State-of-the-art fusion simulation leads three scientists to the 2024 Kaul Foundation Prize

(Press-News.org) Three scientists were awarded the 2024 Kaul Foundation Prize for Excellence in Plasma Physics Research and Technology Development based on their decades of groundbreaking research about how plasma behaves in fusion reactors. 

Choongseok (CS) Chang, Seung-Hoe Ku and Robert Hager of the U.S. Department of Energy’s (DOE) Princeton Plasma Physics Laboratory (PPPL) were recognized “for experimentally validated simulations of turbulence-broadened divertor heat flux widths using the X-Point Included Gyrokinetic Code (XGC),” following decades of research developing comprehensive simulations to model the fusion plasma edge. 

Recently, the scientists – in collaboration with researchers from the Massachusetts Institute of Technology (MIT) and other collaborators working on the DIII-D fusion device at General Atomics – determined that these simulations closely matched experiments in the DIII-D. The research showed that the turbulence doubles the exhaust layer width in conditions similar to what would be found in a commercial-scale reactor such as ITER. This is an important experimental validation that XGC can describe the relevant underlying physics, helping support XGC predictions that ITER could have a much broader exhaust footprint than what has been predicted by present tokamak data. 

This simulation code has been critical for a great deal of research that has advanced fusion science. The code simulates the whole volume of a tokamak plasma, especially the edge region of the magnetically confined plasma that includes the area where magnetic field lines cross, which is known as the X-point. This area is particularly important to study because of its reduced confining magnetic field strength, which can allow plasma particles to escape confinement. XGC is widely considered one of the best such codes available and is used by researchers worldwide on some of the planet’s most powerful computers. 

“This work has brought great honor to the Lab,” said PPPL Director Steve Cowley when he presented the award at PPPL’s annual State of the Lab address. “This high-performance computing exascale project code, developed at our Lab, is also being honored by the U.S. Secretary of Energy with the prestigious Secretary’s Honor Award as part of the broader exascale computing initiative.”

Using very powerful hardware, exascale computers can perform one quintillion (or a billion billion) calculations per second, which makes them faster than the most powerful supercomputers currently used. 

Each recipient of the annual Kaul Foundation Prize receives $7,500 in recognition of their scientific achievement. The prize was established with funds from the late PPPL Director Ronald C. Davidson’s 1993 Award for Excellence in Science, Education and Physics. It honors outstanding contributions to research in plasma physics. 

Notably, the 2024 winners used XGC to determine critical details about how ions and electrons escape the core plasma during fusion when the plasma is confined by magnetic fields inside a tokamak. Their highly sophisticated simulation is for ITER, the multinational fusion facility under construction in France. The simulation suggests that a key region of the ITER wall should not get as hot as once feared based on the experimental data from present-day tokamaks.

“We would like to thank the national and international XGC team members. For the divertor heat load research, credit goes to the ITER Organization collaborators led by Alberto Loarte and Richard Pitts; PPPL, DIII-D, Alcator C-Mod, National Spherical Torus Experiment and Lawrence Livermore National Laboratory participants in the 2016 DOE Office of Science national theory milestone activities that led to the discovery of the ion leakage and turbulent electron loss physics that are responsible for plasma exhaust. We also thank the joint PPPL/DIII-D research team led by Alessandro Bortolon for the XGC application on DIII-D edge plasma and Darin Ernst of MIT for collaborating with us to simulate his experiments in ITER-like conditions, which turned out to be ideal for validating the XGC simulations,” said Chang. 

“We hope to see more experimental validations on other tokamaks. We also would like to thank the tremendous support from the DOE program managers, DOE computer centers and PPPL management, which made the research possible.”

Choongseok (CS) Chang

After graduating with a doctoral degree in physics from the University of Texas at Austin in 1979, Chang was a senior scientist at General Atomics in San Diego before holding a tenure position at the Korea Advanced Institute of Science & Technology (KAIST). He later moved back to the U.S. and worked at the Courant Institute of Mathematical Sciences at New York University (NYU) before ultimately landing at PPPL in 2011.

Chang’s long career in plasma physics has focused on solving key theoretical challenges to make fusion a reliable source of electricity. Chang has spent decades leading multidisciplinary teams of physicists, applied mathematicians and computer scientists — including those who use artificial intelligence — to lead the development of XGC and simulate the extremely complex environment inside fusion reactors. 

“Plasma is not a single physics phenomenon. Several physics interact together. But that was a very difficult theory to develop,” said Chang. Additionally, the problem was multiscale, meaning it needed to be studied at multiple levels of detail. 

“Fortunately, I had a hunch in the late 1990s that computers would become more and more powerful so we could solve these problems,” Chang said. He recruited “a few brilliant students” to work on this important task. Among them, Ku was the main workforce. At the time, creating models that considered multiple physics simultaneously was considered nearly impossible. But Chang and the talented team – especially Ku – persisted. Ultimately, the work developing the necessary computer codes that could realize his multiphysics vision would receive substantial funding from the DOE and recognition from major U.S. computer centers. The success of this work eventually led Chang to resign from his positions in Korea and at NYU to fully dedicate himself to the XGC project and scientific discovery at PPPL. More young and talented physicists joined the development team and raised the code to a higher capability level. Among them, Hager became another distinguished developer and physics researcher.

One of the most rewarding aspects for Chang is seeing his younger group members become successful computational physicists in their own right. His advice to young physicists is to think big. “Don’t be afraid to attack challenging and ambitious scientific problems,” he said.

Seung-Hoe Ku

Ku has been a research physicist at the Lab since 2011, following Chang’s move. He received his doctoral degree in physics from KAIST in 2004. 

Ku has been deeply involved in the research and development of the XGC code for decades, starting from when he was a graduate student at KAIST. Ku was effectively the sole person writing an initial version of what would one day become the backbone of XGC while he was still a graduate student. 

“This has been a lifelong pursuit,” Ku said. He has seen the code through many iterations, moving it from a two-dimensional code into three dimensions and adding code to include turbulence, for example.

“When I extended it to 3D, a few people came on board to help with code performance,” Ku said. Now, many people around the world are working on XGC, with Ku and Hager focusing on managing the core of the code. 

Ku has been interested in physics since middle school. In high school, he also developed an interest in coding. With some friends, Ku wrote what he describes as a precursor to the popular video game Angry Birds. “You throw the ball, and then it calculates the trajectory,” Ku said. “At the time, it was just for fun. But I think that’s my first physics simulation of particles.”

Ku would like to thank his wife, Haehyun Nam, for her patience.

Robert Hager

Hager received a doctoral degree in plasma physics jointly from the Technical University of Munich and the Max Planck Institute for Plasma Physics in 2011. The following year, he came to PPPL as a postdoctoral researcher. Hager has been working with Chang and Ku on XGC ever since. He became a core developer and is now a research physicist at the Lab. 

“Winning the Kaul Prize is confirmation that what we’ve been doing all those years actually makes sense and produces good results,” Hager said. People sometimes question why he would work so hard on a code that is so complex it can only run using the world’s most powerful computers. “Now, finally, I think more people are seeing our results and realizing we can reproduce what people are seeing in experiments and get better insights,” Hager said.

In addition to being one of the main authors and managers of XGC, Hager is also responsible for training and supporting XGC users worldwide.

Hager says the field was definitely the right choice for him. “As a scientist, you sometimes have long stretches where nothing seems to work. But when you find a solution, you understand something new, and that is so rewarding. I also like the technical aspect, tinkering with computer tools.”

Like many in the field, Hager was initially drawn to plasma physics because of the environmental aspect of clean energy from fusion. However, there were also personal factors: His position at Max Planck brought him closer to his girlfriend, who he would later marry. “I would like to thank my wife, Sofia, my Ph.D. supervisor Klaus Hallatschek and everyone who helped make XGC what it is today.”

END