Perchlorate Free Pyrotechnic Composition

and its Application in M115A2 Ground Burst Simulator and

M116A1 Hand Grenade Simulator
Gary Chen, Mark Motyka, James Wejsa
U.S. Army RDECOM-ARDEC, Picatinny Arsenal, NJ 07806, USA

ABSTRACT

A family of perchlorate-free pyrotechnic compositions for military flash-bang training rounds and
non-lethal munitions have been developed and proven out. The fuel and oxidizer pre-blends of the
compositions are separately prepared and loaded into the item, completely eliminating the need for
dangerous manual operations. This initiative is part of DoD efforts to mitigate the potential risk to the
military readiness associated with the recent EPA regulation on restricted use of perchlorate. Excessive
use of perchlorate in military munitions has led to ground and surface water contamination. Perchlorate
contamination in drinking water is linked to the blocking of iodide from entering the thyroid gland, and
thereby interfering with production of thyroid hormone. The current fielded M115A2 Ground Burst
Simulators and M116A1 Hand Grenade Simulators account for the majority of potassium perchlorate
used by US Army and thus were selected as the perchlorate-free technology demonstration platform.

INTRODUCTION recommendation and established an official

reference dose (Drinking Water Equivalent
The Federal Environmental Protection Level) of 24.5 ppb, a daily human exposure
Agency (EPA) has identified the perchlorate level that is not expected to cause adverse health
anion (ClO4-) as a ground and surface water effects. As part of DoD efforts to mitigate the
contaminant due to its high solubility, excessive use of perchlorate in munitions
persistence, and potential effects on human systems, the ARDEC Pyrotechnic Research and
health. Perchlorate exposure has the potential of Technology Branch initiated a product
interfering with iodide absorption by the thyroid improvement program in 2004 to develop and
gland. A preliminary estimate of the current prove-out a perchlorate-free flash bang
DOD ordnance inventory indicates that over 250 composition for use in M115A2 and M116A1
different munitions types contain perchlorates. simulators.
These ordnances utilize ammonium perchlorate
(AP) in rocket and missile propellants and FORMULATION DEVELOPMENT
potassium perchlorate (KP) in pyrotechnic
simulator, delay, incendiary, illumination, gas Identification of Potential Fuel/Oxidizer
generation, and tracer compositions. High levels Systems
of perchlorates were recently found in the
ground water of the Aberdeen Proving Ground Various organic and metallic fuels as
(APG)/FTX Ordnance Center and School. well as nitrate-base oxidizers were identified as
Perchlorates were also detected in some drinking potential ingredients for perchlorate-free
water systems around the country. The formulations. The toxic or halogen-containing
excessive pechlotate levels at Aberdeen were oxidizers (barium nitrate, Teflon, etc.) were not
mainly attributed to the potassium perchlotate considered due to environmental issues. The
used in the flash charge of M115A2 Ground NASA Glenn Research Center Chemical
Burst Simulators and the M116A1 Hand Equilibrium with Applications (CEA)
Grenade Simulators. In February 2005, the EPA Computational Program was initially used to
followed the NRC (National Research Council)

269
calculate the flame temperature, gas/liquid the mix. The results show that the aluminum
fractions and fuel/oxidizer ratio of each fuel level at 40% yielded the best performance
formulation candidate. The results suggest that in quickness and peak pressure for both oxidizer
aluminum / nitrate and magnesium / nitrate systems.
systems would provide the highest visual
brightness based on the calculated flame Various types of aluminum powder were
temperatures. Concerns with powdered also tested in the closed bomb to investigate
magnesium’s history of hydrogen out-gassing in their impact on the ballistic behavior of sample
the presence of moisture and high unit cost led compositions. The evaluated candidates include
to the selection of the aluminum fuel system as spherical (X-65, X-80), flake (German
the best perchlorate-free candidate to go- Balckhead), and conventional (atomized 101Al)
forward. aluminum powders. For Al/KN system the
study focused on the medium particle size
The selection of the type of nitrate for oxidizer with an average of 34 microns because
the oxidizer system was mainly based on the the fine oxidizer yielded a very low fill density
level of hygroscopicity of each material under of below 0.6 g/cc. For Al/SrN system, both the
various environments. Vulnerability to fine and medium size oxidizers are considered
moisture is considered detrimental to the long usable due to strontium nitrate’s high theoretical
term storage stability of most pyrotechnic density as demonstrated in the following
compositions. Hygroscopicity tests were samples:
conducted on dried samples of potassium nitrate, Non-vibrated
sodium nitrate, and strontium nitrate at both Formulation Fill Density, g/cc
75% and 90% relative humidity under ambient 604 - Al 25μm / KN 34μm 0.81
temperature. The moisture absorption data, 605 - Al 25μm / KN 9μm 0.53
collected over 2000 hours, for these nitrates in 609 - Al 8μm / KN 9μm 0.58
reference to potassium perchlorate were plotted
in Figures 1 and 2. As shown, the potassium 603 - Al 25μm / SrN 49μm 0.96
nitrate had the greatest moisture resistance under 606 - Al 25μm / SrN 10μm 0.70
both conditions followed by strontium nitrate. 610 - Al 8μm / SrN 10μm 0.77
In contrast, sodium nitrate absorbed a
significantly higher percent of moisture at both Among the various types of powdered
humidity levels and thus was not considered for aluminum, it was found that the formulations
further study. 604 (Al/KN system) and 603 (Al/SrN) system
performed the best. Both were formulated with
Fuel Level and Type 40% of 25 micron flake aluminum, 5% sulfur
and 1% boric acid. The standout ballistic
A 50-cc closed bomb was used to performance for 603/604 formulations is
generate the combustion product’s pressure-time attributed to the flake aluminum’s high surface
correlations for each potential formulation in area of approx. 9 m2/g, comparing to approx.0.7
order to predict ballistic behavior. Initial studies m2/g for 65X spherical aluminum. The
focused on the peak pressure, rise time and atomized aluminum, 101Al (26 microns), also
ignitibility of aluminum / potassium nitrate performed poorly due to low surface area.
(Al/KN) and aluminum / strontium nitrate
(Al/SrN) compositions at various fuel/oxidizer Sulfur Level
ratios. In this effort, each sample composition
was prepared with 9 micron X-65 spherical Sulfur is a secondary fuel and also an
aluminum powder, 9-10 micron nitrate, 5% excellent ignition facilitator due to its low
sulfur and 1% boric acid. A small amount of melting point. The amount of sulfur used in the
sulfur was added to facilitate ignition and formulations will also impact the ballistic
increase quickness. The boric acid was used as performance, which is closely related to the
a neutralizer to enhance the storage stability of

270
 ~75% Humidity

14

12

10
% Weight Change

8
KNO3
NaNO3
6 Sr(NO3)2
KClO4

0
0 500 1000 1500 2000 2500

-2
Time (hours)
Figure 1: Moisture absorption for oxidizers at 75% relative humidity

~90% Humidity

80

70

60
KNO3
% Weight Change

NaNO3
50 Sr(NO3)2
KClO4

40

30

20

10

0
0 500 1000 1500 2000 2500
Time (hours)
Figure 2: Moisture absorption for oxidizers at 90% relative humidity

271
system’s sound intensity and photometric composition 604. Two levels of M5 (0.25% and
output. Previous closed bomb testing indicates 0.5%) were used in a closed bomb study for
that Al-KN system without sulfur had ignition comparison with the baseline 604 composition
and low peak pressure problems. As a result, the without any anti-caking agent. It was found that
optimal sulfur level was investigated for the M5 not only reduced the dry mixing time but
Al/KN and Al/SrN systems using the 34 micron also increased the peak pressure of the 604
nitrate. Closed bomb data show that the composition by 20% and 25% respectively for
formulations utilizing 5% to 10% sulfur burned 0.25% and 0.5% Cab-O-Sil. The pressure traces
smoothly with a quick rise time. of these tests can be seen in Figure 4.

A separate study was setup as a MANUFACTURING PROCESS

prototype platform to determine the optimal DEVELOPMENT
sulfur level for the Al/KN system with regard to
sound level and photometric output. 13 grams Full Up Hardware Assembly
of mix was loaded into 20-gage shotgun
cartridges and initiated with a percussion primer. All parts on the M115A2 and M116A1
Results showed that the best sound level and were glued together using an epoxy of EPON
photometric output was achieved when using 828 and Epikure 3125 in a 2:1 ratio respectively.
mixes with a sulfur content of 5-15%. For this For M115A2 the pre-assembled whistle tube,
study the fuel level was set at 50% with composition and quick match were epoxied into
atomized 101Al aluminum. the top sleeve. Next, the inner tube that holds
the flash-bang charge was epoxied into the top
Variation of Flash Charge Weight sleeve. This preassembly was then placed into a
140oF oven for at least one hour to accelerate the
Study was conducted with a 50 CC epoxy curing. The charge mix was loaded after
closed bomb to determine the sample sizes of the epoxy had fully cured and the end cap and
Al/KN and Al/SrN compositions that would bottom sleeve were epoxied in place. The final
generate an equivalent pressure to 0.5 grams of assembly was placed back into the oven to
the current M115A2 / M116A1 flash charge complete curing. The safety match and igniter
(aluminum / potassium perchlorate) designated came preassembled and glued to the top disk for
as 473B (standard). Results showed that 0.75 to the M116A1. The ignition assembly and inner
0.875 grams of 604 composition is required to tube were epoxied into the top sleeve and the
achieve an equivalent level. The 603 bottom disk was epoxied into the bottom sleeve.
composition yielded somewhat lower peak The ends of the M116A1 round were filled with
pressure than the 604 composition at equivalent epoxy to strengthen them. The two halves of the
sample weights. This suggested that the 603 M116A1 were then placed into the oven to
requires a higher sample amount in order to shorten the curing time. Once the epoxy had
provide a similar peak pressure as 473B. Figure fully cured the charge mix was loaded into the
3 contains the closed bomb pressure-time traces top half of the round and the bottom half was
for Al/KN and Al/SrN systems in reference to epoxied on. The final assembly was then placed
the standard. back into the oven.

Anti-caking and Processing Agent Two Part Loading and In-Round Mixing

Pyrotechnic dry mixes have a tendency The ability to mix the developed
of caking during loading, assembly and storage, perchlorate-free formulations in round at the
especially in high humidity conditions. The selected charge weight was studied. A two part
mixes were also prone to adhere to the walls of loading and mixing method was used. The first
processing equipment. To prevent this problem part consists of the blended primary fuel
an anti-caking and processing aid, Cab-O-Sil aluminum powder and secondary fuel sulfur.
(Grade M5), was selected for use in the

272
Figure 3: Pressure profiles of Al/KN and Al/SrN systems at various charge weights

Figure 4: Pressure profiles of Al/KN system showing the effect of Cab-O-Sil

273
The second part consists of nitrate oxidizer, Table 1 on the following page shows the
Cab-O-Sil (M5), and boric acid. The oxidizer various 604-based configurations (4 each for the
was pre-dried and screened before blending with M115A2 and M116A1) that were tested in the
Cab-O-Sil and boric acid. The fuel pre-blend mixing study as well as the observations made.
was loaded into the charge container first
followed by the oxidizer pre-blend on the top. The rounds were tumbled for 15 minutes
The round was fully assembled, sealed, and then the machine was stopped to check the
mixed for an hour. condition of the mixing. After the first 15
minutes of mixing, the formulations with Cab-
To assist the observation of mixing O-Sil appeared to be visually well mixed where
progress, two sizes of clear polycarbonate tubes as separation could still be observed in the other
were made that have the same length and inner two mixes. The rounds were placed back in the
diameter as the inner tubes of the M115A2 and machine and tumbled for another 45 minutes.
M116A1. The powders were loaded into the After mixing for a full hour, seven of the eight
round and the ends were taped over. The full rounds appeared to be visually well mixed.
rounds were placed in foam holders. These Non-mixed powder could still be seen in the
holders kept the rounds at a ~30o angle while M115 round with 70 grams of 604 without Cab-
tumbling. Figure 5 is an example of the clear O-Sil. Results were photographed in Figures 7-
tube model. Figure 6 shows the mixing device 14.
with eight rounds loaded.
FULL -UP SYSTEM PROVE OUT

Formulation Definition

Two candidate formulations 603 and

604 defined below were selected for
M115A2/M116A1 full-up system prove-out.
The particle sizes are at 50% point using a laser
diffraction instrument. The aluminum powder is
a German Blackhead version.

wt %
Aluminum 40.0
Figure 5: Clear polycarbonate tube model Mil-A-512 Type 1, Grade B
for mixing study
Oxidizer 53.5
Strontium Nitrate
Mil-S- 20322B, Grade A
-or-
Potassium Nitrate
Mil-P156B, Class 3

Sulfur 5.0
Mil-S-487, Grade C

Boric Acid 1.0

Commercial Grade

Cab-O-Sil 0.5
M5 Grade
Figure 6: Tumbling machine

274
 Table 1: Mixing study matrix and observations
Number Contents Observations
15 minutes 60 minutes
Some white still
1 70 gm 604 Not well mixed visible
Some white still Visually well
2 50 gm 604 visible mixed
Visually well Visually well
3 70 gm 604 w/ 0.25% Cab-O-Sil mixed mixed
Visually well Visually well
4 70 gm 604 w/ 0.5% Cab-O-Sil mixed mixed
Visually well
5 33 gm 604 Not well mixed mixed
Visually well
6 25 gm 604 Not well mixed mixed
Visually well Visually well
7 33 gm 604 w/ 0.25% Cab-O-Sil mixed mixed
Visually well Visually well
8 33 gm 604 w/ 0.5% Cab-O-Sil mixed mixed

Figure 7: M115A2 70g 604 Figure 8: M115A2 50g 604

Figure 9: M115A2 70g 604 Figure 10: M115A2 70g 604

w/ 0.25% Cab-O-Sil w/ 0.5% Cab-O-Sil

275
 Figure 11: M116A1 33g 604 Figure 12: M116A1 25g 604

Figure 13: M116A1 33g 604 Figure 14: M116A1 33g 604
w/ 0.25% Cab-O-Sil w/ 0.5% Cab-O-Sil

Sound and Photometric Output Performance 170 dB). Prior to sound level measurements, the
sound meter was calibrated with a Bruel & Kjaer
Mixes were pre-blended and dried prior type 4231 sound calibrator. In these
to loading into the system hardware. For each measurements, only the peak sound intensity
formulation two different charge weights were level was measured for each round. The
loaded: 60 and 70 grams for the M115A2 luminous flux emitted from each round was
Simulator and 30 and 33 grams for the M116A1 measured with a photometric based light
Simulator. The charge weights were determined measurement system. A calibrated International
from the closed bomb study. The standards Light Silicon photo detector, with a matched
contain 40 grams 473B for the M115A2 and 20 photometric filter, was used to measure the
grams 473B for the M116A1. The test matrix is
summarized in Table 2. Table 2: Perchlorate free full-up system
prove out test matrix
Each sample group was tested at
M115A2
ambient temperature. Cold testing was
Std. 604 604 603 603
conducted for 70 gram and 33 gram groups only.
60g 70g 60g 70g
The sound and the light measurement systems
Ambient 5 5 5 5 5
were placed approximately 50 feet from the
samples to satisfy geometric requirements and to Cold, -65F 5 5
protect the measurement systems from the M116A1
shockwave produced by the energetic rounds. Std. 604 604 603 603
The peak sound intensity was measured with a 30g 33g 30g 33g
Bruel & Kjaer 2238 Mediator sound meter Ambient 5 5 5 5 5
equipped with a 0.25 inch microphone (rated to Cold, -65F 5 5

276
 Table 3: Summary of visible light output and sound intensity measurements
M115A2
Average of 5 Std. 604-60g 604-70g 603-60g
rounds Ambient Ambient Ambient Cold Ambient
Integrated
photopic output
(Cd*sec) 1.20E+05 9.88E+04 1.59E+05 1.16E+05 2.87E+05
Peak photometric
output (Cd) 3.32E+07 8.85E+06 1.09E+07 8.74E+06 2.03E+07
Sound intensity
(dB) @ 50ft 155.8 149.1 150.4 149.1 147.1
M116A1
Average of 5 Std. 604-30g 604-33g 603-30g
rounds Ambient Ambient Ambient Cold Ambient
Integrated
photopic output
(Cd*sec) 6.35E+04 5.03E+04 6.81E+04 6.61E+04 1.55E+05
Peak photometric
output (Cd) 3.32E+07 8.67E+06 5.92E+06 5.97E+06 1.41E+07
Sound intensity
(dB) @ 50ft 151.6 150.9 148.5 146.8 144.8

visible light generated by the rounds. The standard. The data for 70 gram and 33 gram 603
current produced by the silicon photo detector groups are not available. In summary, the 604
was converted to a voltage with an Ithaco model outperformed the 603 in sound intensity while
1211 current pre-amplifier. Depending on the underperformed in photometric output. Test
brightness of the rounds, the amplifier gain used results are summarized in Table 3.
in these measurements ranged from 10-3 A/V to
10-5 A/V. The voltage versus time data was Fragmentation
collected with both an oscilloscope and a
Labview™ based computer data acquisition Another finding from the system prove-
system. Post experiment analysis consisted of out was that a minimum of 60 gram (M115A2)
calculating the peak luminous flux, the and 30 gram (M116A1) of 603 or 604 mix were
integrated flux, rise time to peak flux, and the required to fragment the flash charge cardboard
pulse duration. These quantities were calculated housing bodies and sleeves into a few large
for each sample with an algorithm written with pieces. Fragmentation was improved by
the Labview™ programming language. increasing the amount of flash charge in the
system as demonstrated in the sample groups
Test results show that the integrated with 70 gram and 33 gram 604 mix. In general,
photometric output for the groups with 60 gram the 604 performed slightly better than the 603 at
(M115A2) and 30 gram (M116A1) 604 mix are the same charge weight with respect to the
approximatly 20% below that of the standard fragmentation. The fragmented pieces from
group. This suggests a higher amount of 604 each test group were collected in plastic bags
charge is required to match the standard and photographed for comparison in Figures 15,
performance, as confirmed by the data for the 16, and 17.
groups with 70 gram and 33 gram 604 charge
weights. It was also found that the groups with
60 gram and 30 gram 603 had over twice
amount of integrated visual output and
somewhat lower sound intensity than the

277
Figure 15: Fragmented housing bodies for 603 mix in M115A2 and M116A1

Figure 16: Fragmented housing bodies for 604 mix in M115A2

Figure 17: Fragmented housing bodies for 604 mix in M116A1

278
 CONCLUSION free-flowing oxidizer pre-blend. The fuel and
oxidizer pre-blends were then separately loaded
Two perchlorate-free flash compositions into the item for final system assembly and
have been developed and proven out. The mixing in a tumbler.
formulation consists of a high surface area flake
aluminum as the primary fuel, sulfur as the A clear polycarbonate tube with a
secondary fuel, and potassium nitrate or similar system configuration was used to
strontium nitrate as the oxidizer. Low levels of optimize the critical powder mixing parameters,
boric acid and silicon dioxide were added to such as mixing time, homogeneity, fill volume,
improve the product storage shelf life, etc. During the system prove-out, various
processing, and mixture homogeneity. The sample groups were fabricated and tested to
fuel/oxidizer ratios and particle sizes were correlate the amount of flash charge with the
optimized through a parametric study, using the sound level, photometric output, and
NASA CEA computer program to simulate the fragmentation level. It was determined that 60-
thermal properties and a 50 cc closed bomb 70 grams of 604 or 603 mix for M115A2 and
model to determine the ballistics of each 30-33 grams of 604 or 603 mix for M116A1
formulation. In an effort to completely were required to achieve satisfactory
eliminate the need for dangerous manual fragmentation and sound level performance. It
operations, a two part (fuel and oxidizer pre- was also found that the Al/SrN 603 formulation
blends) in-round loading and mixing method yielded over twice amount of the photometric
was developed to manufacture the perchlorate- output of the Al/KN 604 system and the existing
free M115A2 and M116A1 Simulators. The perchlorate counterpart. The 604 formulation
fuel pre-blend consists of flake aluminum and a was considered the best perchlorate-free
small amount of sulfur as initiation facilitator. candidate based on the overall performance,
The potassium nitrate or strontium nitrate was material cost, and storage stability.
mixed with Cab-O-Sil and boric acid to form a

REFERENCES

[1] Chen, G.; Motyka, M.; Poret, J.; “Perchlorate Free Flash Band Compositions and Method Making
Same for Pyrotechnic Training Rounds,” Patent Application Serial No. 10/907206, 24 March, 2005.

[2] Parker Bodine, S.; “Memorandum: Assessment Guidance for Perchlorate,” United States
Environmental Protection Agency, 26 January 2006.

[3] Shidlovshy, A. A.; Principles of Pyrotechnics, 1964.

[4] Conkling, J. A.; Chemistry of Pyrotechnics, Marcel Dekker, Inc., New York, 1985.

279
280