I'm working with a system of 192 atoms. I've already relaxed the structure (with ISIF=2 and ISMEAR=0) and achieved the required accuracy. Then, as per the instructions in the VASP manual, I ran the SCF calculation. After that, I copied the obtained CHGCAR file to the DOS directory and attempted to run the calculation. However, every time I receive the following error message:
"WARNING: The chargedensity file is incomplete. ERROR: The charge density could not be read from the CHGCAR file for ICHARG>10."
I have double-checked that the file is copied to the correct directory, and it is complete (I can see the augmentation occupancies for all 192 atoms. Additionally, I have verified the dimensions (NGX, NGY, NGZ, NGXF, NGYF, and NGZF) in the CHGCAR file, and they are all accurate. Can you please assist me with this issue? The INCAR file is given below.
ISTART = 1 ISPIN = 2 ICHARG = 11 LREAL = A ENCUT = 520 PREC = Normal ISMEAR = -5 NELM = 60 NELMIN = 5 EDIFF = 1.0E-05 LORBIT = 11 NEDOS = 1000 NSW = 0 IBRION = -1 POTIM = 0.5
Hey Trivanni Yadav ,
According to your given information of INCAR, I think everything is fine except few points:
1. You can try with the same smearing for the relaxation and DOS calculation to get consistent results. Here you used Gaussian smearing (ISMEAR=0) for relaxation and ISMEAR=-5 for DOS calculations.
2. Try with the default POTIM tag.
3. Please check the memory for writing the CHGCAR as you are using the systems with 192 atoms. (This may not be a reason but please check this point too.)
Good luck !
Best,
Radhey
Thank you for the comment.
I've tried with ISMEAR = 0 and default POTIM tag. The warning is still the same :(
Can you specify lit bit more, about checking the memory thing? Sorry, I've never encountered with such weird issue that's why, I'm bit confused. Thanks in advance :) Radhey Shyam Yadav
Hi Trivanni Yadav
Please check once whether ENCUT is the same in SCF and DOS calculations.
Payal Wadhwa Thank you for your response. ENCUT is same in both of calculations.
I use ICHARG = 10 instead of 11
But now another error pops up:
ERROR FEXCF: supplied exchange-correlation table is too small, maximal index : 3855
Then I use ICHARG = 1, this is what I get (attached Screenshot)
Hi Trivanni Yadav
I believe this error might be related to some memory issues. You have 192 atoms, i never dealt with more than 110 atoms, so it might be possible whatever the number of cores you are using, and the memory assigned to them, might not be sufficient for your calculations.
Try to do scf by increasing the number of processors. Let's see whether CHGCAR incomplete error goes away or not in DOS calculations.
Also Trivanni Yadav , can you use smaller ENCUT, as that will reduce the memory requirement? From your screenshot, it is also clear that you are not using NCORE, you can also try with that, it may also help.
Thanks Payal Wadhwa I did some changes, ENCUT = 450, NCORE =8
Run SCF calculation and copied CHGCAR & CHG file in my DOS directory. But the error still persists.
Total cores = 112
Hi Trivanni Yadav
I recently faced the same error as you. then I found that LCHARG = .FALSE. was there in my scf calculations. Please check whether it is the case for you too?
If yes, then change it to true, and use this CHGCAR for DOS calculations.
Hope it will help!
Payal Wadhwa Thank you! But I have already tried this before, it didn't work as well. I guess this has to do with memory allocation. Though, I have tried everything, nothing works. Will probably look for other options.
I really appreciate your help!
- Badges
- Science topic
- Similar topics
- Counseling Psychology
- Confusion