flip operation in fix deform + fix npt

Dear all,

I have been dealing with the same Problem. Fix NPT in combination with fix deform starts oscillating after the first flip. Unfortunately the oscillations start increasing in amplitude with every subsequent flip. Two flips are usually enough for the potential energy and temperature in my system to become unstable. The only way I have found to deal with this problem is to restart fix deform periodically. After each restart the oscillations go away until the next flip. Fix NVT in combination with fix deform is long term stable.

Best regards,

Jan Mees

To Jan and Thomas - did you both see Aidan’s recent

post on this issue? He does not think it is a code problem

or dynamics, but an issue with the way your are interpreting

the tilt factor (or strain) when you are changing the

box length (via NPT) in the same dimension the tilt is

being applied. Does this address your Qs?

When I read the earlier emails, I didn’t pick up on that

point. I thought you were only changing the box length

in the tranverse dimensions, e.g. y when applying xy tilt.