Top Page > Browsing

relation between scf.energycutoff and scf.Ngrid Date: 2013/09/15 16:44 Name: dchrobak   <dariusz.chrobak@us.edu.pl> Dear All, I'd like to ask you about relationship between scf.energycutoff and scf.Ngrid input file parameters. I encountered that if you fix scf.Ngrid to certain value, the change of scf.energycutoff does not influence the result of calculations (Utot). However, if you are going to make a several experiments with different lattice constant (keeping scf.Ngrid constant) you deal with different "Used cutoff energy (Ryd) for 3D-grids". From the other side, we know that obtained result depends on energy cutoff parameter. Concluding, it seems that keeping constant scf.Ngrid one can obtain the energy v. lattice constant curve as a result of calculations with different energy cutoff. Am I right? BR, DCh.

Page: [1]

Re: relation between scf.energycutoff and scf.Ngrid ( No.1 ) Date: 2013/09/17 23:10 Name: T. Ozaki Hi, > Concluding, it seems that keeping constant scf.Ngrid one can obtain the energy v. lattice > constant curve as a result of calculations with different energy cutoff. Am I right? You are right. The cutoff energy varies in fact when you keep the number of grids by using scf.Ngrid, and change the lattice constant. Since it is known that sudden change of grid points tends to produce discontinuity in the energy curve, we recommend to keep the number of grids rather than keeping the cutoff energy when the total energy vs. lattice constant are calculated. Regards, TO

Page: [1]