ELECTROMAGNETIC PROPERTIES OF DETONATING EXPLOSIVES · ELECTROMAGNETIC PROPERTIES OF DETONATING EXPLOSIVES G. P. Chambers1, R. J. Lee1, T. J. Oxby2, W. F. Perger2, and A. B. Kunz2

ELECTROMAGNETIC PROPERTIES OF DETONATING EXPLOSIVES

G. P. Chambers1, R. J. Lee1, T. J. Oxby2, W. F. Perger2, and A. B. Kunz2

1Energetic Materials Research and Technology Department, NAVSEA Indian Head Division, 101 Strauss Ave, Indian Head MD 20640-5035

2Department of Electrical Engineering, Michigan Technological University, Houghton, MI 49331

Current theories of reaction processes suggest that changes in electronic band structure and radiation producing dipole oscillations occur during shock loading of an energetic crystal prior to detonation. To test these theories, a broadband antenna, capable of measuring polarization, was employed to observe shock-induced electromagnetic radiation from a crystalline explosive, RDX. The frequency spectra from these experiments were analyzed using time/frequency Fourier methods. Changes in conductivity resulting from this shock loading were also measured at the opposite end of the crystal from the shock source. A four-point-probe arrangement was used to eliminate errors involving lead resistance. This arrangement uses two leads and a fast discharge circuit to pass current through the crystal interface at the time conductivity begins to change in conjunction with the arrival of the shock wave. Also reported are corresponding light (observed with a high-speed electronic camera) and sub-microwave emission observed during the passing of the shock wave in the RDX crystal prior to detonation.

INTRODUCTION

There exists motivation from both theoretical considerations and previous experiments to pursue both the measurement of electromagnetic fields radiated from and the conductance of energetic materials under shock. Recently, work by B. Kunz’s group 1 has indicated that shock energy is converted to lattice vibrations, which in turn give rise to molecular vibrations, electronic excitations and collective excitations. As the electronic energy gap of an energetic crystal destabilizes during shock induced deformation, metallization and reaction onset occur, a model advocated by J. Gilman2. We have therefore devised a series of experiments to measure the frequency spectrum

through the radio-frequency band of electromagnetic emissions, as well as the conductance of energetic materials near the onset of shock-induced reaction.

EXPERIMENTAL ARRANGEMENT

The experimental arrangement consisted of an RDX crystal (cube, nominally 9-mm on each edge) sandwiched in a Teflon holder with a 10-mm wide groove down the middle to allow back-lit photography of the crystal. A 19-mm diameter Pentolite pellet initiated by an RP-80 detonator provided a planar shockwave into the crystal through a variable PMMA gap. CTH3 simulations

were used to confirm the planarity of the shockwave with respect to the front face of the crystal. Four conducting pins were placed in a line against the crystal opposite from the donor charge to record voltage and current in a four-point probe arrangement. Each pin was 0.75-mm in diameter and set 1.5-mm between centers. The total spacing was 5.33-mm. The crystal rested on a nylon disk from which the pins protruded to make contact with the crystal surface. Silver paint was used to provide an ohmic contact between the pins and the crystal. The circuit is similar to that used by Weir et al.4, consisting of a 100-nF capacitor charged by 205-Volt battery. The capacitor was connected across the two outside-pins to deliver current when the crystal interface began to conduct. No current flowed through the circuit until the RDX sample changed from an insulator to a conductor upon arrival of the shock wave. The capacitor was placed just behind the nylon disk to minimize inductance and thereby insure a fast rise time for the current. Voltage was measured across the two inner pins. Both measurements were made using a LeCroy oscilloscope at 500 ps/point. The circuit was calibrated using thin wafers of copper (a good conductor) and n-doped Silicon (a typical semiconductor). RF radiation was measured with two bipolar antennae positioned nominally 19 mm from the central axis of the crystal. They were mounted on a single substrate and aligned vertically and horizontally with respect to the shock propagation. Signals were recorded using separate LeCroy oscilloscopes at 125 ps/point. For shots involving the detection of RF radiation, water was used to surround the sample and thereby prevent signals from air ionization. The detonator was fired into both Teflon and Lexan to determine the sensitivity of the antenna to noise during its operation alone. While the antennae were found to be sensitive to the currents generated by the firing pulser, signals did not appear in the window between onset of the shockwave into the crystal (~7.5 µs) and the initiation of reaction at the crystal/nylon interface (>10.5 µs). Other control shots were fired with the detonator and pentolite donor into water, to

determine the radio pickup onto the antennae as well.

RESULTS 1

2

3

4

5

6

7

8

FIGURE 1. High-speed photograph of shocked RDX crystal exhibiting shock-induced luminescence. Fig 1 shows a high-speed camera record of a shock wave from a pentolite pellet entering an RDX crystal. Frame 1 starts at 7.5 µs after firing of the detonator pulse. The interframe time is 400 ns, with a duration of 40 ns per frame. Light emission can be seen at 8.7 (frame 4) and again at 9.5 µs (frame 6). The light emission appears to be the result of passage of the shockwave through the sample and is occurring prior to detonation. By 8.7 µs, light from the pentolite detonation has died out, so reflection of this light through the crystal is not the source of the light emission. Concurrent with light emission, we observed RF signals on the antenna temporally correlated with the onset of light emission. The raw voltage data from this measurement can be seen in Fig. 2. Signals at 8.7 and 9.6 µs are temporally correlated with the light emission observed in Fig. 1. Furthermore, the strength of the signals tends to correlate with the amount of light emission occurring at the same time. The RF signal at 8.7 µs appears weaker than the signal at 9.6 µs, consistent with the degree of light emission in the high speed photograph, which appears stronger at the later time, 9.5 µs. This agreement tends to support the view that the signals are correlated to some physical phenomena.

-0.04

-0.03

-0.02

-0.01

0

0.01

0.02

0.03

0.04

7.5 8 8.5 9 9.5 10 10.5 11

volta

ge

time (µs) FIGURE 3. Voltage trace observed directly on the antennae as a function of time.

Fig. 3 shows spectrally analyzed signals from the RF, generated through time/frequency Fourier analysis of the antennae signals. The signal at 9.6-µs spectrally resolves into a series of discrete frequencies at 0.5, 0.43, and 0.37-GHz. At 8.7-µs, lines appear at 0.1 and 0.25-GHz. These frequencies are in the sub-microwave regime, and hence are not likely due to molecular rotational spectra which occur at microwave frequencies, in the tens or hundreds of GHz range. A conductivity measurement is shown in Fig. 4 using the four-point probe test. This figure shows the change in conductivity as a function of time, with time zero marked from the discharge of the detonator circuit. The resistance (an indication of changing conductivity) was obtained by taking the ratio of the voltage and current signals from the four-point probe. This figure shows the RDX crystal begins to become conductive at about 8.4 µs as a result of a planar 120-kbar incident shock. At this point, the four-point-probe’s capacitor begins to discharge causing a current to flow through the sample. Resistance begins to stabilize at a plateau of ~100 ohms, beginning at about 8.8 µs. Given the inner probe electrode spacing of 2.5-mm, it may be assumed that resisitivity was on the order of ~400 ohm/cm2. This value is well within the semiconductor regime. No reaction was observed prior to 9-µs in the camera record of this test.

(a)

(b)

FIGURE 4. Comparison of spectrally analyzed antennae signals occurring at 8.7 µs (a) and at 9.6 µs (b).

-100

-50

0

50

100

150

200

250

300

8.2 8.3 8.4 8.5 8.6 8.7 8.8 8.9 9

Res

ista

nce

(ohm

s)

TIme (µs) FIGURE 5. Change in RDX resistivity as a function of time for an incident shock wave.

DISCUSSION In previous work by Dick5 , luminescent emissions from a shocked PETN crystal were interpreted as being due to shock induced decomposition products NO2 and possibly NO6. Further, these authors noted that such emissions may be related to the explosive sensitivity of the crystal. Visible light is known to be produced by electronic transitions, however, without spectroscopy, we cannot distinguish between the possibility that it is correlated with excited states of shock decomposition products or due to radiative emissions from the RDX molecule as a whole. One possible explanation for the source of RF emission is excitation of plasma oscillations within the crystal. Such plasma modes do not exist in insulators, but can be excited in metals or semiconductors. Since the shock wave transforms RDX into a semiconductor during its passage through the crystal, it is possible to excite such plasmon oscillations within the crystal. These oscillations may be the source of the RF emissions detected with the antennae. The correlation of RF emissions with luminescent emissions, and the correlation of the latter with sensitivity, suggests a relationship between the RF field and reaction initiation in RDX. This mechanism may need to be considered in MD simulations if an accurate prediction of sensitivity is to be obtained. Lastly, conductivity changes have been observed as a result of the passage of the shockwave through the crystal, prior to onset of reaction. This was predicted by Gilman2. This represents a ten order of magnitude change in conductivity of a molecular RDX crystal as a result of the incident shockwave. Such changes have previously been measured in the ionic crystal KCl at kbar shock pressures. These changes ultimately push the organic crystal towards metallization. A metallized organic crystal therefore has electrons available to participate in intermolecular reactions, which may either facilitate or be a necessary condition for detonation chemistry to occur. Thus, our results may provide support for both models of reaction onset. The detection of RF

emissions from the crystal during shock loading directly supports theoretical work by Kunz et al1 while conductivity changes towards metallization support models of Gilman for intermolecular processes.

SUMMARY AND CONCLUSIONS

Shock-induced light emission from an RDX crystal prior to detonation has been observed. These emissions corresponded with shock-induced electromagnetic radiation. A change in crystal conductivity as a function of time was also observed prior to and during the detonation process. Future parametric studies, using the techniques developed here, will help broaden our understanding of the roles these phenomena play in initiation.

ACKNOWLEDGMENTS

Discussions with R. Doherty and funding from the IHD Internal Research Program are gratefully acknowledged. REFERENCES 1 Zwitter, D. E., Kuklja, M. M., and Kunz, A. B., Shock Compression in Condensed Matter, Snowbird Utah, June 1999 2 J. Gilman, Shock Compression in Condensed Matter, Amherst MA, July 1997 3 R.L. Bell et al. “CTH User’s Manual and Input Instructions” version 2.00 Sandia National Laboratories (Sept 1995) 4 W.J. Nellis, S.T. Weir, C Mitchell Physical Review B, Vol. 59. P. 3436 5 J.J. Dick, R.N. Mulford, W.J. Spencer, D.R. Pettit, E. Garcia, D.C. Shaw, J Appl. Phys. 70 p. 3572 (1991)