Skip to main content

Full text of "DTIC AD0872306: Thermal Analyses Studies on Candidate Solid JPL Propellants for Heat Sterilizable Motors"

See other formats


UNCLASSIFIED 



AD NUMBER 

AD872306 

NEW LIMITATION CHANGE 
TO 

Approved for public release, distribution 
unlimited 


FROM 

Distribution authorized to U.S. Gov't, 
agencies and their contractors; Critical 
Technology; Jul 1970. Other requests shall 
be referred to National Aeronautics and 
Space Administration, Attn: JPL, Pasadena, 
CA. 

AUTHORITY 

USANWC ltr, 30 Aug 1974 


THIS PAGE IS UNCLASSIFIED 














AD Bo.- AD 87 23 06 


NWC TP 4258 



PROPELLANTS FOR HEAT STERILIZABLE MOTORS 



C-T) 

c- *, 


bv 


Jack M. Pakulak, 
and 

Edward Kuletz 
Propulsion Development 


I 

Department 


i ‘ 

J 

A) 




ABSTRACT. Times to deflagration of solid JPL propellant grains 
varying in size and geometry have been measured, and the results have 
been analyzed in light of the thermal explosion theory. In addition, 
laboratory-scale thermoanalytical techniques have been used to study 
the procedural chemical kinetics of the decomposition reactions of 
representative solid JPL propellants. The results of these studies 
indicate that, for the propellants considered, estimates of thermal 
deflagration times can be made on the basis of laboratory-scale experi¬ 
ments. The effects of sterilization heat cycling tests on solid JPL 
propellant grains varying in size and geometry have also been studied 
and the results are herein reported. 



NAVAL WEAPONS CENTER 

CHINA LAKE, CALIFORNIA * JULY 1970 


DISTRIBUTION STATEMCNT 

THIS DOCUMENT IS SUBJECT TO SPECIAL EXPORT CONTROLS AND EACH TRANSMITTAL TO 
FOREIGN GOVERNMENTS OR FOREIGN NATIONALS MAY BE MADE ONLY WITH PRIOR APPROVAL 
OF THE NAVAL WEAPONS CENTER 






NAVAL WEAPONS CENTER 

AN ACTIVITY OF THE NAVAL MATERIAL COMMAND 


M. B. Etheridge, CAPT, U8N..Commtndei 

H. G. Wlieon.Technical Dlreotor (Acting) 


FOREWORD 


The study described in this report was undertaken aa a result of 
a request by the Jet Propulsion Laboratory to the U. S. Naval Ordnance 
Test Station (now Naval Weapons Center) via letter SML-31 of 14 December 
1965. They were concerned about the potential thermal hazard or "cook¬ 
off" problem during the heat sterilization of solid propellant grains. 

Thi9 study was supported by JPL Purchase Order Nos. Z-351290 
(propellant degradation studies) and Z-351291 (heat sterilization tests). 

As requested by the Jet Propulsion Laboratory, the original manu¬ 
script was reviewed for technical accuracy by Dr. Joseph Wenograd of the 
University of Hartford, West Hartford, Connecticut. The revised manu¬ 
script was reviewed for technical accuracy by Warren W. Oshel. • 


Released by 

CRILL MAPLES, Head 

Quality Assurance Division 

15 July 1970 


Under authority of 
G. W. LEONARD, Head 
Propulsion Development Department 


NWC Technical Publication 4258 


Published by . 

Collation . 

First printing .... 
Security classification 


. Propulsion Development Department 

Cover, 36 leaves, DD Form 1473, abstract cards 

. 295 numbered copies 

. UNCLASSIFIED 













CONTENTS 


Introduction . 1 

Thermal Explosion Studies . 1 

Analysis of Thermal Explosions . 1 

Expej'iir.antal Techniques for Isothermal Analysis 

of Deflagration Times . 6 

Results of Deflagration Time Measurements . 7 

Thermoanalytical Studies . 11 

Analysis of Thermoanalytical Data. 11 

Experimental Thermoanalytical Techniques . 14 

Results of Thermoanalytical Studies. 15 

Effect of Size on Heat Sterilizable Solid Propellant Charges ... 17 

Conclusions. 20 

References. 67 

Nomenclature . 68 


iii 


























NWC TP 4258 


INTRODUCTION 


The adoption of sterilization requirements for spacecraft designed 
to land on extra-terrestrial bodies has necessitated significant re- 
evaluations of specifications regarding thermal cycling. In the area of 
solid propellant technology it has been necessary to formulate propel¬ 
lants with sufficient thermal stability to withstand repeated and pro¬ 
longed exposures to temperatures as high as 135'C. Propellants designed 
to meet the required specifications are required not only to maintain 
their mechanical and ballistic integrity through sterilization cycles, 
but also to be free of any hazards associated with thermal explosions. 

The present study was undertaken (1) to assess the degree of hazard 
associated with the use of a series of candidate solid Jet Propulsion 
Laboratories (JPL) propellant formulations, (2) to investigate the appli¬ 
cation of small-scale laboratory experiments, limited larger-scale field 
tests, and known scaling laws to predict the onset of thermal explosions, 
(3) to investigate those aspects of ammonium perchlorate propellant de¬ 
composition which lead to thermal explosions, and (4) to observe the 
mechanical-physical change on heat cycling of solid JPL propellant grains. 


THERMAL EXPLOSION STUDIES 


ANALYSIS OF THERMAL EXPLOSIONS 

In the sterilization of a rocket motor, it is desired to bring the 
body of the propellant to some high temperature and to maintain it at 
this temperature for a specified period of time. The temperature and 
time cycle must be selected in such a manner that the motor will be 
reasonably free of viable organisms. It is also necessary to assure 
.?* temperature is not so high or the time so long that the motor 
wiii deflagrate spontaneously. For this reason, it is necessary to con¬ 
sider the factors which govern the onset of thermal explosions and eval¬ 
uate the thermophysical and chemical properties of the propellants which 
are important in thermal explosion theory. With such information in 
hand it will be possible to design sterilization cycles for propellants 
which aie free from thermal explosion hazards. 


The phenomenon being considered involves the 
pellant grains. Energy is liberated in the inter! 
grain due to slow chemical decomposition. If this 
ated from the grain the temperature of the propell 
ing the reaction rate further. A balance point is 
heat generated is equal to the heat liberated. Th 
described by a critical temperature (T ) for a giv 
ticular geometry. When the decomposition kinetics 


self-heating of pro- 
or of the propellant 
energy is not liber- 
ant will rise, increas- 
reached at which the 
is balance point is 
en propellant in a par- 
are presumed to be 


1 















NWC TP 4258 


governed by zero-order kinetics the condition of a heated propellant 
will be described by the differential equation 


-XV T + pc 


(£)- 


pQAe 


-E/RT 


( 1 ) 


Properties of this equation and its solution under a variety of conditions 
have been described before (Ref. 1, 2, 3, 4, 5). The decomposition of a 
small amount of propellant can generate a large amount of heat. Conse¬ 
quently, the amount of propellant decomposed just prior to thermal defla¬ 
gration can be quite small. For this reason the zero-order model is 
quite appropriately applied to a number of explosion situations. 

In the sterilization of a rocket motor the sample is heated to some 
reasonably high temperature for a fairly prolonged period of time. Under 
these conditions it is appropriate to examine the properties of Eq. 1 
under steady-state conditions of QT/3t) = 0. When the resulting equa¬ 
tion is solved and rearranged it is found that there is a temperature T m , 
(maximum surface temperature) which will permit the maintenance of a 
steady-state temperature within the body of the solid. If the surface 
temperature Tj is such that T]^ > T m an explosion will result. __ 
then there will be no thermal explosion according to this zero-order 
model. T m is given by: 


If T 1 < T m 


T = 
m 


2.303R log 



( 2 ) 


The values of the parameter 6 for the slab, the solid cylinder, and the 
solid sphere are 0.88, 2.00, and 3.32 respectively. Some values of 6 
for hollow cylinders are given in Table 1 (Ref. 2). This critical tem¬ 
perature T m is a key parameter used by Zinn and Mader (Ref. 4) and Zinn 
and Rogers (Ref. 5) and others to describe and predict thermal explosion 
properties. It was found (Ref. 4) and (Ref. 5) according to the model 
ot Eq. * with the condition that the surface temperature Ti > T m then 
the explosion time, t e , is given by m 


- e . F /l_ _ E\ 

T \ T m T l) 

where the dimensionless time t /t is a function of (E/T - E/T ). 

e ml 7 ' 


(3) 


k 


ilikUfiii 





TABLE 1. Shape Factors for the Follow Cylinder. 


Radius 

ratio 


0,025 

0.05 

0.10 

0.15 

0.20 

0.25 

0.30 

0.35 

0.40 


Shape 


Same temperature 
inside and outside 


3.008 

3.320 

3.892 

4.492 

5.171 

5.966 

6.921 

8.091 

9.556 


0.45 


.50 


.55 


.60 


.65 


11.43 

13.88 

17.19 

21.81 

28.55 


factor, 6 

Adiabatic inside, 
constant temperature outside 

2.003 

2.020 

2.077 

2.172 

2.306 

2.483 

2.713 

3.004 

3.384 

3.878 

4.530 

5.413 

6.646 

8.434 


.70 


.75 


.80 


.85 


.90 


.95 


38.91 

56.10 

87.73 

156.1 

351.3 

1405 


11.17 

15.69 

23.92 

41.57 

91.50 

358.4 


1.00 


©o 


oo 



















NWC TP 4258 


Here t is the thermal time constant of the sample. 


2 

a 

a 


- where (ct 



(A) 


In the case of a solid cylinder, it was found (Ref. 4) that 

(E/T - E/T.) - 1.6 when t /r - 1. 
ml e 

Thus, when t “ x 
* e 


=- - + Y" (E in cal/mole) (5) 

m l 

A similar expression for the sphere is 


+ |- (E in cal/mole) (6) 

m A 1 

When t 4 T, Fig. 3 of Ref. 4 gives the appropriate value of X in the 
equation: 



Equations 3 through 7 permit the calculation of T m for a given explosive 
malarial from one or from a series of measurements of t e (explosion time) 
as a function of (ambient or surface temperature), provided a value 
of E is known. Once E and T m are known, values of t e for any imposed 
value of T]_ can be computed. These considerations apply as long as the 
zero—order model is effective and the amount of reactant consumed prior „ 
to deflagration is negligible. 

At low temperatures near or even below T m , the deflagration time 
(t e ) can be much longer than the thermal time constant (t). Under these 
conditions it becomes necessary to apply a first-order model since re¬ 
actant depletion effects become significant. Under these conditions 

the right hand side of Eq. 1 becomes p QAwe is necessary to 

consider 

dv/dt - -„:>- E/RT ( 8 ) 


Where w is the mass fraction unreacted at time, t. 


4 









NWC TP 4258 


However, when t e >> 1 the propellant sample may be regarded as approxi¬ 
mately isothermal and Eq. 8 may be integrated (Ref. 5) to give 


t - . MlzfL 
e A -E/RT 
Ae 


(9) 


where f 

A -E/RT 
Ae 


is 

is 


the fraction reacted at the time of deflagration, 
the specific rate constant k 


Since 


k = 


ln(l-f) 

t 

e 


( 10 ) 


This equation permits the calculation of the specific reaction rate 
constant from deflagration time data if f is known. The value for f 
can be estimated from the k determined via DTA data. 

The value of k can be calculated from deflagration data in two ways, 
depending on the test conditions. If the test is conducted at a temper¬ 
ature near or at the critical temperature of the test sample, the data 
under steady-state conditions with constant T (Eq. 1) can be integrated 
and as an approximation yields: 


K = 


4 X 


QP a 


( T c - h) 


an 


where T c is the center temperature (Ref. 1). The second technique in¬ 
volves an adiabatic condition where Eq. 1 is described for a no-heat 
transfer condition from the center of the propellant to some other point 
in the grain. This condition is given by Eq. 12 


k 


Q Vdt/ 


( 12 ) 


which measures the rate of temperature change with time at the geometric 
center of the sample. Under these conditions f can be calculated from 
Eq. 10 and with either Eq. 8, 11 or 12. 

The temperature region of interest in sterilization studies is the 
region below T m . This is the region where the zero-order theory predicts 
no deflagration, and deflagrations occur only as a result of first-order 
reaction effects when a significant quantity of the material has reacted. 


5 





F 


NWC...1P 4258 _ 

The deflagration studies are carried out in the temperature range above 
T m in order to provide a continuity for the study of effects in the lew 
temperature, t e >> T, isothermal range. 

EXPERIMENTAL TECHNIQUES FOR ISOTHERMAL ANALYSIS 
OF DEFLAGRATION TIMES 

The propellants used in this study were of the PBAA - MAPO - AP 
type. Forty-eight solid and tubular, right-circular cylindrical propel¬ 
lant grains of various sizes and two 60-pound motors were prepared by 
JPL. All the propellant grains used fo.r che deflagration and heat ster¬ 
ilization studies were of propellant batch JS-7 identified as HSP-X6 
made with batch 5134 oxidizer and an equivalence ratio, PBAA to MAPO, of 
1.5:1. The size, shape, and number of cylindrical grains used is shown 
in Table 2. 


TABLE 2. Number and Dimensions of Samples (JS-7) 


Charge size (OD x length, inches) 


Charge shape 

1 x 2.5 

3 x 7.5 

5 x 12.5 

10 x 25 

Solid cylinder 

5 

5 

5 

1 

Tubular (d/D = 0.333) 

5 

5 

5 

1 

Tubular (d/D = 0.500) 

5 

5 

5 

1 


In isothermal analysis as customarily performed at the Naval Weapons 
Center (NWC), China Lake, Calif., propellant samples ranging in diameter 
from 0.1 to 12 inches with an L/D of one or greater can be processed. 
Since the sample size varies considerably, the type of oven used also 
varies. 

For small samples (< 0.25 inch diameter) a quartz test tube (0.3 
inen ID) in a tube-type furnace (Serial 634304, Marshall Products Co., 
Columbus, Ohio) is used. For all other samples up to 6 inches in diam¬ 
eter, standard cook-off ovens (8.5 inch ID) are used. These are made 
from 3/8-inch thick stee.l pipe, 8.5 inches in diameter, and 4 feet long. 
The pipe is covered with heating elements and is insulated within a 
transite box, 19 inches wide, 25.5 inches high, and 51.5 inches long. 
Lightweight fiberglass plugs approximately 6 inches long are used in 
each end of the oven. They are a tight fit to minimize convection heat 
losses, but release readily for the hot gases formed during deflagration. 
Samples above 6 inches in diameter are contained within a scaled-up 
standard cook-off oven. 




hh ^ii hh i n iii h wnH flP 



r 




M 

i 

H 

1 


NMC TP 4258 


Another kind of oven can be used in these studies. This oven is 
made of thick-walled aluminum tubing (from 0.5 to 1 inch thick) with 
dimensions to fit the sample being tested. A ribbon-type heater, about 
1,000 watts, is wrapped on the outside of the tube. 

For a test, the oven is preheated to a selected temperature before 
placing the sample in the oven. However, it is not practical to preheat 
an aluminum tube-type oven before introducing the sample. With larger 
test samples (>1 inch diameter), the sample is instrumenced with thermo¬ 
couples (cross-sectional area at the center of the grain length) and 
wrapped in aluminum foil and then located centrally in the oven. The 
temperature of the oven is controlled with a time-proportioning action 
power supply. The power supply for the scaled-up (12 inch diameter 
samples) cook-off oven is of the on-off type with a variable range of 
about ±1°C. 

The test is monitored remotely with a 6- to 24-point temperature 
recorder until deflagration occurs, or until the test is terminated. 

The data from this study are plotted as log time versus the reciprocal 
absolute temperature. The data from the instrumented samples on each 
test show the rate of temperature rise in the sample, and the occurrence 
of exotherms before deflagration or before the termination of the test. 
After a series of tests are completed at different oven temperatures, a 
plot of log time to the exotherm or deflagration versus the reciprocal 
absolute oven temperature is made. This type of plot shows a relation¬ 
ship between the tests performed until a point in temperature is reached 
where the line curves to approach some critical point in temperature 
below which deflagrations or exotherms would approach an isothermal 
condition. Generally, a test is designed for a deflagration or an exo¬ 
therm to take place within about 2 weeks. 


RESULTS OF DEFLAGRATION TIKE MEASUREMENTS 

Most of the grains listed in Table 2 and one of the 60-pound motors 
were subjected to an evaluation of deflagration time as a function of 
temperature. Plots of the deflagration data on eight grains are shown 
in Fig. 1 through 10. All the data are summarized in Table 3. The time 
required for the surface of the grain to reach oven temperature was 
measured for each grain studied. This warm-up time was subtracted from 
the total elapsed time-to-deflagration. This gave a time-at-temperature 
or deflagration-time which was used in subsequent evaluations of the 
data. For comparison purposes, four grains were tested in the aluminum 
tube ovens described above. The warm-up time is very short in this type 
of oven. 

All deflagration-time data (minus the warm-up time) was plotted as 
deflagration time versus 1/°K for the grains tested. The graphic display 
of this reduced data for the 1-inch diameters is given in Fig. 11, 3-inch 
diameters in Fig. 12, 5-inch diameters in Fig. 13, and 10-inch diameters 


i 

i 


7 






NW' 



TABLE 3. Cook-Off Data for Various Propellants 
With an L/D Ratio of 2.5. 


Designation 
JPL BD - 

Diam., 
in. 

Grain 

type a 

Oven 

type 

m 

Total 
time to 
cook-off, 
hr 

Time to 
cook-of f 
at temp., 
hr 

79/2 

i 

Solid 

Air 

19 3.3 

11.6 

9.6 

78/3 

i 

Solid 

Air 

198.9 

7.0 

5.2 

78/2 

i 

Solid 

Air 

201.7 

6.5 

4.3 

79/3 

i 

Solid 

Air 

219.4 

2.9 

1.2 - 0.1* 

79/7 

i 

Tub.-1/3 

Air 

215.6 

3.0 

1.2 - 0.2* 

79/6 

i 

Tub .-1/3 

Air 

226.7 

1.7 

0.65 - 0.2* 

78/5 

i 

Tub .-1/2 

A1 

207.2 

NCO b 

— 

79/4 

i 

Tub.-1/2 

Air 

218.9 

2.0 

0.9 - 0.2* 

79/5 

i 

Tub.-1/2 

Air 

223.9 

1.8 

0.73 - 0.2* 

71/5 

3 

Solid 

Air 

160.0 

NC0 c 

— 

81/2 

3 

Solid 

A1 

165.6 

NCO d 

— 

77/2 

3 

Solid 

Air 

166.1 

62.5 

53.0 

76/2 

3 

Solid 

Air 

177.8 

35.7 

24.7 

80/4 

3 

Solid 

A1 

172.2 

33.6 

33.6 

77/4 

3 

Tub.-1/3 

Air 

183.3 

23.0 

17.0 

76/4 

3 

Tub.-1/3 

Air 

183.3 

22.0 

17.0 

81/4 

3 

Tub .-1/3 

A1 

187.8 

16.5 

1( .5 

71/6 

3 

Tub.-1/3 

Air 

190.6 

11.5 

7.0 

81/3 

3 

Tub .-1/2 

Air 

190.6 

11.8 

6.3 

80/3 

3 

Tub.-1/2 

Air 

196.7 

9.8 

3.8 

76/3 

3 

Tub.-1/2 

Air 

201.7 

8.0 

2.5 

68/3 

5 

Solid 

Air 

164.4 

84.9 

70.9 

83/2 

5 

Solid 

Air 

165.6 

58.7 

43.2 

69/6 

5 

Solid 

Air 

177.2 

35.2 

18.7 

70 / 6 

5 

Solid 

Air 

176.7 

40.5 

23.5 

82/2 

5 

Solid 

Air 

182.8 

32.0 

12.0 

69/4 

5 

Tub.-1/3 

Air 

167.2 

63.3 

52.3 

68/2 

5 

Tub.-1/3 

Air 

176.7 

34.8 

22.8 


8 









TABLE 3. (Continued) 


Designation 
JPL BD - 

Diam., 
in. 

Grain 

type 3 

Oven 

type 

Oven 
temp., 

°C 

Total 
time to 
cook-off, 
hr 

Time to 
cook-off 
at temp., 
hr 

82/4 

5 

Tub.-1/3 

Air 

182.8 

24.6 

14.1 

68/6 

5 

Tub.-1/2 

Air 

173.9 

32.8 

20.3 

82/3 

5 

Tub.-1/2 

Air 

182.2 

28.0 

16.0 

81/4 

5 

Tub.-1/2 

Air 

191.1 

7.3 

2.8 

71/2 

10 

Solid 

Air 

153.3 

__e 

— 

77/6 

10 

Tub.-1/3 

Air 

156.7 

151.5 

106.5 

78/9 

10 

Tub .-1/2 

Air 

156.7 

129.5 

94.5 

79/9 

10 

Tub .-1/3 

Air 

162.8 

74.4 

— 

82/6 

10 

Tub.-1/3 

Air 

168.9 

67.7 

37.7 

87/1 

10 

Tub.-1/2 

Air 

168.9 

86.7 

36.7 

Motor^ 

12 

Tub.-1/2 

Air 

163.3 

86.4 

72.4 


a Tub . = Tubular grain, ratio ■ inside diameter/outside diameter. 


b No cook-off. A 5.6°C exotherm after 2.4 hours and a 14.4°C exo¬ 
therm after 5.1 hours. 

No cook-off to 116.7 hours, oven temperature then increased 16.7°C, 
cook-off occurred at 5 hours. 

^No cook-off to 167.4 hours. Cook-off in 70.4 hours at 176.7°C. 
(Temperature raised to 176.7°C after 167.4 hours.) 

Q 

Cook-off at 77.5 hours on second cycle. 

^L/D ratio = 1. 

* 

Time in hours allowed for AP crystal transition to take place. 


in Fig. 14. The experimental critical temperatures (Table 4) were deter¬ 
mined from the above plots on each size via Eq. 4 and 5 using the indi¬ 
cated 1/Tj_ values shown on Fig. 11 through 14. The experimental critical 
temperature corresponds to a sudden change in the above plots of defla¬ 
gration time versus temperature. In the plot of 10-inch diameter grains, 
there was not a sufficient number of grains available for testing to show 
this effect. Since only ie deflagration test was made using a grain 
cast in motor hardware, a reduced time plot (i.e.. Fig. 11 through 14) 
could not be accomplished. The critical temperature determined from this 
single test and Eq. 4 and 5 was 155°C. 






TABLE 4. Critical Temperature Data on 
JS-7 Propellant (6 - 2.00). 


Diameter of Grain 
(inch) 

T , °C a 
m 

T , °C b 
m 

1 

212.2 

215.0 

3 

181.7 

186.7 

5 

168.9 

173.9 

10 

152.2 

157.2 

12 

148.3 

155.0 

30 

128.9 

(136.7)* 


^ased on DTA kinetic data. 


Based on deflagration data. 

* 

Based on experimental data from Fig. 15. 


A plot of these values for the experimental critical temperature 
versus diameter is given in Fig. 15. Also included is a plot of the DTA 
predicted critical temperature. The experimental and predicted critical 
temperature values are close, but the experimental critical temperature 
value indicates a slightly higher activation energy (34.2 kcal per mole) 
than that predicted by DTA (33.3 kcal per mole). The values for T m 
obtained from these experiments are compared with those obtained from 
DTA studies in Table 4. The data indicate that the first DTA exotherm 
controls the deflagration. The reduced data plots for each size show 
several different conditions concerning these grains. One is that these 
deflagration tests show an isothermal condition cannot exist in a tubular 
grain without having a large heat sink available to maintain a constant 
surface temperature. The aluminum foil used inside the tubular grains 
was not sufficient to maintain constant surface temperature and the in¬ 
side surface of these grains approached an adiabatic condition. In such 
a condition, the shape factor does not change greatly from that of a 
solid. A shape factor of 2.0 was used in all of these calculations. 

The other feature is that at relatively low temperatures, where 
Tl < Tm such that deflagration times are much longer than the reduced 
time (Eq. 4), the chemical reaction or decomposition can be regarded as 
approximately Isothermal and the deflagration times could be determined 
by Eq. 9 and 10. At a temperature where > T m> the isothermal approxi¬ 

mation would not apply and the deflagration times would be shorter. 

Data from grains that deflagrated at temperatures below their critical 
temperature were used to calculate the fraction reacted, f, at the time of 
deflagration. This was accomplished by using Eq. 9, 10, and 11. The value 
obtained for f was 0. 073. The data calculations are given in Table 5. 


10 











NWC TP 4258 


TABLE 5. Extant of Reaction, /, Calculations at 
the Time of Deflagration. 


Test no. 

BD - 

Grain size, 
in. 

t /a 3 , 
sec/cm 3 

T c “ Ti, 

°C 

f 

Oven 

type 

68/3 

5 

6.33 x 10 3 

1.94 

0.107 

Air 

83/2 

5 

3.86 x 10 3 

1.66 

0.057 

Air 

70/6 

5 

2.1 x 10 3 

2.22 

0.042 

Air 

69/6 

5 

1.67 x 10 3 

2.22 

0.034 

Air 

80/4 

3 

8.24 x 10 3 

1.66 

0.12 

A1 

77/2 

3 

13.1 x 10 3 

0.55 

0.065 

Air 

76/2 

3 

6.12 x 10 3 

1.66 

0.089 

Air 


Average value of f = 0.073. 


Since Eq. 9 is independent of mass, a plot of time to deflagration 
after attaining test temperature versus this test temperature was deter¬ 
mined for all grain sizes tested below their critical temperature. The 
plot is shown in Fig. 16. The slope of this plot has a value of 26.2 
kcal per mole for the activation energy. 


THERMOANALYTICAL STUDIES 


ANALYSIS OF THERMOANALYTICAL DATA 


The ability of a propellant to survive the sterilization environment 
is dependent in large measure on the kinetics of its thermal decomposition 
reactions. If appropriate rate parameters were known they could be useo 
together with the equations described above to make a proper assessment 
of the susceptibility of a given propellant grain to thermal explosion. 

Two small-scale laboratory techniques, differential thermal analysis 
(DTA) and thermogravimetric analysis (TGA) have been employed in this 
study to measure the decomposition parameters of JPL propellants. These 
are particularly appropriate since they measure quantities (heat evolution 
and weight loss) which are particularly related to thermal explosions, and 
because they are conveniently and rapidly applicable to solid propellants. 
As such, these techniques would be most valuable for the purpose of screen¬ 
ing candidate propellant formulations if their relationship with thermal 
explosion behavior could be established. 


11 


'JUH.iUlil.1 : :rt:J * JU 







NWC TP 4258 


Differential thermal analysis involves continuous measurement of 
the temperature difference between a small sample (usually 10 to 75 mg) 
and a thermally inert reference material, as a function of the sample 
or reference temperature and/or time, as both the sample and reference 
material are heated simultaneously at a predetermined rate. When a 
reaction occurs, changes in the heat content and in the thermal proper¬ 
ties of the sample are indicated by a deflection or peak which shows on 
the graphic record. If the reaction proceeds at a rate varying with 
temperature (i.e., possesses an activation energy), the position of the 
peak varies with the heating rate, providing that other experimental 
conditions are maintained fixed (Ref. 6 and 7). Five or more DTA runs 
are made at different selected constant heating rates for each sample. 

The decomposition rates associated with the exotherms are determined 
in the following manner. The activation energy and frequency factor are 
determined by the modified variable heating rate method (Ref. 2 and 8) 
in which is plotted, for a particular exotherm, the log-heating rate over 
the absolute peak temperature squared versus the reciprocal absolute 
peak temperature. The absolute peak temperature is taken with respect 
to the reference temperature rather than the sample temperature. In the 
thermal stability studies, the exotherms occurring in the lower tempera¬ 
ture ranges of a thermal pattern are evaluated first. With this type of 
graph the activation energy (E) calculations are made, based on the equa¬ 
tion (Ref. 6): 


d log (*/T^) 
d (1/T rp ) 


E 

2.303 R 


Once the above values for E are known, the frequency factor (A) (sec ) 
can be calculated from Eq. 14, assuming the order of reaction to be 
unity. 

_ . -E/RT 

-- Ae P (14) 

R T 

rp 


The determination of the specific rate constant k (sec ) can be made 
using the E and A values and the Arrhenius rate equation: 


, . -E/RT 2.303 , o\ 

* “ Ae “ t l0 * [rj 


12 










we TP 4258. 


Values for k can be determined at selected temperatures. The specific 
rate constant can be used in estimating the percent reacted for a given 
endothermic or exothermic peak. The temperature (T) is absolute, time 
(t) is in seconds, with C Q the initial concentration, and Cj the concen¬ 
tration after a specific time and at a specific temperature. The actual 
concentrations need not be known for calculations since first-order or 
pseudo first-order reactions are assumed. 

A second method for obtaining kinetic data from a DTA thermal pro¬ 
file is the area method proposed by Borchardt and Daniels (Ref. 9) and 
Reed, et al (Ref. 10). The method is based upon Eq. 16, where the heat 
capacity of the sample is assumed to be quite small compared with the 
heat of reaction. 


k 


KT 


(16) 


where AT is the temperature difference between the sample and reference 
material (deviation from the base-line) and is proportional to the rate 
of reaction at the sample temperature. The quantity |A t | is the differ¬ 
ence between the area under the entire thermogram peak and the area trace- 
out up to the particular temperature T. The value of |A t | is assumed to 
be proportional to the amount of unreacted material. The activation 
energy and frequency factor are determined from Eq. 15. By using the 
area method, kinetic parameters can be determined from a single OTA pat¬ 
tern, providing the exotherms do not overlap. 

Inherent in the development of both these methods of treating DTA 
results is the assumption that the rate of reaction is independent of 
time. The low temperature decomposition of ammonium perchlorate and of 
ammonium perchlorate propellants does appear to be time dependent and 
the application of the Kissinger method and the area method should be 
regarded as procedural kinetic rate measurements. Because of the exis¬ 
tence of an "induction period," and because the "induction period" may 
vary with temperature, the kinetic results should be regarded as only 
approximate. 

The TGA technique is also applicable to the study of propellant 
decomposition. Thermogravimetric analysis consists of the continuous, 
or frequently repeated measurement of the weight, or change-in-weight 
of a specimen as it is subjected to a temperature program. Generally, 
it will include automatic and continuous recording of weight, or change- 
in-weight, as the sample is subjected to a constant temperature, or to a 
selected uniform, or changing, dynamic heating rate. The temperature 
programming and/or control is usually of the same general nature as for 
differential thermal analysis and for isothermal analysis. 


13 








NWC TP 4258 


EXPERIMENTAL THERMOANALYTICAL TECHNIQUES 

The differential thermal analysis apparatus used in these studies 
was designed and fabricated at NWC. This apparatus, which is operational 
to approximately A50°C, is described as follows: 

The sample (usually 10 to 75 mg) is massed about the tip of the 
thermocouple probe which is housed within a glass reaction tube^-. The 
reaction tubes are constructed of Pyrex, 3/8 inch diameter and approxi¬ 
mately 5 Indies long, with a 10/30 ground glass female joint on the 
upper end. A glass sidearm is attached to the midpoint of the reaction 
tube. The sidearm extends at a 45 degree angle for about 2.25 inches, 
then horizontally for about 3 inches. This sidearm can be employed ad¬ 
vantageously in a number of applications, e.g., (1) addition of glass 
beads without contamination of the ground glass joints, (2) for making 
runs under dynamic gas flow systems, (3) for containment of toxic reac¬ 
tion products by utilizing a balloon or by exhausting the products to an 
exhaust port via a flexible tube, (4) for maintaining the dry-box con¬ 
dition of the sample within the reaction tube by the use of a balloon 
or a drying tube, and (5) for drawing a partial vacuum on the sample. 

The thermocouple probes, which have ground glass male joints, are 
placed within the reaction tubes; the ground glass joints give a secure, 
leak-free connection. The reaction tube is filled with glass beads 
(0.05 - 0.1 mm diameter) through the sidearm of the tube to completely 
cover the sample and probe tip. A second reaction tube, containing the 
inert reference probe, is filled only with the glass beads. A third 
reaction tube contains a thermocouple probe which is arranged in a manner 
similar to the inert reference tube, and the output from this thermocouple 
is used to record the actual reference temperature simultaneously with 
the differential thermal temperature from the sample and reference thermo¬ 
couples. All three reaction tubes are filled with glass beads up to 
identical levels, approximately 0.5 inch. The three reaction tubes are 
inserted into their respective wells in the aluminum heating block (2.25- 
inch diameter x 3-inch long cylinder); the cylinder is placed within a 
pint-size Dewar flask, which affords protection against rapid heat loss 
and yields a more uniform heating rate. 

The heating unit, a tube-type 140-watt cartridge heater, is located 
within the center of the aluminum block. It is connected to and con¬ 
trolled by a power proportioning temperature programmer (Model 240M-00, 
Hewlett-Packard Co., Palo Alto, California). The programmer senses the 
temperature of the block by means of a thermocouple located within the 
block, midway between the two sample wells. The differential output 
from the sample-inert reference thermocouples is transmitted to the 
microvolt amplifier (DC microvolt-ammeter, Model 425A; Hewlett-Packard 
Co., Palo Alto, California) and then fed into a two-pen potentiometric 
recorder which produces a thermal pattern (thermogram) as the temperature 
of the system is increased in a linearly controlled manner. However, the 


^A schematic drawing of a aimilar-type tube with slightly varying 
dimensions is shown in NOTS TP 2748. 


14 







HWC If 


I 

i 



amplifier is not necessary if a properly equipped Honeywell Model 194 
recorder is used. The recorder can be adjusted to give a zero reading 
at any portion of the chart, thus both endothermic and exothermic reac¬ 
tions can be recorded adequately. On the same chart, the temperature 
of the system is recorded simultaneously with the thermogram. This per¬ 
mits the direct determination of any temperature level on the DTA thermo¬ 
gram and, since the chart speed is known, the true heating rate can be 
determined easily. This type of DTA equipment, with minor modifications, 
has been in use for over 6 years and has produced several thousand 
thermal patterns. 

For TGA studies, a Cahn elecrobalance is used to continuously weigh 
a sample (usually 50 to 100 mg) as a function of temperature and/or time. 
This balance (Ventron Instrument Corp., Cahn Division, 7500 Jefferson 
Street, Paramount, California) is a null-point instrument in which a 
light source photocell is the detector and an electromagnetic D’Arsonval 
movement supplies the restoring force. The loop gain of the servo system 
is in excess of 1,000, so that the actual beam deflection under load is 
very small, and the balancing torque is essentially equal to the sample 
torque. The torque motor used in the balance is as linear as precise 
weights and precision potentiometers can determine. Thus, the balancing 
current is a direct measure of the sample weight to an accuracy of better 
than ±0.1% and a precision of better than ±0.01% of full scale sample 
weight. 

Any change in sample weight causes the balance beam to move. This 
motion is detected by the change in photocell voltage because of the 
movement of the shutter attached to the beam, which interrupts a light 
source. The photocell voltage is amplified and then applied to the coil 
attached to the beam. The coil is in a magnetic field and current pass¬ 
ing through it exerts a force on the beam, restoring it to a null position. 
The coil current is thus a measure of sample weight. 


RESULTS OF THERMOANALYTICAL STUDIES 

Propellants JS-1, JS-2, JS-3, and JS-7 were studied in the DTA 
apparatus. The identification and the equivalence ratio, PBAA to MAP0, 
for the four propellant samples were HSP-X4 (1:3), HSP-X5 (1:1), HSP-X6 
(1.5:1) and HSP-X6 (1.5:1), respectively. Cup type thermocouples were 
used for these materials in the representative thermograms shown in 
Fig. 17. Point-junction type th raocouples were used for obtaining the 
kinetic runs for the four propel.’'-its as shown in Fig. 18 through 29. 

The results of the kinetic study are plotted in Fig. 30 through 33 
respectively, for JS-1, JS-2, JS-3, and JS-7. The JS-1 gave very erratic 
results, even on a repeat DTA study. This may be due, in part, to the 
overlapping of the first and second exotherms. The activation energy as 
determined by Eq. 14 was about 15 kcal per mole. This is a very lew value 


15 










NWC TP 4258 


for E and may be due to the high concentration of MAPO present in the 
propellant. The JS-2 propellant had an activation energy of 33.2 kcal 
per mole as determined by Eq. 1A for the first exothermic peak while 
the JS-3 propellant had a value of about 27 kcal per mole. The study 
on JS-2 and JS-3 propellants was limited to results from the DTA tests. 
The JS—7 propellant was studied more in detail than the other propel¬ 
lants since JPL had selected this propellant for the deflagration and 
heat sterilization test grains. For the JS-7 propellant, the activation 
energy for the first exotherm was 33.3 kcal per mole, using Eq. 1A. The 
area method (Eq. 15 and 16) for this exotherm gave 31.A kcal per mole, 
which used only the peak value of each DTA. The plot is shown in Fig.' 3A. 
The area method results are similar to the results obtained by using the 
variable heating rate method. The area method results are given for 
comparison purposes only and were not used in predicting the critical 
temperature. 

In predicting the critical temperature from the DTA data using 
Eq. 2, certain values had to be obtained for the density, heat capacity, 
and thermal conductivity. The density, p, was given by JPL as 1.76 g/cm 3 . 
The heat capacity, c, was estimated at 0.3 cal/g. The thermal conduc¬ 
tivity, A, was measured from 3- and 5-inch diameter solid grains of the 
propellant when they were being heated in an air-type oven. The equation 
used for determining the thermal diffusivity (<*) is given as: 



where: 


dT 

— = heating rate at the surface of the grain (°C/sec) and 


A = cpa 


The average value for A was 0.00053 cal/(cm)(sec)(°K). The predicted 
temperature values for different sizes and shapes of JS-7 propellant are 
given in Table A. 

Weight-loss studies were also performed on small samples (~100 mg) 
of JS-7 in the temperature range of 193 to 232°C. The sample would 
initially lose about 0.7% of its weight at the selected test temperature. 
After this initial loss, the sample would lose weight slowly as would be 
expected, but after a time depending on the test temperature, the weight 
loss rate would increase very rapidly. This temperature dependency of 
the weight loss rate was plotted as an Arrhenius function using Eq. 15. 
This plot is shown in Fig. 35. The two weight loss steps in Fig. 35 
show about the same activation energy of 3A to 35 kcal per mole but 


16 
















differ in rate of decomposition by about a magnitude of 10. This would 
indicate that the decomposition paths are the same in each case, but 
that the faster or third weight-loss step is either dependent on the 
second weight-loss step or on some other time delay function. Both 
weight-loss steps appear to have about the same decomposition path as 
the first DTA exotherm also shown in Fig. 35. 


EFFECT OF SIZE ON HEAT STERILIZABLE 
SOLID PROPELLANT CHARGES 


An experimental program was carried out to determine the number of 
135°C heat cycle that would produce cracking or voids in propellant 
charges of various sizes and shapes. One cycle would consist of the 
time to heat tne slowest-heating portion of the charge to 134°C (as mea¬ 
sured by the thermocouple in the control sample) plus 53 hours at 135 ± 
1°C. The heat cycle time would not include the cool-down period. Visual 
inspection and possibly photographic coverage would be made on each 
charge prior to and after each cycle. 

The test (heat) cycle, for each diameter size, was determined by 
employing a charge that contained a thermocouple which was located within 
the geometrical center of the propellant mass. The ovens were of the 
same types used for the deflagration studies. The setup was monitored 
closely to determine the time lapse (dwell time) from the moment of 
sample insertion into a preheated oven to the time when the internally 
located thermocouple reached 134°C. The internal temperature of the 
charge was then maintained at 135 ± 1®C for 53 hours. It required 2 to 
3 cycles to obtain the correct operational parameters. Subsequent cycles 
were then run at the same operational oven temperature for identical time 
periods. Usually, a second sample, same diameter but without an internal 
thermocouple, was run under the established operational parameters. 

The 3-inch diameter propellant charge required 6 to 7 hours of dwell 
time in a preheated oven before the internally located thermocouple 
reached 134°C; the dwell times for the 5- and 10-inch diameter charges 
ranged from 29 to 30 and 45 to 53 hours, respectively. It was determined 
that che duration of the test cycles for the 3-, 5-, and 10-inch diameter 
charges would be 60, 82.5, and 104 hours, respectively; the duration 
included dwell time plus 53 hours at 135 ± l'C. The test cycle was termi¬ 
nated by withdrawing the charge from the oven. This vas accomplished by 
employing a sled-like device that held the charge and also pulled forth 
an oven closure-port when the charge was removed remotely at the comple¬ 
tion of the cycle. After the charge was removed from the oven, it was 
allowed to approach the ambient air temperature. For the 3-inch diameter 
charges, the internal temperature was reduced quickly to 42-50°C after 
only 3 hours and to 37°C after an additional 2-hour period. However, 










NWC TP 4258 


for the 10-inch diameter charges much longer cooling periods were re¬ 
quired, as the larger mass afforded better insulation. The internal 
temperature of the 10-inch sample was reduced to 88-90°C after 10 hours 
and to only 63-66°C after an additional 6 hours. Observations were made 
on the samples at the completion of each test cycle. In addition, X-ray 
pictures were taken of the 12-inch motor after the first and second 
cycles, and they were sent to the Jet Propulsion Laboratory. The results 
of the cycling tests on the 3-, 5- and 10-inch diameter samples are sum¬ 
marized and depicted graphically in Fig. 36. 

Because of the mass effect, it requires more time for the center of 
a larger solid grain than for the center of a smaller grain to reach 
134°C when both are placed simultaneously within a preheated 134 to 135°C 
oven. Therefore, the durations of the test (heat) cycles, used in pre¬ 
paring Fig. 36, were adjusted to compensate for this difference. First, 
the temperature-time data obtained from a thermocouple located within 
the geometrical center of the charge was plotted (Fig. 37 through 39). 

Then, a tangent was drawn for each curve, as illustrated. The intercept 
of the tangent with the sterilization temperature level gave the time 
correction value for each diameter. The test (heat) cycle periods of 
60, 82.5, and 104 hours were adjusted to give 54, 69, and 75 hours for 
the 3-, 5-, and 10-inch-diameter charges, respectively. The adjusted 
values were utilized in preparing the data presented in Fig. 40. This 
plot shows the apparent empirical relationship between the diameter and 
the time to Che appearance of cracks or voids when the propellant charges 
are held at the sterilization temperature level. Figure 40 also shows 
the period (not adjusted) when cracks were observed first. 

A 60-pound, 12-inch diameter motor, with a d/D ratio of 1/3 and a 
L/D ratio of 1, and containing JS-7 propellant, was cycled until the 
motor failed. The propellant thermocouples (Fig. 41) were inserted at 
Che nozzle end and then fed out through the igniter opening. Before the 
start of the first cycl , both the igniter and the nozzle ends were 
sealed and then pressure tested at about 20 psia for leaks. The period 
of cycling for the motor was monitored by thermocouple No. 2 located at 
the inside surface of the propellant grain (Fig. 41) . During the first 
test cycle, che oven temperature had to be raised carefully to avoid 
"over-shooting" the target temperature of 135 ± 1°C inside the motor. 

On the second cycle test, the oven temperature had to be increased by 
1°C ao that the lower limit would be reached. Apparently, some self¬ 
heating occurred during the first test cycle. The initial oven controller 
temperature was set at 138°C, then during the second test cycle the oven 
controller setting was raised to 143°C, and this setting was maintained 
for the remaining tests. The controller temperature oscillations ranged 
between 0.6 to 0.8°C. The third and fourth cycles were the same in re¬ 
gards to the overall length of time and temperature. During the latter 
two cycles, the motor required a dwell time of 36 hours within the pre¬ 
heated oven before the inside surface of the propellant grain acquired a 
temperature of 134°C. 


18 




At the completion of each test cycle the motor was removed from the 
oven and was allowed to cool to the ambient air temperature. The tern - 
perature at the inside surface of the propellant grain was reduced to 
93°C after 5 hours, to 66°C af.er 8 hours, and to 38°C after a total of 
20 hours. 

The propellant in the motor was cracked badly after the first test 
cycle (Fig. 42 and 43). An attempt was made to determine when the pro¬ 
pellant first started to crack, from an examination of the temperature 
record, but this could not be done. The condition of the propellant 
remained about the same throughout the second and third cycles (Fig. 44). 
After completion of the fourth cycle, it was observed that the propellant 
in the motor had bulged and swelled (Fig. 45). The outside motor case 
did not appear to have been affected by the bulging propellant. This 
motor was returned to JPL. 

In a deflagration study done previously on a 12-inch diameter motor 
with a similar propellant type, it was determined that this motor had a 
critical temperature of about 155°C. This would mean that at about 155°C 
or higher, self-heating would have an effect on the time to deflagration. 
Below 155°C, the self-heating effect would diminish and the time to de¬ 
flagration would follow a first-order decomposition path under Isothermal 
conditions. In the sterilization temperature region, some degradation 
effects were apparent in the instrumented 12-inch motor when it waa 
observed that bulging and swelling occurred during the fourth sterili¬ 
zation cycle. 

The thermal stability of an ammonium perchlorate propellant is very 
important in this study as the propellant must maintain its mechanical 
and ballistic integrity after being subjected to the rigorous steriliza¬ 
tion cycling procedure. A detailed kinetic study was not performed on 
the ammonium perchlorate itself as it has been studied and reported by 
numerous investigators. However, inhouse quality control studies have 
indicated that the purity of the oxidizer is a determining factor when 
thermal stability characteristics are evaluated. Even though high purity 
ammonium perchlorate could be used, the PBAA binder might still catalyze 
the low temperature decomposition of the ammonium perchlorate. Therefore, 
the binder should be as inert as possible in this temperature region of 
sterilization. This is brought out when the different types of propel¬ 
lants used in other studies were examined. It was possible to show the 
effect of various ingredients on the critical temperature as shown in 
Fig. 46. In this figure, the data for the prediction of the critical 
temperature were taken from DTA and isothermal analysis studies. The 
reference line at the 3-inch diameter grain was picked arbitrarily since 
most deflagration studies were conducted with 2- to 5-inch diameter sam¬ 
ples. The dividing lines shown between different propellant Ingredients 
are only approximate temperature limits and they are based on information 
found with the propellants listed. For comparison, the data on the JS-7 
propellant is included. The figure shows that to gain more stability, 




NMC TP 4258 


the long chain polymers and fluorocarbons are likely candidates for 
binder material. For the oxidizer, high purity ammonium perchlorate, 
specially stabilized ammonium perchlorate, or potassium perchlorate 
would be necessary. 


CONCLUSIONS 


This study has shown that kinetic data from DTA and isothermal 
analysis studies can be used to predict the thermal stability of propel¬ 
lants. These techniques can also be used to predict the thermal stability 
of explosives, pyrotechnics, and other materials that exhibit exothermic 
reactions, provided the kinetic parameters can be determined by thermal 
analysis. The results of a prediction can be given in precise terminol¬ 
ogy. e.g., terms that relate geometry (mass), time, and temperature. 

This will eliminate the possibility of performing tests under any set 
of conditions, tending to yield almost any type of data desired. It is 
possible, however, to relate and correlate results from data obtained 
by different persons with different types of instrumentation. 


Of the four propellants studied with the polybutadiene type binder, 
the JS-7 propellant appears to be the most thermally stable. The kinetic 
studies have shown a difference in the thermal stabilities of four pro¬ 
pellants that were very similar in composition. This shows that small 
changes in formulations can directly affect the thermal stability and 
indirectly affect the aging characteristics of propellants. 

A 60-pound 12-inch diameter JS-7 propellant grain, cast in motor 
hardware, underwent four sterilization temperature test cycles. Although 
the propellant cracked badly during the first cycle, it was not until the 
fourth cycle was completed that it was possible to observe any noticeable 
bulging and swelling of the propellant grain. The outside motor case did 
not appear to have been affected by the bulging propellant. The cracking 
of the propellant was a physical type action (possible post-curing) while 
the bulging was the result of chemical decomposition reactions. 


20 









FIG. 1. Thermal Profile for a 1-lnch Diameter x 2.5-Inch Long Solid 
Cylinder of JPL Propellant (BD 78/3) in a 198.9°C Oven. 


21 








2 50 


•C 

200 


150 




2. Thermal Profile for a 1-Inch Diameter x 2..5-Inch Long Tubular 
Cylinder of JPL Propellant (BD 79/7) in a 215.6°C Oven. 




TIME, HR 


NWC TP 4258 



FIG. 3. Thermal Profile for a 3-Inch Diameter x 7.5-Inch Long 

Solid Cylinder of JPL Propellant (BD 77/2) in a 166.l'C 
Oven. 


23 












TIME, HR 


DEFLAGRATION 



FIG. 4. Thermal Profile for a 3-Inch Diameter x 7.5-Inch Long 
Tubular Cylinder of JPL Propellant (BD 76/4) in a 
183.3°C Oven. 



























TIME , HR 


NWCTP4258 



l/*K X 10 s 


Fig. 6. Thermal Profile for a 5-Inch Diameter x 12.5-Inch Long 
Tubular Cylinder cf JPL Propellant (BD 68/2) in a 
176.7°C Oven. 









FIG. 7. Thermal Profile for a 10-Inch Diameter x 25-Inch Long 


Tubular Cylinder of JPL Propellant (BD 82/6) in a 
168.9°C Oven. 













































NWC TP 4258 























I/°K X 10' 


FIG. 14. Plot of Deflagration Data on 10-Inch 
Diameter JPL Propellant Grains. 














DIAMETER, IN. 


NWC TP 4258 


250 


200 


150 



FIG. 15. Comparison of Predicted and Experimental T m 
for Various Diameters of JPL Propellant 
Grains. 


35 







TIME AT TEMR HR 


NWC TP 4258 


°C 

250 200 !50 



l/*K X 10 s 

FIG. 16. Deflagration Times Under Isothermal 
Conditions Where T n > for Various 
Diameters of JPL Propellant Grains. 











NWC TP 4258 


SAMPLE: JS-I, 2, 3, 7 

RUN NO. 723 

HEATWG RATE: 7* C/MIN 

SAMPLE WEIGHT: 

NO l-JS-l 108 mg 
NO. 2-JS-2 10.3 mg 

NO. 3-JS-3 9 6 mg 

NO. 4-JS-7 10 9 mg 


1 





















SAMPLE JS-I 

pun no.; 5-io-z 

HEATING RATE rC/MIN 

WEIGHT: 37.3 mfl 


316.6‘C 



I_ 

100 


_I— 

350 

•C 


259 4*C 



225 


_1 _ 

250 

•C 


275 


FIG. 18. DTA Thermal Pattern for JPL Propellant, Batch JS-1 


38 












.—.. 















FIG. 20. DTA Thermal Pattern for JPL Propellant, Batch JS-1. 


40 











306.6 



FIG. 21. DTA Thermal Pattern for JPL Propellant, Batch JS-2. 

























23. DTA Thermal Pattern for JPL Prope 




















-FIG. 25. DTA Thermal Pattern for JPL Propellant, Batch JS- 









SAMPLE 



DTA Thermal 
















27. DTA Thermal Pattern for JPL Propellant, Batch JS- 















395 5*C 



28. DTA Thermal Pattern for JPL Propellant, Batch 


















































FIG. 32. Plot of DTA Data on JPL Propellant, Batch JS-3 















Ut OTA EXOTHERM 
E = 33.3 KCAL/MOLE 
WT a 20 mg 



FIG. 33. Plot of DTA Data on JPL Propellant, Batch JS-7 


















1ST EXOTHERM 
E - 31 4 KCAL/MOLE 
AREA METHOD 



FIG. 34. Plot of DTA Data on JPL Propellant, Batch JS-7 


















NWC TP 4258 










































280 


NWC TP 4258 



57 


FIG. 37. Determination of Correction Factor for 3-Inch Diameter JPL Propellant Charge. 












320 



o 

O 

o 

O 

o 

O 

o 

s 

o 

o 

o 

o 

O 

o 

o 

o 

u, 

o 

CD 

<0 

+ 

CM 

o 

CD 

* 

CM 

o 

CD 

<D 

* 

CM 



wr) 

CM 

CVJ 

CM 

CM 

CM 

* 



“ 









do iJW31 


58 


Determination of Correction Factor for 5-Inch Diameter JPL Propellant Charge. 















Obi 



39. Determination of Correction Factor for 10-Inch Diameter JPL Propellant Charge. 















DIAMETER. IN. 


NVC TP 4258 



FIG. 40, First Appearance of Cracks During Cycling Tests 
on JPL Propellant Grains. 


60 



















i 


i 

t 


i- 

i 


! 



62 


FIG. 62. Nozzle-End View of a 60-Pound, 12-Inch Diameter 
Motor After First Test Cycle. 










FIG. 43. Internal View of a 60-Pound, 12-Inch Diameter 
Motor After First Test Cycle. 





LHL 131030 

FIG. 44. Nozzle-End View of a 60-Pound, 12-Inch 
Diameter Motor After Second Test Cycle. 


64 














002 



FIG. 46. Temperature Ranges for Various Ingredient-:- Use 










REFERENCES 


Longwell, P. A. "Determination of Self-Heating Reaction Kinetics 
Data Applicable at Low Temperatures," Tech. Memo. 865. Aerojet- 
General Corporation (July 1965). 

U. S. Naval Ordnance Test Station. Calculation of Critical Tempera¬ 
ture and Time-to-Explosion for Propellants and Explosives, by P. A. 
Longwell, China Lake, Calif., NOTS, March 1961. (NAVWEPS Report 
7646, NOTS TP 2663). 

The Thermal Decomposition Characteristics of Explosives (U), 
by C. D. Lind. China Lake, Calif., NOTS, February 1962. (NAVWEPS 
Report 7798, NOTS TP 2792), CONFIDENTIAL. 

Zinn, J.. and C. L. Mader. "Thermal Initiation of Explosives," 

J. APPL. PHYSICS, Vol. 31, No. 2 (1960), p. 323. 

Zinn, J., and R. N. Rogers. "Thermal Initiation of Explosives," 

J. PHYS. CHEM., Vol. 66, (1961), p. 2646. 

Kissinger, H. E. "Reaction Kinetics in Differential Thermal Analysis 
ANAL. CHEM., Vol. 29 (November 1957), p. 1702. 

Thompson, D. S., R. L. Reed, L. Weber, and B. S. Gottfried. "Differ¬ 
ential Thermal Analysis and Reaction Kinetics," IND. ENG. CHEM. 
FUNDAMENTALS, Vol. 5, No. 2 (May 1966), p. 286. 

U. S. Naval Ordnance Test Station. Predicting Propellant Safe-Life 
(U), by Jack M. Pakulak, Jr., China Lake, Calif., NOTS, 11 October 
1961. (NAVWEPS Report 7775, NOTS TP 2756), CONFIDENTIAL. 

Borchardt, H. J. and F. Daniels. J. AM. CHEM. SOC., Vol. 79, 

(January 1957), p. 41. 

Reed, R. L., L. Weber, and B. S. Gottfried. "Differential Thermal 
Analysis and Reaction Kinetics," IND. ENG. CHEM. FUNDAMENTALS, Vol. 4 
Ho. 1 (February 1965), p. 38. 









NOMENCLATURE 


A 

AP 

a 


C 

o 

C 1 

d/D 

dT 

dt 

DTA 

E 

F 

j 

k 

L/D 

MAPO 

PEAA 

Q 

R 

T 

t 

T 1 

T 

c 

t 

e 

TGA 

T 

m 

T 

rp 

AT 


Frequency factor, sec 
Ammonium perchlorate 
Radius, cm 

Remaining area of DTA peak 
Heat capacity 
Initial concentration 


Concentration after a specific time at a specific temperature 

Ratio, inside diameter/outside diameter 

Heating rate at the surface of the grain, °C/sec 


Differential thermal analysis 
Activation energy, cal/mole 

Function depending on geometry and the initial temperature 

The fraction reacted 

Specific rate constant, sec' 1 

Ratio, length/diameter 

Tri(methylaairidinyl) phosphine oxide 

Polybutadiene acrylic acid 

Heat of reaction, cal/g 

Gas constant, 1.987 cal/(mole)(°K) 

Absolute temperature 
Time, seconds 

Maximum surface temperature 


The center temperature 


Time to exotherm or deflagration, sec 


Thermogravimetric analysis 
Critical temperature, °K 

Absolute temperature of peak (°K) with respect to the reference 
temperature 

Differential temperature 



68 








NMC TP 4258 


w Mass fraction unreacted at time, t 

2 

* Thermal diffusivity, cm /sec 

<P Heating rate, °C/min 

A Thermal conductivity, cal/(cm)(sec)(°K) 

V* Laplacian operator 

P Density, g/cm^ 

6 Shape factor 

T Dimensionless time 


69 












Sfvuntx ClussiftCd!i 'i 


DOCUMENT CONTROL DATA • R & D 1 

Sr .-r \:hc at ton of title %•! :tar • .tnJ mdexir.,- i'ino. , .i'i,.n rmf !;r enter* d «!irn '.’i^ o.eroll report i » , Inisilied/ | 

1 Oh J’N* live » C T v t v . Corp-orate au.'her 

Naval Weapons Center 

China Lake, California 93555 

*'•. t* f P'C Ml SECUh r v C L A ss> (ifil IOM 

UNCLASSIFIED 

it f.BC^P 


:• REPOS’ TiTiC 


THERMAL ANALYSES STUDIES ON CANDIDATE SOLID JPL PROPELLANTS FOR HEAT 
STERILIZABLE MOTORS 


* 3ESCRi°Tivt NO'ES ' TV pc of repot I a no inc !v>:vt dates / 


5 Avj ThOP.Si fFttSt name, middle initial, last name: 

Jack M. Pakulak, Jr. and Edward Kuletz 


6 REP- OR T O A T E 

_ July 19 70 

T. total MO OF bases 

70 

' b. MO O* - REPS 

10 

t« CCn’PaC t’CR SfiAN* NO 

JPL Purchase Order Nos. Z-351290 and 
Z-351291 


■irfl. OR'G-NA TOR'S RfPCR - NcMBERJl 

NWC TP 4258 

c. 

d. 


86. otm£r report nO'Si ' Anv other number* I ft at may be a a signed 
this repo't) 

IC PI * T Ri BU T IQS JTATEMtS* 

THIS DOCUMENT IS SUBJECT TO SPECIAL EXPORT CONTROLS AND EACH TRANSMITTAL TO FOREIGN 

GOVERNMENTS OR FOREIGN NATIONALS MAY BE MADE ONLY WITH PRIOR APPROVAL OF THE NAVAL 

WEAPONS CENTER. 

Sy p °LtMfNtABl NO ’ C » 

W JPONSORlNC MIU-Ahl ACTW.T. 

National Aeronautics & Space Administration 
Jet Propulsion Laboratory 

California Institute of Technology 

■ • a »BJTR»C’ 





Times to deflagration of solid JPL propellant grains varying in 
size and geometry have been measured, and the results have been 
analyzed in light of the thermal explosion theory. In addition, 
laboratory-scale thermoanalytical techniques have been used to study 
the procedural chemical kinetics of the decomposition reactions of 
representative solid JPL propellants. The results of these studies 
indicate that, for the propellants considered, estimates of thermal 
deflagration times can be made on the basis of laboratory-scale experi¬ 
ments. The effects of sterilization heat cycling tests on solid JPL 
propellant grains varying in size and geometry have also been studied 
and the results are herein reported. 











Thermal Analysis 
Kinetic Studies 
Solid Propellants 
Heat Cycling 
Degradation 
Critical Temperature 
Cook-Off 


DD ,?o*“..1473 

(PAGE 2) 


(BACK)