QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: ams_nanolab on February 14, 2017, 15:51

Title: Total energy and energy cut-off
Post by: ams_nanolab on February 14, 2017, 15:51

Is there any correlation between total energy of a system and the energy cutoff used in the simulations. Say I want to calculate the total energy of a large (50-60 atoms) supercell would there be any difference if I use a cutoff of 300eV in one case and 600eV in other for my LDA-PZ calculations?

Title: Re: Total energy and energy cut-off
Post by: Ulrik G. Vej-Hansen on February 14, 2017, 16:22

Just to be sure, you are referring to the density mesh cutoff, right?

The short answer is yes, there is a correlation, though if 300 eV is converged, then the difference to 600 eV should be insignificant. However, note that the default unit is Hartree, not eV. Those values are fairly low for DFT, where the default value is 75 Hartree.

Title: Re: Total energy and energy cut-off
Post by: ams_nanolab on February 15, 2017, 16:34

By converged you mean the error is reduced below a certain minimum tolerance value right? Or something else?

Title: Re: Total energy and energy cut-off
Post by: Ulrik G. Vej-Hansen on February 15, 2017, 19:50

Yes, that is exactly what I mean :)