calculating elastic constants

[verners]

Dear users,

I was wondering if there was an option for calculating the elastic constants without calculating the Hessian (as with IBRION=6 and ISIF>3).

[karthik112490]

Well. You could manually impose strains on the cells to calculate the constants. For eg. in a cubic symmetry, if you were calculating c11, apply a stain in one direction keeping the others constant. Calculate c11 by U= U0 + 0.5*c11*(Strain^2)*volume. C44 can be done by applying a stain on the top face. ( ie the lattice vector for the one direction becomes 1.0 0.0 c ( where c is the strain)). Calculate B either by manually applying stains in all directions or by calculating from the Murghnaghn equation of state. 3B = c11 + 2c12 . Hope this helps

[maartendft]

Or just use stress-strain directly, from the stress tensor in VASP. No need to use the energy!