Hi Lennart,

Thanks for your replying!

And about my simulation, I separated it into three stages. First was a 1 ns long simulation in NVT-hoover ensemble, and then restarted it as a 500 ps simulation in NPT-hoover ensemble, the above two stages were used to equilibrate the system. After that, I ran a 5 ns simulation to get the trajectory by restarting the previous simulation, which was the third stage.

Actually, I got this average internal pressure (23.11227 atm) after the whole 6.5 ns simulation. The average internal pressure was -79 atm right after the second stage. So, is it because the fluctuation of internal pressure being so big ( about 400 ) that there's no need to expect the value to be exactly 1 atm ? Or simply because 6.5 ns is not long enough to reach 1 atm?

Thanks for further suggestions.