In my simulation of metallic glass formation, I need to increase the temperature from 300 to 2300K at the rate of 10^13 K/sec with NPT . But I am not able to understand that how I can set this rate of temperature increase? I have get following code line from literture
fix 1 all npt temp 300 2300 0.1 iso 0.0 0.0 0.2
run 10000
unfix 1
But i cant understand the exact calculation.
I tried some of them but my simulation is not reaching to 2300 and it goes only upto 60 to 100 K.
Please help me to short out this problem
What is your timestep value? The system likely does not have time to adjust with only 10000 timesteps performed.
You can decrease your timestep value and increase the timesteps performed while keeping the same rate.
10^13 K/s is 0.01 K/fs, so either of the following would achieve that rate from 300 to 2300 K
Timestep : 1 fs, timesteps : 200,000
Timestep : 0.1 fs, timesteps : 2,000,000
2 Likes
As Steve had mentioned in hist tutorial talk, using a Nose-Hoover thermostat is not a good approach to quickly transfer kinetic energy since it operates globally. At this point, since this is only for melting, it actually does not matter to have the statistical mechanical correctness of the Nose-Hoover algorithm. Thus a combination of fix nph (for pressure) with fix langevin may be better. You still need to correctly compute your heating rate as @tjbarrett pointed out. Even fix nve + fix press/berendsen + fix langevin should work.
2 Likes