Three-dimentional kinetic modelling of edge plasmas

In recent years, edge-plasma modelling has become one of the most challenging tasks in plasma physics, the reason being that edge plasmas exists in any plasma device and can strongly influence, or even define, the overall discharge performance. This topic is especially important for designing and optimizing present-day and next-generation magnetic confinement fusion devices, such as ITER and DEMO. A number of problems to be tackled require three- dimensional fully kinetic treatment. The aim of the proposed project is to develop a realistic three-dimensional kinetic model of fusion-relevant edge plasmas, including nonlinear dynamics of plasma, impurity and neutral particles, as well as plasma-surface interactions. The corresponding numerical code, BITL, will be installed and run on a massively parallel platform, thus enabling simulation of large-scale plasma systems with finest resolution in time and space (down to electron gyro-motion and plasma oscillations). Development of this model will require collection of atomic, molecular and plasma-surface interaction data, development of corresponding optimized Monte Carlo routines, and implementation of a highly scalable fast three-dimensional Poisson solver. This (to our knowledge) unique self-consistent model will be applied to studying a number of relevant fusion- related edge-plasma problems. Among others, we intend to answer a long-standing question: What is the structure of the edge plasma in front of a rough wall surface in the presence of different cross-field drifts and of an arbitrary magnetic-field configuration. Moreover, we intend to develop reproductive and predictive models of divertor-plasma simulators such as Pilot-PSI and Magnum-PSI, NAGDIS-II and others. These efforts will be helpful to understanding a number of unexplained experimental observations, predicting the actual particle and heat loads to the plasma-facing components, and estimating the impurity net sputtering rates in realistic conditions. We expect that the results obtained in this project will contribute to further optimization of the design of both ITER and post-ITER machines (especially DEMO). The three-dimensional kinetic code to be developed under this project will represent a powerful tool for realistically studying fully and partially ionized plasmas. It will be applicable in practically any branch of plasma physics ranging from low-temperature laboratory plasmas all the way to astrophysical ones. We intend to perform this project in collaboration with our colleagues in Austria, the EU, Japan and the USA.

Controlled thermonuclear fusion has a perspective to be an environmentally friendly, intrinsically safe and potentially unlimited energy source. In the magnetic confinement thermonuclear fusion reactor the energy is generated in an extremely hot plasma, the fourth state of the matter, of up to 100 million degrees of Celsius captured by the strong magnetic field. The temperature of the plasma in front of the reactor walls should not exceed few tens of 1000. Therefore, plasma has to be cooled down very effectively during its motion to the wall, which is a complex process. Correspondingly, the study of plasma properties near the wall represents one of the top important tasks in fusion plasma research. Due to the complexity of the problem, the main tools for this study represent sophisticated numerical codes. The absolute majority of such codes use so called fluid model of plasma operating on averaged Velocity Distribution Functions (VDF) of plasma particles. This approach leaves out a number of important properties of the plasma edge. The solution is to develop more complex numerical tools operating directly on VDF so called kinetic codes. The aim of the given project was to i) develop a three-dimensional kinetic code for simulation of fusion-relevant edge plasmas, and ii) study properties of boundary plasma using a kinetic model. Such code, BIT3, has been developed. It incorporates few hundreds of different types of elastic and inelastic nonlinear interactions between plasma, neutral particles and the wall of the plasma device, and enables the development of probably the most complete plasma edge models. The code has been successfully tested on more than 104 processors on the Marconi supercomputer. The development of BIT3 required the introduction of a number of know-hows in numerical optimization, in calculation of atomic and plasma-surface interaction data and their implementation into the kinetic codes, as well as in the method of kinetic plasma modelling. Parallel with the BIT3 development, we have performed the modelling of the fusion relevant plasma edge with already existing tools (codes BIT1 and BIT2), developed by us under previous FWF projects. We have simulated existing fusion plasma devices, so called linear divertor simulators (e.g. PSI-2) and tokamaks (e.g. JET), as well as the next generation tokamak ITER. Among the main findings of this study I will mention the following two. We demonstrated that during the optimized operations of the tokamak, the main carriers of the power to the heat resistant wall elements, divertors, represent not the thermal plasma particles, as it was previously assumed, but so called super-thermal electrons. They originate from the upstream plasma and flow downstream in a collisionless manner. Another important finding is the new type of plasma edge diffusive sheath, which will appear in next generation tokamaks and might significantly change plasma edge properties and plasma-wall (divertor) interaction there.