Local linear gyrokinetic simulations : an overview of nuclear fusion, instabilities, and improvements

Fusion is a promising technology that has developed into a multi-disciplinary endeavor. Simulations of plasmas are critical to explore regimes of magnetically confined fusion in a way that would be unfeasible experimentally. Furthermore, simulations are useful to understand and interpret experimental observations. The gyrokinetic model is a computationally-tractable model for simulating the small scale turbulence that determined confinement in tokamaks. Gyrokinetic Electromagnetic Numerical Experiment (GENE) and QuaLiKiz (a quasilinear gyrokinetic tokamak transport model, QLK) are two codes that are capable to solve instances of gyrokinetics. While GENE is very comprehensive in terms of physics capabilities, it is much more computationally expensive. QuaLiKiz attempts to lighten the computational load by making several simplifying assumptions. This thesis shows results of modifications proposed on QLK to improve accuracy in a wider parameter space. I will compare simulations from GENE to the main branch of QLK as well as a modified branch and show that the modifications improve QLK’s precision in relation to GENE. Specifically, there is better handling of relatively large α [subscript MHD], and less overestimation of maximum growth rates.