Parametric approach to linear induction accelerator design

Past work on the design of linear induction accelerators has centered on the development of computer codes to analyze accelerator designs, using the current filament method. While these filament models are a very valuable tool for evaluating the performance of an induction launcher design, they provide little insight into the selection of dimensions, materials, and operation points for accelerators with interesting performance. Described in this paper is a parametric approach to defining effective accelerator designs. This method uses a computer optimization routine to iteratively seek out effective designs. The optimization routine is forced to search within a parameter space restricted to interesting and realistic parameters such as size, weight, voltage, and temperature rises. A filament model is used as the filter for the optimizer. Several linear induction accelerators have been designed using this method. The accelerators designed all used a switched capacitor power supply. While the run time of this code on The University of Texas' CRAY XMP-24 computer is moderately long, the resulting designs have good predicted performance. With realistic power supplies and materials, accelerator efficiencies in the 20 to 40% range were easily obtained. This paper describes the effect of armature diameter, length-to-diameter ratio, and weight, as well as other parameters, on the optimum accelerator design