Summary

The thermoplastic explosive (TE-T7005) was developed as a Gen-

eral Purpose (GP) Insensitive High Explosive (IHE) candidate due to a

number of factors including: low small scale sensitivity characteristics;

low processing cost; theoretical high performance; re-meltability (with

associated economic and environmental bene®ts); and potential

endothermic characteristics during cook-off. Theoretical high perfor-

mance and excellent cook-off characteristics were veri®ed with sub-

sequent large scale tests

(1±4)

. This paper will report on large-scale

fragment impact sensitivity test results for the composition TE-T7005.

Fragment cubes measuring 1.27 cm61.27 cm61.27 cm were ®red at

and impacted two separate test units (loaded with the explosive TE-

T7005) at an average velocity of 8670 ft=s (2643.6 m=s). Each reaction

was judged to be a brief burning that was not sustained. No blast

pressures from the reaction of the test units were detected, indicating

that no signi®cant reaction occurred. The reaction response was

somewhat milder compared with HTPB analogs of TE-T7005

(5)

.

1. Test Description Discussion

TE-T7005 explosive was cast into a split mold ®xture

having an interior diameter comparable to the test unit

interior diameter. The test unit is known as a modi®ed

Naturally Fragmenting Test Unit (NFTU). The NFTU is

manufactured from mild steel and consists of a right circular

cylinder with exterior dimensions of 8 in.616 in.

(20.32 cm640.62 cm) and has a wall thickness of 0.375 in.

(0.95 cm). Additional end con®nement was provided by

bolting steel end plates (with a diameter of 8 in. and

thickness of 0.25 in.) at both ends of the NFTU (Fig. 1).

After the TE-T7005 explosive was loaded into the casting

®xture (which actually had a height of 22 in. (55.88 cm))

and allowed to cool and solidify, the explosive was removed

from the mold. The explosive was then X-rayed through

two mutually perpendicular, transverse axes (0



and 90



) to

verify that voids did not exist over a length of at least 17 in.

Once radiographic inspection veri®ed that no signi®cant

voids existed in the charge, the explosive was placed in a

wooden miter box. The explosive was then manually cut

with a coarse 7 bit=inch hand saw (using the X-ray ®lm as a

guide to insure that cutting excluded any voids) to a length

of 16 in. (40.64 cm). A piece of 20 grit sand paper was

adhered to a 12 in.612 in.60.5 in. (30.5 cm630.5 cm6

1.3 cm) piece of plywood and both of the explosive's ends

were smoothed to ensure that ¯at surfaces would be exposed

to the booster and bottom closure plate. Prior to insertion of

the explosive charge into the NFTU case, a 1=4 in. diameter

hole was drilled into the end of the NFTU and the NFTU

and explosive charge were both weighed. The NFTU was

then coated on the interior with RTV (R-81) to which an

Large-Scale Fragment Impact Sensitivity Test Results of a Melt

Castable, General Purpose, Insensitive High Explosive

Theodore S. Sumrall*

Sverdrup Technology, Inc., Niceville, FL 32578 (USA)

Untersuchungsergebnisse der Emp®ndlichkeit gegen Splitterauf-

schlag im Groûversuch eines schmelzgieûfaÈhigen, allgemein ver-

wendbaren, unemp®ndlichen Hochleistungssprengstoffs

Der thermoplastische Sprengstoff TE-T7005 wurde entwickelt als

moÈglicher allgemein verwendbarer, unemp®ndlicher Hochleistungs-

sprengstoff (IHE) im Hinblick auf zahlreiche Ein¯uûgroÈûen ein-

schlieûlich geringe Emp®ndlichkeit im Kleinmaûstab, niedrige

Herstellungskosten, hohe theoretische Leistung, Umschmelzbarkeit

(verbunden mit oÈkonomischen und unweltfreundlichen Vorteilen) und

potentiell endothermes Verhalten waÈhrend des Cookoff. Hohe theo-

retische Leistung und ausgezeichnete Cookoff-Eigenschaften wurden

bestaÈtigt durch anschlieûende Groûversuche

…1ÿ4†

. In der vorliegenden

Arbeit wird berichtet uÈber Versuchsergebnisse der Emp®ndlichkeit

gegen Splitteraufschlag im Groûversuch bei Composition TE-T7005.

SplitterwuÈrfel wurden abgeschossen in der GroÈûe 1.27 cm6

1.27 cm61.27 cm und beaufschlagten zwei getrennte Versuchsauf-

bauten (beladen mit dem Sprengstoff TE-T7005) bei einer mittleren

Geschwindigkeit von 8670 ft=s (2643,6 m=s). Jeder Versuch wurde

gepruÈft auf ein kurzes Auf¯ackern, das nicht anhielt. Es wurden keine

Druckwellen aus der Reaktion im PruÈfstand nachgewiesen, ein

Beweis, daû keine wesentliche Reaktion stattfand. Die Reaktion-

semp®ndlichkeit war etwas geringer im Vergleich zu HTPB-Analogen

des TE-T7005

(5)

.

ReÂsultats d'eÂtudes de sensibilite aÁ l'impact d'eÂclats lors d'un essai

aÁ grande eÂchelle sur un explosif aÁ haute puissance, insensible,

coulable par fusion, aÁ usage geÂneÂral

L'explosif thermoplastique TE-T7005 a eÂte deÂveloppe en tant

qu'eÂventuel explosif aÁ haute puissance insensible aÁ usage geÂneÂral

(IHE) compte tenu d'un certain nombre de facteurs y compris la faible

sensibilite aÁ petite eÂchelle, les couÃts de fabication reÂduits, la puissance

theÂorique eÂleveÂe, la refusibilite (lieÂe aÁ des avantages eÂconomiques et

eÂcologiques) et le comportement endothermique potentiel pendant

l'eÂchauffement. La puissance theÂorique eÂleveÂe et les proprieÂteÂs

d'eÂchauffement excellentes ont ensuite eÂte con®rmeÂes par des essais aÁ

grande eÂchelle

…1ÿ4†

. La preÂsente eÂtude fait eÂtat de reÂsultats expeÂri-

mentaux concernant la sensibilite aÁ l'impact d'eÂclats lors d'essais aÁ

grande eÂchelle avec la composition TE-T7005. On a tire des eÂclats

cubiques de taille 1,27 cm61,27 cm61,27 cm sur deux montages

expeÂrimentaux seÂpareÂs (chargeÂs de l'explosif TE-T7005) aÁ une vitesse

moyenne de 8670 ft=s (2643.6 m=s). On a aÁ chaque fois constate une

reÂaction caracteÂriseÂe par une breÁve combustion qui ne dure pas. Dans

le stand, on n'a pas deÂtecte d'onde de souf¯e avant la reÂaction, ce qui

prouve qu'aucune reÂaction importante n'a eu lieu. La reÂactivite eÂtait

leÂgeÁrement plus faible que celle de produits HTPB analogues au TE-

T7005

…5†

.

* Correspondence author

# WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1999

0721-3115/99/0306±0030 $17.50‡:50=0

30

Propellants, Explosives, Pyrotechnics 24, 30±36 (1999)

appropriate amount of Dibutyl Tin Di-Laurate (DBTDL)

was added as a cure agent=catalyst. The explosive charge

was also liberally coated with the same RTV mixture and

then inserted into the NFTU case. The hole previously

drilled into the end of the NFTU permitted air and excess

RTV to escape as there was minimal clearance between the

explosive and NFTU wall. Once the RTV was cured, the

charge was thus ®rmly bonded into the NFTU case and the

excess RTV was trimmed from the exterior of the case and

exposed explosive charge. The exposed end of the explosive

was then covered with aluminium foil, securely taped, and

shipped to the explosive testing facility where the end plates

were attached and the charges were subjected to fragment

sensitivity testing.

Fragment sensitivity tests were conducted in accordance

with MIL-STD-2105B by subjecting the modi®ed NFTUs

to multiple, random impacts of 250 grain steel cubes that were

launched byanexplosive charge. Thelaunchingchargewasan

8-in. by 8-in. by 32-in. block of Comp-B explosive (57 kg) that

was designed to accelerate the cubes to a velocity of at least

8300 ft=s. Figures 2 and 3 illustrate the test con®guration and

photographic coverage for the test.

2. Theory and Procedure

2.1 Theory

The fragment impact test is designed to determine the

sensitivity of an explosive composition to high velocity

impact and assessment of subsequent shock to detonation

transfer (SDT). Traditional PBX development encourages

the use of a ``soft pliable'' binder based upon the theory that

a soft binder will absorb impact shock and help dissipate

energy which might otherwise contribute to the initiation of

the explosive. Earlier work with the experimental explosive

(TE-T7005) resulted in signi®cantly improved sub-scale

sensitivity characteristics

(1)

. However, the thermoplastic

binder used to manufacture TE-T7005 is quite hard (similar

to TNT) and is not soft like traditional HTPB based PBX

compositions

(3)

. Theoretically, therefore, the TE-T7005

should be more sensitive to fragment impact, compared to a

composition manufactured with similar solids but employ-

ing a curable ``soft'' binder such as HTPB.

2.2 Procedure

The fragment impact test requires that a certain minimum

area of explosive be shocked to a certain minimum pressure.

For projectile impact scenarios, this requires a certain

minimum projectile velocity, referred to as the ``critical

impact velocity.'' The larger the area impacted, the lower

the critical impact velocity. Therefore, impactors with the

lowest convexity (¯at faces) are most effective at initiating

detonation by an SDT mechanism. However, ¯at-faced

impacts upon barriers are more likely than edge or corner

impacts to shatter the projectile so that subsequent pene-

tration and shock-energy transfer capabilities may both be

reduced

(6)

. However, by accelerating the fragments to a

minimum velocity of 8300 ft=s (2529.8 m=s), a maximum

shock transfer is accomplished. When a fragment strikes an

explosive, four responses can result: detonation; explosion;

burning; or no reaction. A detonation is de®ned as a high

order reaction which pierces a 1=2 in. thick mild steel wit-

Figure 1. NFTU assembly for fragment impact test.

Propellants, Explosives, Pyrotechnics 24, 30±36 (1999)

Fragment Impact Sensitivity Test Results of IHE 31

ness plate with a hole that is approximately the same dia-

meter as the acceptor charge (the NFTU in this case) and is

classi®ed as an IM ``Class-I Reaction''. An explosion is a

lower order violent high pressure reaction that can damage

the test stand, the case material and explosive, and throws

large fragments of case material or explosive  50 feet

(15.24 m). In an explosion, the witness plate does not sus-

tain damage and is classi®ed as an IM ``Class-II'' or ``Class-

III Reaction''. A sustained burn reaction consists of ener-

getic material ignition and burning until all of the explosive

is consumed. This usually takes a number of minutes

depending on the type of explosive, the mass of the

explosive and level of con®nement. Burning is classi®ed as

an IM Class-IV or Class-V reaction depending on the

severity of the explosion.

Prior to test setup, each unit was X-rayed through two

mutually perpendicular, transverse axes. Each modi®ed test

unit with additional end con®nement was placed on a 1-ft

Figure 2. Fragment impact test setup.

32 Theodore S. Sumrall

Propellants, Explosives, Pyrotechnics 24, 30±36 (1999)

by 1-ft by 2-in. steel dent plate that had been placed on a

wooden test stand. The longitudinal axis of the test unit was

oriented vertically and its geometric center was approxi-

mately 51 in. above the ground plane. Two metal banding

straps were used to secure the test unit and the dent plate to

the test stand. One 22-gauge steel witness panel was placed

near the test unit to collect fragment velocity data with high

speed 16 mm cameras. Another 22-gauge steel witness

panel was placed behind the test unit to monitor test unit

debris. A 2-ft by 2-ft by 7=8-in. steel witness panel was

placed at a 2-ft standoff distance from the test unit, and

®berboards were positioned behind the panel to catch it. The

launching charge (125 lb. of Comp-B explosive) was placed

on a test stand at a nominal distance of 16 ft from the test

unit. The launching charge was adjusted to the same ele-

vation as the test unit, and the fragment mat was then

attached to the launching charge. An electrical sensor was

attached to the fragment mat to detect ®rst motion and the

signal from this circuit was recorded on each camera and

data recorder to provide a common data reference. An

electronic velocity screen was placed in front of the test unit

or wrapped around the test unit to collect fragment velocity

data as shown in Figure 2. Figure 4 shows a test unit

arranged on the dent plate and the test stand. Figure 5 shows

a test unit in relation to the test structures, and the launching

charge and fragment mat assembly. Test events were

documented using a VHS video cassette recorder in con-

junction with a color, closed-circuit television system. The

video cassette record was annotated with the date and time

of each test. The launching charge was then detonated using

a J-2 blasting cap and a 1-in. diameter by 1-in. length CH-6

booster pellet. Upon completion of the test, test unit debris

was recovered, and the size and location of each piece of

debris was noted.

Fragment velocity data was collected via high speed

photography. The time base was established by counting the

number of frames between the ®rst light caused by the

initiation of the launching charge and the ®rst light caused

by impact of the fragments on the test unit, and then

adjusting the time base by the length of time required for

the detonation to propagate through the booster and

launching charge to accelerate the fragments. Fragment

velocities were calculated with the following equation

V ˆ

D

F=R ÿ T

where: V ˆ velocity of fragments, D ˆ distance between

fragment mat and surface of test unit, F ˆ frames on photo-

Figure 3. Camera ®elds of view for fragment impact test.

Figure 4. NFTU arranged on test stand for fragment impact test.

Propellants, Explosives, Pyrotechnics 24, 30±36 (1999)

Fragment Impact Sensitivity Test Results of IHE 33

graphic ®lm between time launching charge was initiated

and time fragments impacted test unit, R ˆ rate of frames on

photographic ®lm, and T ˆ time for detonation to propagate

through booster and launching charge.

The time for the detonation to propagate through the

booster and the propelling charge was established as fol-

lows:

T ˆ

Booster Length

Booster Detonation Rate

‡

Launching Charge Length

Launching Charge Detonation Rate

T ˆ

0:0254 m

8550 m=s

‡

0:8128 m

7840 m=s

ˆ 107 ms

3. Data

Fragment impact tests were conducted on explosive TE-

T7005 in modi®ed NFTUs (S=N 92-16 and S=N 92-17) on

9=16=92. Fragment impact tests were also conducted on

explosive PBXW-124 (S=N 91-25 and S=N 91-29) and

PBXW-126 (S=N 94-3 and 94-4), HTPB analogs of TE-

T7005. As originally formulated, the only difference

between PBX-125 and PBX-126 was the RDX particle

size

(9)

. Data for the PBX compositions are presented for

comparative purposes.

3.1 TE-T7005 Test Results

NFTU S=N 92-16 was impacted by six fragments. The

average velocity of the fragments was 8697 ft=s (Fig. 6).

The reaction was judged to be a brief burning reaction that

was not sustained (Type-V). The test unit ruptured and

approximately 75% of the explosive was scattered about the

test site. Approximately 25% of the explosive remained in

the case debris (Fig. 7). No blast pressure from the reaction

of the test unit was detected at the blast gauge locations.

NFTU S=N 92-17 was impacted by ®ve fragments. The

average velocity of the fragments was 8642 ft=s. The reac-

tion was judged to be a burning reaction. Some of the

explosive that was ejected from the case and some of the

explosive that remained in the case debris burned (Fig. 8). A

large quantity of the explosive that was ejected from the

case did not react (Fig. 9). No blast pressure from the

reaction of the test unit was detected at the blast gauge

locations.

3.2 PBXW-124 Test Results

NFTU S=N 91-25 was impacted by four fragments. The

average velocity of the fragments was 8502 ft=s. The reac-

tion was judged to be a burning reaction. The additional

con®nement hardware remained attached to the case of the

NFTU. The case remained as an assembly, but had two

cracks along the impact points. The explosive material was

consumed in the burning reaction. The test unit debris was

found in the immediate test area, within 21 ft. of the pre-test

location. No blast pressure from the reaction of the test unit

was detected at the blast gauge locations.

NFTU S=N 91-29 was impacted by four fragments. The

average velocity of the fragments was 8673 ft=s. Four large

areas of explosive byproducts were found. The case was

split along the impact points and a large amount of liner

material was observed in the case. The test unit debris was

found in the immediate test area, within 55 ft. of the pre-test

location. The case and end plate assembly were found at a

distance of 26 ft. of the test site and the closure plate was

found at a distance of 24 ft. of the test site. No blast pressure

from the reaction of the test unit was detected at the blast

gauge locations.

3.3 PBXW-126 Test Results

NFTU S=N 94-3 was impacted by three fragments. The

average velocity of the fragments was 8579 ft=s. The reac-

Figure 5. Fragment impact test unit in relation to test structure.

Figure 6. Fragment impact test results of explosive TE-T7005 (SN

92=16).

34 Theodore S. Sumrall

Propellants, Explosives, Pyrotechnics 24, 30±36 (1999)

tion was judged to be a burning reaction and was initiated

immediately upon fragment impact. Two additional steel

cubes grazed the case such that their contribution to the

reaction was considered negligible. The case, the end plate,

and the closure plate were found as an assembly (along with

the additional end con®nement hardware) at a distance of

12 ft. from the pre-test location. Explosive burning was

observed for approximately 1.5 minutes and all of the

explosive was consumed in the reaction. No blast pressure

from the reaction of the test unit was detected at the blast

gauge locations.

NFTU S=N 94-4 was impacted by four fragments. The

average velocity of the fragments was 8548 ft=s. The reac-

tion was judged to be a burning. The case and the end plate

were found as an assembly (along with the additional end

con®nement hardware) at a distance of 9 ft. from the pre-test

location. Explosive burning was observed for approxi-

mately 1.2 minutes and all of the explosive was consumed

in the reaction. No blast pressure from the reaction of the

test unit was detected at the blast gauge locations.

A summary of results for fragment impact tests are

summarized in Table 1.

4. Discussion

This research project essentially involved replacing the

traditional HTPB=IPDI cured binder (used to manufacture

PBX-124, PBX-125 and PBX-126) with a proprietary

thermoplastic binder (TTB-531)

(1,3)

. The experimental

explosive tested above (TE-T7005) contained solids which

bracketed the solids loading for PBX-124=125=126, and

therefore, the essential difference between the thermoplastic

composition and the HTPB versions was the binder type.

Table 2 details the composition of the PBX compositions

and the thermoplastic analog, TE-T7005.

While the theory which supported manufacture of PBX

compositions with ``soft and pliable'' binders (such as

HTPB) appears to be logical, test results indicate that the

explosive composition manufactured with the harder binder

(TE-T7005) was actually less sensitive than the composi-

tion manufactured with the softer binder type. Other factors

than physical properties may be entering into the equation

and the following theories are offered as possible explana-

tions for the observations noted.

Figure 7. Fragment impact test results of explosive TE-T7005 (SN

92=16).

Figure 8. Fragment impact test results of explosive TE-T7005 (SN

92=17).

Figure 9. Fragment impact test results of explosive TE-T7005 (SN

92=17).

Table 1. Summary of Results for Fragment Impact Tests

Explosive

Type

NFTU

S=N

# of Hits

Fragment

Velocity

Reaction

(Type)

TE-T7005

92-16

6

8697 (ft=s)

Burn (V)

TE-T7005

92-17

5

8642 (ft=s)

Burn (V)

PBXW-124

91-25

4

8502 (ft=s)

Burn (V)

PBXW-124

91-29

4

8673 (ft=s)

Burn (V)

PBXW-126

94-3

3

8579 (ft=s)

Burn (V)

PBXW-126

94-4

4

8548 (ft=s)

Burn (V)

Propellants, Explosives, Pyrotechnics 24, 30±36 (1999)

Fragment Impact Sensitivity Test Results of IHE 35

(1) As the binder is thermoplastic in nature, it is possible

that the binder is absorbing impact energy and thermal

energy from the high speed fragments (which is both

kinetic and thermal in nature) by ®rst undergoing a

melting reaction. Energy which might be imparted unto

the energetic solids (such as RDX) could be imparted

®rst unto the binder and the binder would then absorb

some of this energy (in a thermal melting nature) and

only the remaining energy would be transmitted to the

energetic solids.

(2) The end of mix viscosity of the thermoplastic explosive

TE-T7005 was typically less than 2 kP as measured by a

Brook®eld rheometer. This contrasts with end of mix

viscosities as high as 20 kP or greater for the PBX

formulations. The solids are therefore coated much

better by the thermoplastic binder which, by virtue of

its' lower viscosity, is much more able to enter into the

pores of energetic materials and thus eliminate=

decrease sites for hotspot formation.

5. Conclusions

The data demonstrate that TE-T7005 is a highly impact

insensitive explosive composition. Earlier data demonstrate

higher performance characteristics relative to the HTPB

analogs and even superior performance characteristics

relative to a number of highly sensitive high explosives

(5)

.

Additional data reveal very low thermal sensitivity relative

to the HTPB analogs

(4)

.

An analysis of the post-test photographs reveal that while

all reactions were classi®ed as ``burn reactions'', TE-T7005

reacted in a much milder manner (relative to the HTPB

analogs) and much un-reacted explosive was recovered

after fragment impact. It is not surprising that the HTPB

formulations (PBXW-124 and PBXW-126) almost com-

pletely burned when subjected to fragment impact testing

because HTPB was originally designed to be a solid rocket

motor fuel. It is surprising, however, that the TE-T7005

explosive did not burn more violently because earlier

research indicated that the aluminium used to manufacture

TE-T7005 was much more reactive than the aluminium

used to manufacture the PBX analogs such as PBX-109

(8)

.

6. Recommendations

Other PBX compositions (i.e. PBX-109) should be

manufactured with the TTB-531 thermoplastic binder

replacing the conventional HTPB binder and then subjected

to fragment impact testing to determine if the thermoplastic

binder system can reduce the shock sensitivity of much

more sensitive compositions.

7. References

(1) T. S. Sumrall and W. H. Graham, ``Formulation of a Melt Castable

General Purpose Insensitive High Explosive'', Journal of Japan

Explosive Society 58, 2 (1997).

(2) T. S. Sumrall and W. H. Graham, ``Melt Castable PBXW-124=125

Development'', IM Technology Symposium, Williamsburg, VA,

15±18 June, 1992.

(3) U.S. Patent Application, Docket No. 1090.6.13.

(4) T. S. Sumrall, ``Large Scale Thermal Sensitivity Results of a Melt

Castable General Purpose Insensitive High Explosive'', 23rd

International Pyrotechnics Seminar, Tsukuba, Japan, October,

1997.

(5) W. P. Burgess, ``Report of Vulnerability and Performance Tests of

ABF Candidate Explosives''. (U); Explosion Dynamics Branch,

NSWC, Dahlgren, VA; July, 1990.

(6) H. R. James, ``TTCP-WAG-11 Bullet=Fragment Protocol'', AWE

(Foulness), UK, 1992.

(7) MIL-STD-2105B, ``Hazard Assessment Tests for Munitions'', 12

January, 1994.

(8) T. S. Sumrall, ``Sub-Scale Ingredient Screening to Predict Burn

Rate and Performance of an Insensitive Aluminized General

Purpose Explosive'', 23rd International Pyrotechnics Seminar,

Tsukuba, Japan, October, 1997.

(9) L. T. Wilson, D. R. Reedal, and B. M. Simpson, ``Comparison of

PBXW-126 and PBXC-129 For Use in Large Fragmentation

Warheads'', Insensitive Munitions and Energetic Materials Tech-

nology Symposium, Tampa, FL, October, 1997.

Acknowledgements

Gratitude is expressed to Thiokol Corp. and the US Government

who funded research; development; scale-up; and advanced testing,

and who authorized publication of this information in the open lit-

erature. This scale-up effort was conducted at the Huntsville Division

of Thiokol Corp. where the author was employed as the Principal

Investigator until closure of that facility. Testing was conducted at

NSWC=Dahlgren, VA. Photographs courtesy of NSWC

(4)

.

(Received August 28, 1997; revised April 4, 1998; Ms

45=97 rev)

Table 2. Composition of PBX Formulations

Ingredient

Vendor

(PBX-124)

(TE-T7005)

(PBX-126)

Wt.% and

Wt.% and

Wt.% and

Diameter (mm)

Diameter (mm)

Diameter (mm)

HTPB ‡ E702

4.85 ‡ 0.05

N=A

4.44 ‡ 0.05

IPDI ‡ TPB

0.46 ‡ 0.01

N=A

0.45 ‡ 0.01

IDP ‡ Lecithin

7.23 ‡ 0.4

N=A

6.65 ‡ 0.4

TTB-531

Thiokol

N=A

12

N=A

Al Powder

Reynolds

20=18

23=17

26=18

AP

Kerr McGee

20=200

20=200

20=200

NTO

Olin

27=250

25=250

22=250

RDX

Holston AAP

20=4

20=4

20

36 Theodore S. Sumrall

Propellants, Explosives, Pyrotechnics 24, 30±36 (1999)