Hello dear colleagues! I want to find the best parameters o methodology to relax a bulk structure (cubic material) using the commercial DFT software called VASP at the ground state. I have been reading carefully the manual of vasp, but some doubts are still coming to my mind.
Firstly, I performed a loop of single calculations to determine the best ENCUT and KPOINT, and subsequently using those parameters, I tried to relax the structure in two different ways (relaxation I, relaxation II). After relaxation II, the energy of the cell was the lowest but the lattice parameter became distorted. (i.d. a=b~c). Am I wrong??? How do you do it??
All opinions are welcome!
Hi Julio Cesar Gonzalez
In 1st relaxation method, you are changing only the position of atoms (ISIF = 2), not the cell size. If you opt method 1, then you need to see the change in ground state energy as a function of the lattice parameter, like you did for ENCUT and KPOINTS. The one having the minimal energy, will give you the optimized lattice parameter. You can first change 'c' by keeping some optimal guess on 'a', then once 'c' is optimized, you can keep that value of 'c' to optimize 'a'. At the end, you can do a final relaxation with optimized values of a and c. In addition, I suggest you to use ISMEAR = 0 here too.
Second method works well for 3D systems, but not 2D, as ISIF = 3 allows to change of the position of atoms as well as the cell parameter. just running one calculation, you will get optimized cell parameters here. But, you need to check whether convergence criterion is reached or not, if not, increase NSW to 100. And if your system is not much complicated, you can decrease the SIGMA value, as 0.1 is a little high for metals. you can keep it as 0.05.
Hope this helps!
Since you dont want the cell shape to change for your cubic system, you can first relax the volume using ISIF=7, then do only ionic relaxation using use ISIF=2 for that fixed volume.
Tapasendra Adhikary Payal Wadhwa Thanks for your contribution!
- Badges
- Science topic
- Similar topics
- Filing