failure in DFT calculation

37 views
Skip to first unread message

J Giaccai

unread,
Feb 26, 2024, 7:57:10 PMFeb 26
to NWChem Forum
I have recently come across an error message that I'm not sure how to troubleshoot.  

I'm using DFT to find the orbitals/energies/etc of a non-covalently bound dimer system.

I understand that my calculation is not reaching the convergence criteria, but am not sure what I can modify/relax to help reach convergence.  I've attached an output file.  The input file can be seen echoed at the beginning of the output file.  Unfortunately, since I'm trying to see the orbitals, I have a pretty huge output file, so I've provided an edited version without the extra orbital print information to keep it a bit more manageable.  I can find a way to upload the giant file if that is needed.  

I have had some disk-write issues recently when running, but this seems to be a separate issue.  Although it is breaking just when I would expect it to write a Gaussian cube file.

Thanks, Jennifer
2648206_10_edit.log

Edoardo Aprà

unread,
Feb 27, 2024, 12:52:11 PMFeb 27
to NWChem Forum
Your DFT calculation was almost converged. You simply need to increase  SCF cycles (maxiter).
Another suggest is to switch to direct SCF since you seem to be running out of disk space.

dft
  ...
  maxiter 199
  direct
end

One more suggestion: you might want to switch to a spherical basis set instead of the default cartesian one, since it will make the calculation more numerically stable and faster.

BASIS spherical
 * library 6-31+G* 
END
Reply all
Reply to author
Forward
0 new messages