Hi Fateme Molajafari

First, I want to say that depending on how big your system is, it may take days or even weeks for your calculation to converge.

In order for us to help you, can you share your INCAR and POSCAR and tell us how long your calculations have been running?

To see the evolution of your GO and if it is on its way to converge, here us a little trick for you. On your terminal, run the command:

grep entropy= OUTCAR

it will show you all the calculated energies after each iteration to this point. Then plot the iteration (x axis) vs energy (y axis) in a spread sheet. If your graph seems to end up in an horizontal line, chances are there is nothing wrong with your calculation.

Dear Domoina Holiharimanana

Thank you for your reply. My system is pretty big with 290 atoms. All the forces are below 0.02 eV/Angstrom, still, it won't stop. It has been running for 250 ionic steps.

Please find below my INCAR:

PREC = Accurate

ENCUT = 600

ALGO = All

ISMEAR = 0

SIGMA = 0.05

ISIF = 2

IBRION = 1

NFREE = 20

NELMIN = 8

NELM = 150

EDIFF = 1E-7

EDIFFG = - 0.03

NSW = 25

POTIM = 0.015

NWRITE = 1

LCHARG = .FALSE.

LWAVE = .FALSE.

LAECHG = .FALSE.

It's strange strange that your job has been running for 250 ionic step, because in your INCAR file, NSW=25 which means once 25 ionic step is reached, the calculation just ends even if it did not converge. It is possible that the 250 are electronic steps, not ionic steps.

Also your ENCUT is pretty high which makes your GO computationally costly.

190 atoms is huge, and I am not sure DFT can handle that many atoms

What type of system are you working on? A MOF? COF? A biomolecule?

Can you attach your POSCAR as well?

Domoina Holiharimanana

I meant 250 ionic steps in total. Each time after 25 ionic steps, I used CONTCAR as a POSCAR and continued the Job.

My system is H-ZSM5 (2 Al) zeolite

Fateme Molajafari

I have a system with 101 atoms with 36 of them held fix and my GOs ran for about 6-8 days on average, so I cannot say there is anything wrong with your calculations. For the system I worked on, my NSW was 500 which is enough ionic steps to get things done in a week or so.

I am not familiar with porous systems but here's my two cents:

• I attached the SI and INCAR file of the following paper

  Article Structural Basis of CO2 Adsorption in a Flexible Metal-Organ...

  You can look at their parameters and compare them with yours
• I do believe your ENCUT is way too high. If you do not wanna do a cutoff energy convergence test, a safe bet is to use ENCUT = max(ENMAX)x1.3 Your max(ENMAX) is probably O, which is 400. your ENCUT should be 520
• I think your POTIM is also too small. You may wanna increase it to 0.05 or even 0.1
• Your NSW is also not large enough. If you look at the NSW they used in the attached files, their NSW is 999

Also, I want to reiterate that some calculations can run for a very long time, even an entire month and depending on the number of atoms(or electrons) in your system, that's perfectly normal.

To get your job done quicker, something you may want to do is increasing the number of cores you are using for your calculations and maybe 2 nodes if you can.

Good luck

Dear Domoina Holiharimanana

Thank you very much for your Help!

I should use indeed smaller ENCUT and higher POTIM.

Thanks!