dual approach towards better characterizing uncertainties in density reconstruction using deep learning based surrogates to accelerate Bayesian reconstructions on the one hand and using variational inference and Bayesian machine learning on the other hand.

A spatially resolved air-wedge shearing interferometer and shadowgraph diagnostic provides measurements of electron density with a resolution of similar to 40 mu m. A similar to 100-ns-long, monoenergetic electron bunch at 19.8 MeV and a current of 1.4 kA (8.5 x 10(14) e(-)) heats 100-mu m-thick aluminum (Al) foils in a 1-mm-spot to T-e similar to 1 eV. A 5-ns-long, similar to 60 mJ, frequency doubled Nd:YAG laser probes the dense Al plasma. Electron densities up to 10(20) cm(-3) are resolved; the maximum resolvable density is limited by opacity, transmission, and spatial fringe density achievable with the detector. This diagnostic provides measurements of the total phase shift, transmission, and electron density. Several measurements at different time slices provide the ability to determine the velocity of the leading edge of the shadowgraph and compare it to the motion of different density shells. These measurements are also compared to radiation hydrodynamics simulations. A rough quantitative agreement is shown between the hydro simulations and the measurements; there are differences in the exact density distributions. Published under an exclusive license by AIP Publishing.

Temperature and density measurements of range-thin pure aluminum foils heated by an intense, relativistic, monochromatic electron beam have been performed for the first time. Electron density measurements are obtained from a 532-nm air-wedge shearing interferometer and indicate a large, long-lived volume of dense plasma. The plasma grows to extend more than 5 mm off the target face, and lasts more than 500 ns from the initial deposition of energy. Spatially-and temporally-resolved visible spectroscopy normal to the target foil provides a complimentary diagnostic of the plasma plume temperature and density. The spectrometer measures the Al-I 3p–4s resonance lines, which were consistently observed to be optically thick. Stark broadening of the Al-I doublet, after correcting for self-absorption effects, yields electron densities >4x10 17 cm -3 and electron temperatures around 1 eV. Initial atomic kinetics calculations are presented to threshold plasma temperatures based on the Al-I line profiles. A spectroscopic-quality radiation transport model is developed to post-process results of a radiation-hydrodynamics simulation of the energy deposition and subsequent hydrodynamic expansion of the plasma, including estimating the optical depth of the Al-I lines.

Spatially and temporally resolved visible absorption spectroscopy is performed on sodium D-lines present as surface contaminants on an expanded dense aluminum plasma plume. An 80-ns FWHM, intense, relativistic electron beam deposits 5.4 J into a 100-μm-thick Al foil, which isochorically heats and subsequently hydrodynamically expands the material through the warm dense matter state and into a classical-like plasma state, with a coupling parameter of approximately 0.2 and a degeneracy parameter of approximately 270. The Na contamination, carried along with the expanding plume, shows saturated absorption features in the dense Al continuum for λ> 450 nm. X-ray photoelectron spectroscopy and laser-induced breakdown spectroscopy confirm Na is a surface contaminant with an atomic concentration of ∼0.1% when interrogating identical foil samples. A spectroscopic-quality radiation transport model is used to post-process 2D hydrodynamic simulations to interpret the plasma conditions based on the measured Na 3p-3s doublet line profiles. A sodium number density of 3×1015 cm−3 best matches the experimental spectra, which originate from a dense surface plasma with ne=3.0±0.8×1018 cm−3.

Cu foils, $200\phantom{\rule{4pt}{0ex}}\ensuremath{\mu}\mathrm{m}$ in thickness, were heated in two stages by a $\ensuremath{\sim}100$-ns-long monoenergetic electron bunch at 19.8 MeV and a current of 1.7 kA ($8.5\ifmmode\times\else\texttimes\fi{}{10}^{14} {e}^{\ensuremath{-}}$) in a 2-mm-spot to ${T}_{e}\ensuremath{\sim}1$ eV. After 45 ns of isochoric heating, the pressure in the foil builds up to $g20$ GPa (200 kbar), it begins to hydrodynamically disassemble, and a velocity spread is measured. Near the end of the electron pulse, the 1550 nm probe is cut off or absorbed. Photonic Doppler velocimetry measurements were made to quantify the expansion velocity, hydrodynamic disassembly time, and pressure of the foil prior to cutoff. Measurements indicate foil motion begins the instant electrons pass through the foil and continues until the particle velocity approaches the ambient sound velocity of Cu and the bulk density exceeds the critical density of the probe. Once the density of the plasma drops below the critical threshold and begins reflecting again, an expansion velocity of the classical plasma is also measured, similar to the point-source solution.

beam energy of 19.8 MeV. The principal goal of this project is to characterize these angular distributions to determine the optimal location to deploy the soft X-ray spectrometer. In addition, a proof-of-principle design will be presented. The ultimate goal of the spectrometer is to obtain measurements of the plasma temperature and density to benchmark equation-of-state models of the warm dense matter regime.

We have developed a general, three-dimensional method to locate sources of earth-directed radiation that takes into account the flight paths of reflected signals. While time-of-arrival algorithms exist for locating radio sources using line-of-sight propagation paths, radio sources originating above the surface of the Earth will not necessarily emit strong power along direct paths to satellites. A combination of direct and ground-reflected pulses or only ground-reflected signals from such sources could be received by satellite-borne sensors. The work presented here applies to satellite detection of subionospheric radio sources in a vacuum environment with ideal reflection off the Earth's surface. Because satellites are not static receivers, their configuration with respect to a radio source is not always optimal. Therefore, a statistical study is performed using 1,000 randomized configurations of 24 satellites in middle Earth orbit. For each configuration, the radio source latitude and longitude is fixed, and its altitude is varied from 1 to 97km. An analytic direct-path algorithm using the five satellites nearest the radio source provides an initial guess for the radio source latitude and longitude. We find that, using this approximation, the mean error in the calculated nadir location has a maximum value of 134m for the 97-km source altitude. The coordinates of the initial guess are used to define the origin of a grid within which an all-points search refinement is performed. Using this procedure, the overall maximum mean error in radio source position is found to be on the order of the computational grid size.

This paper reports on synthetic transmission results from Lasnex [Zimmerman and Kruer, Comments Plasma Phys. 2, 51 (1975)] radiation-hydrodynamics simulations of opacity experiments carried out at Sandia National Laboratories' recently upgraded ZR facility. The focus is on experiments utilizing disk targets composed of a half-moon Fe/Mg mixture tamped on either end with 10-μm CH and an additional 35-μm beryllium tamper accessory on the end facing the spectrometer. Five x-ray sources with peak power ranging from 10 to 24 TW were used in the simulations to heat and backlight the opacity target. The dominant effect is that the beryllium behind the Fe/Mg mixture is denser and more opaque than the beryllium unshielded by metal during the times of greatest importance for the transmission measurement for all drives. This causes the simulated transmission to be lower than expected, and this is most pronounced for the case using the lowest drive power. While beryllium has a low opacity, its areal density is sufficiently high such that the expected reduction of the measured transmission is significant. This situation leads to an overestimate of iron opacity by 10%–215% for a photon energy range of 975–1775 eV for the 10-TW case. It is shown that if the tamper conditions are known, the transmission through each component of the target can be calculated and the resulting opacity can be corrected.

Atmospheric electromagnetic pulse (EMP) events are important physical phenomena that occur through both man-made and natural processes. Radiation-induced currents and voltages in EMP can couple with electrical systems, such as those found in satellites, and cause significant damage. Due to the disruptive nature of EMP, it is important to accurately predict EMP evolution and propagation with computational models. CHAP-LA (Compton High Altitude Pulse-Los Alamos) is a state-of-the-art EMP code that solves Maxwell s equations for gamma source-induced electromagnetic fields in the atmosphere. In EMP, low-energy, conduction electrons constitute a conduction current that limits the EMP by opposing the Compton current. CHAP-LA calculates the conduction current using an equilibrium ohmic model. The equilibrium model works well at low altitudes, where the electron energy equilibration time is short compared to the rise time or duration of the EMP. At high altitudes, the equilibration time increases beyond the EMP rise time and the predicted equilibrium ionization rate becomes very large. The ohmic model predicts an unphysically large production of conduction electrons which prematurely and abruptly shorts the EMP in the simulation code. An electron swarm model, which implicitly accounts for the time evolution of the conduction electron energy distribution, can be used to overcome the limitations exhibited by the equilibrium ohmic model. We have developed and validated an electron swarm model previously in Pusateri et al. (2015). Here we demonstrate EMP damping behavior caused by the ohmic model at high altitudes and show improvements on high-altitude, upward EMP modeling obtained by integrating a swarm model into CHAP-LA.

Electromagnetic pulse (EMP) events produce low-energy conduction electrons from Compton electron or photoelectron ionizations with air. It is important to understand how conduction electrons interact with air in order to accurately predict EMP evolution and propagation. An electron swarm model can be used to monitor the time evolution of conduction electrons in an environment characterized by electric field and pressure. Here a swarm model is developed that is based on the coupled ordinary differential equations (ODEs) described by Higgins et al. (1973), hereinafter HLO. The ODEs characterize the swarm electric field, electron temperature, electron number density, and drift velocity. Important swarm parameters, the momentum transfer collision frequency, energy transfer collision frequency, and ionization rate, are calculated and compared to the previously reported fitted functions given in HLO. These swarm parameters are found using BOLSIG+, a two term Boltzmann solver developed by Hagelaar and Pitchford (2005), which utilizes updated cross sections from the LXcat website created by Pancheshnyi et al. (2012). We validate the swarm model by comparing to experimental effective ionization coefficient data in Dutton (1975) and drift velocity data in Ruiz-Vargas et al. (2010). In addition, we report on electron equilibrium temperatures and times for a uniform electric field of 1StatV/cm for atmospheric heights from 0 to 40km. It is shown that the equilibrium temperature and time are sensitive to the modifications in the collision frequencies and ionization rate based on the updated electron interaction cross sections.