Open FG-TUM opened 4 years ago
I am not sure if your proposed way works in all cases as there are still some explosion heuristics (and maybe other legacy thermostats) enabled which influence the velocity of the particles. I stumbled upon this when doing a simulation in the two phase region. Here is a quick and dirty plot of several quantities sampled over time. The "explosions" were detected at steps ~1100 and ~5100: You can see the jumps in velocity and total energy. The colors just indicate different sampling methods.
Is your feature request related to a problem? Please describe. Currently, there is no intuitive way to set up an NVE ensemble. (See below how to do it currently)
Describe the solution you'd like
~Describe alternatives you've considered~
Additional context
100% conservation of energy will not happen due to numerical imprecision.
Currently (a2c2482) to set up an NVE ensemble configure your XML with the following: