Hello,
I am trying to relax a slab heterostructure that is very difficult to converge (even for a single self-consistent step). From what I can tell in the output, it seems like the forces on atoms are oscillating a huge amount from ionic step to ionic step, or even increasing. Unfortunately the OUTCAR is too large to attach here, but hopefully the output file also gives some information about the steps.
I attach the input file as well as the output file in case anyway has a suggestion for minimization algorithm, step size, or anything else that I should try in order to more efficiently converge to a relaxed structural state. As an FYI,I imagine a suggestion will be to turn off spin-orbit coupling for the relaxation, but if at all possible I'd prefer not to do this, since it seems to change the relaxed parameters quite a bit in previous tests.
Thank you in advance for your help, let me know if I can provide anymore information.