Christopher L. Rousculp

Electrical Design and Operation of the Phelix Pulsed Power System

The Precision High Energy-Density Liner Implosion Experiment (PHELIX) is a pulsed power driver capable of delivering multimegampere currents to cylindrical loads. The PHELIX hardware includes novel design features to provide a high-energy conversion efficiency of approximately 10-MA output current per megajoule of stored energy. This is achieved by a rail-gap switched low-inductance Marx design (resistively damped) driving a multifilar air-core pulse transformer. The Marx output cables form the toroidal transformer that is an integral part of the disc line and removable load cassette assembly. The transformer and disc line uses conformal insulation methods and does not require replacement; after each shot, the transformer is completely reusable. Load cassettes can be easily exchanged to facilitate experimental variation. PHELIX is selfcontained within its own transport container and Faraday cage that can be moved from the maintenance building to the Los Alamos Neutron Science Center 800-MeV proton accelerator facility to perform multipulse proton radiography. This paper details the electrical and mechanical design of the Marx and multifilar transformer assemblies as well as presenting the operational performance achieved to date.

PHELIX: Design and Analysis of a Transformer-Driven Liner Implosion System

To provide substantial reduction in the size and energy of high-energy-density experiments, we have designed, built, and operated a liner implosion system that is driven by a multiturn-primary, single-turn-secondary, current step-up toroidal transformer. The Precision High Energy-density Liner Implosion eXperiment (PHELIX) pulsed-power driver, which is currently under development at Los Alamos National Laboratory, Los Alamos, NM, can provide >;400 kJ of capacitively stored energy and peak load currents of >;5 MA to implode centimeter-size liners in 10-20 μs, attaining speeds of 1-4 km/s. Diagnosis of scaled-down liner implosion experiments will be performed with the 800-MeV proton radiographic (pRad) system at Los Alamos Neutron Science Center (LANSCE); therefore, PHELIX is designed to be portable with a footprint of only 8 ×25 ft2. The multiframe, high-resolution imaging capability of pRad will be used to study hydrodynamic and material phenomena. Experiments with scaled-down electromagnetic railguns, pulsed high-field magnets, and magnetic flux compression are also under consideration. This paper discusses the overall PHELIX design concept and layout, and details of the electromechanical design needed to ensure repeatable operation.

The PHELIX Liner Demonstration Experiment (PLD-1)

The PHELIX Liner Demonstration Experiment (PLD-1) took place in September of 2010 at Los Alamos National Laboratory. The PHELIX machine consists of a ∼500 kJ single-marx capacitor bank cable-coupled to a toroidal 1∶4 current step-up transformer which delivers multi-Mega-Ampere currents to a cm size load. In this experiment the load consisted of a ∼3 cm radius, 0.8 mm thick, ∼3 cm tall aluminum liner, copper glide planes, a thin polyethylene insulator, and a 0.5 cm thick aluminum return conductor. Two independent channels of fiber optic Faraday rotation measured a peak load current > 4 MA with a pulse width of ∼ 10 µs. Four linear Rogowski coils measured the output current of the 4 marx modules. High-resolution flash X-radiography imaged a stable, highly symmetric and uniform liner 14.5 µs after current start. A 12 channel laser Doppler velocimetry (LDV) system tracked the inside surface of the liner throughout the experiment and showed a peak velocity before impact with probes of ∼ 1 km/s. The LDV probes were arrayed axially as well as azimuthally and confirmed the symmetry of the liner trajectory. Surprisingly, the LDV showed distribution of velocities of the inner liner surface late in time. PLD-1 is the first step towards utilizing the PHELIX pulsed-power system at the Los Alamos proton radiography facility.

The PHELIX pulsed power project: Bringing portable magnetic drive to world class radiography

The PHELIX pulsed power project will introduce magnetically driven hydrodynamics experiments to the Los Alamos National Laboratorys proton radiography facility (pRad). The Precision High Energy-density Liner Implosion eXperiment (PHELIX) has been commissioned at Los Alamos. A small footprint capacitor bank consisting of four parallel connected single-stage marx units (∼500 kJ) is cable coupled to a toroidal, current step-up transformer to deliver multi-Mega-Ampere, ∼10 µs current pulses to cm size cylindrical loads. In a sequence of tests the performance of each component (capacitor bank and transformer) was evaluated and compared to a circuit model. The transformer coupling was observed to be k ∼0.93. The tests culminated in a liner implosion experiment in which an ∼3 cm radius, 0.8 mm thick, ∼3 cm tall aluminum liner was accelerated to a velocity of ∼ 1 km/s. The suite of machine diagnostics included linear Rogowski coils and Faraday rotation for current measurements. The experimental diagnostics include B-dot probes, multi-channel photon Doppler velocimetry (PDV), and single-frame, flash X-radiography to evaluate the performance of the high precision liner implosion. Currently, work is focused on integrating PHELIX into normal operations with the 800 MeV proton radiography facilities. There, high-resolution, high-frame-rate imaging of hydrodynamic experiments will be possible.

Pulsed Power Hydrodynamics: An Application of Pulsed Power and High Magnetic Fields to the Exploration of Material Properties and Problems in Experimental Hydrodynamics

Pulsed-power hydrodynamics (PPH) is an evolving application of low-impedance pulsed-power technology. PPH is particularly useful for the study of problems in advanced hydrodynamics, instabilities, turbulence, and material properties. PPH techniques provide a precisely characterized controllable environment at the currently achievable extremes of pressure and material velocity. The Atlas facility, which is designed and built by Los Alamos National Laboratory, is the worlds first, and only, laboratory pulsed-power system designed specifically for this relatively new family of pulsed-power applications. Atlas joins a family of low-impedance high-current drivers around the world, which is advancing the field of PPH. The high-precision cylindrical magnetically imploded liner is the tool most frequently used to convert electromagnetic energy into the hydrodynamic (particle kinetic) energy needed to drive strong shocks, quasi-isentropic compression, or large-volume adiabatic compression for the experiments. At typical parameters, a 30-g 1-mm-thick liner with an initial radius of 5 cm and a moderate current of 20 MA can be accelerated to 7.5 km/s, producing megabar shocks in medium density targets. Velocities of up to 20 km/s and pressures of > 20 Mbar in high-density targets are possible. The first Atlas liner implosion experiments were conducted in Los Alamos in September 2001. Sixteen experiments were conducted in the first year of operation before Atlas was disassembled, moved to the Nevada Test Site (NTS), and recommissioned in 2005. The experimental program resumed at the NTS in July 2005. The first Atlas experiments at the NTS included two implosion dynamics experiments, two experiments exploring damage and material failure, a new advanced hydrodynamics series aimed at studying the behavior of particles of damaged material ejected from a free surface into a gas, and a series exploring friction at sliding interfaces under conditions of high normal pressure and high relative velocities. Longer term applications of PPH and the Atlas system include the study of material interfaces subjected to multimegagauss magnetic fields, material strength at high strain rate, the properties of strongly coupled plasmas, and the equation of state of materials at pressures approaching 10 Mbar.

Dynamic Friction Experiments at the Atlas Pulsed Power Facility

A Series of dynamic friction experiments has been conducted at the Atlas Pulsed Power Facility. Pulsed currents in excess of 21 MAmps were delivered to a cylindrical liner in about 15 ¿s. The liner was accelerated to km/s velocities and symmetrically impacted a hollow Ta/Al/Ta target. Due to the shock speed difference in Ta and Al, sliding velocities of almost a km/s were achieved at the Ta/Al interfaces. Initial analysis indicates that the machine performed to within a few percent of the design specifications. The primary diagnostic for these experiments was three radiographic lines-of-sight to look at thin gold wires embedded within the Al piece of the target. The magnitude of the displacement and the amount of distortion of the wires near the material interface is used as a measure of the dynamic frictional forces occurring there. Other diagnostics included a single-point VISAR and line-ORVIS to measure the breakout time and velocity on the inside of the target. Also, the Faraday rotation of a laser beam through a circular loop of optical fiber located in the power-flow channel of the experiment is used to measure the total current delivered to the experimental load. Data are being compared to a theoretical dynamic friction model for high sliding velocities. The model is based on molecular dynamics simulations and predicts an inverse power law dependence of frictional forces at very high sliding velocities.

Crenulation-­1 Flash Report

We briefly report on the first of PHELIX driven Crenulation experiments diagnosed with proton radiography. PHELIX is a 300 kJ capacitor bank located at the LANL LANSCE pRad facility. It is capable of delivering a 4 MA, 10 us current pulse to a low inductance cylindrical load. Using a magnetically driven cylindrical liner, we have shocked a concentric tin cylindrical shell to melt-on-release and compared to theory and calculations. The inner surface of the tin cylinder has three sectors of single-mode perturbations, which are subject to Richtmyer-Meshkov Instability (RMI). Recent theoretical work on the EOS of tin has modified both the Hugoniot and the isentropes for release into various states in tabular data. The new multiphase EOS for tin, SESAME 2161, includes the beta and gamma solid phases as well as a liquid phase. It predicts a lower pressure boundary for release to pure solid (~20 GPa) and a higher pressure boundary for release to pure liquid (~35 GPa) than the existing SESAME 2160 table.

Small particle transport experiments in vacuum and gas using pulsed-power Z-pinch liner-on-target drive and diagnosed with proton radiographic imaging

Summary form only given. When a sufficiently strong shock emerges from the free surface of a solid, micron-sized particles may be “ejected” from the tiny defects, grain boundaries or surface inclusions characteristic of real surfaces. If the solid surface bounds a gas or plasma such as an MTF or ICF target, the introduction of surface (perhaps high-Z) material into the gas or plasma may significantly alter its properties and behavior. The formation of ejecta particles has been the subject of both experimental measurements and computational modeling. Simple hydrodynamic drag models have been less than completely adequate, and recent work has explored hydrodynamic (such as Richtmyer-Meshkov) explanations. Less experimental work has been devoted to exploring ejecta particle transport in gas (or plasmas), and most of that work has been done in planar geometries. A new high precision, experiment, called the Damaged Surface Hydrodynamics Experiment, has been developed to explore transport of ejecta particles into gas (or plasma) in converging geometries, diagnosed by high resolution, fast, multi-frame imaging by proton radiography to inform the continuing development of transport models and validate current and future simulations.To provide a high precision, predictable, reproducible, and controllable drive, a pulse-power driven, dynamic, cylindrical, liner-on-target configuration has been developed that is compatible with the Los Alamos proton radiography facility. To further control the initial conditions, preformed, carefully characterized, micron-sized (tungsten) particles were used in place of shock-formed particulate. Initial experiments explored the shock-launched transport of ~1 micron particles from the converging surface into vacuum and into about 10 atmospheres of a low atomic weight (Ar) and higher atomic weight (Xe) gas. The experiment was driven by the PHELIX pulsed-power system utilizing a high-efficiency (k ~ 0.93) transformer to couple a small capacitor bank (U ~ 300 kJ) to a low inductance load in a Z-pinch configuration. The current pulse (Ipeak = 3.7 MAmp, Γι~10 μs) was measured via a fiber optic Faraday rotation diagnostic. The experimental load consisted of a cylindrical Al liner (6 cm diam, 3 cm tall, 0.8 mm thick) and a cylindrical Al target (3 cm diam, 3 cm tall, 0.1 mm thick). The target was coated with a thin (0.1 mm) uniform layer of tungsten powder. The shock-launched powder layer fully detaches from the target and transports as a spatially correlated, radially converging (vr ~ 800 m/s) ring. The powder distribution is highly modulated in azimuth during transport in both the vacuum and gas-filled cases suggesting that radial motion modified by simple drag models are inadequate to describe the transport. Results are compared to 1D and 2D MHD simulations.

Improvements in Ranchero magnetic flux compression generators

Los Alamos “Ranchero” Magnetic Flux Compression Generators (FCGs) have been used to power imploding liner loads. The fundamental FCG design is based on a cylindrical detonation system that expands the armature simultaneously into a coaxial generator volume and has been shown to generate currents as high as 76 MA. Analysis of the 76 MA test results revealed a weakness in the design at the output glide plane. To prevent premature shorting at the output current slot of the generator, the armature/glide plane interface was originally designed to lag the leading edge of the armature. However, 2D-MHD calculations reveal that at very high currents a magnetically driven aneurism develops in this lagging section which reduces the performance of the generator. A new model Ranchero is being developed to correct this weakness and provide enhanced performance. In the new model, the output glide plane is eliminated and the armature is extended along the FCG axis, with its radius increasing along a curve until it reaches the current output slot. A cylindrical detonation system of the type required for earlier designs continues to be used, and the high explosive (HE) in the extended section is detonated by the last point of the cylindrical detonator. The stator of the FCG is contoured, allowing the contact point of the armature to zipper from the input to the output end in the last few μs of flux compression. In addition, the new model Ranchero is intended to use PBX 9501 (9501) for the HE and also remove the smoothing layer, which has been part of all Ranchero HE systems to date. Both of these factors lead to increased performance. 9501 is more energetic than the PBXN 110 used in Ranchero generators to date, and both calculations and experiments have shown that the smoothing layer is not needed when the detonator point spacing is 18 mm. Tests of original model Rancheros using PBXN 110 castable HE, with an imbedded smoothing layer, demonstrated an armature expansion velocity of 3.1 mm/μs. Further tests show that removal of the smoothing layer increases the speed to 3.3 mm/μs, and replacement of the cast PBXN 110 with 9501 without a smoother gives a velocity of 3.8 mm/μs. Designs, concerns, and experimental results facilitating the new model Ranchero are presented. In addition, performance estimates are given for the initial imploding liner tests to be conducted, and further computational details are presented in a companion paper given by C. L. Rousculp and others at this conference.

Update on PHELIX pulsed-power hydrodynamics experiments and modeling

Summary form only given. The PHELIX pulsed-power driver is a 300 kJ, portable, transformer-coupled, capacitor bank capable of delivering 3-5 MA, 10 μs pulse into a low inductance load. Here we describe further testing and hydrodynamics experiments. First, a 4 nH static inductive load has been constructed. This allows for repetitive high-voltage, high-current testing of the system. Results are used in the calibration of simple circuit models and numerical simulations across a range of bank charges (±20 <; V0 <; ±40 kV). Furthermore, a dynamic liner-on-target load experiment has been conducted to explore the shock-launched transport of particulates (diam. ~ 1 μm) from a surface. The trajectories of the particulates are diagnosed with radiography. Results are compared to 2D hydro-code simulations. Finally, initial studies are underway to assess the feasibility of using the PHELIX driver as an electromagnetic launcher for planer shock-physics experiments.