Experimental and computer simulation studies of a pulsed plasma accelerator

Summary form only given. The present abstract describes an on-going effort to provide detailed experimental diagnostics and advanced computational simulations of the behavior and performance of high power plasma thrusters for possible applications in nuclear electric propulsion systems. Electromagnetic acceleration of plasmas for propulsion has long been seen as a means to efficient high specific impulse systems. Our approach is to evaluate simulations results from advanced computational codes with data collected from a laboratory prototype thruster that is designed for accurate diagnostics. The thruster is a coaxial electrode design that discharges plasma through a section of vacuum chamber with flat quartz windows that are used for optical diagnostics. High speed photography has documented successful firing of the thruster, a heterodyne laser interferometer has been used to obtain line-of-site electron number densities near the thruster exit, and Rogowski coils have been utilized to monitor thruster current. Tests with an array of B-dot probes positioned to help characterize the thruster current sheet have commenced. The supporting simulations are being obtained from two computer codes, MACH2 and GEMS. MACH2 is a well-established MHD code based upon the ALE formulation. It has been applied to diverse problems, and its capabilities are well recognized. For the present problem, however, it has several limitations: it is a 2-D code; its grid capabilities are limited to quadrilaterals; and its architecture does not lend itself to parallel computation. Of primary importance in the present application is MACH2´s restriction to the MHD approximation, disallowing magnetic field propagation by wave processes. Therefore, we are extending the electromagnetics capability of the three-dimensional general equations mesh solver (GEMS) code that has been previously used to simulate a steady-state MHD generator to provide a time accurate simulation capability that incorporates m- dern computational methods. This extension has focused on identifying methods for solving coupled electromagnetic/fluid dynamic problems in regions where the MHD approximation fails. We have developed a solution algorithm that does not depend upon the MHD approximation, but solves the complete Maxwell equations at resource levels that appear to be similar to those associated with the magnetic induction equation. Comparisons between results from the present tests and those obtained from MACH2 and GEMS are expected to be available for ICOPS