Speaker
Description
UKAEA (United Kingdom Atomic Energy Authority), Culham Campus, Abingdon, Oxfordshire, OX14 3DB, UK.
The accurate simulation of neutral particles in the tokamak exhaust has been a topic of research across many decades, with the primary numerical approach being Markov Chain Monte Carlo methods, widely used in linear transport problems[1,2]. Historically, neutral-neutral collisions have been treated using the BGK method in order to fit in with the linear transport constraints[3], with reduced models recently gaining traction[4].
Recent work at UKAEA has been focused on performance-portable particle libraries built on top of NESO-Particles[5], with VANTAGE-Reactions[6] providing abstractions and general implementations for plasma-neutral, particle-surface, and neutral-neutral interactions. It is actively being integrated into finite volume (Hermes-3[7]) and finite element (PENKNIFE[8]) codes. Following a less common path to kinetic neutral modelling for tokamaks[9,10,11], we build on methods from rarefied gas dynamics, specifically the Direct Simulation Monte Carlo method[12], with a focus on performance-portability and flexibility in both steady state and time-dependent scenarios. In this talk, we will present the general weighted particle approach used by VANTAGE-Reactions, benchmarking results so far, as well as early progress in development of binary collision support through an event-splitting stochastic weighted particle method[13,14].
This work has been part-funded by the EPSRC Fusion Grant 2022/27 [grant number EP/W006839/1].
References:
[1] D. Reiter, Journal of Nuclear Materials, 196–198 80–89 (1992)
[2] D.P. Stotler et al. “DEGAS 2 neutral transport modeling of high density, low temperature plasmas.” (1997).
[3] V. Kotov, et al., Plasma Physics and Controlled Fusion, 50 10 (2008)
[4] D. V. Borodin et al., Nuclear Fusion, 62 8 (2022)
[5] https://github.com/ExCALIBUR-NEPTUNE/NESO-Particles
[6] https://github.com/UKAEA-Edge-Code/VANTAGE-Reactions
[7] B. Dudson et al. Computer Physics Communications 296 2024
[8] https://github.com/ExCALIBUR-NEPTUNE/PENKNIFE
[9] S. Varoutis et al., Fusion Engineering and Design, 121 13–21 (2017)
[10] S. Q. Korving et al, Physics of Plasmas, 30 4 (2023)
[11] K. Kvist et al., Physics of Plasmas, 31 3 (2024)
[12] G. Bird, Molecular Gas Dynamics and the Direct Simulation of Gas Flows, Oxford University Press (1994)
[13] S. Rjasanow et al., Journal of Computational Physics, 124. 2 243–253 (1996)
[14] G. Oblapenko et al. Journal of Computational Physics 466 (2022)