About how to properly control temperature in Simulated annealing in Gromacs

GROMACS version:
GROMACS modification: Yes/No
Dear all,

I am running a Simulated annealing over a molecule in gromacs.
When I visualize the temperature vs time xvg file I observed that in my .mdp file the maximum temperature was set to 1800K but actually the system was heated until 2000K.

How could I tune this properly in a way that my temperature only increases until 1800K?
For more information I am attaching the .xvg file and my .mdp file used.

Best regards,

Lazaro

md_SA.mdp (2.4 KB) temperature_160ps.xvg (8.7 KB)

That’s expected behaviour. Thermodynamically, you would expect your temperature to fluctuate quite a bit around an average value due to the small number of atoms (e.g. compared to the Avogadro constant). So what you’re seeing is exactly right. For more insight into the theory on how temperature is controlled have, e.g. a look here: https://arxiv.org/pdf/0803.4060.pdf

Thank you very much Cblau. As always very clarifying. Now I understand better.

Kindly,

Lazaro