Explosive pulsed power for driving HPM loads.
The Phillips Laboratory compact explosive pulsed power program is tasked with providing a pulsed power system to power a 500-kV, high-impedance load with a rise time of 100 ns and a pulse width of 1 sec. The general approach taken for this system consists of a compact helical generator driving a pulse-conditioning system. The pulse-conditioning system may be composed of a switched-primary pulse transformer, a direct acting opening switch (either an exploding wire fuse or an explosively formed fuse) or either of the above with a nonlinear resistor to achieve the required pulse shape. Additionally, a peaking capacitor may be required ahead of the load to achieve the required rise time.
These various pulse-conditioning schemes are presently being investigated at the Phillips Laboratory, both computationally and experimentally. In all of the schemes, some form of opening switch will be required, likely a fuse, and experiments are underway to investigate the properties of aluminum and copper fuses. In addition, a full-scale experimental system to drive the pulse-conditioning systems in the laboratory (i.e., without the need for an explosive generator drive) is being constructed.
Explosive pulsed power experiments began at the Phillips Laboratory in the summer of 1995 with joint pulsed magnetohydrodynamic experiments with Sandia National Labs. These experiments spurred the development of an explosive experiment infrastructure at the Phillips Laboratory High Energy Research and Technology Facility (HERTF). The explosive experiments to be described in this paper were begun at the HERTF approximately one year ago. These experiments centered around the use of a simple helical explosive generator to drive further development of explosive experiment infrastructure, develop diagnostic techniques and provide benchmarking data for design and analysis codes.
The Phillips Laboratory HERTF is a large laboratory complex, located in a remote box canyon, which is designed to accommodate explosive experiments of up to 1,000 lb of HE, as well as conventional high-energy pulsed power and HPM experiments. The explosive pulsed power experimental facility within the HERTF includes a 200-kJ, 10-kV capacitor bank for providing seed flux for FCGs. The output of this bank is fed through 50 low-inductance coaxial cables to a detonator-driven switch located near the explosive pad. The current is then fed to the explosive containment structure on the pad. Capacitor discharge units (x-units) for driving detonators are located in concrete culverts near the pad. Diagnostic coaxial and fiber-optic cables terminate in a small diagnostic shed on the pad. Another shed located on the pad houses a Beckman and Whitely 189 high-speed framing camera, which is used for optical diagnosis of explosive events. Electrical and optical signals are fed to a screen room located in the HERTF. This screen room houses the capacitor bank and x-unit controllers, 28 digitizer channels, timing and delay systems and a computer control and data acquisition system. A streak camera records signals from a crushed fiber-optic diagnostic system, which is used to provide a temporal record of the armature impact with the generator stator (this diagnostic will be described in detail below).
The first explosive experiments performed in support of the HPM compact explosive pulsed power program involved the detonation of armatures, which compress the seed magnetic flux within the FCG. These armatures consisted of 6.4-cm i.d. 6061 seamless aluminum tubes approximately 45-cm long. Armatures of varying wall thickness were fired in order to determine the minimum wall thickness consistent with the desired armature performance (factor of two expansion with negligible armature breakup and jetting). After machining, these armatures were annealed to a T3 hardness and were loaded with approximately 7 lb of C4 high explosive. The principle diagnostic employed in these experiments was the high-speed framing camera described above. So called argon candles were used to provide illumination of the armature during expansion. This device consists of a plywood box with a Lexan[R] window, filled with atmospheric-pressure argon and lined with detasheet explosive. The explosive detonation shock ionizes the argon fill, producing a brilliant light pulse.
In addition to the high-speed photography, electrical impact pins and crushed fiber diagnostics were used to provide temporal information about the armature arrival at the factor-of-two expansion point. The electrical impact pins were mounted in a steel anvil below the armature, with the ends of the pins located radially at the desired expansion points. The pins consist of two coaxial electrodes biased to a potential of approximately 300 V. Upon impact, the armature shorts the pin, which results in a pulse being recorded via a multiplexer/biasing circuit. While the impact pins provided reasonable values for both the explosive detonation velocity (i.e., the axial velocity of the self-consistent armature expansion) and the radial expansion velocity (estimated from the Gurney velocity), there is some evidence that the absolute timing of the impact pins is early relative to the other diagnostics. We believe that this may be due to premature shorting of the pins by the air shock wave in front of the armature outer surface.
The crushed fiber-optic diagnostic utilizes the black-body radiation produced by optical fibers under extreme impact pressure. The fibers are arranged parallel to the armature, such that the light produced by the impact is transmitted by the remaining (unimpacted) fiber. The light output is recorded using the streak camera. Typically, four to six fiber signals are recorded using the camera, along with the light output from a pulsed LED, which gives a fiducial timing signal to allow correlation of the crushed fiber diagnostic with the overall experiment timing.
Once the optimal armature wall thickness was determined, magnetic flux compression experiments were performed with simple helical FCGs. These generators used 32- and 64-turn stators with an i.d. of 15.2 cm and a length of 40 cm. The generator initial inductance was 55 and 200 mH, respectively. The generators drove load inductances of 150 to 200 nH, with an initial seed current of from 10 to 25 kA. The diagnostics included the impact pin and crushed fiber diagnostics (with the pins inserted between stator turns and the fibers lying axially along the stator) with the addition of current and voltage measurements. The current diagnostics consisted of load Rogowski(*) coils, while the voltage was measured with resistive dividers across the load. Additionally, the voltage across the last three turns of the generator was measured using a Pearson coil to record the current through a calibrated resistor across the turns.
The current and voltages for a typical shot are shown in Figure 1. These data are for a 32-turn generator seeded with 10 kA. The generator current gain is approximately 35. The crushed fiber impact probe data is shown in Figure 2 along with the dI/dt across the load (i.e., unintegrated Rogowski coil signal). The light output from the crushed fibers in these experiments appeared as discrete pulses, which we assume to correspond to the discrete windings of the stator, and will be discussed further. The horizontal bars on the crushed fiber data represent the uncertainty of the location in tune of the light signal because of the non-zero width of the light pulse. Comparison of the dI/dt signal with the crushed fiber data presented in Figure 2 clearly shows that the generator output voltage decays several turns before the armature reaches the load end of the generator. This may be due to high-voltage breakdown within the generator or breakup of the armature (or, in fact, breakdown induced by armature breakup). In any event, the loss of several end turns of the generator has occurred in nearly every experiment and represents a major loss of current and energy gain in these devices. We are addressing these problems in next-generation FCG designs.
As mentioned above, the crushed fiber diagnostic light output from the generator experiments gave discrete light pulses, with the number of pulses equaling the number of windings in the generator. Additionally, assuming that the pulse spacing is related to the physical turn spacing, the armature impact axial velocity determined from the crashed fiber diagnostic is very close to the C4 detonation velocity.
In order to test the hypothesis that the light pulses correspond to the discreet turns of the stator, an experiment was designed in which the stator was replaced with a series of discrete rings with variable spacing. The light output from the crushed fibers upon detonation was pulsed, with the temporal spacing of the pulses corresponding well with the explosive detonation velocity in the region with large ring spacing (1.27 cm on center). The light output inexplicably disappeared in the region with narrow ring spacing (0.32 cm on center). There were fibers placed at five azimuths around the ring structure. The position versus time for these fibers, shown in Figure 3, seems to indicate that the armature is impacting the ring structure in an asymmetric fashion. This may be due to imprecise alignment of the armature in the fixture or nonuniform expansion of the armature during detonation. Experiments are planned to attempt to resolve this.
Several computer codes are used or are in development in support of the Phillips Laboratory explosive pulsed power program. These include circuit codes such as Microcap and circuit-based generator specific design and analysis codes such as CAGEN. Material interface tracking and explosive detonation capability have been incorporated into MACH2, a two-and-a-half dimensional magnetohydrodynamic code. MACH2 and MACH3, the three-dimensional version under development, are well suited to detailed studies of areas such as armature expansion.
In the Phillips Laboratory explosive pulsed power program, we have successfully conducted armature expansion and helical explosive generator experiments. A capacitor-based simulation facility is under construction for testing pulse-conditioning concepts. The experiments which have been performed to date have achieved the goals of developing explosive pulsed power experiment capability at the Phillips Laboratory, developing diagnostics and providing benchmarking data for computer codes.
Near term experiments will concentrate on the refinement of diagnostics and improvements of armature and, therefore, FCG performance. The next-generation helical FCG design will address the apparent breakdown near the load end of the generator. This generator will be designed to meet the specifications of the ultimate mission.
* Editor's note: Rogowski coils are small probes used to measure the current density distribution in plasma discharges. The Rogowski coil generally consists of a toroidal, air-cored sensor that produces an output voltage which is proportional to the current waveform. A properly designed coil may be used to measure currents ranging from less than an ampere to millions of amperes.
1. Agee, F.J., F.M. Lehr, M. Vigil, R. Kaye, J. Gaudet and D. Shiftier, "Explosively-Driven Magnetohydrodynamic (MHD) Generator Studies," Proceedings of the Tenth IEEE International Pulsed Power Conference, 1995, pp. 1068-1071.
2. Kennedy, J.E., "Explosive Output for Driving Metals," Proceedings of the Twelfth Annual Symposium of the Behavior and Utilization of Explosives in Engineering Design, 1972, pp. 109-124.
3. CAGEN, Caren Co., 12137 Midway Dr, Tracy, CA 95376, (209) 835-0295.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||high power microwave|
|Author:||Lehr, M.; Baca, G.; Chama, D.; Shiffler, D.; Graham, J.; Sommars, W.; Coffey, S.; Englert, T.|
|Publication:||Journal of Electronic Defense|
|Date:||Jun 1, 1997|
|Previous Article:||UAV sensor scrabble: when will it spell EW?|
|Next Article:||Another path to high power microwaves.|