Saturday, October 12, 2024

 SCI-FI-TEK

New AI models of plasma heating lead to important corrections in computer code used for fusion research



Researchers find an effective alternative to overcome modeling limitations using machine learning



DOE/Princeton Plasma Physics Laboratory

Álvaro Sánchez Villar, an associate research physicist at the U.S. Department of Energy’s Princeton Plasma Physics Laboratory 

image: 

Álvaro Sánchez Villar, an associate research physicist at the U.S. Department of Energy’s Princeton Plasma Physics Laboratory, has developed new AI models for plasma heating that increase the prediction speed while preserving accuracy and providing accurate predictions where original numerical codes failed. 

view more 

Credit: Michael Livingston / PPPL Communications Department




New artificial intelligence (AI) models for plasma heating can do more than was previously thought possible, not only increasing the prediction speed 10 million times while preserving accuracy, but also correctly predicting plasma heating in cases where the original numerical code failed. The models will be presented on October 11 at the 66th Annual Meeting of the American Physical Society Division of Plasma Physics in Atlanta.

“With our intelligence, we can train the AI to go even beyond the limitations of available numerical models,” said Álvaro Sánchez-Villar, an associate research physicist at the U.S. Department of Energy’s (DOE) Princeton Plasma Physics Laboratory (PPPL). Sánchez-Villar is the lead author on a new peer-reviewed journal article in Nuclear Fusion about the work. It was part of a project that spanned five research institutions.

The models use machine learning, a type of AI, to try to predict the way electrons and ions in a plasma behave when ion cyclotron range of frequency (ICRF) heating is applied in fusion experiments. The models are trained on data generated by a computer code. While much of the data agreed with past results, in some extreme scenarios the data wasn’t what they expected.

“We observed a parametric regime in which the heating profiles featured erratic spikes in rather arbitrary locations,” said Sánchez-Villar. “There was nothing physical to explain those spikes.” 

New artificial intelligence (AI) models for plasma heating can do more than was previously thought possible, not only increasing the prediction speed 10 million times while preserving accuracy, but also correctly predicting plasma heating in cases where the original numerical code failed. The models will be presented on October 11 at the 66th Annual Meeting of the American Physical Society Division of Plasma Physics in Atlanta.

“With our intelligence, we can train the AI to go even beyond the limitations of available numerical models,” said Álvaro Sánchez-Villar, an associate research physicist at the U.S. Department of Energy’s (DOE) Princeton Plasma Physics Laboratory (PPPL). Sánchez-Villar is the lead author on a new peer-reviewed journal article in Nuclear Fusion about the work. It was part of a project that spanned five research institutions.

The models use machine learning, a type of AI, to try to predict the way electrons and ions in a plasma behave when ion cyclotron range of frequency (ICRF) heating is applied in fusion experiments. The models are trained on data generated by a computer code. While much of the data agreed with past results, in some extreme scenarios the data wasn’t what they expected.

“We observed a parametric regime in which the heating profiles featured erratic spikes in rather arbitrary locations,” said Sánchez-Villar. “There was nothing physical to explain those spikes.” 

“This means that, practically, our surrogate implementation was equivalent to fixing the original code, just based on a careful curation of the data,” said Sánchez-Villar. “As with every technology, with an intelligent use, AI can help us solve problems not only faster, but better than before, and overcome our own human constraints.”

As expected, the models also improved the computation times for ICRF heating. Those times fell from roughly 60 seconds to 2 microseconds, enabling faster simulations without notably impacting the accuracy. This improvement will help scientists and engineers explore the best ways to make fusion a practical power source.

Other researchers on the project include Zhe Bai, Nicola Bertelli, E. Wes Bethel, Julien Hillairet, Talita Perciano, Syun’ichi Shiraiwa, Gregory M. Wallace and John C. Wright. The work was supported by the U.S. Department of Energy under Contract Number DE-AC02-09CH11466. This research used resources of the National Energy Research Scientific Computing Center (NERSC) operated under Contract No. DE-AC02-05CH11231 using NERSC Award FES m3716 for 2023. 


Department of Energy announces $49 million for research on foundational laboratory fusion



Projects address scientific gaps foundational to enabling fusion energy



DOE/US Department of Energy




WASHINGTON, D.C. - As the Department of Energy (DOE) continues to accelerate a clean-energy future that includes fusion technology, a total of $49 million in funding for 19 projects was announced today in the Foundational Fusion Materials, Nuclear Science, and Technology programs.

The purpose of the funding is to reorient the laboratory-based foundational and basic science research programs to better align and support the new FES program vision.  

“The Fusion Nuclear Science Foundational research program, in enabling research and development and furthering research in fusion nuclear science and fusion materials, is vital to addressing critical scientific gaps foundational to enabling fusion energy,” said Jean Paul Allain, DOE Associate Director of Science for Fusion Energy Sciences.

Fusion energy holds the potential to revolutionize the world’s energy supply by providing a virtually limitless, clean, and sustainable power source. Unlike current nuclear power, which relies on splitting atoms (fission), fusion mimics the process that powers the sun by combining atomic nuclei to release massive amounts of energy.  

Fusion produces no long-lived radioactive waste, emits no carbon dioxide, and uses abundant fuels like hydrogen. If harnessed successfully, fusion energy could provide a safe, reliable solution to meet global energy demands while significantly reducing the environmental impact of power generation.

The projects funded under this initiative cover a wide range of cutting-edge research areas, each crucial to the development of fusion energy technology. For instance, scientists are testing new magnet designs that will help control the extremely hot plasma needed for fusion.

Other teams are working on materials that can withstand the damage caused by plasma, ensuring that systems used for maintaining the plasma remain functional and efficient. Some researchers are investigating blanket materials, which are designed to absorb heat from the plasma and turn it into usable energy, while also studying how these materials can be made durable enough to function in such an extreme environment.

Additionally, efforts are being made to improve fuel cycle systems, which help maintain the continuous flow of the fuel necessary for fusion reactions. Finally, advanced structural materials are being developed to construct stronger, more durable components that can endure the harsh conditions inside the fusion systems. Together, these projects aim to advance our understanding and capability in the pursuit of clean, sustainable fusion energy.

The projects were selected by competitive peer review under the DOE Lab Call: Opportunities in Foundational Fusion Materials, Nuclear Science, and Technology.  

Total funding is $49 million for projects lasting up to three years in duration, with $7 million in Fiscal Year 2024 dollars and outyear funding contingent on congressional appropriations. The list of projects and more information can be found on the Fusion Energy Sciences program homepage.

Selection for award negotiations is not a commitment by DOE to issue an award or provide funding. Before funding is issued, DOE and the applicants will undergo a negotiation process, and DOE may cancel negotiations and rescind the selection for any reason during that time.   



No comments:

Post a Comment