I am currently performing a self-consistent field (SCF) calculation in VASP, but unfortunately, the calculation stopped unexpectedly after completing 54 iterations without converging. I would like to know how I can resume the calculation from where it stopped, rather than starting from the beginning.
Hi,
if the calculation is aborted without writing restart information (e.g. the WAVECAR, vaspwave.h5, CHGCAR file, etc), there is no way to salvage the calculation:/
If you know that the calculation will be interrupted, e.g., due to a scheduled outage or limited wall time, you can use a STOPCAR file to intervene beforehand.
If restart files are present you can select different modes of restating. See ICHARG and ISTART.
Does this answer your question?
Marie-Therese
yes i know we can start from previous CHGCAR and WAVECAR file with ISTART=1 nad ICHARGE=11 to read previous data,,i am not clear is it correct or not?my calculation will be correct or not?and remeber it is SCF calculation and terminated after 54 iterations i want to start where it terminated. thanks
Sorry, can I confirm if I understand your question: You have a CHGCAR and WAVECAR which were written after 54 iterations. Now you want to restart and you wonder if it will be the correct SCF result?
If this is your question, restarting a calculation does not always go to the same solution, but usually.
Here is a pathological example: There might be a symmetry preserving and a symmetry broken solution. Whether the SCF calculation converges to one or the other may critically depend on the initial conditions. In either case after reaching convergence, the solution is "correct". It depends on your task if you are looking for the symmetry preserving or symmetry broken solution.
Marie-Therese