band energy change upon lattice deformation

Message

dong913 · #1 Post by **dong913** » Mon Mar 05, 2012 4:56 am

Hi, all
We are doing 3D calculations, and we need to get band energy shift with lattice dilation, but the E-zero may change in each calculation. Anybody can give me suggestions to solve the problem?

Thanks.
Dong

dong913 · #2 Post by **dong913** » Mon Mar 05, 2012 6:00 am

One more question: is the totoal energy in OUTCAR or OSZICAR the absolute energy?

tlchan · #3 Post by **tlchan** » Thu Mar 08, 2012 9:20 am

[quote="dong913"]
is the totoal energy in OUTCAR or OSZICAR the absolute energy?
[/quote]
I think VASP subtracts the atomic energies away from the total energy, and the cohesive energy is printed. The total energy itself is an absolute energy for a neutral system. For a charged calculation, the total energy depends on the boundary condition of the system.

b_lack · #4 Post by **b_lack** » Thu Mar 08, 2012 12:32 pm

it's better to used another language as a controller like bash. Do a single step relaxation (NSW=1) and then calculate the band energy, and copy the everything from that folder to a new folder and replace the POSCAR with the CONTCAR from last result, and repeat single step relaxation again --> band energy again and copy again until the energy different is small or position different is small. And you'll have many forders for each relaxation step contain band energy.