Improving RAPTOR tokamak plasma profile simulations

: neutral beam injection module and automated model parameter optimization

Abstract

The q-profile, also known in literature as the safety factor profile, is an important quantity because it determines the stability as well as the performance of a fusion plasma. A strong degree of control over this q-profile makes the operation of a tokamak fusion reactor more stable (avoidance of plasma instabilities) and efficient (optimal q-profile shape). Control of the q-profile can be achieved through either open-loop (feedforward) control or closed-loop (feedback) control, or a combination of both. For the development and testing of both types of control a fast control-oriented transport code which can simulate the evolution of the safety factor profile in time is crucial. The RAPTOR (RApid Plasma Transport simulatOR) code was developed with this aim in mind. The performance of the open-loop and closed-loop control system strongly depends on the quality of the RAPTOR predictions.During this research the predictions of RAPTOR were improved by adding a fast Neutral Beam Injection module (NBI) and by developing a generic method to estimate the model parameters in RAPTOR.The NBI system is an important external heating and current drive actuator to alter the q-profile of a fusion plasma. It is used at many existing tokamaks and will be an important actuator for ITER. The developed fast Neutral Beam Injection module, based on a pencil beam approach, produces results similar to more complete beam codes for an ITER like scenario.In RAPTOR an adhoc-model is used, instead of a first principle physics model, to describe the electron heat diffusivity in view of computational speed. The structure of the adhoc-model is given by the physics knowledge, and only the unknown physics of which is more complicated and less well understood is captured in its model parameters. During this research on the one hand the adhoc-model was extended in order to better describe physical phenomena such as the sawtooth instability and the degradation of transport due to higher temperature. On the other hand a generic method to estimate RAPTOR's model parameters was developed. For the TCV tokamak in Lausanne it was shown that the developed method is capable of finding the model parameters such that the RAPTOR predictions agree well with measurements.As a result of the work presented in this thesis the RAPTOR code is now equipped with a NBI module that allows RAPTOR to simulate a multitude of tokamaks including ITER. Benchmarks of this fast NBI module showed good agreement with large scale NBI codes, while running significantly faster.Furthermore the extension of the transport model and a newly developed model-parameter estimation routine now results in a better description of the physics and allows for a less adhoc and more automated method to implement RAPTOR on a variety of tokamaks.

Date of Award	30 Nov 2013
Original language	English
Supervisor	F.A.A. Felici (Supervisor 1) & M.F.M. de Bock (Supervisor 2)