Which of the following is the more optimal method to determine crystal structures?

[Leon1999526]

Hello, I have a question about determining the crystal structure of materials.
I have read many papers in which VASP is used for computation of crystal structure, and the structures are identified through the following process: First, Birch-Murnaghan equation is used to determine cell volume, and then parameters determining the ionic positions in the lattice is adjusted to find the structure of lowest energy. I think IBRION=2 and ISIF=3 can also do the same thing but it seems to be less frequently used than the method I mentioned previously. Does ionic relaxation have any significant disadvantage at determining the optimal crystal structure than the other method?

[martin.schlipf]

The more parameters you optimize at the same time, the harder it gets to relax to the global minimum. It can be more robust to relax volume, shape, and ionic positions separately but then you potentially need multiple iterations until everything is converged. For simple systems using ISIF = 3 should be fine.

[Leon1999526]

Thanks. It seems that my system requires multiple steps of convergence.