Code/Code comparison of impurity transport in EMC3 and DIVIMP

Recently, the global impurity migration code WallDYN was coupled to the 3D edge transport code EMC3-Eirene in order to make WallDYN applicable to 3D, non-toroidally symmetric, geometries. As part of the coupling, a new impurity transport module was implemented, which, in contrast to the existing impurity module, takes the inertia of the impurity into account, making it more suitable for heavy impurities like W. This new 'kinetic' impurity transport module uses the same physics as the 2D trace impurity transport code DIVIMP and therefore, to check the correctness of the implementation, a code/code benchmark was performed comparing the transport of Be- and W-impurities in EMC3 and DIVIMP. The comparison was performed for toroidally symmetric plasma that could be handled both by EMC3 and DIVIMP. Both codes were setup such that they would treat the transport of the impurities as similar as possible. Then the re-distribution of Be and W launched from different positions around the poloidal circumference of the first wall was calculated in both codes. To compare the resulting material re-distribution data ($\equiv R_{i,j}$) and judge the significance of any differences, $R_{i,j}$ was used in WallDYN calculations that model the evolution of the first wall due to erosion, deposition, re-erosion and re-deposition. These WallDYN calculations, performed using the $R_{i,j}$ from EMC3 or DIVIMP, showed the same major erosion and deposition areas and it can therefore be concluded that the new kinetic impurity transport module is implemented correctly and now allows us to make DIVIMP-like transport calculations using EMC3 in 3D. (topical issue article)