Hello there !!!
I'm new to VASP and I do hybrid calculation using PBE0 functional. In the relax calculation, I saw a WARNING message of not getting self-consistency and I stuck with some confusion. I start doubting in my input files of not getting the right tags that I had used so far. So, here I upload the INCAR file and a picture of my output. Please kindly help me out of this error.
In addition, I'd be thankful If you'd make me clear on:
- How can we verify ourselves of using right tags for our system, if there is no experimental data?
- How can we choose right functional for our system?
Thanks !
There is no single answer that works for every case, you just have to follow the recommendations of increasing NELM or switching ALGO. However, you are doing a geometric relaxation. It's not always a problem if some of the early geometric steps do not reach electronic self-consistent convergence, as long as they get fairly close to convergence. A very bad electronic solution can lead to atoms getting thrown all over the cell by unrealistically large forces. But if it's close to convergence, and the atoms only move a small amount for the next geometric step, then hopefully the next step will converge properly. The final energy will be correct* as long as the last geometric step reaches self-consistent convergence.
*assuming you are not letting the volume change during relaxation. if you are, you'll need to do a final one-step static calculation ISIF=0 to avoid basis set change error.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publications