Nuclear Fusion Simulations at Exascale - Nu-FuSe
Zusammenfassung der Projektergebnisse
The Nu-FuSE project’s aim was to significantly improve computational modelling capabilities that significantly enhance the predictive capabilities needed to address key physics challenges of a new generation of reactorrelevant fusion systems. Commensurate with “path to exascale” progress in supercomputers, we focussed on three specific scientific areas – fusion plasmas; radiation damage; and the plasma edge – that require extreme scale computing across a range of simulation codes that benefited from interdisciplinary research engaging Applied Mathematicians and Computer Scientists. Scientific Highlight – New core physics findings indicate more efficient burning plasma: For over a decade, both experimental observations and theoretical simulations of turbulent losses of fusion-grade tokamak plasmas have indicated that energy confinement degrades as the size of the tokamak increases in the so-called “Bohm regime.” However, such simulations have also predicted that for sufficiently large tokamaks there will be a turnover point into a “Gyro-Bohm regime,” where the losses become independent of system size. For burning plasma devices such as ITER, it is of key importance that systems can operate in this favorable Gyro-Bohm regime. A modern gyro-kinetic code (GTC-P) capable of utilizing world-leading HPC platforms – such as the “Mira” and “Sequoia” BG/Q supercomputers in the U.S. and the “K-computer” in Japan – has been developed by the Nu-FuSE project to achieve unprecedented phase-space resolution enabled by successfully running on over 1.5 million cores. Associated core physics studies have been carried out with the discovery that the magnitude of turbulent losses in the Gyro-Bohm regime can be up to 50% lower than indicated by earlier much lower-resolution simulations and that the Bohm to Gyro-Bohm transition is much more gradual as the plasma size increases. This finding was made possible only after going to high-resolution, long-time scale simulations needed to achieve the physics fidelity enabled by computing at extreme scales. Scientific Highlight – New plasma edge physics capability enabled and coupled to core model: Modeling the plasma-edge region poses an immense challenge because of the combination of sharp gradients characteristic of the associated density and temperature profiles coupled with geometrically complex (boundary) wall elements. The Nu-FuSE project has implemented a new computational paradigm based on a mesh-free multi-pole algorithm for handling kinetic plasma-wall physics and demonstrated its suitability for path to exascale computing architectures by scaling the code to a record 458,752 cores (1.8M threads) on the BG/Q “JUQUEEN” supercomputer at the JSC in Germany. This provides an important scalable computational means of simulating the burning core plasma dynamics together with the material science of the vessel wall within an integrated modeling framework. Scientific Highlight – New materials simulations identify key catalytic self-healing properties: Commercial fusion reactors will produce energy in the form of very high energy neutrons which can cause enormous radiation damage to the structure of such systems. Since this damage is inevitable, what is needed are a new generation of materials that can self-heal. However, different materials can have hugely different responses to neutron irradiation, and there is no macroscopic theory for these differences. Since experiments with irradiated samples are costly, screening via computer simulations is an essential part of the process. Moreover, with millions of atoms involved, codes that can exploit exascale systems will be essential. One class of materials which show good radiation resistance are the so-called "ODS steels." The Nu-FuSE project has identified the associated reason – a catalytic recombination of topological defects at the interface between iron and yttrium oxide particles. Now that the mechanism is understood, a systematic pathway to improving on existing materials can be devised. Scientific Highlight – New mathematical algorithms developed to facilitate extreme scale computing: A new kinetic theory approach has been developed for gas dynamics based on the Boltzmann Equation. The associated evolution equation follows naturally from the relationship between the kinetic and the gas dynamics description of continuous media. Observables such as density, momentum, and energy flux can then be obtained as the moments of the distribution function. The Nu-FuSE project has shown that the electrodynamics of charged particles can be captured by this new mathematical model – an approach with promise of facilitating applications to extreme scale computing. Since such an approach combines kinematic and field concepts of statistics, the distribution function correctly takes into account the essential vector behavior which drives the evolution of the electromagnetic fields.
