Numerical simulation of metallic surface plasma formation by megagauss magnetic fields

Abstract

Plasma formation on the surface of thick metal in response to a pulsed multi-megagauss magnetic field is being investigated at the University of Nevada, Reno, using aluminum rods that have radii larger than the magnetic skin depth. US and Russian radiationmagnetohydrodynamic codes are being used to help interpret the experimental results such as time of plasma formation and rate of current channel expansion. The best results obtained to date with the UNR code MHRDR use a standard SESAME Maxwell-construct EOS and a Russian resistivity model, and the computed times of formation agree well with the observations across the full range of wire diameters. This leads to the conclusion that plasma formation is an MHD effect and does not involve the non-MHD processes often evoked in other contexts. The computations show that plasma is formed in lowdensity material that is resistive enough to expand across the magnetic field and yet conductive enough that Ohmic heating exceeds expansion cooling as the expanding material undergoes the liquid-vapor transition.193 9781-4244-4065-8/09/$25.00 ©2009 IEEE

Verification and validation adventures in simulating megagauss-magnetic-field-induced plasma formation on thickwire metallic surfaces

“…Therefore, to the extent that our calculations reproduce the observations, it is tempting to say that our computational models are validated and can be used to interpret the experiment. However, as shown previously [9,10], the computed results depend upon choice of equation-of-state and resistivity models, and constitutive models that have agreed with other experiments do not necessarily agree with the Zebra experiments. Because a range of answers can be obtained, we cannot discount the possibility the agreement between the calculations and experiment could be fortuitous.…”

“…The value of co used to match the Lagrangian simulations is comparable to the lowest cell density that occurs in the Lagrangian simulations at the time of plasma formation, whereas the simulations with lower co have cells with a density much lower than occurs in the Lagrangian simulations. Because lower density can lead to smaller oh (as discussed in [9] and [10]), plasma formation occurs earlier in simulations that have a lower density. A priori we would expect the MHRDR computations to become more accurate as co is reduced, and oh arguments would support this conjecture.…”

“…We have previously [9,10] reported Eulerian calculations using a variety of EOSs and a variety of resistivity models and have shown a strong sensitivity to the choice of EOS-resistivity combination. In [9], we compared the surface radiation temperature to the green brightness temperature data that was then available.…”

“…In [9], we compared the surface radiation temperature to the green brightness temperature data that was then available. The data was subsequently found to be in error due to diode saturation effects.…”