deepmodeling / abacus-develop

An electronic structure package based on either plane wave basis or numerical atomic orbitals.
http://abacus.ustc.edu.cn
GNU Lesser General Public License v3.0
162 stars 127 forks source link

Question: about the default value setting of parameter `ecutwfc` #4283

Open kirk0830 opened 3 months ago

kirk0830 commented 3 months ago

Details

Different from softwares which limit users' choice on pseudopotentials to use, for which developers can type in recommended ecutwfc values in their published pseudopotentials. However for community using upf pseudopotentials, this is not possible and reliable. Even as far as I know there are plenty pseudopotentials without such a line like "suggested ecutwfc" and/or "suggested ecutrho".

However sometimes user may forget to set ecutwfc, for cases where "hard" pseudopotentials are used in calculation, the default value 50 Ry may not be enough and result in highly unreliable data. Therefore the default value seems to be case-specific.

How to set the default value of ecutwfc?

Have you read FAQ on the online manual http://abacus.deepmodeling.com/en/latest/community/faq.html

Task list for Issue attackers (only for developers)

WHUweiqingzhou commented 3 months ago

I agree to improve the default value of ecutwfc to 100 Ry for LCAO basis, and ecutrho=4*ecutwfc for sure. As for PW bases, I think 100 Ry is quite large for some simple cases. Maybe we should keep the default value as 50 Ry.

kirk0830 commented 2 months ago

But a more reasonable plan would be, greping ecutwfc that used to generate orbitals and get max to set ecutwfc for LCAO calculation, while set element-by-element for PW. But due to present code structure, both solutions are not plausible.