Argonne takes on fusion’s hardest engineering problem

Atomic simulation of helium clustering in tungsten under fusion neutron damage, revealing embrittlement mechanisms Argonne targets with new composites and OpenMC modelling.

(image courtesy of Argonne National Laboratory)

Tungsten sounds like an odd hero for the fusion story, but right now it’s one of the most important materials in the field. When plasma temperatures climb past 100 million degrees Celsius inside a tokamak, the components facing that plasma don’t last. They crack, become brittle under neutron bombardment, and absorb impurities that contaminate the plasma itself. That degradation problem has quietly held back fusion development longer than almost any other factor.

Argonne National Laboratory is attacking it on two fronts. The first is a software tool called OpenMC, developed at Argonne together with MIT, which uses Monte Carlo methods to simulate how neutrons and photons move through complex reactor systems and damage materials. It predicts, for example, how quickly fuel is consumed or how much radiation damage reactor materials will sustain, enabling virtual experiments before costly prototypes are built. Optimised to run on supercomputers including the Aurora exascale system at Argonne’s Leadership Computing Facility, OpenMC can now simulate entire fusion devices at a level of detail that simply wasn’t possible before. The tool recently received an R&D 100 Award and has rapidly become a go-to resource across national laboratories, universities, and a growing roster of private fusion companies, with strong adoption across the US and Europe.

Because OpenMC is open source, scientists anywhere can both use and improve the code. In May 2025, Argonne hosted the OpenMC Application to Tokamak Neutronics Analysis Meeting in collaboration with EUROfusion, drawing around 90 international participants to focus on applying the code to devices including ITER, SPARC, DEMO, and STEP.

On the materials side, the challenge is well understood even if the solutions aren’t yet settled. Pure tungsten cannot fully satisfy the requirements for plasma-facing components in future high-duty cycle reactors. The extreme environments it must survive include radiation damage exceeding 100 displacements per atom, helium production above 1000 atomic parts per million, steady-state heat loads around 10 megawatts per square metre, and transient thermal shocks in the gigawatt range. Tungsten-based composites fabricated using spark-plasma sintering present promising solutions, and research into dispersoid-strengthened tungsten and tungsten-zirconium composites points toward optimisation paths for future reactor designs.

Together, the neutronics modelling and materials research represent the two sides of the same problem: understanding how components will fail before they fail in a real machine. The goal, in Argonne computational scientist Paul Romano’s framing, is that by the time hardware is built, engineers already understand its key behaviours. That iterative loop between simulation and materials development is where the real progress is being made.

Stay ahead in the fusion revolution explore more breakthroughs from leading innovators in clean energy technology.