“…This is in contrast to our previous findings [22], where significant reductions in the peak stress and work of separation were found. This discrepancy is likely due to the fact that atomic relaxations were allowed in [22], while they were prevented for the validation herein -both for the DFT and classical modelling. In fact, for iron and nickel GBs it has been found that rigid and relaxed GB separation modelling can give contradictory results and that atomic relaxation is necessary to quantitatively capture the impact of impurities on the GB strength [52,78].…”