A computer program that can limit instabilities that reduce the performance of fusion plasmas has been developed by Imène Goumiri, a former Princeton University graduate student and now a researcher at the University of Wisconsin-Madison. The more instabilities there are, the less efficiently doughnut-shaped fusion facilities called tokamaks operate.

Goumiri worked with physicists at PPPL to construct the new model. It applies feedback from sensors for real-time control of the rotation of plasma that fuels fusion reactions, and draws on the fact that rotating different sections of a plasma at different speeds creates a force called “shear” that reduces instabilities.

Goumiri modeled rotation from data collected from PPPL's National Spherical Torus Experiment (NSTX) before it was upgraded. She built the program in MATLAB software and then added it to a predictive model based on PPPL’s TRANSP code, the global standard for analyzing plasma performance. The TRANSP model found the new approach to be effective at controlling rotation.

“This confirmed the validity of our model,” said Goumiri, lead author of a paper describing the process in the journal Nuclear Fusion. Coauthors included Clarence Rowley, Princeton professor of mechanical and aerospace engineering, and David Gates, principal research physicist at PPPL and stellarator physics leader, who served as her academic advisors; and Steven Sabbagh, senior research scientist and adjunct professor of applied physics at Columbia University on long-term assignment to PPPL, a member of her doctoral committee who served as a scientific advisor.

Looking ahead, researchers said this new class of controller programs can be developed from simulations based on experimental data, with no need for additional experiments for calibration. The new method would necessitate fewer experiments and would provide a way to predict requirements for adjusting plasma rotation in future fusion facilities.