Please note change in time and location for this seminar.
Speaker: Brendan C. Lyons, General Atomics
Title: Pulse Design and Digital Twin Capabilities of the FUSE Integrated-Modeling Framework
Abstract: The FUsion Synthesis Engine, FUSE (https://fuse.help), is an open-source, flexible framework for integrated whole-facility modeling of fusion experiments and pilot plants (FPPs). Originally created for tokamak FPP machine design, FUSE has been extended to model the time evolution of both the plasma and plant, supporting simulations in both feedforward and feedback modes with controllers. An overview of the individual models in FUSE and their integrated capabilities will be given. By combining first-principle, AI/ML, and reduced models, FUSE simulates plasma physics, engineering, control, costing, and risk, all with a combination of high speed and high fidelity. This allows for the generation of large databases of individual model results and complete FUSE designs which are used to train neural networks efficiently (e.g., for transport and equilibrium). Furthermore, genetic algorithms are used to perform constrained optimization of FPPs. Time-dependent modeling, including trajectory optimization and DIII-D experimental analysis, will also be shown. In addition, a modern, Grad-Hogan-like solver is being developed in FUSE, which promises a novel capability coupling first-principles-based equilibrium evolution, high-fidelity transport models, and realistic plasma control. In total, FUSE is prepared to provide digital-twin tools for experimental analysis, predict-first pulse design, and FPP optimization, making it a valuable resource for bringing fusion energy to the grid. Opportunities of collaboration as both users and developers will be discussed. Work supported by General Atomics corporate funding.
Bio: Dr. Brendan C. Lyons is a Scientist in the MHD and Disruption Physics branch of the General Atomics Theory and Computational Science Department. His research interests include magnetohydrodynamics (MHD), integrated modeling, and high-performance computing. His doctoral research in the Princeton University Program in Plasma Physics (2014) focused on the coupling of MHD and neoclassical models. He then received a DOE Fusion Energy Sciences Postdoctoral Research Fellowship to perform research at General Atomics, using the extended-MHD code M3D-C1 to model ELM-suppression and mitigation experiments. Since becoming a staff scientist in 2017, his professional research has focused on nonlinear MHD modeling of disruption mitigation. In addition, he has been a key developer of numerous integrated modeling projects including OMFIT, SEGWAY, STEP, and now FUSE. He has also engaged substantially in various fusion strategic-planning and science-policy activities.
Event details: In-person seminars are only available to CU ID holders. At this time, Non-Columbia affiliates and the general public are only invited to participate remotely. Contact [email protected] if you would like the Zoom link for this seminar.