What parameter affect the total energy?

Message

zhuanghl · #1 Post by **zhuanghl** » Wed Jan 18, 2006 3:37 pm

What parameter affect the total energy?Except some key parameters,such as Cutoff energy,K-point,FFT grid.Well,I mean that for example,I set different Mixing parameters which all could make the system to converge(Of course,other parameters are kept the same).Should the total energy in this case be the same or not?
Thanks!

konglt · #2 Post by **konglt** » Wed Jan 18, 2006 4:15 pm

I guess they should be the same.

tjf · #3 Post by **tjf** » Wed Jan 18, 2006 4:34 pm

My understanding is that if the parameters are related to the electronic iteration (as opposed to the expression of the density/wavefunction or the determination of an energy from that density/wavefunction), then changing those parameters does not affect the final energy if the electronic energy converges.

Anyone else care to comment on this?

zhuanghl · #4 Post by **zhuanghl** » Thu Feb 16, 2006 10:34 am

Could someone else give more detail?Thanks!

#5 Post by **admin** » Fri Feb 17, 2006 10:22 am

I recommend to read
1) the online-handbook, chapter 9: "The most important parameters, source of errors"
2) the tutorial-file "accuracy.pdf" of the VASP-wokshop, which you can download from http://cms.mpi.univie.ac.at/vasp-workshop
they summarize all parameters which affect the calculated total energies
(choice of POTCAR, k-mesh, BZ-integration, PREC, ENCUT, ENAUG,...)
Mixing should never affect the converged total energies themselves, only the
convergence rate.