J. V. Parker

Submicrosecond Pulsed Power Capacitors Based on Novel Ceramic Technologies

Capacitor energy density for submicrosecond discharge applications was investigated for capacitors based on the following: 1) polymer-ceramic nanocomposite (PCNC); 2) antiferroelectric (AFE); and 3) paraelectric (PE) ceramic dielectrics. The developmental PCNC dielectric enabled design, fabrication, and testing iterations to be completed relatively rapidly. The PCNC capacitors were nominally 4 nF and were tested to dc potentials of at least 75 kV. The capacitors were then charged from 20 to 48 kV with a dc high-voltage power supply and discharged into a nearly critically damped test circuit of up to 5 pulses/s (pps) repetition rate for lifetime testing. The discharge time was 65 ns. Shot life as a function of the charge voltage was compared for three design iterations. Changes in the manufacturing of the PCNC capacitors have yielded up to 100× improvements in pulse discharge life. The 1-2-kV prototype, nonlinear (antiferroelectric and paraelectric) multilayer ceramic capacitors had zero-voltage capacitance ratings of between 60 and 300 nF. They were charged to their operating voltage and discharged into a nearly critically damped load in 2-6 μs, depending on their capacitance, at repetition rates of up to 75 pps. Their operating voltage for fast, repetitive discharge was determined for lifetimes consistently over 105 shots. Discharge energy densities of 0.27-1.80 J/cc and energy losses of 7.9-36.8% were obtained for the packaged multilayer capacitors with different formulations of nonlinear dielectrics. Increased field-induced strain was correlated with increased permittivity and contributed to the limitations on the operating voltage. Multilayer ceramic capacitors fabricated from AFE and PE ceramic dielectrics have the potential to achieve high energy density owing to their high relative permittivities that vary with applied electric field, assuming they can be scaled up to sufficiently high voltages.

Design, fabrication, and operation of a high-energy liner implosion experiment at 16 megamperes

We discuss the design, fabrication, and operation of a liner implosion system at peak currents of 16 MA. Liners of 1100 aluminum, with initial length, radius, and thickness of 4 cm, 5 cm, and 1 mm, respectively, implode under the action of an axial current, rising in 8 /spl mu/s. Fields on conductor surfaces exceed 0.6 MG. Design and fabrication issues that were successfully addressed include: Pulsed Power-especially current joints at high magnetic fields and the possibility of electrical breakdown at connection of liner cassette insulator to bank insulation; Liner Physics-including the angle needed to maintain current contact between liner and glide-plane/electrode without jetting or buckling; Diagnostics-X-radiography through cassette insulator and outer conductor without shrapnel damage to film.

Development and Testing of a High-Gain Magnetic Flux Compression Generator

The performance of a high-gain FCG is often limited by internal electrical breakdown caused by the high voltage generated during operation. Modern diagnostic techniques provide the opportunity to diagnose internal breakdowns so that generator designs can be improved. This paper describes the internal breakdowns observed in the JAKE FCG developed at the AFRL during the late 1990s. A revision to the stator winding pattern of the JAKE generator has led to improved control of the internal voltage. Designated JILL, the revised generator has substantially better flux transport efficiency, particularly at higher seed current. The techniques employed to design the new stator winding and the results of development testing are presented.

Flux Compression Generator Development at the Air Force Research Laboratory

The Air Force Research Laboratory (AFRL) maintains an extensive capability for the design, analysis, construction and testing of explosive pulsed power (EPP) components. Three flux compression generators (FCGs) were designed as part of an EPP technology development effort sponsored by AFRL and the Defense Advanced Research Projects Agency (DARPA). A secondary-stage, high-current FCG was designed to deliver 10 MA into a nominal load inductance of 80 nH from an initial generator inductance of 1.6 muH that is seeded with 1 MA. We have also developed a coaxial FCG to deliver more than 20 MA into a 2 nH load. The initial flux in the coaxial chamber (60 nH at 1.5 MA) is compressed uniformly using a copper armature, which is simultaneously initiated using a slapper detonator. Either of these two FCGs can be seeded with a third generator design: a high-gain, helical FCG. This model serves as our workhorse generator capable of delivering 2 MA into a 0.5 muH inductive load. It has also been operated into load inductances ranging from 0.1 to 2.0 muH with comparable flux delivery. All experiments are conducted on an explosive test range located on Kirtland Air Force Base [1]. The design effort is supported by powerful computer modeling using CAGEN [2], CALE and MACH2. Design features for all three FCGs are presented in this paper with results from recent explosive tests.

Solid dielectric transmission lines for pulsed power

This paper documents recent work developing solid dielectric transmission lines for sub-microsecond, 100 kV class compact pulsed power systems. Polymer-ceramic nanocomposite materials have demonstrated sub-microsecond discharge capability in parallel plate capacitors and transmission lines [1, 2]. With a dielectric constant of approximately 50, the propagation velocity is 2.5 cm/ns, necessitating lines of several meters length to achieve > 100 ns pulse lengths. By folding the line in a fashion analogous to ceramic multilayer capacitors, the physical length of the line can be significantly shorter than the electrical length. We present the results of an experimental effort to develop a folded transmission line using a polymer-ceramic nanocomposite dielectric. The pulse length was somewhat shorter than expected based on a simple calculation using the geometry and the dielectric constant. Fully 3-D electromagnetic calculations were used to examine the role of the edges in curtailing the pulse length. Dielectric breakdown in this device occurred below the electric field threshold demonstrated in the prior work [1]. Improvements in the large scale fabrication of TiO2 beginning with nanoscale grains have opened the possibility for producing single layer high voltage devices. Given a dielectric constant approaching 140, transmission lines using nano-TiO2 can be considerably shorter than with other materials. Relatively thick, flat sheets of TiO2 have been fabricated for testing up to 50 kV. Several transmission lines, employing a serpentine electrode geometry, have been manufactured and tested. Testing up to several 10s of kV has confirmed the operation of the lines according to the design. As expected, the triple point between the TiO2, electrode, and insulating medium has proven difficult to manage for high voltage operation. Several techniques to mitigate the effects of the triple point, including resistive grading at the edges of the electrodes, are discussed. Fully 3-D electromagnetic modeling is used to examine the effects of electrode geometry and composition on the performance of the lines.

The Explosive Pulsed Power Test Facility at AFRL

The Air Force Research Laboratory has developed and tested a variety of explosive driven pulsed power devices over the past twenty-two years. Testing is performed primarily at a dedicated facility located at Chestnut Site on Kirtland Air Force Base. The facility is described in this paper, including details of recent upgrades.

Fiber-Optic Systems at the Explosive Pulsed Power Test Facility at AFRL

The Air Force Research Laboratory (AFRL) located on Kirtland Air Force Base performs explosive pulsed power experiments [1] - [3]. The large separation distances between the related subsystems of these shots increase the likelihood of inadvertent multiple electrical ground connections. This paper describes some of the fiber-optic devices routinely used during our explosive power tests to mitigate the problems associated with ground loops.

Experimental Studies of an Ultrahigh-Speed Plasma Flow

In 1991, Turchi et al. reported evidence for a 2000 km/s aluminum plasma that originated from the upstream boundary of a wire array armature in a plasma flow switch (PFS). The 2008 article by Turchi et al. posits that if such high Z plasma could instead be composed of deuterium or a deuterium-tritium mixture, then the resultant multi-keV plasma would make an effective target for magnetized plasma compression to fusion conditions. This report documents several experiments executed in an effort to achieve an ultrahigh-speed flow in a deuterium plasma. The first phase of this research concentrated on extension of the earlier work to a lower current system that would emulate the PFS used in series with an imploding liner load. The apparatus was also modified to permit pulsed injection of deuterium gas along the insulated coaxial electrodes between the PFS armature and the vacuum power feed. The experiments met with limited success, exhibiting evidence of a 550 km/s plasma flow which convected a small fraction of the total magnetic field. Two subsequent tests were conducted using foam armatures. In both cases, current prematurely shunted upstream in the vacuum feed. Several possible causes were explored for the shunting of the current. Among the modifications implemented, the gas injection system was altered to increase both the quantity of gas adjacent to the armature while facilitating an increased pressure gradient between the armature and the current feed. A series of low-energy shots were conducted to examine the impact of several proposed design modifications on current delivery to the armature. These experiments demonstrated that the hardware assembled for this investigation was unlikely to forestall breakdown in the injected gas as required by Turchi et al. Nevertheless, two experiments were conducted to evaluate performance with foam armatures. Both experiments exhibited good current delivery to the armature, behaving initially like the low-energy experiments. The magnetic flux convected downstream was greater than in any of the prior experiments, though significant work remains to demonstrate the ultrahigh-speed plasma flow concept.

Adventures in the experimental development of an ultrahigh speed plasma flow

In 1991, Turchi et al. [1] reported evidence for a 2,000 km/s aluminum plasma that originated from the upstream boundary of a wire array armature in a plasma flow switch (PFS) [2]. The 2008 article by Turchi et al. [3] posits that if such high Z plasma could instead be composed of deuterium or a deuterium-tritium mixture then the resultant multi-keV plasma would make an effective target for magnetized plasma compression to fusion conditions. This report documents several exploratory tests executed in an effort to achieve significant energy transfer from a plasma flow switch to a deuterium plasma. The first phase of this research concentrated on extension of the earlier work [1, 2] to a lower current system that would emulate the PFS used in series with an imploding liner load. The apparatus was also modified to permit pulsed injection of deuterium gas along the insulated coaxial electrodes between the PFS armature and the vacuum power feed. In analyzing the armature behavior, the initial conditions used in 2-D axisymmetric MHD simulations to approximate the wire-array/polymer film composite armature resulted in significant uncertainty in the validity of the calculations. This uncertainty confounded efforts to improve the opening switch behavior of the armature. Low density foams, commonly used in other high energy density plasma experiments, were seen as a candidate material for the armature that would facilitate greater fidelity between simulations and the experiment. Two subsequent tests were conducted using foam armatures. In both cases, current prematurely shunted upstream in the vacuum feed. Several possible causes were explored for the shunting of the current. Among the modifications implemented, the gas injection system was altered to increase both the quantity of gas adjacent to the armature while facilitating an increased pressure gradient between the armature and the current feed. A series of low energy shots were conducted to examine the impact of several proposed design modifications on current delivery to the armature. One conclusion of these experiments was that it has been very difficult to forestall breakdown in the injected gas as required by Turchi et al. [3]. Nevertheless, two experiments were conducted to evaluate performance with foam armatures. Both experiments exhibited good current delivery to the armature, behaving initially like the low energy experiments. The magnetic flux convected downstream was greater than in any of the prior experiments, though significant work remains to demonstrate the ultra-high-speed plasma flow concept.

Experiments on field reversed configuration (FRC) formation and their compression using liners

Three types of experiments developing FRC formation, injection, and compression are described: field-compression, FRC formation-translation-capture, and FRC formation—translation—capture—compression. All involve the generation of primarily axial guide and mirror magnetic fields with ∼ 2 Tesla peak fields, using ∼5 ms rise time discharges into 9 pulsed magnet coils surrounding the liner implosion portion of the device. The field compression and FRC compression experiments use 12 MA, 4.5 MJ Shiva Star capacitor bank axial discharges to drive implosions of 30 cm tall, 10 cm diameter, 1 mm thick Al shells or liners. The FRC capture experiments are a pre-requisite for the destructive FRC compression experiments. All FRC experiments use 3 capacitor discharges into a segmented theta coil surrounding the FRC formation region to establish a bias field, accomplish pre-ionization of deuterium gas, and provide the reverse field main theta discharge (∼ 1 MA) which forms the FRC. This is aided by two cusp field discharges. The guide and mirror fields enable translation of the FRC and its capture in the liner interior region. Diagnostics include pulsed power (current and voltage), magnetic field, field exclusion, laser interferometry, imaging and spectroscopy, radiography, and both activation and time-of-flight neutron detection. Design features and operating parameters are guided by 2D-MHD simulations. Supported by DOE-OFES.