A comparison of real SEM tungsten microstructures (left column) with machine learning–generated synthetic microstructures (right) for different values of the model setting parameters. Adjusting the model setting controls how diverse or sharp the synthetic microstructures appear. (Image: ORNL, DOE)

Researchers have developed a model to generate images that serve as synthetic data close-ups of tungsten surfaces under fusion reactor conditions.

Tungsten is a top-choice material for plasma-facing components (PFCs) in fusion machines, so understanding tungsten’s performance is critical to the safety and longevity of component designs.

Researchers at General Atomics (GA) are proposing a breeding blanket made of modular silicon carbide–based components to withstand the intense conditions in a high-power fusion power plant. The GA modular blanket (GAMBL) concept is described in an article published this month in the journal Fusion Engineering and Design, and was introduced by GA in a July 13 press release.

Researchers at the DIII-D National Fusion Facility (DIII-D) are preparing to test a new method that could enable future fusion power plants to withstand the heat and particle flow created by the fusion reaction, General Atomics reported this week.