LAECHG tag to get all-electron density

Message

ssharifz · #1 Post by **ssharifz** » Mon Mar 26, 2007 11:00 pm

Hello,
I would like to get the all-electron density from a PAW calculation. I find that the LAECHG tag set to true, gives core electron density that is strongly dependent on the FFT grid size. As I increase the size of the grid, the charge density seems to approach the correct value (in my case, it should be 74, but is 73.88 at 100x100x500), but starts to oscillate between 72 and 73.5 as I increase the grid size further. Could you please give me more information on the LAECHG keyword and how to converge my calculations?
Thank you very much!

VietBac · #2 Post by **VietBac** » Wed Jul 02, 2008 4:35 am

Dear Vasp admin and users,
I have the same problem with the convergence of the core electrons by using FFT grid. My system is a cluster placed in the quite large cubic supercell. The ADDGRID tag seems to not support the good FTT grid for my system. My system includes 130 core electrons, by playing around FFT grid in INCAR I always get 111, 135 or 190. Eventhough I sometimes changed the ENAUG value also.
I appreciate any help and suggestion.
Thank you

#3 Post by **admin** » Wed Jul 02, 2008 1:40 pm

please check if the k-point sets are converged, increase PREC and decrease the mixing parameters