A quasi-spherical z-pinch may directly compress foam or deuterium and tritium in three dimensions as opposed to a cylindrical z-pinch, which compresses an internal load in two dimensions only. Because of compression in three dimensions the quasi-spherical z-pinch is more efficient at doing pdV work on an internal fluid than a cylindrical pinch. Designs of quasi-spherical z-pinch loads for the 28 MA 100 ns driver ZR, results from zero-dimensional (0D) circuit models of quasi-spherical implosions, and results from 1D hydrodynamic simulations of quasi-spherical implosions heating internal fluids will be presented. Applications of the quasi-spherical z-pinch implosions include a high radiation temperature source for radiation driven experiments, a source of neutrons for treating radioactive waste, and a source of fusion energy for a power generator.
We present the first comprehensive study of high wire-number, wire-array Z-pinch dynamics at 14-18 MA using x-ray backlighting and optical shadowgraphy diagnostics. The cylindrical arrays retain slowly expanding, dense wire cores at the initial position up to 60% of the total implosion time. Azimuthally correlated instabilities at the array edge appear during this stage which continue to grow in amplitude and wavelength after the start of bulk motion, resulting in measurable trailing mass that does not arrive on axis before peak x-ray emission.
Over the last few years, a variety of experiments studying higher photon energy (>4 keV) radiators have been performed, primarily at the Z accelerator. In this paper, the results of experiments designed to study the effects of initial load diameter on the radiated output of stainless steel wire arrays are presented. Stainless steel is primarily iron, which radiates in the K-shell at 6.7 keV. Nested wire arrays from 45 mm initial outer diameter to 80 mm outer diameter were fielded at the Z accelerator. A nested array consists of two wire arrays, with the inner concentric to an outer. All of the arrays fielded for this work had a 2:1 mass and diameter ratio (outer:inner), and the arrays were designed to have the same implosion time. A degradation of K-shell output was observed (pulse shape and power) for the smallest and largest diameter arrays, suggesting a region in which optimal conditions exist for K-shell output. The degradation at small diameters results from the reduced eta value, due to low implosion velocity. Eta is defined as the kinetic energy per ion divided by the energy required to get to the K-shell. At large diameters, a dramatic degradation of output is observed not just for the K-shell, but also for the lower energy X-rays. This may be the result of the low mass required to maintain an appropriate implosion time - there simply aren't many radiators available to participate. One other possibility is that the higher acceleration necessary at large diameters to achieve the same implosion time results in additional instability growth. Also necessary to consider are the effects of interwire gap: due to the limited wire sizes available, the interwire gap on the large diameter loads is large, in one case more than 3 mm. Comparisons of the trends observed in the experiments (radiated yield, pulse shape, and spectra) will be made to calculations previously benchmarked to K-shell data obtained at Z. The reproducibility of the arrays, advanced imaging diagnostics fielded, current diagnostics, and sensitivities of the calculations are also discussed.
The experimental and computational investigations of nanosecond electrical explosion of thin Al wire in vacuum are presented. We have demonstrated that increasing the current rate leads to increased energy deposited before voltage collapse. Laser shadowgrams of the overheated Al core exhibit axial stratification with a {approx}100 {micro}m period. The experimental evidence for synchronization of the wire expansion and light emission with voltage collapse is presented. Two-wavelength interferometry shows an expanding Al core in a low-ionized gas condition with increasing ionization toward the periphery. Hydrocarbons are indicated in optical spectra and their influence on breakdown physics is discussed. The radial velocity of low-density plasma reaches a value of {approx}100 km/s. The possibility of an overcritical phase transition due to high pressure is discussed. 1D MHD simulation shows good agreement with experimental data. MHD simulation demonstrates separation of the exploding wire into a high-density cold core and a low-density hot corona as well as fast rejection of the current from the wire core to the corona during voltage collapse. Important features of the dynamics for wire core and corona follow from the MHD simulation and are discussed.
A technique for manufacturing wires with imposed modulation in radius with axial wavelengths as short as 1 mm is presented. Extruded aluminum 5056 with 15 {micro}m diameter was masked and chemically etched to reduce the radius by {approx}20% in selected regions. Characterized by scanning electron microscopy, the modulation in radius is a step function with a {approx}10 {micro}m wide conical transition between thick and thin segments, with some pitting in etched regions. Techniques for mounting and aligning these wires in arrays for fast z-pinch experiments will be discussed. Axially mass-modulated wire arrays of this type will allow the study of seeded Rayleigh-Taylor instabilities in z pinches, corona formation, wire initiation with varying current density in the wire core, and correlation of perturbations between adjacent wires. This tool will support magnetohydrodynamics code validation in complex three-dimensional geometries, and perhaps x-ray pulse shaping.
The high-pressure response of cryogenic liquid deuterium (LD{sub 2}) has been studied to pressures of {approx}400GPa and densities of {approx}1.5g/cm{sup 3}. Using intense magnetic pressure produced by the Sandia National Laboratories Z accelerator, macroscopic aluminum or titanium flyer plates, several mm in lateral dimensions and a few hundred microns in thickness, have been launched to velocities in excess of 22 km/s, producing constant pressure drive times of approximately 30 ns in plate impact, shock wave experiments. This flyer plate technique was used to perform shock wave experiments on LD{sub 2} to examine its high-pressure equation of state. Using an impedance matching method, Hugoniot measurements of LD{sub 2} were obtained in the pressure range of {approx}22-100GPa. Results of these experiments indicate a peak compression ratio of approximately 4.3 on the Hugoniot. In contrast, previously reported Hugoniot states inferred from laser-driven experiments indicate a peak compression ratio of approximately 5.5-6 in this same pressure range. The stiff Hugoniot response observed in the present impedance matching experiments was confirmed in simultaneous, independent measurements of the relative transit times of shock waves reverberating within the sample cell, between the front aluminum drive plate and the rear sapphire window. The relative timing was found to be sensitive to the density compression along the principal Hugoniot. Finally, mechanical reshock measurements of LD{sub 2} using sapphire, aluminum, and {alpha}-quartz anvils were made. These results also indicate a stiff response, in agreement with the Hugoniot and reverberating wave measurements. Using simple model-independent arguments based on wave propagation, the principal Hugoniot, reverberating wave, and sapphire anvil reshock measurements are shown to be internally self-consistent, making a strong case for a Hugoniot response with a maximum compression ratio of {approx}4.3-4.5. The trends observed in the present data are in very good agreement with several ab initio models and a recent chemical picture model for LD{sub 2}, but in disagreement with previously reported laser-driven shock results. Due to this disagreement, significant emphasis is placed on the discussion of uncertainties, and the potential systematic errors associated with each measurement.
When several kA pulses are passed through single, fine 25 {micro}m diameter wires, the wire material heats, melts, vaporizes and expands. Initially the voltage across--and current through--a wire increases until an abrupt voltage collapse occurs. After this collapse the voltage remains at a relative small value while the current continues to increase. In order to understand how this early time behavior may affect the subsequent implosion, small-scale experiments at Cornell University's Laboratory of Plasma Studies concentrated on diagnosing expanding single wire dynamics. X-ray backlighting, interferometry and Schlieren imaging as well as current and voltage measurements have been employed. The voltage collapse has been attributed to the formation of plasma around the wire and a transfer of current to this highly conducting coronal plasma. Interferometry has confirmed the plasma formation, but the current transfer has only been postulated. Subsequent experiments on the Z-Facility at Sandia National Laboratories have produced impressive x-ray yields etc.
Using intense magnetic pressure, a method was developed to launch flyer plates to velocities in excess of 20 km s{sup -1}. This technique was used to perform plate-impact, shock wave experiments on cryogenic liquid deuterium (LD{sub 2}) to examine its high-pressure equation of state (EOS). Using an impedance matching method, Hugoniot measurements were obtained in the pressure range of 22--100 GPa. The results of these experiments disagree with the previously reported Hugoniot measurements of LD2 in the pressure range above {approx}40 GPa, but are in good agreement with first principles, ab initio models for hydrogen and its isotopes.
Recently an innovative technique known as the Isentropic Compression Experiment (ICE) was developed that allows the dynamic compressibility curve of a material to be measured in a single experiment. Hence, ICE significantly reduces the cost and time required for generating and validating theoretical models of dynamic material response. ICE has been successfully demonstrated on several materials using the 20 MA Z accelerator, resulting in a large demand for its use. The present project has demonstrated its use on another accelerator, Saturn. In the course of this study, Saturn was tailored to produce a satisfactory drive time structure, and instrumented to produce velocity data. Pressure limits are observed to be approximately 10-15 GPa (''LP'' configuration) or 40-50 GPa (''HP'' configuration), depending on sample material. Drive reproducibility (panel to panel within a shot and between shots) is adequate for useful experimentation, but alignment fixturing problems make it difficult to achieve the same precision as is possible at Z. Other highlights included the useful comparison of slightly different PZT and ALOX compositions (neutron generator materials), temperature measurement using optical pyrometry, and the development of a new technique for preheating samples. 28 ICE tests have been conducted at Saturn to date, including the experiments described herein.
Numerical calculations have been performed to investigate the role that load thickness may play in the performance of fast annular z pinch implosions. In particular, the effects of load thickness on the mitigation of the magnetically-driven Rayleigh-Taylor (RT) instability and energy coupling between the load and generator are addressed. using parameters representative of the Z accelerator [R.B.Spielman et al., Phys.Plasmas, 5, 2105 (1998)] at Sandia National Laboratories, two dimensional magnetohydrodynamic (MHD) simulations show that increased shell thickness results in lower amplitude, slightly longer wavelength RT modes. In addition, there appears to be an optimum in load velocity which is directly associated with the thickness of the sheath and subsequent RT growth. Thin, annular loads, which should couple efficiently to the accelerator, show a large reduction in implosion velocity due to extreme RT development and increased load inductance. As a consequence, thicker loads on the order of 5 mm, couple almost as efficiently to the generator since the RT growth is reduced. This suggests that z-pinch loads can be tailored for different applications, depending on the need for uniformity or high powers.
Recent improvements in z-pinch wire array load design at Sandia National Laboratories have led to a substantial increase in pinch performance as measured by radiated powers of up to 280 TW in 4 ns and 1.8 MJ of total radiated energy. Next generation, higher current machines will allow for larger mass arrays and comparable or higher velocity implosions to be reached, possibly extending these result.dis the current is pushed above 20 MA, conventional machine design based on a 100 ns implosion time results in higher voltages, hence higher cost and power flow risk. Another approach, which shifts the risk to the load configuration, is to increase the implosion time to minimize the voltage. This approach is being investigated in a series of experimental campaigns on the Saturn and Z machines. In this paper, both experimental and two dimensional computational modeling of the fist long implosion Z experiments will be presented. The experimental data shows broader pulses, lower powers, and larger pinch diameters compared to the corresponding short pulse data. By employing a nested array configuration, the pinch diameter was reduced by 50% with a corresponding increase in power of > 30%. Numerical simulations suggest load velocity is the dominating mechanism behind these results.
Recent success on the Saturn and Z accelerators at Sandia National Laboratories have demonstrated the ability to scale z-pinch parameters to increasingly larger current pulsed power facilities. Next generation machines will require even larger currents (>20 MA), placing further demands on pulsed power technology. To this end, experiments have been carried out on Saturn operating in a long pulse mode, investigating the potential of lower voltages and longer implosion times while still maintaining pinch fidelity. High wire number, 25 mm diameter tungsten arrays were imploded with implosion times ranging from 130 to 240 ns. The results were comparable to those observed in the Saturn short pulse mode, with risetimes on the order of 4.5 to 6.5 ns. Experimental data will be presented, along with two dimensional radiation magnetohydrodynamic simulations used to explain and reproduce the experiment.
High wire number, 25-mm diameter tungsten wire arrays have been imploded on the 8-MA Saturn generator, operating in a long-pulse mode. By varying the mass load from 710 to 6140 ps/cm, implosion times of 130 to 250 ns have been obtained with implosion velocities of 50 to 25 cn-dys, respectively. These z-pinch implosions produced plasmas with millimeter diameters that radiated 600 to 800 kJ of x-rays, with powers of 20 to 49 TW; the corresponding pulse widths were 19 to 7.5 ns, with risetimes ranging from 6.5 to 4.0 ns. These powers and pulse widths are similar to those achieved with 50 ns implosion times on Saturn. Two-dimensional, radiation- magnetohydrodynamic calculations indicate that the imploding shells in these long implosion time experiments are comparable in width to those in the short pulse cases. This can only be due to lower initial perturbations. A heuristic wire array model suggests that the reduced perturbations, in the long pulse cases, may be due to the individual wire merger occurring well before the acceleration of the shell. The experiments and modeling suggest that 150 to 200 ns implosion time z-pinches could be employed for high-power, x-ray source applications.
The advent of the 20-MA Z accelerator [R.B. Spielman, C. Deeney, G.A. Chandler, et al., Phys. Plasmas 5, 2105, (1997)] has enabled implosions of large diameter, high-wire-number arrays of titanium to begin testing Z-pinch K-shell scaling theories. The 2-cm long titanium arrays, which were mounted on a 40-mm diameter, produced between 75{+-}15 to 125{+-}20 kJ of K-shell x-rays. Mass scans indicate that, as predicted, higher velocity implosions in the series produced higher x-ray yields. Spectroscopic analyses indicate that these high velocity implosions achieved peak electron temperatures from 2.7{+-}0.1 to 3.2{+-}0.2 keV and obtained a K-shell emission mass participation of up to 12%.