My Community

I have a question but it rests on my understanding of the last page (page 26) of "M. MARSMAN, HANDS ON (4): MAGNETISM" @ https://www.vasp.at/vasp-workshop/slides/handsonIV.pdf

The title of that page is: "What can one do when convergence is bad?"

I take 'bad convergence' in this context to mean: the VASP calculation finishes, the time taken is printed to the end of OUTCAR, all the tolerances in INCAR were met... required accuracy was reached etc. etc. etc.

...but the moment information VASP reports is not the correct moment.

This is what is meant by the title for page 26 of "M. MARSMAN, HANDS ON (4): MAGNETISM"

Is my understanding correct?

If yes, then:
How are we supposed to know --as we try the things suggested on page 26 of "M. MARSMAN, HANDS ON (4): MAGNETISM"-- whether we have 'converged' to the correct magnetic solution or not (perhaps the answer is that we don't know... we simply have the answer that our theory, at the given settings, provides... then we must compare to experiment)

Is this the bottom line?

Thanks!

Bad convergence means that particular electronic step cannot fulfill convergence criteria and run until NELM. This repeats in each ionic step until NSW and finally requrired accuracy is not reached.
Converged calculations provide magnetic moments. Correct or not correct - the answer gives the comparison with experiment.