Efficient and robust evaluation of fast particle losses in non-axisymmetric tokamak plasmas

Abstract

We present various techniques that make orbit-following Monte Carlo simulations faster and more reliable when assessing collisionless fast particle losses due to magnetic field perturbations. These techniques are based on identifying various loss channels in constants of motion space using the so-called loss maps. We demonstrate that this allows one to attribute losses quantitatively to different transport mechanisms, increase signal-to-noise ratio when estimating FILD signal and peak power loads, and connect magnetic field structure directly to fast particle losses. Furthermore, we show that collisionless fast particle transport can be treated as an advection-diffusion process where the transport coefficients can be evaluated with the orbit-following method. Applying these techniques has the potential to make orbit-following simulations faster to perform, or to avoid them completely, while making the results more reliable as they become more clearly connected to underlying physics. We demonstrate these techniques for ITER by showing how alpha particle losses are affected by various magnetic field perturbations, estimating ICRH losses in reduced field scenarios, and performing a scan on alpha particle losses as a function of ELM control coil current phases.