Labo : http://irfm.cea.fr/en/index.php
Voir aussi : http://west.cea.fr/en/index.php
In a tokamak plasma, the current profile impacts strongly the turbulent transport, therefore the energy content. Maximizing the energy, hence the D-T fusion rate, is at the heart of research on fusion by magnetic confinement. Current profile is the easiest plasma parameter that can be shaped from the tokamak control room using external magnetic field coils as well as electromagnetic waves. To shape in real time the current profile, improved first principle based turbulent transport codes are needed. Since 2017, for the first time, a neural network regression of a quasilinear transport code is now available and integrated in a real time control suite.
The PhD student will apply these novel tools to
1) reproduce well diagnosed experimental current ramp up of the world largest tokamak, JET
2) validate the embedded turbulence code with respect to nonlinear codes over the, rarely explored, parametric domain of a current ramp up.
The iteration between first principle physics and experiment modelling will be carried out until convergence.
Then, on the WEST tokamak, in Cadarache, experiments will be designed and realized in order to optimize and control the current profile shape leading to improved plasma performances.