Summary

An extensive study of the initiation of nitromethane by shaped-charge

jets has been conducted and is reported in the paper following this one.

In order to accomplish this a great deal of background information and

description of the effect of temperature on failure diameter, measure-

ment of jet diameters, material properties and experimental con®gura-

tion is necessary. The collection of these data is of general interest to

the ®eld of energetic materials research, and is presented here as a

separate paper.

1. Introduction

This paper provides a summary of the analysis techniques

and material parameters used to study the effect of failure

diameter on the initiation of nitromethane by a hypervelocity

jet. The results of that study

(1)

are found in the next article.

Evaluation of a hypervelocity jet initiation threshold criter-

ion requires knowledge of the initiation mechanism as a

function of both the explosive and jet parameters. An

understanding of the individual material properties of the

explosive and shaped-charge, and their dynamic interaction

is therefore appropriate. The explosive failure diameter (d

f

)

is often listed as a characteristic parameter of energetic

materials. The magnitude of the failure diameter is a func-

tion, in part, of the reaction zone thickness

(2)

, temperature,

con®nement, density, porosity, particle size, and concentra-

tion of impurities. It is de®ned as the minimum diameter at

which a detonation wave in the explosive is or becomes

unsustained due to lateral energy losses that exceed the

energy supplied to the shock front through the sonic region

by the reaction of the energetic material.

Shaped-charges

(3)

consist of high explosive (HE) billets

machined to accommodate liners of thin layers of materials,

e.g. glass or metal. The opposing end of the HE cylinder is

machined for a detonator and a booster pellet, which are

used to initiate the HE. The spherical detonation wave

encounters the apex of the liner forcing the liner material to

accelerate and collapse on the centerline, creating a jet. The

leading tip of the jet can have velocities in excess of

9 mmyms, resulting in penetration velocities much faster

than HE sound velocities. The trailing end of the jet, the

slug, travels at much lower velocities. This velocity gradient

initially causes the jet material to stretch until it undergoes

particulation. During particulation the jet necks, and then

separates into multiple particles with different velocities.

The virtual origin

(3)

is de®ned as the apparent point of

origination of these particles. It is the nearly common

intersection point of the extrapolated particle trajectories

back in time. This information is useful in determining the

jet's estimated position relative to a selected time after

initiation of the detonator.

2. Materials and Properties Measurements

Four different materials were used for the tests: nitro-

methane (NM, CH

3

NO

2

); diethylenetriamine (DETA,

NH

2

CH

2

CH

2

NHCH

2

CH

2

NH

2

); boron carbide (B

4

C); and

Cab-o-sil TS-720 fumed silica (TS720, SiO

2

). The funda-

mental detonation properties of NM have been previously

characterized, and the explosive is readily available in

commercial-grade form with low concentrations of impu-

Improved Characterization of Nitromethane, Nitromethane

Mixtures, and Shaped-Charge Jet Properties

D. I. Idar and B. W. Asay

High Explosives Science and Technology, MSC920, Los Alamos National Laboratory, Los Alamos,

New Mexico 87545 (USA)

E. N. Ferm

Hydrodynamic Applications, MS P940, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 (USA)

Verbesserte Charakterisierung der Eigenschaften von Nitro-

methan, Nitromethan-Mischungen und dem Hohlladungsstachel

Eine ausfuÈhrliche Studie uÈber die Initiierung von Nitromethan durch

Hohlladungsstachel ist durchgefuÈhrt worden und wird in der nachfol-

genden Arbeit beschrieben. Um diese zu vervollstaÈndigen ist eine groûe

Menge an Hintergrundinformation, die Beschreibung der Wirkung der

Temperatur auf den Versagensdurchmesser, die Messung des Stachel-

durchmessers, die Materialeigenschaften und der experimentelle Auf-

bau notwendig. Die Sammlung dieser Daten ist von groûem Interesse

fuÈr die Forschung auf dem Gebiet der energetischen Materialien und

wird in dieser separaten Arbeit vorgestellt.

CaracteÂrisation ameÂlioreÂe des proprieÂteÂs du nitromeÂthane, de

meÂlanges nitromeÂthane et du jet de charge creuse

On a effectue une eÂtude deÂtailleÂe sur l'initiation du nitomeÂthane par

un jet de charge creuse, que l'on deÂcrit dans une eÂtude faisant suite aÁ la

preÂsente eÂtude. Pour pouvoir compleÂter cette eÂtude, on a besoin d'une

grande quantite d'informations de fond, de la description de l'effet de la

tempeÂrature sur le diameÁtre rateÂ, de la mesure du diameÁtre de jet, des

proprieÂteÂs du mateÂriau et du montage expeÂrimental. La compilation de

ces donneÂes preÂsente un grand inteÂreÃt pour la recherche dans le

domaine des mateÂriaux eÂnergeÂtiques et est preÂsenteÂe dans cette eÂtude

seÂpareÂe.

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

0721-3115/99/0306±0001 $17.50‡:50=0

Propellants, Explosives, Pyrotechnics 24, 1±6 (1999)

1

rities. At ambient conditions NM is a clear homogeneous

liquid making it ideal for optical measurements and diag-

nostic gauge placement. Previously Engelke and others

demonstrated that the NM failure diameter can be easily

modi®ed by the addition of chemicals

(4,5)

, or particulates

(6)

.

These methods allow for experimentally sampling a range of

failure diameters and evaluating homogeneous and

heterogeneous materials in the same medium.

DETA was used to chemically sensitize NM, and it

effectively decreased the failure diameter (d

f

) by up to a

factor of four at room temperatures. The DETA was obtained

from a single source of technical-grade J. T. Baker, Lot

D13632, H768-07 DETA. All of the NMyDETA mixtures

were tinged a light brown color after preparation. Chemical

degradation was minimized by performing the tests within

an hour of solution preparation.

NM heterogenization was achieved by the addition of B

4

C

particles gelled with TS720. The average particle size of the

B

4

C was determined to be 1.3 mm. The fumed silica was

purchased from the Cabot Corporation, Cab-O-Sil Division,

Tuscola, IL. It is surface-treated SiO

2

with a moisture

content of < 0.5% at 105



C as determined by heating tests

performed by Cabot Corporation. Mixtures were prepared

with the addition of 3 wt% TS720 to the NM, followed by

the B

4

C. All of the NMyB

4

CyTS720 mixtures were dark

gray in color similar to the B

4

C additive. Notable settling of

particulates was avoided by vigorous agitation of the mix-

ture just prior to testing.

3. Failure Diameter Measurements

Failure diameter values were determined using the plate

dent technique. A detonation was determined by the pre-

sence of a characteristic dent in a steel witness plate. Various

diameters of standard-wall Pyrex tubing, 254-mm length,

were used for the tests, with the smallest length to diameter

ratio exceeding thirteen. SE-1-PT detonators coupled with

PBX-9407 pellets (94.0y6.0 wt% RDXyExon 461) and

Composition

B

boosters

(59.5y39.5y1.0 wt%

RDXyTNTyWax) were used to shock initiate the NM and

NM mixtures. Six tubes of different diameters were aligned

inside a foam box for each NM mixture. On occasion several

NM mixtures were tested at the same time, and these are

identi®ed with the same test number. The foam box and a

forced-air heating system were used for temperature control.

Thermocouple temperature measurements were taken

directly in the liquids or in the box. The d

f

values were

calculated as the average of two adjacent tube diameters for

which the detonation propagated and failed. The range

values on the failure diameters, Dd

f

, are reported as half the

difference between the adjacent tube diameters.

3.1 Neat NM

A number of variables can affect HE material character-

istics including composition, con®nement, density, particle

size, porosity, impurities, and temperature. Typically, as the

bulk temperature is lowered in liquid energetic materials,

barring a phase change, the failure diameter value will

increase

(7,8)

. The temperature effect on the NM failure dia-

meter is signi®cant

(7)

, increased temperatures correlate with

decreased failure diameters. We derived a failure diameter

equation as a function of temperature using previously

published data

(9)

.

d

f

ˆ d

fo

e

…TÿT

o

†=ÿ60:913

…1†

where T is the temperature in Celsius, d

f

is the calculated

failure diameter, and T

o

, and d

fo

are our reference tempera-

ture and failure diameter measured for neat NM, 19.3



C and

16.4 1.1 mm, respectively. Equation (1) varies slightly

from the form of the equation we initially used to ®t the data,

and reported in a previous paper

(10)

. The slightly modi®ed

new form shown in Eq. (1) resulted in a smaller change in

the failure diameter and was calculated to be an average

ÿ0.25 mmy



C over the temperature range from ‡19 to

28



C.

3.2 NMyDETA

Additives, such as DETA, also can greatly affect the

failure diameter of NM. Initially several plate dent tests were

performed using selected concentrations of DETA to match

Engelke's results

(4)

; however, these initial data indicated the

DETA we were using had a stronger effect. Thus, we

reproduced the entire data set and checked the precision of

our earlier measurements. Our measured d

f

and corre-

sponding temperature values are given in Table 1. For ana-

lysis and comparison, the NMyDETA failure diameters were

adjusted with a temperature correction to 22



C. We assumed

the same functional temperature dependence of the failure

diameter as in our neat NM results using the average change

of ÿ0.25 mmy



C. These data are also given in Table 1.

Detonations propagated in all six tubes in two of the tests.

The failure diameter values for these are reported as less than

Table 1. Failure Diameters and Ranges for our NMyDETA Mixtures

Test #

NMyDETA

Measured d

f

Measured HE Temp. Corr.

(wt%)

 Dd

f

(mm) Temperature

d

f

 1(



C)

(mm) to 22



C

F-6469 100.0y0.0

16.4 1.1

19.3

15.7

F-6465

99.99y0.01

9.8 0.85

23.6

10.2

B-9845 99.99y0.01

9.7 0.8

19.0

9.0

B-9792 99.99y0.01

9.3 1.4

22.0

9.3

B-9842 99.99y0.01

<10.3

21.0

<10.1

(a)

B-9845 99.98y0.02

9.8 0.8

20.0

9.3

B-9842 99.98y0.02

9.0 1.3

21.0

8.8

F-6465

99.98y0.02

8.5 0.6

22.3

8.6

F-6468

99.98y0.02

8.5 0.6

19.2

7.8

F-6465

99.97y0.03

8.4 0.6

20.0

7.9

B-9845 99.97y0.03

<7.9

20.0

<7.4

(a)

B-9845 99.96y0.04

8.5 0.6

20.0

8.0

B-9845 99.947y0.053

7.4 0.6

20.0

6.9

F-6469

99.801y0.199

4.4 0.5

18.1

3.5

(a)

These data are not shown in Figure 1.

2 D. I. Idar, B. W. Asay, and E. N. Ferm

Propellants, Explosives, Pyrotechnics 24, 1±6 (1999)

the smallest diameter Pyrex tubing. For comparison

Engelke's NMyDETA data

(4,11)

are given in Table 2, and the

same temperature correction was applied to the failure dia-

meter values, that is an average change of ÿ0.25 mmy



C.

Based on the assumption that the failure radius of the

solution was inversely proportional to the one-dimensional

reaction zone length, Engelke derived an equation for the

failure radius as a function of DETA concentration. This can

easily be rewritten in terms of the failure diameter as shown

by Eq. (2). The failure diameters of the NMyDETA mixture

and neat NM are represented by d

f

and d

fo

respectively, and

A is a proportionality constant characteristic of the NM

mixture.

1

d

2

f

ˆ

1

d

2

fo

‡ A‰wt% DETAŠ

…2†

Using this equation we calculated linear regression ®ts for

both our data set and Engelke's NMyDETA data. The neat

NM data points were weighted more heavily for these ®ts

than the NM mixture data based on the following assump-

tion. The inherent experimental errors for the NM mixtures

will be larger due to the potential errors induced by mixture

preparation, i.e. slight errors in the calculation of the additive

concentrations introduced by mass measurement errors, and

a higher probability of introducing additional impurities

other than the additive to the mixture. These ®ts are graphed

with the experimental data in the form of Eq. (2) in Figure 1.

The error bars for each data point were calculated by sub-

stituting the range values, Dd

f

, and failure diameters, d

f

,

given in Tables 1 and 2, into Eq. (3).



2d

f

Dd

f

‡ Dd

2

f

d

4

f

‡ 2d

3

f

Dd

f

‡ d

2

f

Dd

2

f





…3†

A comparison of the slope values, 0.394 for our data with a

correlation coef®cient of 0.986, and 0.169 for Engelke's with

a correlation coef®cient of 0.994, indicates that our DETA

was more effective at reducing the failure diameter. How-

ever, it is not clear from the analyses if this was due solely to

the fact that we used a different source for the DETA, or if

there were other contributing factors.

3.3 NMyB

4

CyTS720

Boron Carbide, B

4

C, and TS720 additives were used to

make a heterogeneous particulate NM mixture. Engelke

(12)

demonstrated earlier that NM's failure diameter could be

reduced with glass beads suspended in a mixture of guar

gum and NM. He showed that the extent of the d

f

reduction

was a function of the glass bead diameter, with beads in the

1±4 mm diameter range exhibiting the largest change. The

average particulate size of our B

4

C was 1.3 mm.

Only temperature measurements of the foam box interior

were obtained in the NMyB

4

CyTS720 plate dent tests, and

all were within 26 2



C with the exception of one. A single

plate dent test with 3 wt% TS720 mixed with NM was used

to determine the effect of the fumed silica alone on the

failure diameter. We measured a failure diameter of

16.5 1.1 mm at 21.1



C, reasonably comparable to the

failure diameter previously obtained for our neat NM test.

Five other plate dent tests were performed with different B

4

C

concentrations in NM. These data and the calculated tem-

perature corrected failure diameters to 22



C are given in

Table 3. Again we assumed the same functional dependence

on temperature as our neat NM results using the average

change as ÿ0.25 mmy



C. For comparison, Engelke's

NMyglass beadyguar gum data

(12)

are given in Table 4 with

Table 2. Failure Diameter Measurements and Ranges for Engelke's

NMyDETA

(4)

Test #

NMyDETA

Measured d

f

Measured HE Temp. Corr.

(wt%)

 Dd

f

(mm)

Temperature

(



C)

d

f

(mm) to 22



C

C-4899 100.00y0.0

16.2 0.4

24.7 1

16.9

C-4898 99.99y0.01

11.11 0.52 25.0 2

11.9

C-4896 99.98y0.02

10.3 0.19 25.0 2

11.1

(a)

99.97y0.03

9.24 0.12

(a)

24 1

(a)

9.7

C-4888 99.96y0.04

8.56 0.53 24.5 1

9.2

C-4887 99.947y0.053 7.54 0.55 25.9 1

8.5

(b)

99.75y0.25

4.2 0.8

(b)

24.0 ‡ 1, ÿ0

(b)

4.7

(a)

Please see footnote (e) of Table 1 of Engelke's NMyDETA work

(4)

.

(b)

Private communication, R. Engelke, 1994

(11)

.

Figure 1. NMyDETA failure diameter data in the form 1=…d

2

f

† as a

function of wt% DETA added to NM, and plotted in the form of Eq. (2)

with the corresponding temperature corrected to 22



C.

Our data are compared with Engelke's NMyDETA

(4)

data. A linear

regression analysis was performed on both sets of data to determine

slope and intercept values. For our NMyDETA data the ®t produced

1=…d

2

f

† ˆ 0:004 ‡ 0:394 [wt% DETA], and for Engelke's data

1=…d

2

f

† ˆ 0:004 ‡ 0:169 [wt% DETA].

Table 3. Measured Failure Diameters and Ranges for NMyB

4

Cy

TS720 Mixtures

Test # NMyB

4

CyTS720 Measured d

f

Measured

Temp. Corr.

(wt%)

 Dd

f

(mm) Box Temp.

d

f

 1(



C)

(mm) to 22



C

C-6451

97.0y0.0y3.0

16.5 1.1

21.1

16.3

C-6521

96.5y0.5y3.0

9.7 0.7

28.0

10.4

F-6454

96.0y1.0y3.0

8.5 0.5

24.7

9.2

F-6453

95.5y1.5y3.0

7.4 0.6

25.8

8.4

B-9841

95.0y2.0y3.0

7.4 0.6

28.0

8.9

F-6452

93.5y3.5y3.0

4.7 1.0

24.1

5.2

Propellants, Explosives, Pyrotechnics 24, 1±6 (1999)

Improved Characterization of Nitromethane, Nitromethane Mixtures 3

same type of calculated temperature corrected failure diam-

eters.

Both sets of these data were ®t using the form given in Eq.

(2) for the different additive concentrations using the same

assumption for weighting the neat NM points more heavily

as described previously in the NMyDETA failure diameter

analysis. These ®ts are graphed with the experimental data in

Figure 2. Error bars for these data were also calculated using

Eq. (3). The data produced by these ®ts, 0.011 slope with a

correlation coef®cient of 0.935 for the boron carbide mix-

ture, and 0.001 slope and 0.916 correlation coef®cient for

the glass bead mixture, indicate that the NMyB

4

CyTS720

mixtures are more effective at reducing the failure diameter

than the glass beads by weight percentage comparison.

A comparison of our chemical NMyDETA versus physical

additive B

4

C data shows that the B

4

C suspension is less

effective than the chemical additive DETA on an equivalent

weight percentage basis. Differences in the data obtained

between the glass bead and B

4

C mixtures could be attributed

either (1) to differences in the number density of each in

mixture, i.e. a difference in the number of available hot spots

for heterogeneous initiation; or (2) to impedance mismatch

differences; or (3) to a function of both variables as

demonstrated with the following arguments.

A comparison of the glass bead and B

4

C number densities

per unit weight can be easily calculated using the following

assumptions: (1) spheroid geometry with an average dia-

meter of 2.5 mm and 1.3 mm for the glass beads and B

4

C,

respectively, and (2) densities of 2.45

(12)

and 2.52

(13)

gycm

3

for the glass beads and B

4

C, respectively. Based on these

assumptions the number density for the B

4

C is calculated to

be almost seven times greater than for the glass beads for

the same weight percentage.

Impedance comparisons are not as easily calculated

because the sound speed data for B

4

C is not readily avail-

able. However, if we assume (1) the same sound speed

characteristics for both the glass beads and B

4

C doped NM

in unshocked and shocked material, and (2) the impedance

differences are only a function of the differences in the

respective densities, then the B

4

C would have an impedance

approximately 3% greater than the glass beads. This slightly

higher impedance would lead to a higher re¯ected shock of

an incident shock wave, and increase the likelihood of

initiation.

Based on the arguments just described, we can safely

assume that both the higher number density of B

4

C parti-

culates on a comparable weight percentage basis to the glass

beads, and the slightly higher impedances would most likely

lead to more signi®cant decreases in the failure diameter of

the heterogeneous mixture.

4. Shaped-Charged Jet Characterization

Viper shaped-charges were used for all of the tests, with

the exception of one. These shaped-charges were readily

available and known to be fairly reproducible. However,

detailed analysis of the jet diameter and velocity under dif-

ferent attenuating conditions was still required. This infor-

mation was used to evaluate velocities under different

experimental conditions, and the virtual origin positions and

times.

The Viper shaped-charge has an outer copper cone diam-

eter of  65 mm, with a cone angle of  42



, and is driven

by LX-14 (95.5y4.5 wt% HMXyEstane). Bare billets were

used for all of our tests, and the HE was initiated with SE-1-

PT detonators and PBX-9407 booster pellets. For the single

test without the Viper, a TOW-IIA tip charge, with a 38-mm

diameter copper cone was used. This extended our data

range by changing the jet diameter with respect to the HE

failure diameter.

The explosively formed copper jet was radiographed at

two different times using either 150 keV or 450 keV ¯ash X-

ray units. Magni®cation was determined by measurements

on a static radiograph of a calibrated copper ®ducial rod. The

X-ray exposure times were recorded on Los Alamos

National Laboratory designed time interval meters, with

relative recording errors of 2±3 nanoseconds.

Jet velocities were calculated from the radiographic and

timing data. The jet velocity at the time of impact was

Table 4. Failure Diameters and Ranges for the NMyGlass Beadsy

Guar Gum

(12)

Test # NMyglassyguar Measured d

f

Measured

Temp. Corr.

(wt%)

 Dd

f

(mm) Box Temp.

d

f

(



C)

(mm) to 22



C

E-4893 98.5y0.0y1.5

17.2 0.6

24.2 0.5

17.8

E-4904 98.0y0.5y1.5

16.1 0.5

23.3 0.5

16.4

E-4903 96.75y1.75y1.50

12.6 1.0

20.0 2.0

12.1

E-4896 95.5y3.0y1.5

10.8 0.7

24.4 0.5

11.4

E-4900 92.5y6.0y1.5

10.8 0.7

24.4 0.5

11.4

E-4902 89.5y9.0y1.5

9.8 0.7

21.0 2.0

9.6

Figure 2. NMyB

4

CyTS720 mixture and Engelke's NMyglassy guar

(12)

failure diameter data in the form 1=…d

2

f

† as a function of wt% additive

added to NM, and plotted in the form of Eq. (2) with the corresponding

temperature corrected to 22



C.

A linear regression analysis was performed on both sets of data to

determine slope and intercept values. For our NMyB

4

CyTS720 data the

®t produced 1=…d

2

f

† ˆ 0:004 ‡ 0:008 [wt% additive], and for Engelke's

data 1=…d

2

f

† ˆ 0:003 ‡ 0:0009 [wt% additive].

4 D. I. Idar, B. W. Asay, and E. N. Ferm

Propellants, Explosives, Pyrotechnics 24, 1±6 (1999)

assumed to be the same as the jet velocity determined from

the radiographic measurements. Subsequent measurements

made at much later times and longer distances in a different

con®guration have demonstrated the jet was actually still

accelerating slightly. The differences in the velocities

ranged from 5% at the high velocities to 10% at the low

velocities. The majority of our data was taken at the higher

velocity range with the smaller amount of error, and at

shorter distances. We therefore assumed that the impact

velocity at the distance in our experiments correlated well

with our previous radiographic measurements. The velocities

obtained from radiographs were used to determine the

experimental virtual origin positions and times which are

given in Table 5. A plot of the velocities as a function of

attenuator thickness is shown in Figure 3 and the data are

given in Table 6 with the corresponding jet diameters. A

closer examination of these data reveals the consistency and

reproducibility of the attenuated Vipers.

Preliminary estimates of the jet diameter were determined

with scale measurements of the radiographic images. These

values, in the range from 1.6 mm±2.4 mm, were only used

for the initial data analysis but are not presented here. We

estimate the largest possible error in this method to be

 0.5 mm (  20±30%). Re®ned measurements were

obtained by analysis of data measured with a Leeds and

Northrop scanning microdensitometer. Scanned jet diam-

eters were obtained at four different locations on the radio-

graphic image: at the widest part of the tip, 5 mm, 15 mm,

and 25 mm back from the tip along the shank. Three scans

were performed at each of these positions, and a mean and

standard deviation value were determined for each. Dia-

meters were determined using a method described by

McAfee

(14)

to determine the effective spot size of the X-ray

source using a calibrated standard and the following equa-

tion:

d

j

ˆ ‰I ÿ …m ÿ I†sŠ=m

…4†

where d

j

is the jet or calibration rod diameter, I is the image

width at the base line, m is the magni®cation factor, and s is

the spot size of the X-ray source.

The scanned images of the calibrated rod did not yield a

constant spot size, especially in the 450 keV calibration, but

had a gentle trend to larger spot sizes for smaller rod dia-

meters. We ®t a linear relation of the image width to the

diameters of the calibration rod for each X-ray setup used.

This method improved the accuracy of the diameter mea-

surement to within  8% with occasional deviations reach-

ing 10% at the extreme edge of the calibration range of  1±

3 mm.

We used the shank diameter of the jet, not the tip

diameter, as the relevant parameter for our correlations.

The shank diameter was de®ned as the average of the three

values measured behind the tip. Because the radiographs are

obtained prior to impact and the jet is stretching in time, the

shank diameter measured with the radiograph is not repre-

sentative of the jet at impact. To estimate the diameter at the

time of impact, the assumption was made that the stretching-

jet segment conserves volume, and that it has the same

stretch rate as the unattenuated jet estimate given in Table 5.

The formula used for estimating the stretched diameter was

d

j

…t

i

† ˆ d

j

…t

x

†



…t

x

ÿ t

vo

†=…t

i

ÿ t

vo

†

p

…5†

Table 5. Calculated and Experimentally Determined Virtual Origin

Positions and Times for the Viper Shaped-Charge Jet at Different

Velocities

Tip

Velocity

Tail

Velocity

Calc. x

vo

(a)

(mm)

Calc. t

vo

(ms)

Expt. x

vo

(a)

(mm)

Expt. t

vo

(ms)

(mmyms) (mmyms)

9.17

9.17

0.0

21.81

9.17

7.17

ÿ85.4

14.35

7.17

5.17

ÿ39.2

20.80

5.17

3.17

ÿ21.3

24.25

3.17

1.17

ÿ3.1

30.00

9.17

7.64

ÿ105.9

11.59

7.64

2.45

ÿ29.2

21.63

(a)

Constant velocity segment tip at 0 mm and tail at ÿ17 mm.

Figure 3. Viper jet velocity as a function of mild steel attenuator plate

thickness.

Table 6. Jet velocities and Diameters Determined from Radiographic

and Timing Data, as a Function of Attenuation Thickness

Attenuation Velocity Diameter

Attenuation Velocity Diameter

(mm)

(mmyms) at Impact

(mm)

(mm)

(mmyms) at Impact

(mm)

6.35

7.54

2.02

57.15

6.02

2.31

6.35

7.85

2.02

57.15

6.05

2.03

25.4

7.45

2.12

57.15

6.08

2.62

38.1

6.91

2.22

63.5

5.99

2.04

44.45

6.47

2.70

63.5

5.89

2.31

50.8

6.20

2.29

82.55

5.04

2.31

50.8

6.31

2.01

88.9

4.78

2.52

50.8

6.39

2.25

88.9

4.66

2.00

(a)

57.15

5.82

2.28

88.9

4.72

2.00

(a)

57.15

6.10

1.69

95.25

4.64

2.19

57.15

6.05

2.34

107.95

4.43

2.61

57.15

6.07

2.24

114.3

4.22

2.00

(a)

57.15

6.06

2.02

133.35

4.13

2.62

57.15

6.01

2.04

133.35

3.89

2.19

57.15

6.09

2.29

158.75

3.46

2.42

57.15

6.06

2.36

184.15

3.10

2.71

57.15

6.02

2.43

184.15

2.91

2.00

57.15

6.03

2.32

209.55

2.81

1.95

57.15

6.13

2.36

234.95

2.49

2.60

(a)

Jet diameters were estimated at 2.00 mm for these covered tests.

Propellants, Explosives, Pyrotechnics 24, 1±6 (1999)

Improved Characterization of Nitromethane, Nitromethane Mixtures 5

where d

j

(t) is the jet diameter at time t, t

i

is the time of

impact, t

x

is the time of the X-ray, and t

vo

is the experimental

virtual origin time corresponding to the velocity of the

attenuated jet given in Table 5.

The accuracy of this extrapolation is a function of the

constant-volume approximation, i.e., it ignores thermal

expansion from plastic work and any void in the rod. It also

assumes that the attenuated jet is stretching at the same rate

as the unattenuated jet, and that the jet is not particulating.

For all of the uncovered test data, jet diameters were deter-

mined with the above method. All the data were used to

generate average values of jet diameters at de®ned jet velo-

cities. This information was used to estimate the jet dia-

meters for the covered test data.

5. Conclusions

Plate dent and temperature measurements have been

obtained to determine failure diameter ®ts for neat NM,

NMyDETA, and NMyB

4

CyTS720 mixtures as a function of

temperature andyor additive concentration. We have shown

that for equivalent wt% concentrations, DETA is more

effective than B

4

C at reducing the failure diameter of nitro-

methane. We have also demonstrated that the B

4

C additive

has a larger effect on the failure diameter of nitromethane

than the glass bead mixture used by Engelke

(12)

. This may be

a result of increased number densities of particles or a dif-

ference in particle impedances.

Radiographic jet diameter measurements were re®ned by

incorporating a microdensitometer scanning technique and

subsequent analysis

(14)

to evaluate the diameter. Diameter

errors were reduced from  20%±30% to a maximum of

 10% or less with this method. These higher resolution

measurements provided for further re®nement of the

threshold criterion.

The d

f

curve ®ts were subsequently used to determine

failure diameters for threshold experiments, and comparison

of homogeneous- and heterogeneous-sensitization. These

data were also crucial to providing a scale for comparison of

the data measured here with other explosives.

6. References

(1) D. J. Idar, B. W. Asay, and E. N. Ferm, ``Hypervelocity Jet Initiation

Threshold Criteria of Nitromethane and Nitromethane Mixtures'',

Propellants, Explosives, Pyrotechnics 24, 7±16 (1999).

(2) A. W. Campbell and R. Engelke, ``The Diameter Effect in High-

Density Heterogeneous Explosives'', Sixth International Sympo-

sium on Detonation, White Oak, MD, 24±27 August, 1976 (Of®ce

of Naval Research ± Department of Navy, Arlington, Virginia),

1976, [Proc.], pp. 642±652.

(3) W. P. Walters and J. A. Zukas, ``Fundamentals of Shaped Charges'',

John Wiley & Sons, Inc., New York (1989).

(4) R. Engelke, ``Effect of a Chemical Inhomogeneity on Steady-State

Detonation Velocity'', Phys. Fluids 23, 875 (1980).

(5) M. Kusakabe and S. Fujiwara, ``Effects of Liquid Diluents on

Detonation Propagation in Nitromethane'', Sixth International

Symposium on Detonation, Coronado, CA, 24±27 August, 1976

(Of®ce of Naval Research ± Department of the Navy, Arlington,

Virginia), 1976, [Proc.], pp. 133±142.

(6) R. Engelke and J. B. Bdzil, ``A Study of the Steady-State Reaction-

Zone Structure of a Homogeneous and Heterogeneous Explosives'',

Phys. Fluids 26, 1210 (1983).

(7) A. W. Campbell, M. E. Malin, and T. E. Holland, ``Temperature

Effects in Liquid Explosive, Nitromethane'', J. App. Phys. 27, 963

(1956).

(8) A. F. Belyaev and R. K. Kurbangalina, ``Effect of Initial Tem-

perature on Critical Diameter of Nitroglycerin and Trinitrotoluene'',

Russ. J. Phys. Chem. 34, 285±289 (1960).

(9) B. N. Dobratz, LLNL Explosives Handbook: Properties of Che-

mical Explosives and Explosive Simulants (Lawrence Livermore

National Laboratory, Livermore, CA, USA (1982).

(10) B. W. Asay, D. J. Pauley, and E. N. Ferm, ``Jet Initiation Thresh-

olds of Nitromethane'', Tenth International Symposium on

Detonation, Boston, MA, 12±16 July 1993 (Of®ce of Naval

Research, Arlington, Virginia), 1993, [Proc.], pp. 104±112.

(11) Personal Communication with R. Engelke, 1994.

(12) R. Engelke, ``Effect of the Number Density of Heterogeneities on

the Critical Diameter of Condensed Explosives'', Phys. Fluids 26,

2420 (1983).

(13) CRC Handbook of Chemistry and Physics; Vol., edited by R. C.

Weast (CRC Press, Inc., Boca Raton, Florida, 1981).

(14) J. M. McAfee, ``Accurate Diameter Measurement of Explosively

Formed Metal Jets and Effective X-ray Source Size'', 1989 Flash

Radiography Topical Symposium, Welches, OR, 1989, [Proc.],

pp. 249.

(Received June 19, 1998; Ms 16y98)

6 D. I. Idar, B. W. Asay, and E. N. Ferm

Propellants, Explosives, Pyrotechnics 24, 1±6 (1999)