New Algorithm Overcomes Obstacle in Fusion Energy Simulation

The era of exascale has brought with it a slew of fusion energy simulation projects, aimed at stabilizing the notoriously tricky – and so far unmastered – clean energy source that would transform the world virtually du jour. on the next day. Supercomputers and deep learning have shown promise in being able to predict the destabilizations that derail fusion reactions and determine the corrections needed to prevent them, but the time required in a fraction of a second remains out of reach, and many Elements of the operation of fusion energy reactors remain, in general, poorly understood. Now, researchers at DOE’s Princeton Plasma Physics Laboratory (PPPL) have developed a new computational method to simulate the motion of free electrons in plasma physics experiments, a major obstacle to simulating fusion energy.

Like so many fusion energy simulations, the algorithm involves a key device called a “tokamak” (shown in the header), which uses magnetism to suspend active plasma particles during the fusion reaction. During this process, electrons moved around the tokamak – called pitch angle scattering – in a way that proved difficult to simulate due to its complexity. The PPPL algorithm skips this hurdle by calculating the probable trajectories of scattered electrons.

“Solving the stochastic differential equation gives the probability of each path that scattered electrons can take,” Yichen Fu, a graduate student at PPPL and lead author of the paper describing the research, said in an interview with John Greenwald of PPPL. “However, the trajectories are probabilistic and we don’t know exactly where the electrons would go because there are many possible paths. But by solving the trajectories, we can know the probability that the electrons choose each path, and know that this allows for more accurate simulations that can lead to better plasma control.

This better control of the plasma would come from changes in the way electrical current is conducted in the plasma and from a better understanding of how pitch angle diffusion can pose hazards to the stability and integrity of devices involved in the reaction. “This technical breakthrough shows the role of [Computational Sciences Department at PPPL], said Hong Qin, senior research physicist at PPPL and co-author of the paper. “One of his goals is to develop algorithms that lead to improved fusion simulations.” PPPL’s ​​Computer Science Department is a relatively new addition, having been launched last year.

This new equation expands on previous research, adding the incorporation of magnetic fields and performing a more rigorous proof of the accuracy of the algorithm. “It gives experimenters a better theoretical description of what’s going on to help them design their experiments,” Qin said. “Previously, there was no working algorithm for this equation, and physicists got around this difficulty by modifying the equation.”

To learn more about this research, read PPPL’s ​​John Greenwald article here and read the research paper here.

About Mariel Baker

Check Also

AMD Releases Latest Consistent Device Memory Mapping Linux Code – Designed for Frontier

Over the past year, we have seen various patches released by AMD engineers with a …