Temperature-Pressure Dependent Elastic Tensors in VASP.

[hszhao.cn@gmail.com]

Dear VASP Forum,

Can VASP's built-in functionalities calculate elastic tensors under specific temperature and pressure conditions? If so, what is the recommended approach?

Thank you,
Zhao

[jonathan_lahnsteiner2]

Dear hszhao,

You can sample elastic constants at finite temperature and pressure from molecular dynamics simulations. Currently there are no implementations which will automatize this for you. But you could for example
run NVT runs where you have a simulation box which is under strain in a certain direction. Then you can use the linear stress-strain relationship to compute the elastic tensor. You have to make a NVT run for every component of the elastic tensor. Maybe this paper is also interesting for you https://www.nature.com/articles/s41598-022-17666-2

In VASP it is possible to get elastic tensors from a finite difference approach. By setting IBRION=6 or 7. Check this wiki page and the pages linked there for further information. This will in principle be a zero Kelvin approach.But with POTIM you can adjust the magnitude of the displacements. So you could model the temperature effects by adjusting POTIM to the average root mean square displacements of the atoms at a given temperature. The root mean square displacement you would have to sample again from a molecular dynamics run. Still this method will be much easier to do compared to sampling the elastic constants directly from MD.

All the best Jonathan