I do know that there are some frustrated systems that plainly can't be optimized to a genuine 0 K state. I'm familiar with the work done by people working on instantaneous normal modes (INM) and inherent structure (IS). I'm getting imaginary frequency normal modes, which I knew was possible. But shouldn't I still see the 0-frequency translational and rotational modes? I've looked carefully and can't seem to find anything on this.
I guess what I was really asking was more of a physical question than a question about GROMACS. Because of the physics of a (locally) harmonic central potential, the Hessian components are not all independent. So I should see 6 singular values, i.e. one eigenmode for each independent translation and rotation of the whole system that leaves the potential unchanged.
Do you happen to know if these singular value normal modes can be shifted to negative energy as a result of error in the Hessian? I guess I should mention that these are collective modes of an entire system with about 80,000 degrees of freedom, about 27,000 atoms. Not modes of a subset of atoms.
Thanks!
- Badges
- Science topic
- Similar topics
- Automotive Engineering
- Machines