Search

Reactor relevant fusion devices will use tungsten (W) for their plasma facing components (PFCs) due to its thermomechanical properties and low tritium retention. However, W introduces high-Z impurities into the plasma, degrading its performance. Different wall conditioning methods have been developed to address this issue, including coating of W PFCs with layers of low-Z material. Wall conditioning by boron (B) powder injection using an impurity powder dropper (IPD) is being studied in WEST. Two series of experiments were conducted since the installation of the new ITER grade full W divertor. During the first series in 2023 ∼ 1 g of B powder was injected in total at a maximum rate of ∼ 58 mg/s, both of which are three times greater than respective values in the initial WEST powder injection experiments. The second series of experiments included injection of B and BN powders for comparison of their effects on plasma performance. The presence of an instantaneous conditioning effect is suggested by visible spectroscopy measurements of low-Z impurity lines and a rollover of total radiated power past an injection rate of ∼ 20 mg/s was observed. Presence of B coating layer formation is supported by the evolution of the average radiance of visible lines of B, W and oxygen (O). To understand B transport, an interpretative modeling workflow is employed, utilizing the SOLEDGE-EIRENE fluid boundary plasma code and the Dust Injection Simulator (DIS) code. Parameters like B perpendicular diffusivity and recycling coefficients are varied to match experimental results to see if the initial assumption of B sticking to the PFCs immediately after the contact with the wall is adequate for correctly modelling its distribution on the PFCs.

A fluid collisional-radiative model for H2 has been implemented in the edge-fluid code UEDGE and compared to the kinetic neutral code EIRENE on a simple, 2D, orthogonal domain with a constant, static plasma distribution. The novel CRUMPET Python tool was used to implement dissociation and energy rate coefficients that consider molecular-assisted processes, binding energy, and radiation due to molecular processes into the UEDGE fluid molecular model. The agreement between the fluid and kinetic molecular models was found to be within 20% when corresponding rates were used in UEDGE and EIRENE for a domain with absorbing boundaries. When wall recycling was considered, EIRENE predicted up to a factor of 2.2 higher molecular densities than UEDGE at T