search

ajkerr0 / kappa

A python package to calculate thermal conductivity across molecular interfaces.
MIT License
6 stars 4 forks source link

Missing handling of zero-modes #64

Open ajkerr0 opened 8 years ago

ajkerr0 commented 8 years ago

The suggestion is to find a central atom in molecules and add a small term to its diagonal hessian matrix terms. The interpretation of this is that the center atom (and this atom only) is attached to some outside spring, or is in its own little potential well. This will prevent the long wavelength, low frequency (basically zero) normal modes that correspond to the translational motion of the molecules.

Maybe write a routine that finds the atom closest to the center of mass? Or geometric center.

ajkerr0 commented 8 years ago

This actually hasn't been a problem in the calculations, but might be the source of the divide by zero errors. They may have been the source of singular linear systems.

ajkerr0 commented 8 years ago

Let's not ignore the possibility of zero-modes corresponding to rotation in addition to translation.

ajkerr0 commented 8 years ago

We have added a method to return the center of mass of molecules. We have increased the stapling constant. Maybe we should pull the trigger on handling these zero modes in that way.

Or we could tell the code to ignore modes that don't oscillate/don't decay.