标题: Question about difference in energy [打印本页] 作者Author: zako 时间: 2020-2-27 18:04 标题: Question about difference in energy Hello everybody,
what's the difference between "Geometry optimisation" and "self-consistent" calculations? In terme of energy. 作者Author: sobereva 时间: 2020-2-27 18:28
The word "self-consistent" is quite ambiguous without context. If you refer to self-consistent field (SCF) calculation, first note that both geometry optimizaiton and SCF require performing iteration, while their natures are quite different. Geometry optimization gradually alters geometry in each iteration and thus electronic energy is also correpsondingly changed, while SCF gradually changes electronic energy while the geometry is kept unchanged. Note that in each step of geometry optimization, SCF process is needed to perform once.作者Author: zako 时间: 2020-2-28 06:10
But in some systems, SCF and geometry optimzation gives same energy but not for others, i don't know why ?作者Author: liyuanhe211 时间: 2020-2-28 10:14
zako 发表于 2020-2-28 06:10
But in some systems, SCF and geometry optimzation gives same energy but not for others, i don't know ...
It will be the same when and only when (1) the geometry for the calculations are already optimized (2) the computation levels are the same (3) the electronic structure converges to the same state (e.g. sometimes one needs to use guess=read) (4) the other miscellaneous settings are the same (e.g. integration grid for DFT, SCF convergence criteria, etc)