Surrogate model of energetic particle transport in reactor-relevant fusion devices

We aim to deliver a robust surrogate model of energetic particle transport in reactor-level fusion devices which is significantly faster and reasonably accurate when compared to existing high-fidelity simulation codes. This model holds crucial importance for the design of a fusion pilot plant, particularly in achieving steady state burning plasma conditions, where the confinement of energetic alpha particles plays a pivotal role. Energetic alpha particles generated as by-product of D-T fusion reactions may give rise to plasma instabilities such as Alfvén eigenmodes (AEs), which can de-confine alphas, leading to inefficient plasma self-heating and erosion of the wall-components. This poses a major challenge for optimizing a Fusion Pilot plant (FPP) design due to a gap in our understanding of the impact of energetic particle (EP) losses on the burning plasma conditions. Our approach begins with the generation of a comprehensive dataset. Simulations, conducted using our existing simulation code across a range of reactor-relevant physical parameters, provide the necessary data for surrogate model development. We will then leverage this dataset to construct a surrogate model of alpha transport using innovative machine learning tools. This general surrogate model will be further incorporated in integrated frameworks such as IPS-FASTRAN framework to perform self-consistent calculations of burn performance in a reactor-level fusion device.