Description
In experimental fusion devices, plasma-wall interaction (PWI) plays an important role in the operation and planning, since the lifetime of plasma-facing components (PFCs) and the impurity concentration in the core are critical aspects of the reactor design. Simulations with the 3D Monte-Carlo code ERO2.0 assess these PWI processes and impurity transport of sputtered particles throughout the scrape-off layer of fusion devices. Local changes in the surface concentration of PFCs are considered by the application of a Homogeneous Mixing Model (HMM) in case more than one chemical element is present as PFC material in the device.
With the original HMM implementation in ERO2.0, a violation of the particle balance is possible: in net deposition conditions, only the topmost surface interaction layer (usually chosen to correlate to the penetration depth of incoming fuel/impurity fluxes, i.e. some nanometres) is considered for surface concentration changes. Once accumulated material exceeds this layer thickness, the excess material is not accounted for. This is in generally only relevant in specific cases of ERO2.0 applications, since ERO2.0 is normally applied in steady-state plasma conditions. In this case, local erosion and deposition zones are not expected to change over time, i.e. deeply deposited material is not likely to re-erode. However, erosion of thin initial layers of material on top of a bulk material or transitions from deposition to erosion zones cannot be accurately described by this model.
An updated version of the HMM is introduced in this work: a flexible buffer layer is introduced, which lies below the surface interaction layer and stores any particles deposited below. Upon erosion of the surface interaction layer, material is first replenished from the buffer layer, before bulk material is considered. This ensures correct particle balance as it was in the original ERO1.0 code. With a flexible initialization method of the buffer layer, an important application of this improved HMM is the modelling of finite boron layers on tungsten. Therefore, the improved model finds application in modelling for in full-tungsten fusion devices like ITER, where regular boronizations are planned.