Impedance modeling of accelerator beams with discontinuous charge density using scattered-field physics-informed neural networks

Abstract

A physics-informed neural network method for solving electrodynamic interaction problems including a relativistic beam of charged particles in particle accelerators has been recently proposed. However, the method still has a limitation on modeling accelerator beams with discontinuous charge density. To remove this limitation, the scattered field formulation is introduced into this method. This approach allows us to model the field of discontinuous beam charge density such as point and ring charges. Its numerical error is shown to be smaller than that of the boundary element method. The presented approach is applied to three different vacuum chambers.