90 Propellants, Explosives, Pyrotechnics 24, 90Ä…94 (1999)
Sensitivity of Solid Rocket Propellants for Card Gap Test
Eishu Kimura and Yoshio Oyumi
Third Research Center, Technical Research and Development Institute, Japan Defense Agency, 1-2-10 Sakae, Tachikawa,
Tokyo 190-8533 (Japan)
 Â
Emp®ndlichkeit von Raketenfesttreibstoffen beim Card-Gap-Test Sensibilite de propergols solides de fusees lors du test Card-Gap
  Â
Der Card-Gap-Test, standardisiert durch die Japan Explosives Le test Card-Gap, standardise par la Japan Explosives Society, a ete
 à  Â
Society, wurde modi®ziert zur Anwendung bei Raketenfesttreibstoffen modi®e en vue d'etre utilise dans des propergols solides de fusees et
und ausgefuhrt zur Bewertung der Emp®ndlichkeit gegen concu pour evaluer la sensibilite a l'effet de choc. Les propergols
È Ë Â Â Á
 Â
Stoûeinwirkung. Die hier getesteten Festtreibstoffe waren haupt- solides testes etaient principalement des propergols composites poly-
È Á Á
sachlich azidhaltige polymere Komposit-Treibstoffe, welche vor- meres a base d'azide, qui contenaient essentiellement du nitrate
  Â
zugsweise Ammoniumnitrat als Oxidator enthielten. Doublebase- d'ammonium en tant qu'oxydant. On a egalement evalue des proper-
Á Â Â
Treibstoffe, zusammengesetzt aus Nitroglycerin und Nitrocellulose, gols a double base composes de nitroglycerine et de nitrocellulose et
Á
und ammoniumperchlorathaltige Komposit-Treibstoffe wurden eben- des propergols composites a base de perchlorate d'ammonium en vue
Á Á Â
falls ausgewertet zum Vergleich mit den azidhaltigen Poly- de les comparer aux propergols polymeres a base d'azide. On a montre
  à   Â
mertreibstoffen. Es wurde gefunden, daû die Emp®ndlichkeit durch die que la sensibilite etait maõtrisee par les proprietes oxydantes. Les
Á Â
Oxidatoreigenschaften beherrscht wurde. AP- und AN-haltige Treib- propergols a base d'AP ou d'AN ont une sensibilite plus faible, les
È Á Â Â Â
stoffe haben eine geringere, HMX-haltige Treibstoffe eine hohere propergols a base de HMX une sensibilite plus elevee; l'addition de
 Á   Â
Emp®ndlichkeit, die Zugabe von NC und TMETN trug zur Ver- NC et de TMETN a contribue a la deterioration de la sensibilite lors du
Â
schlechterung der Emp®ndlichkeit im Card-Gap-Test bei. Gute test Card-Gap. On a obtenue de bonnes relations entre la sensibilite
Á Â Â
Beziehungen zwischen der Emp®ndlichkeit im Card-Gap-Test und der lors du test Card-Gap et le bilan d'oxygene des propergols etudies.
Sauerstoffbilanz der hier untersuchten Treibstoffe wurden erhalten.
Summary the sample because of a relatively lower sensitivity of
propellants compared to that of explosives.
Card gap test, which is standardized in Japan Explosives Society,
was modi®ed in order to apply it to solid rocket propellants and carried
out to evaluate sensitivities against shock stimuli. Solid propellants
tested here were mainly azide polymer composite propellants, which 1.2 Oxygen Balance
contained ammonium nitrate (AN) as a main oxidizer. Double base
propellant, composed nitroglycerin and nitrocellulose (NC), and
Oxygen balance is de®ned as the number of oxygen
ammonium perchlorate (AP)-based composite propellants were also
equivalents per unit of propellant when a sample detonates
evaluated in order to compare with the azide polymer propellants. It is
found that the sensitivity was dominated by the oxidizer character- and all of each atom in the sample ®nally converts into the
istics. AP-and AN-based propellant had less sensitivity and HMX-
following molecules:
based propellant showed higher sensitivity, and the adding of NC and
TMETN were contributed to worse sensitive for the card gap test.
C ! CO2; H ! H2O; N ! N2
Good relationship was obtained between the card gap sensitivity and
the oxygen balance of propellants tested here. Therefore, when a propellant represents CxHyOzNu, the
reaction for the explosion is expressed below.
y u 1 y
1. Introduction CxHyOzNu ˆ xCO2 ‡ H2O ‡ N2 ‡ 2x ‡ z O2
2 2 2 2
The oxygen balance is calculated as follows in an ideal
1.1 Card Gap Test
reaction described above.
y y
Rocket propellants have been required to be insensitive to
1
2x ‡ z 32 16 2x ‡ z
2
2 2
shock such as an unexpected drop at handling, bullet or
ˆ Ä…
12x ‡ y ‡ 16z ‡ 14u 12x ‡ y ‡ 16z ‡ 14u
fragment impacts. Card gap test has an advantage to estimate
quantitative sensitivities of propellant against shock wave.
Since propellant consists of binder, oxidizer and some
The test was originally standardized and conducted usually
additives, ``x'', ``y'', ``z'' and ``u'' in CxHyOzNu are calcu-
for an estimation of sensitivity of explosives(1Ä…4). In this time,
lated from the molecular formula of each ingredient and the
the card gap test of Japanese test standard(5) has been
composition percentage. For example, CxHyOzNu of the
modi®ed to suit for evaluating sensitivity of rocket propel-
propellant composed of oxidizer=binder ˆ 80=20 is calcu-
lants(6). The main modi®cations occurred on the materials of
lated by using the molecular formula of the oxidizer
the gap plate and of the sample holder, poly(methylmetha-
(CaHbOcNd) and the binder (CeHfOgNh), as follows:
crylate) (PMMA) to aluminum and polyvinyl chloride to
x ˆ 0:8a ‡ 0:2e; y ˆ 0:8b ‡ 0:2f;
steel, respectively. It was for the purpose of diminishing a
loss of the energy of shock wave, generated by the donor, into z ˆ 0:8c ‡ 0:2g; u ˆ 0:8d ‡ 0:2h
# WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999 0721-3115/99/0306Ä…0090 $17.50‡:50=0
Propellants, Explosives, Pyrotechnics 24, 90Ä…94 (1999) Sensitivity of Solid Rocket Propellants for Card Gap Test 91
An intensity of detonation become a maximum at around from the insensitive point of view. AP, HMX and NC=NG are
the oxygen balance of 0. The ideal reaction, as mentioned used to compare with AN based propellant.
above, may not occur in the practical detonation and
another kind of gases, such as CO, NO, NO2, were pro-
duced. The gas composition produced in the detonation is
2.2 Calibration of Shock Pressure
strongly effective on the value of the oxygen balance(7).
Kamlet mentioned the oxygen balance showed good
It is necessary to characterize the shock wave intensity
relationship in the impact sensitivity of organic high explo-
through the gap plate. The characteristics are derived from
sives(8). Kamlet used OB100, a measure of oxidant balance.
the shock wave velocity travelling in the aluminum plate.
OB100 was de®ned as 1002d b 2a 2NcooÄ…=(molecular
The method was described in detail elsewhere(6). Figure 1
weight) for a material composition CaHbNcOd with Ncoo as
shows relationship between aluminum length and shock
the number of carboxyl group. Storm and co-worker also
passing time in the detonation of the donor explosive. The
demonstrated correlation of sensitivity with only structural
time is obtained by the least square ®tting to equation
information of the compound, by using the Sensitivity Index
t ˆ a ‡ bL ‡ cL2:
(SI) de®ned as 100d a by2Ä… NcoÄ…y5a ‡ b ‡ c ‡ dÄ…Ä…
for CaHbNcOd with Nco as the number of carbonyl groups(9).
t ˆ 0:148 ‡ 0:145L ‡ 3:98 10 4L2 1Ä…
OB100 and SI are similar to the oxygen balance of the
The shock velocity, symbolized Us, at L(mm) from the top
compounds. Herein, the oxygen balance described in Eq.
of the gap plate is derived by differentiation in Eq. (1):
(*) was selected to perform correlation of sensitivity for the
card gap test with the structural information about solid dL 1
Us ˆ ˆ 2Ä…
rocket propellants.
dt 0:145 ‡ 7:96 10 4L
Then, the particle velocity, Up, in aluminum plate is cal-
culated from the shock Hugoniot for the aluminum used
2. Experimental
here(10) and the measured Us:
Us ˆ 5:35 ‡ 1:34Up 3Ä…
2.1 Samples
The particle velocity in the aluminum at interface of the
The compositions of propellant samples are shown in sample propellant can be extrapolated at L ˆ 0 by the least
Table 1. The binders are based on BAMO=NMMO copoly- square ®tting since the relation between Up and aluminum
mer (BN), GAP and HTPB. AN is chosen as a main oxidizer thickness is linear. The shock pressure, P, is obtained in the
Table 1. Composition of Propellant Samples
Sample BN GAP HTPB TMETN NC NG AN HMX AP Al Others
BN-based composite propellant
1 25 75 PbSt;4.9, CB;0.5
2 25 60 15 FeB;2.0, CuCr;3.0
3 25 50 25 FeB;2.0, CuCr;3.0
4 25 45 30 FeB;2.0, CuCr;3.0
5 25 40 35 FeB;2.0, CuCr;3.0
6 25 30 45 FeB;2.0, CuCr;3.0
7 25 15 60 FeB;2.0, CuCr;3.0
823 77 FeO;3.0, ZrC;2.0
GAP-based composite propellant
9 30 55 15 B;1.0, CB;0.6
10 20 20 60
11 15 15 20 50 CuCr;1.0, PbCi;2.0
12 18 80 2 FeO;1.0
HTPB-based composite propellant=Double base propellant
13 25 60 15 FeB;2.0, CuCr;3.0
14 12 71 17 FeO;2.0
15 37 33 14 BDR1;16.0
16 48 42 BDR2;10.0
BN: 3,3-Bis(azidomethyl)oxetane=3-nitratomethyl-3-methyloxetane (BAMO=NMMO) ˆ 7=3 binder; GAP: glycidyl azide polymer binder;
HTPB: hydroxyl-terminated polybutadiene binder; TMETN: trimethylolethane trinitrate; NC: nitrocellulose; NG: nitroglycerin; AN: ammonium
nitrate; AP: ammonium perchlorate; FeB: Butacene; CuCr: copper chromite; FeO: Iron oxide; ZrC: zirconium carbide; CB: carbon black; PbCi:
lead citrate; BDR1: polyol-based binder; BDR2: urethane-based binder.
92 Eishu Kimura and Yoshio Oyumi Propellants, Explosives, Pyrotechnics 24, 90Ä…94 (1999)
Figure 2. Setup for card gap test.
Figure 1. Relationship between gap length and shock passing time.
chloride (JIS K 6741, VP-30) into steel (JIS G 3452, A32
Eq. (4) with initial propellant density, r, and Us and Up in
carbon steel pipe), sized 42.7 mm outer diameter with 3.5 mm
Eqs. (2) and (3):
thick and 50 mm long. This change is contributed to be a
larger shock pressure traveling in the sample because of more
P ˆ rUsUp 4Ä…
restraint of sample in comparison with polyvinyl chloride
Table 3 shows the shock pressure as a function of the alu-
tube. The witness plate under sample is steel (JIS G 3141,
minum plate thickness.
SPCC), sized 100 mm square with 2 mm thickness.
Three trials are conducted at the same gap length, varying
by 5 mm intervals in the thickness of aluminum plate and
determined whether ``go'' or ``no go''. Detonation of sample
2.3 Card Gap Test
propellant can be judged if cracks or holes were observed on
the witness plate. The critical gap length is determined when
The experimental setup shown in Figure 2 is almost the
detonation of sample was observed in all three trials. The
same as the standard of Japan Explosives Society(5). Pentolite
critical shock pressure is de®ned as the pressure given b y the
(PETN : TNT ˆ 50 : 50, density: 1.62 g=cm3) is used as a
critical gap length.
donor and it is initiated by #6 detonator cap. The material of
the gap plate was changed PMMA (JIS K 6718, ®rst grade) to
aluminum (JIS H 4000, 6061, density: 2.703 g=cm3 ), shaped
3. Results and Discussion
60 mm circle with 5Ä…30 mm thickness. The shock pressure
transmitted to sample propellant through the aluminum plate
An oxidizer in sample propellant dominated sensitivity of
may become approximately twice as large as that of PMMA
the card gap test, as shown in Figure 3. Samples, whose main
because of large density for aluminum in Eq. (4). The
oxidizer was AN or AP, showed lower sensitive to the card
material of the sample holder was also converted polyvinyl
gap test. The properties were independent on their binder,
such as BN, GAP and HTPB. Samples contained AP,
Table 2. Molecular Formula and Oxygen Balance for Samples
Sample Molecular formula Oxygen balance
Table 3. Shock Velocity, Particle Velocity and Shock Pressure in
Aluminum Plate
1 C5.700 H10.825 O6.675 N6.800 0.57676
2 C3.300 H8.425 O3.675 N3.200 0.75317
LUs Up P
3 C3.700 H8.825 O4.175 N3.800 0.70544
mm mm=msmm=ms GPa
4 C3.900 H9.025 O4.425 N4.100 0.68578
5 C4.100 H9.225 O4.675 N4.400 0.66829
50 5.41 0.0457 0.668
6 C4.500 H9.625 O5.175 N5.000 0.63856
40 5.65 0.227 3.47
7 C5.100 H10.225 O5.925 N5.900 0.60364
35 5.79 0.325 5.09
8 C2.484 H7.519 O3.701 N1.506 0.68372
30 5.92 0.426 6.82
9 C2.586 H6.643 O3.351 N3.011 0.61658
25 6.06 0.533 8.73
10 C2.324 H6.362 O3.934 N2.274 0.48299
20 6.21 0.645 10.8
11 C2.943 H6.476 O5.095 N2.302 0.41452
15 6.37 0.763 13.1
12 C1.324 H5.362 O5.095 N1.274 0.30032
10 6.54 0.886 15.7
13 C2.378 H7.280 O3.435 N2.414 0.65935
5 6.71 1.02 18.5
14 C0.853 H4.156 O2.903 N0.726 0.19857
0 6.90 1.15 21.4
15 C3.734 H5.500 O7.695 N3.002 0.18735
L: gap length, Us: shock velocity, Up: particle velocity, P: shock
16 C4.725 H6.397 O8.761 N2.453 0.26179
pressure in aluminum plate
Propellants, Explosives, Pyrotechnics 24, 90Ä…94 (1999) Sensitivity of Solid Rocket Propellants for Card Gap Test 93
though both of the slopes were almost the same. It means that
GAP binder is slightly more insensitive to shock stimuli than
BN binder. The insensitiveness of GAP binder is supported
that the critical gap lengths of AP- and AN-based samples are
0 mm, as shown in Figure 3, though those of BN-binder
samples are 5 mm. HTPB-binder sample were also more
insensitive than BN-binders, since sample 13 was located
at higher oxygen balance compared to sample 2 in Figure
4(a). The positions of the plot for double base samples
propellant had further higher oxygen balance and a different
correlation might exist from that of composite propellants.
No relation could be obtained because of only two samples
tested here.
Figure 4(b) indicates the relation between oxygen balance
and sensitivity of AP-based sample with AN- and HMX-
based ones. The values of oxygen balance for AP-based
samples of the respective binder showed only poor correla-
tion to the critical shock pressures. The reason why this poor
Figure 3. Critical gap length.
correlation is obtained may be that the criterion of ``go'' or
``no go'' determination in the card gap test is detonation of the
sample propellant. As mentioned before, burned-out phe-
however, burned out after the test because of its high nomena were observed in AP-based sample after trials of the
combustibility. It is impossible in this card gap test to gap test because of its high combustibility. AP-based sample
distinguish deŻagration from detonation since the witness was sensitive to initiation and then deŻagration occurred in
plate located under the sample propellant is used in the the shock ignitability test(11).
judgement. No damage on the witness plate existed even in
the case of deŻagration.
The sensitivity got worse by adding of HMX, NC and NG.
Especially, sample 11 showed the critical gap length of
25 mm even it contains 50% AN in the composition. Table
4 summarized the critical gap length and the critical shock
pressure in the card gap test.
Figure 4(a) shows relations between oxygen balance listed
in Table 2 and critical shock pressure. Good relations were
observed for each binder, except for double base propellants.
The sensitivities became worse as oxygen balance increased.
The linear line for GAP-binder samples was shifted to higher
oxygen balance, compared with that for BN-binder samples,
Table 4. Results of Card Gap Test
Sample Lc (mm) Pc (GPa)
1 25 8.73
2 5 18.5
3 10 15.7
4 15 13.1
5 15 13.1
6 15 13.1
7 20 10.8
8 5 18.5
9 0 21.4
10 5 18.5
11 25 8.73
12 0 21.4
13 5 18.4
14 0 21.4
15 20 10.8
Figure 4. Correlation of card gap test with oxygen balance.
16 25 8.73
(a) without AP-based samples,
Lc: critical gap length, Pc: critical shock pressure (b) with AP-based samples.
94 Eishu Kimura and Yoshio Oyumi Propellants, Explosives, Pyrotechnics 24, 90Ä…94 (1999)
(2) K. Hashizume and N. Sasaki, ``Card Gap Test of Industrial High
4. Conclusions
Explosives'', J. Indust. Explos. Soc., Japan 63, 34 (1975).
(3) S. Matsumoto, M. Tanaka, and T. Yoshida, ``Study of Card Gap
AP- and AN-based propellants showed relatively large
Test for Industrial Explosives'', J. Indust. Explos. Soc., Japan 37,
critical shock pressure but HMX- and NC-based propellants 173 (1976).
(4) Y. Hirosaki, T. Ishida, K. Hattori, and H. Sakai, ``Card Gap Test of
were very sensitive in the card gap test. Three different
Emulsion Explosive'', J. Indust. Explos. Soc., Japan 43, 323
propellant binders, such as BN, GAP and HTPB, were
(1982).
evaluated and difference in sensitivity was hardly observed
(5) ``Test Standard of Explosive Sensitivity (IV)'', Japan Explosives
Society (edit), 1995, pp. 92.
among them. It was found that the sensitivity was dominated
(6) Y. Oyumi, E. Kimura, and K. Nagayama, ``Initiation and Deto-
by the oxidizer characteristics.
nation Properties of Azide Polymer Propellants'', J. Explos. Soc.,
In each binder type of propellant groups, linear relation-
Japan 55, 194 (1994).
ships were observed between the critical shock pressure and
(7) S. Nakahara, in: Kayakugaku Gairon (``Study of Explosives''),
Sangyotosho, Tokyo, 1983, pp. 4.
the oxygen balance of the propellant, except AP-based
(8) M. Kamlet, ``The Relationship of Impact Sensitivity with Struc-
propellant. AP-based propellants tended to deŻagrate
ture of Organic High Explosives. I. Polynitroaliphatic Explo-
because of its ignitability. The deŻagration phenomena
sives'', Proc. 6th Sym. (International) on Detonation, (1976), pp.
might not affect the measured value of the critical gap 313.
(9) C. B. Storm, J. R. Stein, and J. F. Kramer, ``Sensitivity Relation-
length in this card gap test. The sensitivity became worse
ships in Energetic Materials'', in: ``Chemistry and Physics of
as the oxygen balance increased in each propellant group.
Energetic Materials'', Kluwer Academic Press, Boston, 1990, pp.
The slope in BN binder propellant was almost the same as
605.
that of GAP-binder propellant. When the oxygen balance was (10) S. P. Marsh, ``LASL Shock Hugoniot Data'', University of
California Press, Berkeley (1980).
the same, GAP-binder propellant was more insensitive than
(11) E. Kimura and Y. Oyumi, ``Shock Ignitability Test for Azide
that of BN one.
Polymer Propellants'', J. Energetic Materials, 16 (2=3), 173Ä…185
(1998).
5. References
(1) M. Iida, S. Fujiwara, and M. Kusakabe, ``Shock Sensitivity of
Explosive Materials. I. Fundamental Experiments on Gap Test'',
J. Indust. Explos. Soc., Japan 33, 291 (1972). (Received May 7, 1998; Ms 12=98)