Dr. Manfred A. Bohn Fraunhofer-Institut fuer Chemische Technologie, ICT

Transcription

1 Characterization of the ageing of TPB bonded MX based insensitive high explosive charge by eat Flow Microcalorimetry (FC), Chemiluminescence, Gap test and Dynamic Mechanical Analysis (DMA) Dr. Manfred A. Bohn Fraunhofer-Institut fuer Chemische Technologie, ICT Presented on NDIA 2009 Insensitive Munitions and Energetic Materials Technology Symposium May 11 to 14, 2009, Ventana Canyon Tucson, Arizona, USA 2009 Bohn / Titel 1

2 Content Composition of high explosive charge (EC) TPB-MX Objectives Basic stability data Time-temperature (ageing) program for EC Sensitiveness by gap-test Glass transition temperature determined by torsion DMA eat generation rate between 120 C and 150 C Chemiluminescence in air between 100 C and 140 C Summary and conclusions 2009 Bohn 2

3 What is EC of type KS 32 KS32 is a polymer bonded high explosive charge (EC), manufactured by EADS-TDW, Schrobenhausen, Bayern, Germany (now part of MBDA) Typical composition Energetic material MX 85.0 mass-% Binder TPB cured with IPDI 13.7 mass-% Plasticizer DOA 1.0 mass-% Antioxidant protected phenol type 0.1 mass-% Curing catalyst DBTL mass-% Additionals mass-% TPB IPDI DOA hydroxy terminated polybutadiene ísophorone diisocyanate di-(iso-octyl) adipate (DiOA) 3 C (C 2 ) 3 C C 2 O C (C 2 ) 4 C O C 2 C (C 2 ) 3 C 3 C 2 5 O O C 2 5 DBTL di-(n-butyl) tin dilaureate (DnBTDL) (C 3 C 2 C 2 C 2 ) 2 Sn [OCO(C 2 ) 10 C 3 ] Bohn 3

4 Objectives To age the material accelerated up to the equivalent loads with regard to in-service loads To age at the two extremes under air and under oxygen-free atmosphere (here under argon) To get information about the Arrhenius parameters valid at accelerated and in-service temperatures. The minus centigrade region was not included because the chemical and diffusion processes are very slow there and are expected of minor to no influence. To look at some properties if they can be used as ageing monitor 2009 Bohn 4

5 Basic stability data of EC of type KS32 Autoignition temperature determined in Wood metal bath 5 C/min heat rate 244 C 20 C/min heat rate 243 C olland Mass Loss test at 110 C 0-8 h 0.02 % 0-72 h 0.01 % 8-72 h % ( limit value: 2 % ) Vacuum Stability Test, 100 C, 40h 2.5g 0.03 ml/g ( limit value: 1.2 ml/g ) 2009 Bohn 5

6 Time-temperature or ageing program for KS Bohn 6

7 Accelerated ageing at thermal equivalent loads The best way to perform accelerated ageing on a material for prediction purpose is to do it in such a way that the thermal load at in-service will be covered. This is named equivalent load ageing. The equivalent load must be calculated from given or demanded in-service loads for the test conditions at elevated temperatures. The in-service loads can be assumes as isothermal over a set usetime as 20 years or stored at time-temperature profile as experienced during missions For this one needs a temperature parameterization of properties of the material What to do if you do not known them at the beginning? Recommended procedure Assume a ageing at low activation energies in a two step mechanism. Such a two-step mechanistic ageing can be simulated quite reasonably with the generalized van T off rule 2009 Bohn 7

8 Equivalent load relation between test time - test temperature and in-service time - in-service temperature For chemically caused ageing processes the following formula has proven by experience as suitable (Ea about 80 to 120 kj/mol, temperature range 30 C to 90 C) Generalized formulated van T off rule t E [a] = t T [d] F (T T T E ) / T F / d t E time in years at temperature T E t T F test time in days at test temperature T T reaction rate change factor per 10 C temperature change rate decreases with temp. decrease / rate increases with temp. Increase T T test temperature in C T E in-service (environmental) temperature in C T F temperature interval for actual value of F, here T F is always 10 C For details and worked out examples with two mechanisms see Manfred A. Bohn Prediction of equivalent time-temperature loads for accelerated ageing to simulate preset in-storage ageing and time-temperature profile loads. Paper 78, 40th Conference of ICT, Bohn / 8

9 Ageing scheme for KS32 one month: 365/12 = 30.4 days; ageing equivalent of 20 years at 25 C is 24.3 months at 50 C years at 25 C -> ageing temp. [ C] in days in mo in days in mo in days in mo in days in mo in days in mo in days Bohn / The factor of 2.5 is assumed for KS32, this gives a conservative estimate of storage times to reach preset ageing times at 25 C; means longer storage times than with factor

10 Test on sensitiveness by gap-test 2009 Bohn 10

11 Scheme of ICT-Gap-test, 21mm 21mm X mm Electric Detonator Donor Charge Plexiglas Plexiglas Tube Donor with one booster X charge: diam. 21 mm, length 21 mm RDX / wax / graphite (94.5 % / 4.5 % / 1 %) pressing force 50 kn detonator in central bore: T-CU-30-T11U 40mm Acceptor Charge Actual length of KS32 samples were 42 mm 21mm Witness Plate (Aluminium) 10 mm thick 2009 Bohn 11

12 Assembly of charges in 21 mm ICT-gap-test 2009 Bohn 12

13 Ageing of gap-test samples Left: Ageing in air. Full air access by removing of the not sealed stopper two to three times per week. Right: Ageing in argon, applied with a glove box. Full sealing (grease in ground surfaces) and securing the stopper Bohn 13

14 Results of 21mm gap-test in detonation pressure deton. pressure [kbar] 44.4 kbar go region 21mm gap-test at ICT aged at 80 C over 12d, 24d, 36d in air at 80 C over 24d in argon go, argon 36 days at 80 C correspond to 15 years at 25 C go, air no go, air go, argon no go, argon kbar go, air 14% reduction in go 38 kbar no go no go, argon no go, air 35 kbar 15% reduction in no go time [d] Bohn 14

15 Glass transition temperature determined by torsion DMA 2009 Bohn 15

16 DMA instrument ARES TM sample mounting, measurement in torsion mode 2009 Bohn 16

17 Samples for DMA measurements ageing procedure visible influence on samples 2009 Bohn 17

18 Shear modulus and tan(δ) different definitions of glass transition temperature T G Temperature sweep with measurement at 3 frequencies: 0.1 z, 1 z and 10 z 3.E+09 G', G'' [Pa] dg'/dt [Pa/ C] aged at 90 C, 30d, Argon, Pr1 10 z T G in tan(δ) = C tan(δ) [-] E+09 G' tan(δ) E E E+09 5.E+08 T G in G'' = C T G in dg'/dt = C dg'/dt G'' 30d G' 30d G'' Tg in G'', C 30d dg' Tg in dg', C 30d td Tg in td, C T [ C] E Bohn 18

19 Shear modulus and loss factor tan(δ) aged sample, 90 C over 30 days 1.E+10 G', G'' [Pa] dg'/dt [Pa/ C] aged at 90 C, 30d, Argon, Pr1 10 z tan(δ) [-] E+09 G'' tan(δ) 30d G' 30d G'' 30d dg' 30d td E+08 G' E E+06 dg'/dt T [ C] 1.E Bohn 19

20 Shear modulus and loss factor tan(δ) aged at 90 C under argon, 10d, 20d, 30d 1.E+10 1.E+09 1.E+08 G', G'' [Pa] G'' G' tan(δ) 90 C, Argon 10 z aged over 10d, 20d, 30d 30d G' 20d G' 10d G' NP1 G' NP3 G' 30d G'' 20d G'' 10d G'' NP1 G'' NP3 G'' 30d td 20d td 10d td NP1 td NP3 td tan(δ) [-] E E T [ C] 1.E Bohn 20

21 DMA results for glass transition temperatures - 90 C in argon, 10d, 20d, 30d T G [ C] T G from maximum in tan(δ) argon, 90 C, 10z T G = 0.95 C baseline samples aged samples Tg in td [ C] Tg in G'' [ C] Tg in dg'/dt [ C] M-Tg-td-NP M-Tg-td M-Tg-G''-NP M-Tg-G'' M-Tg-dG'-NP M-Tg-dG' T G from maximum in G'' T G from maximum in dg'/dt T G = 0.88 C T G = 0.85 C -70 sample NP1 NP2 NP3 NP mean 90 C, 10d 90 C, 20d 90 C, 30d 2009 Bohn 21

22 FMC- and CL-measurements FMC CL eat Flow Microcalorimetry Chemiluminescence 2009 Bohn 22

23 O 2 -Attack sites on TPB binder chain resulting effect can be cross-linking C C C C. O - O. + - O - O. C C C. C The allylic hydrogen atoms are very susceptible for oxygen attack because the resulting radical is resonance stabilized C C C C. + O - O. - O - O. C. C C C + C 2009 Bohn / 23 C C C C. C C Recombination of two radicals on two neighbouring chains - cross-linking occurs

24 eat Flow Microcalorimetry (FMC) Used instrument TAM II (high temp. version), TA Instruments (formerly Thermometric AB), equipped with 4 microcalorimeter inserts (channels) high long term baseline stability, better than 1 µw over 20 days Measured is heat generation rate (GR) dq/dt GR is proportional to chemical reaction rate isothermal from 120 C to 150 C atmosphere: argon and air EC sampled in Duran-vials inserted in astelloy-c22 ampoules sample masses around 1.0 to 1.25 g 2009 Bohn 24

25 Measurement results (FMC) dq/dt [µw/g] GR and G of EC (TPB bonded MX) in air and argon at 120 C Q [J/g] air air sharp bend argon dq-ec-air-1 dq-ec-air-2 dq-ec-ar-1 dq-ec-ar-2 zero Q -EC-air-1 Q -EC-air-2 Q -EC-Ar-1 Q -EC-Ar argon time [d] Bohn 25

26 Measurement results (FMC) dq/dt [µw/g] sharp bend GR and G of EC (TPB bonded MX) 130 C, in air comparison with MX alone Q, EC dq-ec-air-1 dq-ec-air-2 dq-mx-air zero Q -EC-air-1 Q -EC-air-2 Q -MX-air Q [J/g] dq, EC dq, MX Q, MX time [d] Bohn 26

27 Measurement results (FMC) dq/dt [µw/g] dq-ec-air-1 dq-ec-air-2 zero Q -EC-air-1 Q -EC-air-2 GR and G of EC (TPB bonded MX) 145 C, in air Q [J/g] sharp bend time [d] Bohn 27

28 Data evaluation (FMC) times to sharp bend in GR dq/dt [µw/g] eat generation rate of EC (TPB bonded MX) 120 C in air dq-ec-air-1 dq-ec-air-2 zero dq, EC-2 8 dq, EC d d 4 0 time [d] Bohn 28

29 Results of sharp bend evaluation (FMC) Arrhenius parameterization of reciprocal times to sharp bends in GR curves ln(1/ t kdq [1/d]) 150 C evaluation of time t kdq to reach the sharp bend in GR curve GR of EC measured in air 145 C from GR (dq/dt) curves T [ C] ln(1/ t kdq [1/d]) t kdq [d] C C Ea kdq = ± 3 kj/mol 130 C lg(z kdq [1/d]) = ± R 2 = C /T [1/K] -2 Ea kdq [kj/mol] ± lg(z kdq [1/d]) ± 037 R Bohn 29

30 Chemiluminescence - instrument used instrument from ACL Instruments Inc., Switzerland Measured is light impulses per time per sample mass Signal is proportional to the chemical reaction rate isothermal conditions from 100 C to 140 C atmosphere: air, 30 ml/min sample masses around 30 mg by reaction enthalpy excited oxygen Detector 2009 Bohn 30

31 Measurement results on EC (CL) CL [cts/s/mg] 140 C Chemiluminescence CL-rate of EC (MX-TPB) in synthetic air, flow 30 ml/min CL-100 C CL-110 C C CL-120 C CL-130 C 5000 CL-140 C C C C time [d] Bohn 31

32 Data evaluation (CL) times to sharp bend in CL rate CL [cts/s/mg] Chemiluminescence of EC (TPB bonded MX) 120 C, in synthetic air CL [cts/s/mg] d 100 time [d] Bohn 32

33 Results (CL) - reciprocal times to sharp bend Arrhenius parameterization of reciprocal times to sharp bends in CL curves ln(1/ t kcl [1/d]) 130 C Ea kcl = ± 2 kj/mol lg(z kcl [1/d]) = ± 0.22 R 2 = evaluation of time t kcl to reach the sharp bend in CL-rate curve CL-rate of EC measured in air 120 C 110 C -4 1/T [1/K] CL-rate curves T [ C] ln(1/t kcl [1/d]) t kcl [d] ( ) not used Ea kcl [kj/mol] ± 2 lg(z kcl [1/d]) ± 0.22 R C This result agrees with the corresponding evaluation of GR data: Ea kdq = ± 3 kj/mol lg(z kdq [1/d]) = ± 0.37 R 2 = Bohn 33

34 Kinetic modelling of FMC-data and CL-data 1 Start: radical formation at binder material + O = O Y + OO Y - formation of hydroperoxides by oxygen attack 2a Chain propagation: forms further radical sites at binder Y + O = O YOO Y - + OO Y + OO Y - + YOO Y + YOO formation of species, which can attack the starting material and re-formation of attacking species 2b Chain branching: increasingly autocatalytically effective OO O + O 2 YOO YOO + YO + O YOO YO + O 3 Radical neutralization by antioxidant Stabilization + A Y + A Y YO + A YO + A Radical neutralizing (trapping) by forming a more O + A O + A stable radical A The three main steps are approximated by three reactions, each summarizing the individual reactions in each step Y + YO + A Y A A YOA Consumption of antioxidant A Bohn 34

35 Kinetic model: autocat. reaction + side reaction (= stabilizing reaction) A + Ox A + B AO + B k 1 k 2 k 0 B + C + S 2 B + C + S AO - B (- (- (- R,1 R,S R,2 ) ) ) primary radical formation chain branching / autocat. decomp. stabilizing reaction by AO Model for FCM-data: Q: main: first order + autocatalytic; minor: exponential a ( 1 exp( kqo t) )... Q(t, T) = OF Q + Q(te) kq1(t) + kq2(t)... + (1 a) 1 ( ) ) kq2(t) + kq1(t) exp kq1(t) + kq2(t) t Approximation by three reactions; B stands for all autocatalytically effective species Model for CL-data: Ic: main: first order + autocatalytic; minor: exponential Ic(t,T) = OF Ic ( 1 exp( k t) ) a Ico + Ic(te)... + (1 a) 1 k Ic2... (T) + k Ic1 k Ic1 (T) + k (T) exp (T) Ic2 ( k ) ) Ic1(T) + kic2(t) t 2009 Bohn 35

36 Kinetic modelling of FMC-data 2009 Bohn 36

37 Kinetic modelling of(fmc) - results primary radical formation Chain branching / autocatalytical decomposition AO - stabilizing Ea Qx [kj/mol] lg(z Qx [1/d]) k Q1 [1/d] k Q2 [1/d] k Qo [1/d] ± ± ± ± ± ±0.34 R 2 [-] SD 2 [ln(1/d) 2 ] Bohn 37

38 Results of kinetic modelling FCM data ln(k Qx [1/d]) k Qo (minor), AO consumption heat generation of EC (MX-TPB) modelling with model 'Q: main: first order + autocat.; minor: exponential' Ea Qo = kj/mol, R 2 = k Q2 (autocat.) Ea Q2 = kj/mol, R 2 = k Q1 ( intrinsic), start of oxidation Ea Q1 = kj/mol, R 2 = C 145 C 140 C C 130 C 1/T [1/K] Bohn 38

39 Kinetic modelling of CL-data 2009 Bohn 39

40 Results of kinetic modelling integrated CL data primary radical formation Chain branching / autocatalytical decomposition AO - consumption k Ic1 [1/d] k Ic2 [1/d] k Ic0 [1/d] Ea Icx [kj/mol ± ± ± 9.2 lg(z Icx [1/d]) ± ± ± 1.22 R 2 [-] SD 2 [ln(1/d) 2 ] Bohn 40

41 Discussion of kinetic modelling CL data 10 6 ln(k Ic [1/d]) k Ic2 (autocat.) chemiluminescence of EC (MX-TPB) description of integrated CL-intensity Ic by model 'Ic: main: first order + autocat.; minor: exponential' 2 k Ico (minor), AO consumption Ea Ic2 = kj/mol, R 2 = C 130 C 120 C 110 C 100 C Ea Ico = 97.8 kj/mol, R 2 = k Ic1 (intrinsic), start of oxidation At 140 C some non-linear effects occur, means contributions to signal, which are not present at lower temperatures. Ea Ic1 = kj/mol, R 2 = /T [1/K] Bohn 41

42 Results (CL) slopes of Ic during induction period reaction rate constants k Ic describing the part of stabilizer action and its consumption k Ic is obtained from the slopes of the integrated CL-intensity curves before the main oxidation peak (autocatalysis) T [ C] k Ic [cts/mg/d] R E E E E E E E+08 Ic [cts/mg] 130 C 120 C 110 C Chemiluminescence EC (MX-TPB) in synthetic air integrated CL-rate shown up to 200 million counts per mg Ic-100 C Ic-110 C Ic-120 C 100 C E E+08 Ic-130 C Ea Ic [kj/mol] 98.5 ± 5.6 lg(z Ic [cts/mg/d] ± 0.75 R E+07 time [d] 0.0E Bohn 42

43 Results (CL) slopes of Ic in induction period reaction rate constants k Ic describing the part of stabilizer action and its consumption ln(k Ic [cts/mg/d]) 140 C Arrhenius plot of rate constants k Ic obtained from slope of Ic curves before main oxidation peak C C Ea Ic = 98.2 ± 5.6 kj/mol lg(z Ic [cts/mg/d]) = ± 0.75 R 2 = C C /T [1/K] Bohn 43

44 Oxygen uptake of TPB-IPDI binder Occurs temperature dependent at different mechanisms 120 to 130 kj/mol Switch temperature: 80 C 70 to 80 kj/mol 2009 Bohn / 44

45 Ageing of TPB bonded EC using a two-step mechanism For details and worked out examples with two mechanisms see: Manfred A. Bohn, Paper 78, 40th Conference of ICT, time [d] T E = 25 C t E = 25 years two stage mechanism ageing of TPB-bonded EC Ea 2 =85 kj/mol T S = 80 C Ea 1 =130 kj/mol (T S, t S ) only Ea 2 Ea2 = 85 Ea1 & Ea2 Ea1 = 130 F = 2.8 composed 10 1 Ea 1 & Ea 2 curve, shifted to cross point at (T S, t S ) t S = 44 d F = 2.8 extrapolation 0.1 composed curve only Ea Bohn / 45 T [ C]

46 Summary and conclusions Basic stability The stability test data show that EC of type KS32 is a very stable material. Ageing program The ageing times calculated with F = 2.5 in the factor extrapolation method seem appropriate, because the values of relevant ageing processes (AO consumption, time to sharp bend) are in the range 70 to 130 kj/mol. F=2.5 is therefore a conservative extrapolation. Gap-Test A decrease of go detonation pressure was found with ageing: After 36 days at 80 C (corresponding to 15 years at 25 C) a 14% decrease resulted. This seems independent of surrounding atmosphere. DMA (in torsion) No essential shifts in glass transition temperature could be detected. Very slightly shifts between 0.5 and 1.7 C to positive values have been caused by ageing. No significant difference between ageing in air and in argon Bohn 46

47 Conclusions FMC and Chemiluminescence (CL) eat generation rate and Chemiluminescence (CL) measurements under oxidizing atmosphere revealed a kind of induction period stabilizer (antioxidant) is active In the curves well-defined bends occurred With CL it could be shown that the end of induction period coincides with the well-defined bends in FMC After AO consumption autocatalytic oxidation / decomposition starts Both methods give similar Arrhenius parameters of the reciprocal times to the bends of about 130 kj/mol AO consumption rate is measured directly with CL, activation energy of 98 kj/mol was found in kinetic modelling and in direct evaluation of integrated CL rate AO consumption rate and time to end of AO activity can be used as ageing monitor. The oxygen uptake rate has shown to follow a two-step mechanism with switching (turn-over) temperature Ts at about 80 C For extrapolation below Ts a second process with much lower activation energy must be taken into account, about 65 to 85 kj/mol. Above Ts: oxygen uptake proceeds with higher activation energy: 120 to 130 kj/mol 2008 Bohn 47

48 Thank you for your attention Questions? 2008 Bohn 48