I’m simulating a membrane with a solute and I have been using in my simulation tcoupl = nose-hoover and Pcoupl = Parrinello-Rahman and I have received this note:
NOTE 1 [file NPT.mdp]: leapfrog does not yet support Nose-Hoover chains, nhchainlength reset to 1.
I read somewhere that the system may not be ergodic with this condition. Is it true?
What are the implication to reset the nhchainlength to 1?
Is it a benign note?
Thank you in advance.
The are only ergodicity issues with the NH thermostat when the number of degrees of freedom is very small. For a membrane system there is no issue at all.
@hess
I have been searching online for an hour or so, I could not find a quantitative limit, in terms of number of degrees of freedom, for Nose-Hoover thermostat to be efficient (in sampling terms). Any references or insights in this regard?
Nose-Hoover without a chain is an issue with a handful of degrees of freedom. Most atomistic simulations have thousands of degrees of freedom, there there is no issue at all.
If you are really worried, use the Bussi thermostat instead. There there are no such issues.