Element Version TL A Sede 23 Figure 3 4 View of Setup Versi n Il isse e A At A e 23 Figure 3 5 Heating Element Version III Schematic mnn nnnnennnnonennnen nanna 24 Fig re e 26 Figure 3 7 Expanded Tetlon with D
101. ze of the propellant samples used in this research and most in service rocket motors eliminating possible detonations 13 Upon ignition the combusting propellant emits high temperature gases increasing the pressure while transferring heat to the unburned propellant raising its temperature As a result a chain reaction occurring in the unburned propellant causes an exothermic chemical reaction repeating the process which creates a flame front or combustion wave The propagation of the combustion wave is dependent on the propellant composition structure pressure temperatures both initial and flame and other variables During the combustion three phases exist consisting of solid solid propellant mixed liquid and solid and gas combustion gases and premixed flame The mixed phase region of liquid and solid is a condensed phase above which resides the hot reactive gases which form the luminous flame 14 Assuming a conventional flame structure on the surface of the propellant the AP and HTPB decompose and form NH and HC1O Oxidizing reactants form flamelets which when combined with the fuel reactants produce a diffusion flame structure The final gas products include HCI CO CO H20 and N2 However it is important to note that the point of ignition can occur anywhere inside or on the surface of the propellant 1 3 Literature Review Historically most large scale tests are performed in order to categorize energetic materials acc

University of Illinois at Urbana

Contents

Download Pdf Manuals

Related Search

Related Contents