Hi,
I am trying to calculate the viscosity of mixing system with CO2 and some molecular at various Tempaeratures and Pressures.
First, I tested the water box, and the result seemed to be acceptable.
Then I changed to CO2 system, even only CO2, the the value of 1/viscosity obtained by g_energy was not a bit correct. it fluctuated considerably from 1E+04 to 1.5E+05. I tested the cos_acceleration from 0.001 to 0.025, but somtimes it shown: “srun: error: nid00114: task 0: Segmentation fault
srun: Terminating job step 14252474.1”.
Could anbody help me to give some advice to calculate viscosity, pls? I appreciate it very much.
As you observed, the viscosity calculation with gmx energy converges very slowly and is very sensitive to the parameters you set, so using the cos_acceleration is definitely the way to go.
The segmentation fault indicates that the forces / velocities during your simulation are too high, so there are a few things you could try (if you didn’t already)
use a different thermostat v-rescale with a very low tau value
use a larger simulation system
use a well-equilibrated initial structure, making sure that the system density is correct by running some NPT equilibration beforehand