Skip to content

ConvexHull is very slow to compute hull energies

Qhull_crash_2023-11-29.log

I have 6-component FCC alloy. Using the DFT-relaxed energies with about ~ 800 structures that should span the full composition space, I get this error. I'll make a table of the compositions and energies to share. In the meantime, where to should I look to resolve this?

More info: I'm using tight SCF tolerances (1e-9) and some of the structures are very unstable and have relaxed very far.
I was hoping the use the CHull to plan my next set of training structures in this composition space.

To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information