Global Flux-Driven Simulations of Ion Temperature-Gradient and Trapped-Electron Modes Driven Turbulence with an Improved Multithreaded Gyrokinetic PIC Code

Performance of tokamak fusion plasmas is heavily linked to the radial heat and particle transport, which is known to be mainly produced by turbulence driven by micro-instabilities. Understanding such processes is thus of key importance for the design and operation of future fusion reactors such as ITER. One of the most commonly used tool to study turbulent transport in the very hot core of tokamaks is gyrokinetic theory, which aims at reducing the physical complexity by averaging out the fast gyromotion, while keeping its averaged effect. The \orb{} program is a numerical code solving the gyrokinetic Maxwell-Boltzmann system of equations to study various phenomena such as turbulent transport. It is a global particle-in-cell code using a finite element representation with B-spline basis function for the fields and various noise control schemes.

The main objective of this work is to improve the performance and physical model of \orb{} in order to be able to carry out simulations relevant for the Tokamak à Configuration Variable (TCV) at the Swiss Plasma Center. This is done in three major steps.

First, the code undergoes a complete refactoring and optimization process with the aim of increasing its maintability and performance. Using a simple test bed being a simplification of the PIC method while retaining the key elements, various algorithm and parallelization schemes are developed. Furthermore, it is shown that one of the most critical kernels of the PIC algorithm, the charge deposition, can be accelerated by a factor 2.4 in \orb{}.

Second, the hybrid electron model, which is used to simulate certain electron modes at a lower numerical cost than the fully kinetic electron model, needs to be corrected. Indeed, in this model, trapped electrons are kinetic while passing electrons have an adiabatic response. This is problematic because it does not satisfy the ambipolarity condition and produces spurious sources of particles and momentum. A corrected model in which the flux-surface averaged contribution of the passing electrons is treated as kinetic is developed and implemented in \orb{}. As a first step, an approximation of the flux-surface average consisting of a Fourier filter retaining only the $n=m=0$ mode is tested. Linearly, both the proper flux-surface average and its $n=m=0$ approximation lead to an accurate description of the geodesic acoustic mode dynamics. Nonlinearly, the $n=m=0$ model exhibit interesting physical features such as sharp radial structures at the mode rational surface $q=1$ and a transport regime transition.

Finally, a moment-conserving heat source is implemented in the \orb{} code to allow for flux-driven simulations. The heat source conservation properties are verified and it is shown that conservation of density, parallel momentum, and zonal flows is satisfied up to machine precision. A mode-switching'' approach consisting in using an effective heat source computed from a gradient-driven simulation as an input for the flux-driven run is also tested and consistent results are obtained between gradient-driven, flux-driven, and mode-switched'' simulations. Finally, the heat source is characterized, along with variations, using a 2D phase-space diagnostics. It is shown that in some cases, the use of this fixed heat source can lead to a sampling problem causing an artificial source/sink of moments.

Collisions in Global Gyrokinetic Simulations of Tokamak Plasmas using the Delta-f Particle-In-Cell Approach

Thibaut Vernay

The present work takes place within the general context of research related to the development of nuclear fusion energy. More specifically, this thesis is mainly a numerical and physical contribution to the understanding of turbulence and associated transport phenomena occuring in tokamak plasmas, the most advanced and promising form of magnetically confined plasmas. The complexity of tokamak plasma phenomena and related physical models, either fluid or kinetic, requires the development of numerical codes to perform simulations of the plasma behaviour under given conditions defined by the magnetic geometry as well as density and temperature profiles. The studies presented in this work are based on electrostatic kinetic simulations, taking advantage of a reduced kinetic model (the gyrokinetic model) which is particularly suitable for studying turbulent transport in magnetically confined plasmas, in effect solving an approximate form of the Vlasov equation for the distribution function of each species (electrons, ions) along with a reduced form of the Poisson equation providing the self-consistent electric fields. The main tool of this work, the gyrokinetic ORB5 code making use of numerical particles according to the Particle-In-Cell (PIC) method, has been upgraded during this thesis with different linearized collision operators related to both ions and electrons. The BIRDIE code, enabling to study collisional effects on the evolution of Langmuir waves in an unmagnetized plasma, has been written in order to serve as a test-bed for the collision operators ultimately implemented in ORB5. Some essential algorithms related to collisional simulations have been jointly implemented, such as the two-weight scheme which is extensively described in this work. The collision operators in ORB5 have been further carefully tested through neoclassical simu- lations and benchmarked against other codes, providing reliable levels of collisional transport. Together with different procedures controlling the numerical noise, the collision operators have then been applied to the study of collisional turbulent transport in two different regimes, the Ion-Temperature-Gradient (ITG) regime and the Trapped-Electron-Mode (TEM) regime re- quiring a trapped electron kinetic response. Although not dominant in core tokamak plasmas, collisional effects nevertheless lead to interesting modifications in the turbulence behaviour which are not captured by the often considered collisionless gyrokinetic models. The so-called coarse-graining procedure, a noise-control algorithm which is suitable for collisional gyrokinetic simulations with particles, is shown to enable carrying out relevant simulations over many col- lision times. Consequently, reliable conclusions regarding turbulent transport in the presence of collisions could be drawn in this thesis. Namely, the turbulent transport in the ITG regime is found to be enhanced by ion collisions through interactions with so-called zonal flows as- sociated to axisymmetric modes, while it is reduced by electron collisions in the TEM regime through electron detrapping processes. The zonal flow dynamics in collisionless and collisional ITG turbulence simulations is studied, emphasizing the limitation of the zonal flow level due to Kelvin-Helmoltz-type instabilities. Additionally, some purely collisionless issues related to tokamak physics are discussed, such as the finite plasma size effects in TEM-dominated regime which are found to be important in non-linear simulations but unimportant in linear simu- lations. The role of zonal flows in temperature-gradient-driven TEM turbulence saturation is confirmed to be weak, in agreement with previous studies. Finally, a realistic global gy- rokinetic simulation, accounting for a proper TCV tokamak magnetic equilibrium and related experimental profiles, has been successfully carried out thus demonstrating the relevance of the ORB5 code for predictions related to physics of real tokamaks. A good agreement with GAM experimental measurements is indeed obtained.

EPFL2013

Development and application of Eulerian and Particle-in-Cell gyrokinetic codes for studying the effect of non-adiabatic passing electron dynamics on microturbulence

Julien Stanislas Pierre Dominski

In tokamak fusion plasmas, micro-turbulence transport is known to be the cause of large losses of heat and particles. The present work deals with the study of electrostatic micro-turbulence transport driven by instabilities of essentially two types: the ion temperature gradient (ITG) modes and the trapped electron modes (TEM). The plasma is described within the gyrokinetic framework, which permits to save computational resources compared to the classical Vlasov kinetic description. In gyrokinetic simulations of fusion plasmas, the passing electrons are often assumed fast enough so that they respond instantaneously to the electrostatic perturbations. In this case, their response is computed adiabatically instead of kinetically. The main advantage is that this simplified model for the electron response is less demanding in computational resources. This assumption is nonetheless incorrect, in particular near mode rational surfaces where the non-adiabatic response of passing electrons cannot be neglected. This thesis work focuses on the study of this passing electron non-adiabatic response, whose influence on microturbulence is studied by means of numerical simulations carried out with the gyrokinetic codes GENE and ORB5. In the first part of this thesis work, the response of passing electrons in ITG and TEM microturbulence regimes is studied by making use of the flux-tube version of the GENE code. Results are obtained using two different electron models, fully kinetic and hybrid. In the hybrid model, passing particles are forced to respond adiabatically while trapped are handled kinetically. Comparing linear eigenmodes obtained with these two models enables one to systematically isolate fine radial structures located at corresponding mode rational surfaces, clearly resulting from the non-adiabatic passing-electron response. Nonlinear simulations show that these fine structures on the non-axisymmetric modes survive in the turbulent phase. Furthermore, through nonlinear coupling to axisymmetric modes, they induce radial modulations in the effective profiles of density, ion and electron temperature and zonal flows

E \times B

shearing rate. Finally, the passing-electron channel is shown to significantly contribute to the transport levels, at least in our ITG case. Also shown is that the passing electrons significantly influence the

E \times B

saturation mechanism of turbulent fluxes. Following this study in flux tube geometry, the influence of the non-adiabatic passing electron response near mode rational surfaces is further studied in global geometry with the global gyrokinetic code ORB5, in which a new field solver is implemented for the gyrokinetic quasi-neutrality equation valid at arbitrary wavelength, overcoming the former long wavelength approximation made in the original version of the code. A benchmark is conducted against the global version of the gyrokinetic code GENE, showing very good agreement. Nonlinear simulations are carried out with the new solver in conditions relevant to the TCV tokamak, with the physical deuterium to electron mass ratio (

m_i/m_e=3672

) and are compared to simulations carried out with heavy electrons (

m_i/m_e=400

). The particular spectral organization of the passing electron turbulent flux and its dependence on the radial profile of the safety factor are revealed. In particular, the formation of short-scale transport barriers is studied near low-order mode rational surfaces. Results show that quantitatively correct nonlinear fully-kinetic simulations of tokamak transport must be carried out in a full torus and with the physical mass ratio.

EPFL2016

Gyrokinetic particle-in-cell global simulations of ion-temperature-gradient and collisionless-trapped-electron-mode turbulence in tokamaks

Sébastien Jolliet

The goal of thermonuclear fusion research is to provide power plants, that will be able to produce one gigawatt of electricity. Among the different ways to achieve fusion, the tokamak, based on magnetic confinement, is the most promising one. A gas is heated up to hundreds of millions of degrees and becomes a plasma, which is maintained – or confined – in a toroidal vessel by helical magnetic field lines. Then, deuterium and tritium are injected and fuse to create an α particle and an energetic neutron. In order to have a favorable power balance, the power produced by fusion reactions must exceed the power needed to heat the plasma and the power losses. This can be cast in a very simple expression which stipulates that the product of the density, the temperature and the energy confinement time must exceed some given value. Unfortunately, present-days tokamaks are not able to reach this condition, mostly due to plasma turbulence. The latter phenomenon enhances the heat losses and degrades the energy confinement time, which cannot be predicted by analytical theories such as the so-called neoclassical theory in which the heat losses are caused by Coulomb collisions. Therefore, numerical simulations are being developed to model plasma turbulence, mainly caused by the Ion and Electron Temperature-Gradient and the Trapped-Electron-Mode instabilities. The plasma is described by a distribution function which evolves according to the Vlasov equation. The electromagnetic fields created by the particles are self-consistently obtained through Maxwell's equations. The resulting Vlasov-Maxwell system is greatly simplified by using the gyrokinetic theory, which consists, through an appropriate ordering, of eliminating the fast gyromotion (compared to the typical frequency of instabilities). Nevertheless, it is still extremely difficult to solve this system numerically due to the large range of time and spatial scales to be resolved. In this thesis, the Vlasov-Maxwell system is solved in the electrostatic and collisionless limit with the Particle-In-Cell (PIC) ORB5 code in global tokamak geometry. This Monte-Carlo approach suffers from statistical noise which unavoidably degrades the quality of the simulation. Consequently, the first part of this work has been devoted to the optimization of the code with a view to reduce the numerical noise. The code has been rewritten in a new coordinate system which takes advantage of the anisotropy of turbulence, which is mostly aligned with the magnetic field lines. The overall result of the optimization is that for a given accuracy, the CPU time has been decreased by a factor two thousand, the total memory has been decreased by a factor ten and the numerical noise has been reduced by a factor two hundred. In addition, the scaling of the code with respect to plasma size is presently optimal, suggesting that ORB5 could compute heat transport for future fusion devices such as ITER. The second part of this thesis presents the validation of the code with numerical convergence tests, linear (including dispersion relations) and nonlinear benchmarks. Furthermore, the code has been applied to important issues in gyrokinetic theory. It is shown for the first time that a 5D global delta-f PIC code can achieve a thermodynamic steady state on the condition that some dissipation is present. This is a fundamental result as the main criticism against delta-f PIC codes is their inability to deal with long time simulations. Next, the role of the parallel nonlinearity is studied and it is demonstrated in this work that this term has no real influence on turbulence, provided the numerical noise is sufficiently low. This result should put an end to the controversy that recently occurred, in which gyrokinetic simulations using different numerical approaches yielded contradictory results. Finally, thanks to the optimization of the code, the gyrokinetic model has been extended to include the kinetic response of trapped-electrons, in place to the usual adiabatic (Boltzmann) approximation. For the first time, global TEM nonlinear simulations are presented, and the role of the zonal flow on heat transport is analyzed. This study will help in acquiring some knowledge on the less-known TEM turbulence (as compared to ITG). In conclusion, this thesis is one of the main steps of the development of ORB5, which is now a state-of-the-art gyrokinetic code for collisionless ITG and TEM turbulence, and has brought several contributions to the understanding of these phenomena.

EPFL2009