Numerical modelling of metallic melt damage in contemporary tokamaks and future fusion reactors
Abstract
The European Union recognizes fusion energy as the most promising long-term solution for clean virtually unlimited energy, with the construction of ITER and engineering design of EU-DEMO being short- and mid-term objectives of the EUROfusion roadmap. The provision of plasma-facing components (PFCs) with sufficient lifetime constitutes one of the major technological challenges in the development of magnetic confinement fusion reactors. In ITER, the PFC integrity is primarily threatened by fast transient power loading owing to magnetohydrodynamic (MHD) instabilities such as edge-localized modes (ELMs), vertical displacement events (VDEs) or major disruptions (MDs). In DEMO, the PFC heat load capability could be surpassed during regular plasma ramp-up/-down and H-L transitions, major disruptions or vertical displacement events. The extreme heat loads will generate metallic melt that will be subject to strong plasma-induced volumetric forces that will displace the material, causing large-scale surface deformations that may severely compromise power handling. Moreover, off normal events can be accompanied by the production of relativistic intense runaway electron (RE) beams that may lead to deep melting and PFC explosions. Hence, the reliable modelling of melt and explosive events constitutes a key topic in contemporary fusion research.
MEMENTO, developed and maintained by the Complex Plasmas Group of the SPP division of KTH, is the only code capable of simulating the self-consistent motion of induced metallic melts, thus providing realistic surface deformation profiles. MEMENTO solves the incompressible resistive thermoelectric MHD set of equations within the shallow-water and magnetostatic approximations coupled with the heat convection-diffusion equation. The MEMENTO predictive capability has been successfully tested in multiple dedicated experiments that achieved PFC melting in a controlled manner and has been near-exclusively employed for PFC melt predictions in ITER & DEMO. Benchmarking experiments concerned different plasma loads, tokamaks, materials, geometries, cooling types and electrical connections.
We request for computational resources necessary for: • further MEMENTO benchmarking against experiments on sustained melt bridging of castellated tungsten PFCs: (i) the WEST L-mode leading edge experiment with active cooling, (ii) the ASDEX Upgrade H-mode sloped experiment with inertial cooling; • predictive MEMENTO runs for different ITER scenarios: (i) toroidal gap edge melting by ELMs, (ii) melt damage on the divertor dome and outer baffle by unmitigated downward current quenches, (iii) transient melt damage on first wall panels; • predictive MEMENTO runs for different DEMO scenarios (focusing on the W upper limiter damage due to VDEs); • predictive runs of the Geant4-MEMENTO chain for different ITER scenarios with refined RE input from the JOREK code (varying loaded energies, loading times, PFC thickness, cooling type); • predictive runs of the Geant4-MEMENTO chain for different SPARC scenarios with crude RE input (scans of loaded energies, loading time, pitch angle, energy distribution); • further benchmarking of the Geant4-COMSOL chain for RE-induced brittle failure modelling against accidental RE-driven explosions of boron nitride in WEST; • first benchmarking of a Geant4-LSDYNA chain for RE-induced fragmentation modelling against deliberate RE-driven explosions of graphite in DIII-D;
