This thesis focuses on the improvement of the Hermite-Fourier spectral method for solving kinetic plasma problems. In the first part of the thesis a novel dynamically adaptive techniques for changing the number of Hermite modes and Hermite basis are presented. Preconditioning of the problem and use of high-end scientific toolkit PETSc are discussed. The technique of changing the number of modes and preconditioning are believed to reduce computational time, while a change of basis improves robustness and numerical convergence of the method. The second part of the thesis focuses on paralellization strategies and performance analysis of the code implemented in the Fortran programming language.

Plasma physics governs the area of interactions between charged particles. As 99% of the visible universe is in a plasma state, it is an important topic in astronomy and space physics, where we already at an altitude of 60 km reach the plasma environment surrounding our planet in the form of the ionosphere. The search for fusion, the source of power for the sun, as well as industrial use have been the main topics for earth bound plasma reasurch. A plasma is composed of charged particles which interact by the electromagnetic force. In the kinetic description, via the Vlasov-Maxwell equations, the system is described in terms of probability distribution functions for each particle species, expressed in terms of particles position and velocity. The particles interact via self-consistent fields as determined by Maxwell's equations. For understanding the complex behaviour of the system, we need numerical solvers. These come in two flavours, Lagrangian methods, dealing with the moving around of synthetic particles, and Eulerian methods, which solve the set of partial differential, Vlasov and Maxwell equations. To perform the computations within reasonable time, we need to distribute our calculations on multiple machines, i.e. parallel programming, with the best possible matching between our computational needs and the need of splitting algorithms to adapt to our processing environment. Paper I studies electron and ion beams within a Lagrangian and fluid model and compare the results with experimental observations. This is continued with studies of a full kinetic system, using an Eulerian solver, for a closer look at electron-ion interactions in relation to ionospheric observations, (Papers II and IV). To improve the performance of the Eulerian solver it was parallelised (Paper III). The thesis is ending with the possibility to observe ultrahigh energy neutrinos from an orbiting satellite by using the Moon's surface as a detector Paper V.

No description available.

L'amélioration du code Emedge3D (code de bord électromagnétique) est abordée sous plusieurs axes. Premier axe, des innovations sur les méthodes numériques ont été mises en oeuvre. L'avantage des méthodes de type semi-implicite est décrit, leur stabilité inconditionnelle permet l'augmentation du pas de temps, et donc la diminution du nombre d'itérations temporelles requises pour une simulation. Les avantages de la montée en ordre en espace et en temps sont détaillés. Deuxième axe, des réponses sont proposées pour la parallélisation du code. Le cadre de cette étude est proche du problème général d'advection-diffusion non linéaire. Les parties coûteuses ont tout d'abord été optimisées séquentiellement puis fait l'objet d'une parallélisation OpenMP. Pour la partie du code la plus sensible aux contraintes de bande passante mémoire, une solution parallèle MPI sur machine à mémoire distribuée est décrite et analysée. Une bonne extensibilité est observée jusque 384 cœurs. Cette thèse s'inscrit dans le projet interdisciplinaire ANR E2T2 (CEA/IRFM, Université Aix-Marseille/PIIM, Université Strasbourg/Icube). / The main goal of this work is to significantly reduce the computational cost of the scientific application Emedge3D, simulating the edge of tokamaks. Improvements to this code are made on two axes. First, innovations on numerical methods have been implemented. The advantage of semi-implicit time schemes are described. Their inconditional stability allows to consider larger timestep values, and hence to lower the number of temporal iteration required for a simulation. The benefits of a high order (time and space) are also presented. Second, solutions to the parallelization of the code are proposed. This study addresses the more general non linear advection-diffusion problem. The hot spots of the application have been sequentially optimized and parallelized with OpenMP. Then, a hybrid MPI OpenMP parallel algorithm for the memory bound part of the code is described and analyzed. Good scalings are observed up to 384 cores. This Ph. D. thesis is part of the interdisciplinary project ANR E2T2 (CEA/IRFM, University of Aix-Marseille/PIIM, University of Strasbourg/ICube).

Investigations of soft matter using ultrashort high intensity pulses have been made possible through the advent of X-ray free-electrons lasers. The last decade has seen the development of a new type of protein crystallography where femtosecond dynamics can be studied, and single particle imaging with atomic resolution is on the horizon. The pulses are so intense that any sample quickly turns into a plasma. This thesis studies the ultrafast transition from soft matter to warm dense matter, and the implications for structural determination of proteins.                    We use non-thermal plasma simulations to predict ultrafast structural and electron dynamics. Changes in atomic form factors due to the electronic state, and displacement as a function of temperature, are used to predict Bragg signal intensity in protein nanocrystals. The damage processes started by the pulse will gate the diffracted signal within the pulse duration, suggesting that long pulses are useful to study protein structure. This illustrates diffraction-before-destruction in crystallography. The effect from a varying temporal photon distribution within a pulse is also investigated. A well-defined initial front determines the quality of the diffracted signal. At lower intensities, the temporal shape of the X-ray pulse will affect the overall signal strength; at high intensities the signal level will be strongly dependent on the resolution. Water is routinely used to deliver biological samples into the X-ray beam. Structural dynamics in water exposed to intense X-rays were investigated with simulations and experiments. Using pulses of different duration, we found that non-thermal heating will affect the water structure on a time scale longer than 25 fs but shorter than 75 fs. Modeling suggests that a loss of long-range coordination of the solvation shells accounts for the observed decrease in scattering signal. The feasibility of using X-ray emission from plasma as an indicator for hits in serial diffraction experiments is studied. Specific line emission from sulfur at high X-ray energies is suitable for distinguishing spectral features from proteins, compared to emission from delivery liquids. We find that plasma emission continues long after the femtosecond pulse has ended, suggesting that spectrum-during-destruction could reveal information complementary to diffraction.