Page 1 of 1

Getting a fully relaxed or optimized result

Posted: Thu Jun 22, 2023 5:59 pm
by subhomoy_gupta
For a structure to be fully relaxed or optimised, I need to get a single set of iterations of 'DAV:' in the OSZICAR file. But, even though I am using the new CONTCAR positions in the POSCAR while running it again and again, I am unable to get this single set of iterations. Can anybody please suggest what I should do?

Re: Getting a fully relaxed or optimized result

Posted: Thu Jun 22, 2023 7:30 pm
by fabien_tran1
Your description of the problem is not really clear. Furthermore, you need to provide more information about your calculation, as explained at forum/viewtopic.php?f=4&t=17928

Re: Getting a fully relaxed or optimized result

Posted: Fri Jun 23, 2023 3:21 pm
by subhomoy_gupta
What I am trying to ask is this: I am trying to converge a particular system in my case which is FeNi. It is a ferromagnetic system and at first, I need to converge a supercell of FeNi system with 54 atoms. I am using IBRION = 2 and ISIF = 3 to achieve complete volume relaxation. In order to understand when the system has completely relaxed, I was told by my supervisors that unless and until I get a single set of 'DAV:' lines in the OSZICAR file, my system has not been fully relaxed, instead it has converged to a local minima. My question is, how mandatory is it to achieve this single set of DAV: lines in the OSZICAR file? Also, if I do not get this, are my calculatins completely wrong?
Also, what parameters or tags do I need to change in my INCAR file in order to get this single set of these lines in OSZICAR?
I am attaching my INCAR, OUTCAR, POTCAR, KPOINTS and OSZICAR files to shed some more light on my question.
Feni_files.zip

Re: Getting a fully relaxed or optimized result

Posted: Fri Jun 23, 2023 3:49 pm
by fabien_tran1
Your calculation seems to have finished properly, since at the end of the OUTCAR it is written "reached required accuracy - stopping structural energy minimisation". If you grep for "Iteration" in OUTCAR you can see that three ionic steps (the first number) were necessary to achieve structural relaxation.

This is not true that structural relaxation is reached only when a single "DAV:" (or "SDA:" or "CGA:") is obtained. One of these is printed at each iteration of the electronic (not structural) relaxation. For instance, during the last (third) ionic step, 10 electronic iterations were necessary.

Re: Getting a fully relaxed or optimized result

Posted: Fri Jun 23, 2023 7:13 pm
by subhomoy_gupta
I was not asking whether a single DAV: iteration but only 1 set of the DAV: iterations. In my case, I have got aroung three such sets. What I mean is, at the end of this each set, the Total energy (E0), magnetization (mag) are printed at the end of each iteration. But, since there are three such sets of DAV: iterations in my case, there are three such lines, so among the three lines at the end of each set (where E0, mag are mentioned) which of the values should I choose?

Re: Getting a fully relaxed or optimized result

Posted: Fri Jun 23, 2023 8:21 pm
by fabien_tran1
You have to consider the results obtained at the end of the calculation of course. That means the values printed at the end of OUTCAR. If you compare the three total energies, you can see that the last one is the most negative (it corresponds to the structure obtained after completion of structure relaxation).