Abstract: FUSE is an innovative computational framework designed for the holistic simulation of fusion power plant facilities. Developed in the Julia programming language, it offers unprecedented performance—delivering results 1000 times faster than traditional frameworks. FUSE integrates first-principle analysis, machine learning, and reduced models to address the intricate physics and engineering challenges critical to achieving net electric power generation from fusion.
The framework evaluates a broad range of power plant designs by incorporating essential plasma physics processes like equilibrium, transport, heating & current drive, and stability as well as facility aspects like neutronics, structural mechanics, and magnet coil design. Its core strength lies in constrained multi-objective optimization, which allows for exploring economic and engineering trade-offs in power plant design. Notably, simulations within FUSE have revealed cost-effective configurations using different types of superconductors and tokamak designs, enhancing our understanding of cost drivers in fusion technology.
FUSE also excels in trajectory optimization, enabling dynamic simulations that predict operational states throughout the lifecycle of a fusion plant. This capability is vital for designing efficient and sustainable fusion power facilities, promising to advance the field significantly.
Bio: Since 2020, I have contributed to advancing computational plasma physics at General Atomics, notably through co-founding the FUSE project, which delivers rapid design simulations of fusion power plants. My work on integrated modeling by working on the PRO-create and STEP module within the OMFIT framework advanced the capabilities to predict energy confinement from small tokamaks to large fusion power plants.
Some time ago, in 2016, I was at DIFFER for a research internship focused on creating a zero-dimensional model for CO2 plasma reactors, marking the first step in my journey.
Go to the Events page.