My Community

For structural relaxation (ISIF=3) for the unit cell, there are two different ways in terms of the initial positions of atoms in the unit cell: atoms on the high symmetry points, and atoms on the positions slightly adjusted from the high symmetry positions. For the two cases, the relaxed structure and total energy are a little bit different with each other. As the very important step for subsequent calculations, I am wondering which method is more reasonable, or depends? Which parameter should be used to judge the reasonability? Thank you for your help.

You are right with two ways in terms of the initial positions of atoms.
May be the question should be asked in a different way.
Is it necessary to allow the variation of the cell shape when
the symmetry of the structure is well known (from XRD analysis)?

Yeah, should. Thank you.

I have a related question: when you check my another question about structural relaxation, you will find that all the forces in the y direction
are 0, while forces in the x and z directions are too big and not converged. Is this case has something to do with the symmetry? Thank you.