In a paper published July 25, in the journal Scientific Reports, researchers from Tri Alpha Energy (TAE), the world’s largest private fusion company, and Google have jointly furthered the understanding of physics and computational design, applying a new approach to solving highly complex problems regarding plasma behavior.
TAE and Google have a multi-year partnership that dates back to 2014. The two organizations joined forces in the hopes that machine learning algorithms could advance plasma research, bringing us closer to fusion power, a clean energy source. The complexity of the research into plasma, a gas in an ionized state, so hot that the atoms are blown apart into their constituent nuclei and electrons, led to TAE turning to Google for help.
Producing nuclear fusion
Producing nuclear fusion on Earth involves the fusion of a deuterium atom (an isotope of hydrogen, containing a single proton and neutron but no electron) with a tritium (a different isotope of hydrogen with one proton and two neutrons and two electrons outside) one, writes Digital Journal’s science guy, Tim Sandle.
The enormous amount of data generated in the research had so many variables that TAE’s researchers needed some kind of advanced computing networks to help wade through it all. But even Google didn’t have all the computational resources needed, reports New Atlas.
“The reality is much more complicated,” explains Ted Baltz, from Google’s Accelerated Science Team, “as the ion temperature is three times larger than the electron temperature, so the plasma is far out of thermal equilibrium, also, the fluid approximation is totally invalid, so you have to track at least some of the trillion+ individual particles, so the whole thing is beyond what we know how to do even with Google-scale computer resources.”
Fusing man and machine
Google’s team came through with a solution, though. Fusing man and machine, so to speak, the solution is called the “Optometrist Algorithm.” It takes it name from the person who does an eye examination. The algorithm sorts through a multitude of simulated settings that would keep the plasma under check.
Human technicians can then select the best candidates for using in their experiments, rather than going through the time-consuming process of testing each one. “So we boil the problem down to let’s find plasma behaviors that an expert human plasma physicist thinks are interesting and let’s not break the machine when we’re doing it,” says Baltz.
In the study published in Scientific Reports, The new algorithm unexpectedly gave the TAE research team a 50 percent reduction in energy loss rate and a concomitant increase in ion temperature and total plasma energy in its field-reversed configuration plasma generator.
“Results like this might take years to solve without the power of advanced computation to rapidly scale our understanding of the complex properties of plasma,” says Michl Binderbauer, TAE’s President, and Chief Technology Officer.
Divertor vessels, which are located at both ends of the C-2U device, house the plasma guns and bias electrodes – the key instruments for plasma stability control.
Tri Alpha Energy is not the only company involved in high profile fusion research. The UK’s Tokamak ST40 reactor was fired up for the first time in April 2017 and should reach a temperature of 100 million degrees Centigrade sometimes next year, the temperature required to trigger nuclear fusion.
