[CP2K-user] Energy Drift per Atom

[Omid Shy]

Hi,

when running AIMD (using the GPW method), there is an 'ENERGY DRIFT PER 
ATOM' value printed for each timestep in the output file.

   1. Relative to what this energy drift is calculated?
   2. Should this value remain constant during an NVE ensemble simulation, 
   and should it be close to a specific value?
   3. What about an NVT ensemble?


Thanks,
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20210714/6eb9ef55/attachment.htm>