Damping Parameter for Liquid Argon

I am trying to run NEMD to calculate thermal conductivity using fix langevin command. However, I am struggling with finding the optimum damping parameter for liquid argon. Is there a systematic way to determine this constant?

How does it manifest that you do not have the optimum parameter?

Whenever I change the damping parameter, the temperature profile changes completely. I found that at certain damping factor for a given temperature, the temperatures in all regions are very stable. However, when I run the simulation with the same damping factor at different temperature, the results are completely off.