question on VASP manual -- simulated annealing
Posted: Wed Jul 10, 2013 5:32 am
Hi VASP masters and users,
I'm trying to use the simulated annealing (SA) method to get the ground state geometry. I feel the instruction in the VASP manual on this method not quite clear to me, particularly, the sentence below:
"The temperature should be decreased only slowly. This can be done by decreasing the temperature (TEBEG) in the INCAR file and using the Nose thermostat."
My understanding is that SA calculation should include two steps. In the first step, increase the temperature from, say, 0K to a desired temperature (e.g., above melting temperature by 50K). In the second step, decrease the temperature slowly to 0K. I guess, the sentence quoted above is talking about the second step. If so, TEBEG in the second step can't be freely tuned, right? To slowly decrease the temperature, I thought NSW should be increased, while TEBEG should be fixed.
I think there might be some key physics that I didn't understand. Please correct me. A lot of thanks!
I'm trying to use the simulated annealing (SA) method to get the ground state geometry. I feel the instruction in the VASP manual on this method not quite clear to me, particularly, the sentence below:
"The temperature should be decreased only slowly. This can be done by decreasing the temperature (TEBEG) in the INCAR file and using the Nose thermostat."
My understanding is that SA calculation should include two steps. In the first step, increase the temperature from, say, 0K to a desired temperature (e.g., above melting temperature by 50K). In the second step, decrease the temperature slowly to 0K. I guess, the sentence quoted above is talking about the second step. If so, TEBEG in the second step can't be freely tuned, right? To slowly decrease the temperature, I thought NSW should be increased, while TEBEG should be fixed.
I think there might be some key physics that I didn't understand. Please correct me. A lot of thanks!