Modeling dust in a tokamak
Modeling dust in a tokamak lead image
Components of fusion reactors facing the plasma can be eroded and create an accumulation of contaminants in the form of dust inside the reactor chamber. It is important to understand how the dust, if mobilized, may impact the reactor’s functioning.
Smirnov et al. have developed a modeling technique to study the time-dependent effects of dust on plasma in tokamaks. They couple two numerical simulation codes, DUSTT and UEDGE, to simulate dust trajectories as well as the evolution of background plasma parameters in the presence of this dust. Their modeling results for an ITER-scale tokamak show a number of unexpected features.
Their model followed transient injection of tungsten dust formed at the plasma edge where plasma exhaust comes into contact with the divertor. Besides the hydrogen ions, neon impurities were also included for managing the escaping power load to the reactor components.
The simulations showed that while dust ablation was localized, ionized tungsten could spread through the whole plasma volume.
“In contrast to the localized dust impurity source profiles, further propagation of the introduced impurities has global character,” said author Roman Smirnov.
The investigators observed that the fraction of impurity in the core increased in all simulations, reaching high enough levels that it could affect fusion conditions. Furthermore, the dust perturbed the power load to tokamak components; an initial reduction was followed by a significant increase.
“This unexpected result highlights the complex dynamics that tungsten dust injection in the presence of other impurities can produce,” said Smirnov.
Impurity seeding will be an important tool for reactor power management. Since dust could significantly perturb fusion plasma, it should also be taken into consideration.
Source: “Time-dependent modeling of dust outburst into tokamak divertor plasma,” by R. D. Smirnov and S. I. Krasheninnikov, Physics of Plasmas (2020) The article can be accessed at https://doi.org/10.1063/5.0009767 .
