Hi everyone, does someone know the work principles that VASP(The Vienna Ab initio simulation package) calculate elastic constant? I have checked many literature and they applied a matrix to unit cell, but I don't know if they are used by current version VASP(5.3.3). When I check OUTCAR, I can only find the elastic constants, but I cannot find where is the process, i.e, the matrix applied and the energy of strained lattice. Can someone tell me how to find those data, or introduce some literature about this?
Dear Yixian,
please check the description of IBRION=6 in the VASP manual:
"Finally, IBRION=6 and ISIF>=3 allows to calculate the elastic constants. The elastic tensor is determined by performing six finite distortions of the lattice and deriving the elastic constants from the strain-stress relationship. The elastic tensor is calculated both, for rigid ions, as well, as allowing for relaxation of the ions"
If you use this method you will find calculations for 6 different displacements and corresponding calculation in the OUTCAR file
A different version of this approach would be to use the universal independent coupling strain (ULIC) method - needs fewer strained states and can save some computation time, however you need to create strained configurations and determine Cij yourself. It is describe here:
R., Zhu J., Ye H. Calculations of single-crystal elastic constants made simple. Comput. Phys. Commun. 2010;181(3):671–675
Best, Holger
Dear Holger,
Thank you for your answer.
The initial purpose of this question is that I hope to plot a strain-energy curve by the data of OUTCAR. This is often a part of the papers when VASP had no function to calculate the elastic constants and strain tensor must be applied manually.
However, I have found that the calculation in current version VASP has been separated into cell relax (SYMMETRIZED ELASTIC MODULI) and ion-relax(ELASTIC MODULI CONTR FROM IONIC RELAXATION) part, and the total elastic constant is given by sum of them. Then it makes no sense to plot these curve, since the final result is the sum of them, neither of them can unilaterally represent the calculation.
I will try another method you mentioned.
Again, thank you very much!
Sincerely, Yixian
Dear Yixian,
there are two general approaches for elastic constants. One is via the stress-strain relationship, the other via strain-energy relationship. So in your case you would aim for the second approach which is not the one implemented in VASP. A good description of this method can e.g. be found in:
David Holec, Martin Friák, Jörg Neugebauer, and Paul H. Mayrhofer
Phys. Rev. B 85, 064101
Best, Holger
Dear Holger,
Many thanks to your suggestions!
I will read the articles you recommend.
Sincerely, Yixian
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering