BEAMS3D Validation and Benchmarking on HPC systems
This page documents various efforts to validate and benchmark the BEAMS3d code.
Table of Contents
Benchmarking
As a benchmark test a slowing down to the VMEC last closed flux surface run is performed based on a W7-X beam deposition run with full port model for the high mirror configuration. The simulations starts with 2^18 particles per beam per energy for 2 hydrogen beams so 6x2^18 particles are followed. For each device the scans are in powers of two in terms of nodes 32-64-128-256-512 (where available).
Validation
Various publications document the validation of the BEAMS3D model.
- McMillan, M. and Lazerson, S.A. "BEAMS3D: Neutral beam injection model." Plasma Phys. and Control. Fusion 56, 095019 (2014)
- Cole, M. D. J. et al. "Comparative collisionless alpha particle confinement in stellarator reactors with the XGC gyrokinetic code" Phys. Plasmas 26, 032506 (2019)
- Lazerson, S.A. et al. "Validation of the BEASM3D neutral beam deposition model on Wendelstein 7-X" Nuclear Fusion 60, 706020 (2020)
- Lazerson, S.A. et al. "Modeling and measurement of energetic particle slowing down on Wendelstein 7-X" Nuclear Fusion 61, 096006 (2021)