Researchers from the University of Sydney are modelling a form of nuclear fusion to understand how it can one day be used as a source of energy.

Dr Ben Thornber says "Inertial Confinement Fusion provides a potential pathway towards utilisation of fusion power as an energy source." This method of producing fusion involves using lasers to heat a small capsule of the Hydrogen isotopes Deuterium and Tritium to the required pressure and temperature for it to implode and fuse the atoms into Helium.

Fusing two lighter atoms into a heavier one releases huge amounts of energy, but the pressure and temperature required to get there are similar to those found in the centre of the sun. Working with such extreme conditions makes physical experiments very difficult, so simulations on supercomputers are used to assist the experimental research being performed around the world.

Dr Ben Thornber, along with Dr Markus Flaig, is using Raijin to model the interaction between the various layers of the capsule as they gets compressed in the experiment. "The NCI facility is critical to our research. We use up to sixteen thousand cores on Raijin to generate our results."

Often, mixing between the capsule's layers stops the core from getting hot enough to fuse. He says, "Our research aims to understand how the instabilities develop and transition to turbulence, and thus inform the future design of the fuel capsules."

"We run computations with up to 16 billion points in total," says Dr Flaig, "This enables us to develop simplified models which substantially improves our understanding of mixing and the subsequent impact on the fusion process."

Even with that many points, it is not always easy to include all the physical details to look into. The challenge, says Dr Thornber, "is to define a case which includes all of the key physics involved in the real problem, but that we can solve on today's supercomputers. This involves a judicious simplification of the problem to its core essence."

Powering our homes with nuclear fusion power plants is still decades away, but work like this helps us understand how we are going to get there while learning a lot about nuclear physics and fluid dynamics in the process.