CDS&E: Moment Approach to Multiscale Plasmas

Plasma, one of four fundamental states of matter, can be found everywhere in the cosmos, from the laboratory to the solar atmosphere, to supernova remnants and distant galaxies. This award supports a computational study of how energy in plasmas is distributed and dissipated in systems where collisions between charged plasma particles are infrequent. The project will develop reduced models that leverage novel machine learning techniques to make simulations more efficient, and will integrate them into an open-source simulation code applicable to problems in astrophysics, space weather modeling, and fusion energy development. This multidisciplinary project straddling the fields of plasma physics, computational physics, and machine learning will offer a valuable experience to undergraduate and graduate students. Project's outreach activities will involve presentations at universities and local high schools in coordination with the Wonders of Physics program at the University of Wisconsin-Madison. The primary methodology used by this project will be based on the recently developed moment-hierarchy model for magnetized plasmas at arbitrary collisionality. The model will be integrated into an open-source, machine-learning-ready simulation code that incorporates both traditional neural networks and more advanced physics-informed neural networks. The methodology will enable a transition from simplified models with ad-hoc closures and reduced collision operators to a first-principles framework. This approach will alleviate the computationally intensive nature of micro-scale kinetic simulations and enhance the ability to simulate and understand weakly collisional astrophysical and laboratory plasmas. The work will be performed in collaboration with École Polytechnique Fédérale de Lausanne in Switzerland, the University of California, Los Angeles, the Massachusetts Institute of Technology, and Princeton University. The numerical results will be validated using experimental facilities at the Wisconsin Plasma Physics Laboratory (WiPPL). This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.