[gmx-users] About Temperature Calculation

Ane Xu xzp02 at mails.tsinghua.edu.cn
Thu Apr 1 16:22:01 CEST 2004

Dear Xavier:
>    No solvent, no temperature coupling,
No solvent: not good for a good control of temperature.
No temperature coupling: you need a equilibration periode where you
actually control the temperature till it reaches the value you want. In
NVE the E=Epot+Ecin, the equilibration in needed for the equipartition
of the Epot and the Ecin that will (in theory) be conserved for the rest
of the simulation. What is happening is that the Ecin you give to the
system is transformed into Epot and this will be as long as the
equipartition is not reached 
at the
temperature you want.
I would try Stochastic Dynamics. It is probably the best for what you (I
presume) want to do.
     So, the 0K equilibrium potential (what I got from "em") is not the
"quilibrium" one in the 300K or 500K case, and if I wanna got a system
at 300K, the only way is to do "SD" or "MD" with higher initial
temperature (which results in a equilibrium 300K state). Is it right?
>just a NVE simulation with
in your file you have gen_temp = 500  !!! That explains why you have a 
temperature of 500 K
    I test a case of 500K (which results in 250K finally) after 300K, I
forgot to explain, I am sorry for that.



More information about the gromacs.org_gmx-users mailing list