Data Grids for Large-Scale Fusion Simulations

70796S02-II Waveguide breakdown is the limiting factor in obtaining the large acceleration gradients needed for the Next Linear Collider. Simulation of this phenomenon in three dimensions would likely enable researchers to find configurations with higher breakdown thresholds. However, because current three-dimensional codes are not parallel, they take excessive computer time to run. This project will adapt an existing plasma simulation code, which is parallel and solves Maxwell¿s equations with charged particles and fluids, to simulate breakdown in waveguides. Modifications will include: (1) adding the effects of nonplanar boundaries, emission of particles from the walls, resistive boundaries, and lossy dielectrics; (2) enhancing the specification for both the domain decomposition and geometric objects; and (3) improving user friendliness through visualization tools and, a preparser. In Phase I, nonplanar boundaries, particle emission, and resistive boundaries were prototyped. Results were shown to agree with theoretical expectations. In Phase II, the above features will be made robust and enhanced: nonplanar boundaries will be extended to three dimensions, new particle emission algorithms will be incorporated to study multipacting, resistive walls will be enhanced, lossy dielectrics will be added, the geometric specification of domain decomposition and three-dimensional objects will be improved, and user friendliness will be improved. A simulation of the microwave breakdown problem will be performed. Commercial Applications and Other Benefits as described by awardee: A tool for waveguide simulations would be made available to the high-power microwave community. Such a tool would enable the optimization of power throughput and also enable calculations of waveguide loading.