LA-8647

z

*

1

1

Jet Initiation of Explosives

—..——

DO NOT CIRCULATE\I

PERMANENT RETENTION Ii

REQUIREDBY CONTRACT ;

L%? LOS ALAMOS SCIENTIFIC LABORATORYPost Office Box 1663 Los Alamos, New Mexico 87545

An Affirmative Action/Equal Opportunity Employer

Edited by Sharon L. Crane

Photocomposition by Alice Creek

DISCLAIMER

TM report vms prcptied as an uwunt of work sponsoced by an agency of the Umtcd Swtcs Governmcnl. Nelthcr the United Stxtes Government nor any a~cncy thereof, nor any of their cnlployccs,m~kes my vmranty, cmpres$or implied, or assumes any Icfzd kbihty or rcsponubitity for !he xtiL’ur-ac’y, completeness, or usefulness of any information, apparatus, product, or Process discioscd, or ICP-resenl$ that its use would not inftingc privalcly owned rights. Rcfercnc’e herein to any SIWCCICtom.mcccial product, process, or scrvicc by lrade name, trademark, manufxturer, or otherwise. dots not

necessarily constitute or imply its endorsement, rccommcndatkan, or favoring by the United StatesGovenrmcnt or any agency thcrmf. The views and opinions of authors e~prcssed hetein do not nec-essarily state or reflect those of the United States Government or any agency lhermf.

UNITED STATES

DEPARTMENT OF ENERGY

CONTRACT W-7405 -ENG. 36

Jet Initiation of Explosives

.

.

.- -.

LA-8647

UC-45Issued: February 1981

Charles L. MaderGeorge H. Pimbley

ABOUT THIS REPORTThis official electronic version was created by scanning the best available paper or microfiche copy of the original report at a 300 dpi resolution. Original color illustrations appear as black and white images.

For additional information or comments, contact: Library Without Walls Project Los Alamos National Laboratory Research LibraryLos Alamos, NM 87544 Phone: (505)667-4448 E-mail: lwwp@lanl.gov

●✎

●

.

JET INITIATION OF EXPLOSIVES

by

Charles L. Mader and George H. Pimbley

ABSTRACT

The initiation of propagating detonation in PBX 9502 and PBX 9404 by jets ofcopper, aluminum, and water is numerically modeled using the two-dimensionalEulerian hydrodynamic code 2DE with the shock initiation of heterogeneous explosiveburn model called Forest Fire.

The Held experimental critical condition for propagating detonation is the jet velocitysquared times the jet diameter. In PBX 9502, the jets initiate an overdrhen detonationsmaller than the critical diameter, which either fails or enlarges to greater than thecritical diameter while the overdriven detonation decays to the C-J state.

In PBX 9404 the jet initiates a detonation that propagates only if it is maintained bythe jet for an interval sutlicient to establish a stable curved detonation front.

L INTRODUCTION

The process of shock initiation of heterogeneousexplosives has been modeledl for many shock sensitivitytests using the Forest Fire rate model and the reactivehydrodynamic codes SIN, 2DL, and 2DE. The basicmechanism of heterogeneous explosive shock initiation isshock interaction at density discontinuities producinglocal hot spots that decompose and add their energy tothe flow. Forest Fire models the gross features of thisflow. Calculations have modeled the effect of drivers orprojectiles producing effectively one-dimensional shockwaves in explosives with pressure less than the C-Jdetonation pressure and with varied durations.* Sympa-thetic detonation, the detonation of nearby explosiveobjects by the blast from a primary explosion, has beennumerically modeled using the Eulerian reactivehydrodynamic code 2DE with Forest Fire burn rates.2The gap sensitivity test, where the shock from a standarddonor explosive is transmitted to a test explosive throughan inert barrier (gap), has been numerically modeled forboth Los Alamos National Laboratory and Naval

Ordnance Laboratory tests.3 The minimum primingcharge test, where the least mass of booster explosive isdetermined that will induce propagating detonation in atest charge, has been modeled by Forest’ using the SINcode with Forest Fire. The PBX 9501 initiation by steelcylinders and spheres has been studied by Cort and Fu,susing the 2DE code with Forest Fire. These shockinitiation experiments depend upon the magnitude andduration of the initiating shock wave being sufllcient forbuildup to detonation. When a detonation occurs, itpropagates throughout the explosive if the chargediameters are larger than the failure diameter.

The failure diameter is the smallest explosive chargediameter that can maintain a propagating C-J detona-tion. The failure diameter is smaller if the charge isconfined. The failure diameter of unconfined PBX 9502is 0.9 cm and of PBX 9404 is 0.12 cm. This largedifference is caused by the shock initiation properties ofthe explosives and can be modeled using the Forest Fireheterogeneous shock initiation burn model as describedin Ref. 1. Corner turning in a propagating detonation isalso a function of the shock initiation properties, as

1

shown in Ref. 1 for detonations propagating around airand metal corners and in Ref. 6 for detonations propa-gating from spherical initiators.

Heterogeneous explosive desensitization with a shockwhose pressure is too low and too short in duration tocause propagating detonation can cause failure of apropagating detonation in unshocked explosive to con-tinue propagating when the detonation front arrives atthe previously shocked explosive. This process has beennumerically modeled in Ref. 7. The Forest Fire burnrate is determined primarily by the initial shock pressurepassing through the explosive.

We conclude that the heterogeneous shock initiationprocess dominates not only the buildup to detonation,but also how the detonation propagates when side or rearrarefactions are affecting the flow.

The type of sympathetic detonation that we investigatein this report is the explosive initiation by a nearbyshaped charge with a metal cone that forms small-diameter, high-velocity projectiles.

The process was investigated first by Held.8 He used ashaped charge to drive a 1.5-mm-thick, 60° copper cone.The resulting copper jet was shot into steel plates ofvaryi~g thickness to obtain different exit jet velocities.The jet sizes and velocities were measured using velocityscreens and x-ray flash photography, The jets werepermitted to interact with an explosive, and the criticalvelocity was determined for initiation of propagatingdetonation by various diameter jets.

Held observed that his data could be correlated byassuming that the critical jet value for explosive initiationwas the jet velocity squared (Vz) times the jet diameter(d). If the velocity is in millimeters per microsecond andthe jet diameter is in millimeters, he reported that thecritical V2d was about 5.8 for copper jets initiating 60/40RDX/TNT at 1.70 g/cm3.

Similar experiments have been performed by A. W.Campbell at the Los Alamos National Laboratorywith copper jets initiating PBX 9404 (94/3/3HMX/nitrocellulose/tris-~-chloroethyl phosphate at1.844 g/cm3) and PBX 9502 (95/5 TATB/Kel-F at1.894 g/cm3), and his data is summarized in Table I.Critical V2d values are 127 + 5 mm3/~s2 for PBX9502, 16 * 2 mm3/~s2 for PBX 9404, 37 mm3/us2 for75/25 Cyclotol, and 29 mm3/~s2 for Composition B-3.

If the pressure-particle velocity is matched, as shownin Fig. 1 for copper at 5.0 mm/~s shocking inert PBX9502, the pressure match is 680 kbar; at 6.0 mrn/~s thepressure match is 920 kbar. Since the C-J pressure ofPBX 9502 is 285 kbar, the copper jet must initiate a

TABLE I

CAMPBELL EXPERIMENTAL DATA

Copper Jet Tip JetVelocity Diameter Vzd

(mrr2/~s) (mm) Result (mm3/~s2)

PBX 95024.594.474.654.524.945.565.555.815.93

5.8s5.965.845.924.153.953.953.803.85

Critical 127 * 5Detonation 124Fail 119Detonation 126Fail 121Fail 101

Fail 122

Fail 122

Detonation 128

Detonation 135

PBX 9404 Critical 16+2

2.87 2.04 Detonation 16.8

2.51 2.25 Fail 14,2

2.76 2.11 Fail 16.0

3.07 1.95 Detonation 18.4

3.07 1.95 Detonation 18.4

.*

-.

strongly overdriven detonation. However, the jetdiameter (-4 mm) is less than half the failure diameter(9.0 mm) of unconfined PBX 9502.

Also shown in Fig. 1, the pressure match. is 340 kbarfor PBX 9404 shocked initially by copper at 3.O mm!psand is 250 kbar at 2.5 mrn/Ls. The run distance forPBX 9404 at 250 kbar is about 0.5 mm so detonationwould occur very quickly. The jet diameter (-2 mm) islarger than the unconfined PBX 9404 failure diameter(1.2 mm).

We shall examine the dominant mechanisms of jetinitiation of PBX 9404 and PBX 9502 using the ForestFire burn rate and the 2DE code.

11. NUMERICAL MODELING●

The two-dimensional reactive Eulerian hydrodynamic .code 2DEl was used to describe the reactive fluiddynamics. The Forest Fire description of heterogeneous “shock initiation was used to describe the explosive burn.The HOM equation of state and Forest Fire rateconstants for PBX 9502 and PBX 9404 were identical to

. .

.“

.

1.6—

T 1~“:~

w ‘“2-Ea(n I,o—wmn

0.8 –

0.6 —

0.4 –

0.2 –

o0 1 2 3 4 5 6 7

PARTICLE VELOCITY (mm/p )

Fig. 1.The pressure-particlevelocityHugoniotsfor copper, PBX9502, and PBX 9404 with the reflected shock states forcopper with initial free-surface velocities of 6.0, 5.0, 3.0,and 2.5 mm/ps.

those described in Ref. 1. The Pop plots are shown inFig. 2, and the Forest Fire rates are shown in Fig. 3.

The jet was described as a cylinder of radius 10 or20 cells and 40 cells long with the appropriate initialvelocity. The cell size was 0.02, 0.01, or 0.005 cm,depending upon the jet radius. The BPCJ term in ForestFire was set at 1.5 to permit overdriven detonationa andthe rate was limited to ezo.The viscosity coefficient usedfor PBX 9502 was 0.25 and for PBX 9404 was 0.75.SutTcient viscosity to result in a resolved bum wasnecessary just as in the detonation failure calculationsdiscussed in Refs. 1, 6, and 7.

An uncertainty in the calculations is the approxima-tion of the jet as a cylinder of uniform velocity collidingend-on with the explosive. Although the jet is actuallymany small metal pieces, our calculations indicate thatthe critical conditions are determined by the first piece ofabout the same length as its radius. Side rarefactionsdominate the flow after the reflected shock wave travelsone diameter length back into the jet.

I 1 t 1 I I I I , I I I t

..-.10 100 1000

Pressure (kbar)

Fig. 2.The run to detonation distance as a function of the shockpressure,

One also needs to consider the proper initial condi-tions for the jet. Since it was formed by being shockedand then raretied, it will have a residual temperaturegreater than ambient and density less than the initialdensity of the jet material.

Estimates of the copper residual state can be made byassuming the material is at the same state as if it hadachieved its final velocity by a single shock and then hadrarefied to one atmosphere. The residual temperature ofcopper initially shocked to 830 kbar and then raretied toa free-surface velocity of 3.0 mm/~s is 768 K; theresidual density is 8.688 g/cm3, comparable with theinitial density of 8.903. Calculations were performedwith copper initial conditions of 8.903 g/cm3 and 300 Kand of 8.688 g/cm3 and 768 K. Since the changedequation of state results in only slightly changed ex-plosive shock pressure, the calculated results were in-

sensitive to the copper jet initial conditions.Another uncefiainty is how accurately the Forest Fire

burn rate can be extrapolated to overdriven detonations.

3

,-10 100 1000

Pressure ( kbar)

Fig. 3.The Forest Fire decompositionrates as a function of shockpressure.

Since the overdrive decays rapidly and the side rarefac-tions quickly reduce the failure region pressure into thepressure range where the Pop plot was determined, theaccuracy of the extrapolated bum rates is not importantto the determination of the critical conditions for estab-lishing propagating detonations.

The calculated PBX 9502 results are presented inTable II. The effects of jet composition, diameter, andvelocity were examined. The calculated profdes for a4-mm-diam copper jet with a 7,0-mm/#s initial velocityare shown in Fig. 4 and a 5.O-mm./~ velocity in Fig. 5.

The copper jet initiates an overdriven detonationsmaller than the critical diameter, which enlarges togreater than the critical diameter of self-conihed PBX9502 when shocked by a 7.O-mm/~s copper jet. Whenthe PBX 9502 is shocked by a 5.O-mm/~s copper jet, thedetonation is decayed by side and rear rare factionsbefore it enlarges to the critical diameter.

As shown in Table II, the numerical results bound theCampbeii experimental critical V2d of 127 for copper

jets. The shock pressure sent into the explosive dependsupon the jet material, so the aluminum jet must have agreater velocity or diameter than the copper jet to furnish

an equivalent impulse to the PBX 9502. One thenexpects that the critical V2d is larger for aluminum thanfor copper and larger still for a water jet. Theseconclusions are confirmed by the calculations with the

predicted V2d for aluminum being 325 + 75 and forwater about 800.

The calculated PBX 9404 results are presented inTable HI. The effects of jet composition, diameter, andvelocity were examined. The calculated profiles for a2-mm-diam copper jet with a 3.O-mm/~s initial velocityare shown in Fig. 6, and a velocity of 2.0 mm/lus inFig. 7. The copper jet initiates detonation, which propa-gates only if it is maintained by the jet long enough toestablish a stable curved detonation front.

As shown in Table 111,the numerical results boundthe Campbeii experimental critical V2d of 16 for copperjets initiating propagating detonation in PBX 9404. Thecalculated results suggest a lower value for V2d but oneas accurate as the data and the approximations of thecalculations. Again, the calculated, critical V2d is largerfor aluminum jets (70 + 20) than copper jets and largerstii.ifor water jets (150 * 50)0

TABLE II

PBX 9502 CALCULATIONS

Jet JetJet Diameter Velocity

Material [mm) (mdus) Result V2d

CopperCopperCopperCopper

CopperCopper

AluminumAluminumAluminum

Aluminum

WaterWater

Water

4.0 4.04.0 S.o4.0 6.04.0 7.0

8.0 4.08.0 5.0

4.0 7.04.0 8.04.0 10.0

8.0 7.0

4.0 6.04.0 10.04.0 14.0

-,

Faiied 64Failed 100Marginal 144Propagated 196

Marginal 128Propagated 200

Failed 196Failed 256Propagated 400

Propagated 392

Failed 144Failed 400Marginal 784

●

.

4

r

III. CONCI

PBX 9502

2.0- mmRadiuscopperCylinder

ISOBARS

0.34, 1 I

PBX 9502

“!

0.326 0.642 0.956 1.274 I .590

DISTANCE (cm)

Fig. 4.The initial profde and the pressure and mass fraction profdes for a 4-mmdhun copper jet of7.0 mm/Ms shocking PBX 9502 at 0.75 ps. The isobar contour interval is 20 kbar and the massfraction contour is 0.05. The pressure profile along the axis is also shown.

.USIONS wave was of suf%cient strength and duration to build upto detonation. The propagating detonation was assured

The initiation of propagating detonation in PBX 9502 by the large geometry. In near-critical jet initiation,

and PBX 9404 by jets of copper, aluminum, and water however, a prompt detonation of the explosive results,

has been numerically modeled using the two-dimensional which builds to a propagating detonation only if the

Eulerian hydrodynamic code 2DE with Forest Fire. shock wave produced by the jet is of sutllcient magnitude

Shock initiation by jets near the critical V2d contrasts and duration. Therefore, jets may produce significant

with other shock initiation studies. In the latter, if amounts of energy from an explosive even below the

detonation occurred, it was because the initiating shock critical value of V2d required for propagating detonation.

5

I,ISOBARS

,lMi 9.cloook o: M1.uostt

...:

I 5.0 nmlps

MF

/..

0326 Q642 0.958 I 274 I 590DISTANCE (cm)

Fig. 5.The pressure and mass fraction proffles for a 4-mmdiamcopper jet of 5.0 mnd~s shocking PBX 9502 at 0.90 vs.The isobar contour interval is 20 kbar and the massfraction contour is 0.05. The pressure profde along the axisis also shown.

TABLE III

PBX 9404 CALCULATIONS

Jet JetJet Diameter Velocity

Material (mm) (mm/~s) Result v%

CopperCopperCopper

CopperCopper

Copper

AluminumAluminum

AluminumAluminumAluminum

haterWaterWater

2.0 2.02.0 2.52.0 3.0

4.0 2,04.0 2.5

1.5 3.0

2.0 3.02.0 6.0

4.0 3.04.0 4.04.0 6.0

4.0 3.04.0 6.04.0 7,0

FailedMarginalPropagated

PropagatedPropagated

Marginal

FailedPropagated

FailedMarginalPropagated

FailedMarginal

812.518

1625

13.5

1872

3664

144

‘ 36144

Propagated 196

IV. ACKNOWLEDGMENTS

The authors gratefully acknowledge the contributions,encouragement, and data of Arthur W. Campbell, thesupport of Larry Hantel, and the contributions ofCharles Forest, Allen Bowman, and James Kershner.We also acknowledge the helpful discussions withDr. M. Held of MBB-Schrobenhausen.

b

.

.

PBX 9404

I.O-mm Radius

Coppor

Cylinder I “.“.“.“.

I ..I .:.[ISOBARS

IDS Ilk: .?.0000f-01 MI CROSECO?+OS

I PBX 9404 I

I 3.0 mmlps IMF

0.34 -L

:z 0.23 -

wE

z 0.12 -wwun 0.01

-().l(j~0.1605 0.2395 0.3185 0.3975

DISTANCE (cm )

Fig. 6.The initial profde and pressure and mass fraction profiles for a 2-mm-diam copper jet of 3.0 mn@shocking PBX 9404 at 0.20 KS.The isobar contour intervals is 50 kbar and the mass fraction contouris 0.05. The pressure profde along the axis is also shown.

\

7

IIi!: SOOM 01 MICROSECONDS.. .

—-. ,

.“.

:

:

. . . . . . . . . . . . . . . . . .

ISOBARS

0s

I 2.0 mmIpsMF

‘.o.,o~0.0025 0.0815 0.1605

DISTANCE

0.2395 0.3185 0.3975

(cm)

Fig. 7.

The pressure and mass lkaction profdes for a 2-mm-diam copper jet of 2.0 mm/IM shoc~ng pBx 9404 * 0s25 W.The isobar contour interval is 50 kbar and the mass fraction contour is 0.05, The pressure profile along the axis isalso shown.

.b

-.

.

8

4

I,*

REFERENCES

1. Charles L. Mader, Numerical Modeling of Detona-tions (University of California Press, Berkeley, 1979).

2. Allen L. Bowman, James D. Kershner, and CharlesL. Mader, “Numerical Modeling of SympatheticDetonation,” Los Alamos National Laboratory re-port LA-7989 (1979).

3. Allen L. Bowman, James D. Kershner, and CharlesL. Mader, “A Numerical Model of Gap Test:’ Los

Alamos National Laboratory report LA-8408 (1980).

4. Charles A. Forest, “Numerical Modeling of theMinimum Priming Charge Test:’ Los Alamos Na-tional Laboratory report LA-8075 (1980).

5. G. E. Cort and J. H. M. Fu, “Numerical Calculationof Shock-Induced Initiation of D~tonations,”JANNAF 1980 Propulsion Systems Hazards Meet-ings, Monterey, California, October 27-31, 1980.

6. Charles L. Mader, “Numerical Modeling of In-sensitive High Explosive Initiators,” Los AlamosNational Laboratory report LA-8437-MS (1980).

7. Charles L. Mader and Richard Dick, “ExplosiveDesensitization by Preshocking,” in Combustion andDetonation Processes, International Jahrestagung,June 27-29, 1979 (Karlsruhe, BundesrepublikDeutschland, 1979), pp. 569-579.

8. M. Held, “Initiating of Explosives, a Multiple Prob-lem of the Physics of Detonation,” Explosivstoffe 5,98-113 (1968).

.

9

,

*

.

.Domest !c NTIS

Page Range Price Price code

0014325 s s .00 A020264s0 6.00 A03

051475 7.00 A04

076.100 8.00 A05101.125 9.00 A06126.150 10.00 A07

Printed in the United Slates of Ameriu

Available fromNational Technical Information Service

US Department OFCommerce

S28S Port Royal RoadSpringfield, VA 22161

Microfiche S3.50 (AO1)

C.amestic NTIS Domestic NTISPage Range Price Price Code Page Range Price Price Code

1s1.175 $11.00 A08 301-32S S17.00 A14176-200 12.00 A09 326.3S0 18.00 AIS201-22s 13.00 AlO 3s1-37s 19.00 A16226.250 14.00 All 376400 20.00 A172S1-27S 15.00 A12 40142S 21.00 A18276.300 16.00 A13 426450 22.00 A19

Domestic NTIsPage Range Price Price Code

4s147s S23.00 A20476-500 24.00 A21S01.52S 2s.00 A22S26.S50 26.oo A23

S51-S7S 27.00 A24576-600 28.00 A2S601.uP t A99

tAdd S1.00 for each additional 25.page increment or portion thereof from 601 pages up,

--.,

,’

.

c1,. J

dL-;—

.

h