First predictive simulations for deuterium shattered pellet injection in ASDEX Upgrade

Abstract

First simulations of deuterium shattered pellet injection (SPI) into an ASDEX Upgrade H-Mode plasma with the JOREK MHD code are presented. Resistivity is increased by one order of magnitude in most simulations to reduce computational costs and allow for extensive parameter scans. The effect of various physical parameters onto MHD activity and thermal quench (TQ) dynamics is studied and the influence of MHD onto ablation is shown. TQs are obtained quickly after injection in most simulations with a typical duration of 100 microseconds, which slows down at lower resistivity. Although the n=1 magnetic perturbation dominates in the simulations, toroidal harmonics up to n=10 contribute to stochastization and stochastic transport in the plasma core. The post-TQ density profile remains hollow for a few hundred microseconds. However, when flux surfaces re-form around the magnetic axis, the density has become monotonic again suggesting a beneficial behaviour for runaway electron avoidance/mitigation. With 10 21 atoms injected, the TQ is typically incomplete and triggered when the shards reach the q=2 rational surface. At a larger number of injected atoms, the TQ can set in even before the shards reach this surface. For low field side injection considered here, repeated formation of outward convection cells is observed in the ablation region reducing material assimilation. Responsible is a sudden rise of pressure in the high density cloud when the stochastic region expands further releasing heat from the hot core. After the TQ, strong sheared poloidal rotation is created by Maxwell stress, which contributes to re-formation of flux surfaces.

The JOREK non-linear extended MHD code and applications to large-scale instabilities and their control in magnetically confined fusion plasmas

“…In this regime, the TQ is typically triggered when shards reach the q = 2 surface, but it can be triggered even before for large and slow pellets. This behaviour has been observed in ASDEX Upgrade deuterium SPI simulations [263] and is also typically observed for any tokamak when simulating the injection of shattered pellets containing impurities [80].…”

“…This consideration is important for comparisons or extrapolations between machines of different sizes. It explains partly the different behaviour found in JET [79,87] and ASDEX Upgrade [263] deuterium SPI simulations. Second, simulations in which the resistivity is artificially increased (which is often done for numerical reasons) should be considered with caution because the current decay time will be artificially shortened.…”

Comparing spontaneous and pellet-triggered ELMs via non-linear extended MHD simulations

“…The JOREK code is a versatile MHD code (see section 4 of reference [19] for an overview) that is used to non-linearly solve the reduced and full MHD equations. The reduced MHD model with various extensions [20,21] has been used for this work.…”

“…However, the field has evolved rapidly since the publication of that article. A recent overview of ELM and ELM control simulations with JOREK is given in section 5 of [19].…”

Comparing spontaneous and pellet-triggered ELMs via non-linear extended MHD simulations

“…The JOREK code is a thoroughly tested MHD code (see Section 4 of Ref. [19] for an overview) that is used to non-linearly solve the reduced and full MHD equations. The reduced MHD model with various extensions [20,21] has been used for this work.…”