Comparison and Evaluation of Numerical Techniques and Physico-Chemical Algorithms for the Simulation of an Iodine and Xenon Powered Ion Thrusters

Abstract

An electrostatic ion thruster, modelled on Busek’s BIT-3 [5], is simulated numerically using the open-source software “Starfish”. It is assumed that the thruster is in vacuum conditions propelling a CubeSat in low Earth orbit. Iodine is chosen here as the propellant under test. The results from modelling this relatively recent new fuel are compared to those of the standard propellant of xenon. The plasma in the ion thruster and the associated electric fields are simulated using a particle-based kinetic code in which the hybrid approach of Particle in Cell and Direct Simulation Monte Carlo methods has been used. In modelling these flows, elastic and inelastic collisions can occur involving charge and momentum exchanges. Such collision models use a number of assumptions, e.g., concerning the collisional cross-section area, and in this paper, we present results where the physico-chemical modelling is improved reducing the level of assumptions used. Results are also presented concerning the numerical methods used for the iterative convergence scheme, stochastic sampling, and the importance of the constraints for the mesh size and timestep. It is found that the most appropriate timestep is one which enables both the CFL condition and the highest frequency to be captured. The mesh size affects the choice of the solver being used; the largest the cell sizes the greater the assumption of quasi-neutral flow and thus areas of non-neutrality (such as in the surrounding shealth may be treated inadequately. The subsequent effects on the plume and the thrust produced of using different approaches in modelling the electron temperature distribution are also evaluated to produce a more rigorous modelling methodology.