Theoretical and Computational Fusion Edge Plasma Science
The development of future fusion reactors, whether tokamaks or stellarators, requires precise predictions regarding the stability of plasma operation and the intensity of plasma-wall interaction (PWI). The goal of theoretical fusion physics at IFN-1 is to mathematically describe the plasma-wall system and enable reliable calculations and predictions. However, this problem is characterized by a high degree of complexity, as a multitude of electromagnetic, fluid dynamic, kinetic, atomic physics, chemical, and surface physics processes interact, which also occur on very different temporal and spatial scales.
In practice, such a complex system can only be studied using computer-aided simulations, as a purely analytical description is not possible in most technologically relevant cases. However, if one were to attempt to use models based exclusively on elementary physical principles (first-principles models, single-particle simulations), even the fastest and most modern supercomputers would not be powerful enough. Simplifications, approximations, and additional assumptions are therefore always necessary to make numerical models of plasma physics and plasma-wall interactions practically usable. By comparing simplified and detailed models, conducting sensitivity studies, and—above all—comparing results with experimental observations and measurements, the IFN-1 develops suitable models that are then used for the research and design of specific reactor types.
One of the subproblems is the interaction between neutral particles and the plasma, which consists mainly of charged particles. This interaction is described by the so-called Boltzmann equation, which can be efficiently solved using Monte Carlo methods. This is also the case with the EIRENE code, which is maintained and further developed at IFN-1. Although this code is also used independently, e.g., for calculating wall erosion caused by fast atoms, it is primarily employed as a module within numerical software packages for integrated simulation.
An example of such an integrated package is the B2-EIRENE code, frequently used in fusion research, which describes the plasma boundary layer in tokamaks using a 2D approximation. Magnetized charged particles are treated using the fluid approximation, while neutral particles are modeled using the aforementioned kinetic Monte Carlo model. The first version of B2-EIRENE was largely developed at IFN-1 and later officially transferred to the ITER Organization, where it is now being continued and expanded as the SOLPS-ITER code—still with the support of IFN-1.
Another important research activity of IFN-1 in this area is the EMC3-EIRENE code. The EMC3 program also uses a Monte Carlo method to solve 3D fluid equations describing charged particles and is therefore also suitable for calculating plasma flows in 3D magnetic field configurations, e.g., in stellarators.
The use of state-of-the-art supercomputers is essential for performing calculations with these comprehensive software packages. In collaboration with the Supercomputing Center at Forschungszentrum Jülich (JSC), these numerical tools are continuously adapted and further developed for PWW applications.
In parallel with developments focused on numerical applications for stellarators and tokamaks, the aforementioned codes are also used for simulations of linear plasmas, such as those generated in the PSI-2 experiment at IFN-1. On the one hand, this enables a detailed numerical representation of the linear plasmas, which are used to investigate PWW processes in reactor-relevant operating ranges. On the other hand, the use of the same models as for tokamak and stellarator plasmas allows for comparability and verification of the underlying theories and methods.
Specific aspects of the work on these numerical tools include the development of plasma kinetic models for atomic and molecular processes in hot plasma, the derivation of reduced shock radiation models, the implementation of kinetic models for Coulomb interactions, and the detailed modeling of complex surface structures in plasma-wall interactions, with the use of machine learning methods to accelerate simulations also being investigated.
EMC3-EIRENE simulation of the electron and ion temperature profiles and the Mach number in the edge region of the ITER tokamak.
