Paul Steen

High Yield Argon Z‐pinch Results with a Large Diameter Nozzle

We modified our original 12 cm diameter double‐shell gas puff nozzle to include an on‐axis jet with a large diameter throat and an independent plenum to allow a large fraction of the total mass to be contained in the central region (r=0–1.5 cm). By judicious selection of pressures for the jet and the two shells, we were able to double the Argon K‐shell yield from ∼10 kJ to > 20 kJ with a 3.5 MA current drive and implosion time of ∼205 ns, equivalent to the yield produced at 100 ns implosion time, but with half the pulse‐width, for radiated K‐shell power up to 2 TW.The radiation produced by gas originating in each of the three plenums was distinguished by the use of a chlorine tracer introduced sequentially into each plenum. We thereby deduce that 65% of the K‐shell radiation is produced by gas originating in the jet, 30% from gas originating in the inner shell and only 5% from gas originating in the outer shell.The flexibility of the hardware was further exercised by selectively evacuating one of the thre...

Interfacing of PLIF Measurements with 2D-MHD Calculations

Summary form only given. We have recently employed data obtained from laser-induced fluorescence (PLIF) measurements to provide the initial density distribution for 2D-MHD calculations of argon Z-pinches. In previous studies those data were derived from hydrodynamics calculations. The measured data introduce a feature not present in the calculated initial conditions, namely, the presence of sub-millimeter density fluctuations. In view of the key role played by the Rayleigh-Taylor instability in these plasmas, particularly as shorter wavelengths grow more rapidly, it is necessary to understand the importance of this feature. To address the issue we have performed MACH2 calculations and compared results using the raw, measured data and compared them with cases where the short wavelength components were filtered out. It appears that longer wavelength excitations, produced by the macroscopic density variations (~10 mm) dominate much of the pinching process. With a uniform initial distribution perturbed at a given wavelength, the shorter wavelength behavior would be expected to grow most rapidly. With a non-uniform system where the swept-up material shows long wavelength dominance, the sub-millimeter variations do not grow to large magnitude although they do modify the long wavelength behavior and indirectly, to the compressions achieved near axis. While the short wavelengths do modify plasma compression, the effect on compression saturates after a small number of smoothing iterations

Boundary Conditions for Real Gas Puff Loads

Summary form only given. We have developed a model for treating, in a consistent manner, the measured gas puff distributions used in Z-pinches. This is important in order to model the Z-pinch dynamics in a way that does not introduce numerical excitations that interfere with the development of unstable Rayleigh-Taylor modes. The initial gas density distribution, measured by laser-induced fluorescence (PLIF), exhibits two-dimensional features even near its outer (radial) edges. For the 2D-MHD calculations performed using MACH2 it was found necessary to implement the model in order to apply the electromagnetic boundary conditions consistently. Without doing so, non-physical modes could be excited, leading to errors in the modeling of the Z-pinch implosion. In prior treatments the electromagnetic wave driving the plasma load has been assumed to be uniform across the discharge gap with electric and magnetic fields varying only in the direction tangential to the conducting walls of the vacuum feed. We have found that by applying an open circuit forward-going wave condition at a position removed from the plasma-vacuum interface and by inserting a vacuum transition region that permits the EM field to transform from a 1D to 2D character it is possible to treat the boundary conditions consistently. We illustrate the results obtained with calculations of the pinching process and compare the magnitude and wavelength changes brought about by the use of this method

12 cm diameter gas puff density profile measurements using planar laser induced fluorescence

Summary form only given. The technique of Planar Laser Induced Fluorescence (PLIF) has been applied to characterize the initial gas distributions produced by 12 cm diameter nozzles, as used on recent Decade Quad and Double-EAGLE Ar Z-pinch experiments. This information is required to understand and simulate the implosion dynamics. A 266 nm, /spl sim/5 ns, /spl sim/5 cm line-focused laser beam passes through the Ar gas puff, which is mixed with a few percent acetone by volume. A gated intensified CCD camera views the fluorescence of the excited acetone molecules from the side. The acetone emission has a short, intense pulse, coincident with the exciting laser pulse, and a weak, slowly decaying phosphorescence. The gas density profiles are obtained from the fluorescence images recorded coincident with the laser. Flow velocities of the gas puff are obtained by capturing the images of the phosphorescence at several delay times relative to the laser pulse. The distribution of the clusters is determined by comparing the PLIF signal with 2-D Rayleigh scattering from the gas puff. These measurements are compared with computer simulations to benchmark and refine the 2-D fluid dynamics models.

Bremsstrahlung reflex triode modeling and optimization

Summary form only given, as follows. A nuclear effect testing requires a large-area flash X-ray sources with specific X-ray endpoint energy to properly simulate responses in a wide variety of materials present in the test object. Improvements in radiation hardness have led to requirements for higher dose and fluence. In response, a 300-keV pinch reflex triode was recently developed on the DTRA PITHON simulator. Although the pinch reflex triode has good reproducibility at > 500 keV endpoint, the standard deviation in dose and fluence degrades to about 40% at the 300-keV endpoint. To improve the reproducibility, an experimental parametrical study was carried out where different configurations were tested and parameters such as the A-K gap were optimized. Simulations were performed in order to reproduce the trends observed experimentally and to evaluate the validity of our reflex triode model This paper describes the modeling work performed to understand the electron and ion dynamics in the diode. To simulate the relativistic planar diode, we use the electromagnetic particle-in-cell code LSP2 coupled with the Integrated Tiger Series3 electron/photon transport code. Experimental and simulation results are compared and discussed.

Calculations of X-ray yields in annular argon gas puff experiments

Summary form only given. We have compared the measured yield and timing of x-ray pulses in Double Eagle 2D radiation-MM calculations performed using MACH2. For the twin shell masses involved, the plasma was treated as optically thin to the K-shell radiation and two collisional radiative (CR) models were used to assess their ability to predict output. We found that calculations of both timing and yield of radiation above 1 keV gave reasonably accurate agreement with measurements if (1) the spatial resolution of the plasma during the peak compression and radiation output pulse was adequate and (2) the magnetic field gradients near the plasma-vacuum interface during implosion were sufficiently accurate. For an Eulerian code such as MACH2, such calculations, while possible, are not always straightforward to carry out. Practical values of the pseudovacuum resistivity in the absence of special treatment, may lead to compressed plasma densities that are too low and temperatures that are too high, resulting in radiative yield predictions that fall below measured values. We discuss differences in the spectra calculated using the two CR models and compare them with experimental data.

Decade quad monolithic POS modeling

The decade quad (DQ) was initially fielded as a large area bremsstrahlung (LAB) source. This utilized the four decade modules (DM) triggered simultaneously but each driving a separate hard X-ray bremsstrahlung radiation source (BRS). Recently, water convolute hardware was installed that combines the power from the four Decade modules to drive a soft X-ray plasma radiation source (PRS). The water convolute configuration also enables the use of a monolithic plasma opening switch (MPOS) for driving a common BRS or PRS load. The work reported here was performed in support of the use of an ACE 4 type POS on DQ. The modeling included equivalent circuit and DELTA-CREMIT magneto-hydrodynamic (MHD) simulations. The circuit analysis extracted equivalent circuit parameters from ACE 4 data and applied them to the planned DQ MPOS-BRS configuration. The MHD analyses evaluated strategies to optimize the DQ MPOS.

Modeling of a shell-on-shell nozzle and comparison with laser induced fluorescence and interferometry measurements

Summary form only given. Double shell nozzles have provided improved X-ray yields for Z-pinch implosions. Numerical gas flow calculations, including the interactions between the two plumes of a double Argon puff, were performed and are reported here. 2D gas density profiles as a function of time were obtained using the NOZZLE code, which solves the Navier-Stokes equations for the supersonic transient flow all the way from the plenum, including the effects of the gas release valve motion. The results of this numerical calculations are compared with interferometry and laser fluorescence measurements obtained for double shell nozzles used recently for experiments on Double-Eagle, Saturn and Z (see companion papers by B. H. Failor et al. and H. Sze et al.).

Implementation of the ACE 4 plasma opening switch on the Decade Quad

Summary form only given, as follows. This paper reports on the initial implementation of a monolithic plasma opening switch (MPOS) on the Decade Quad (DQ). The recent installation of a water convolute combining the power flow from the four DQ modules to drive a plasma radiation source (PRS) enables a MPOS configuration for driving a bremsstrahlung radiation source (BRS). The conduction phase requirement for the DQ MPOS is roughly 8 MA for 300 ns. Numerical simulations and experimental results on ACE 4 suggest that the dominate parameter controlling the POS operation is the conduction phase I*T product. The ACE 4 POS design, characterized at an I*T product of 2.4 MA-/spl mu/s, was installed and tested on the water-convoluted DQ configuration. The work reported here will contrast the observed performance of the unoptimized ACE 4 POS on DQ with that expected based on scaling from ACE 4. Optimization of the DQ monolithic POS is planned for a later phase of the present work.

Magnetic diffusion effects on electrically driven flux compression schemes

Summary form only given, as follows. Power amplification in vacuum utilizing electrically driven magnetic flux compression has been introduced and researched recently for efficient drive of z-pinches and other loads operating at sub-100 ns times and very high magnetic fields (tens of megamperes at radii below 1 cm). We present the results of 1 D MHD calculations, in the 1 microsec regime, corresponding to the ECF2 pulsed power generator being built at CEG, France, and the 100 ns regime, corresponding to the Z machine at SNL. These calculations purport to assess the role of magnetic field diffusion through the plasma armature. To evaluate numerical accuracy the equations are solved both in Lagrangian and Eulerian reference frames. We utilize analytic approximations for the equation of state, resistivity and opacity of the Al armature, and compare with results obtained employing SESAME tables. The magnetic field diffusion equation is solved using a mixed finite element algorithm employing an edge-based electric field vector representation.