Hydrogen Mass Repartitioning -> NVE energy loss due to rigidBonds

From: Maximilian Ebert (max.ebert_at_me.com)
Date: Mon Jun 14 2021 - 17:00:40 CDT

I am experimenting with HMR and observed at 2fs time step a constant energy loss in an NVE ensemble when the mass is repartitioned to set the hydrogen mass to 3.024 AMU. The mass comes from the neighboring heavy atom. When I turn off rigidBonds (which I have set to all by default) for the same system the hamiltonian is fluctuating but conserving energy. In contrast, running an NVE ensemble with the same starting coordinates with default mass and 2fs, rigidBonds can be set to all without the loss of energy.

Can anyone explain why SHAKE looses energy with non-standard masses in the NVE ensemble?

Thanks,
Max

This archive was generated by hypermail 2.1.6 : Fri Dec 31 2021 - 23:17:11 CST