Reproducing Elastic Constants


#1

Hi,

I"m trying to reproduce the elastic constants for Aluminum. I downloaded the “VASP” package but I don’t think this is enough to reproduce the elastic properties.

Is there any means by which I can find out how the elastic properties were calculated? I downloaded the source materials but they did not seem sufficient/correct for elastic properties.

Can anyone point me to how I can reproduce the elastic modulus?

PS I have a general method set up on my end, but my results don’t agree with those of the materials project. Based on the literature, the materials project results seem more correct than my own, but I can’t seem to reproduce them…


#2

You can find details on the implementation of the current batch of elastic properties here. Additionally, it’s often non-trivial to set up, but our implementation is explicitly available in our open-source atomate software, which we use to organize and automate all of our simulations now.

Note that Al is notoriously difficult to get correct. In our own database, we have a somewhat customized calculation in that we increased the k-point cutoff significantly (I believe to 40000 kpts/atom) to find an appropriately converged tensor. At some point, we’re hoping to have a “calculation details” page for calculations like these which show explicitly the structure and the intermediate results relevant to a given derivative property, but it’s not available yet.

Please follow up here if there’s anything that should be clarified further.


#3

Thanks this worked out well. Another thing we realized was that you should increase the smearing for Aluminum. The default smearing of 0.2eV is rather low and you can increase to 1.0eV, substantially easing the KPOINT convergence. (This is the value the VASP people recommend for aluminum as well http://cms.mpi.univie.ac.at/vasp/guide/node159.html)