Kinetic Terms

[cw-tan]

Hi, I'm pretty new to Julia and DFTK but I was hoping to use DFTK for orbital-free benchmarking. I scanned through the documentation but wasn't sure if it had what I wanted.

After getting the results from a self-consistent field calculation, is there a way to access the positive-definite kinetic energy density as defined here, the kinetic energy contribution to the stress, and the Kohn-Sham potential? Thank you!

[antoine-levitt]

Hi and thanks for your interest!

• The kinetic energy density you can compute it rather easily from the orbitals. See eg https://github.com/JuliaMolSim/DFTK.jl/blob/master/src/postprocess/current.jl for a related computation. If you do code it up please contribute it to this package, it could be useful to other people (eg for meta-GGAs)!
• The kinetic contribution to the stress you can get by only keeping the Kinetic term in the new_model here: https://github.com/JuliaMolSim/DFTK.jl/blob/master/src/postprocess/stresses.jl
• The Kohn-Sham potential we have, it's the function total_local_potential(ham::Hamiltonian) defined here: https://github.com/JuliaMolSim/DFTK.jl/blob/master/src/terms/Hamiltonian.jl

All of this should be relatively easy, but you do have to look at the code a bit. The examples should help you get the structure. If you're stuck let us know and we can help.

May I ask what you want to do exactly? There's an open issue for Thomas-Fermi here: https://github.com/JuliaMolSim/DFTK.jl/issues/122

[antoine-levitt]

Or for the stresses, even simpler: just change the energies.total line in compute_stresses to energies.kinetic

[cw-tan]

Great, thanks! I literally installed Julia yesterday but I'm eager to have a go at it! I'm just hoping to inspect these quantities from Kohn-Sham DFT to possibly give some insight for the development of kinetic energy functional approximations.

[antoine-levitt]

OK, happy discovery :-) I'm closing the issue for bookkeeping purposes, feel free to ask for help, here or on slack.