[gmx-users] A question on Berendsen NVT ensemble

[David van der Spoel]

On Wed, 1 May 2002, Lianqing Zheng wrote:

>I also got problem with NPT emsemble using NH and Andersen methods.
>Actually this is why I came to try Berendsen method. My code is running
>well with NH's NVT, but becomes unstable for NPT ensemble. The symptom is
>the frequency of oscillation of pressure (or volume) becomes bigger and
>bigger and finally the program crashes. Reducing time step (I tried 0.1fs)
>or enlarging mass of volume seems not to solve this problem.
>
>What do you think about this? Thanks.
You should set the tau_p larger probably.

Now if I remember correctly you had a SiO2 crystal that was fully
interconnected. This will never melt in a normal simulation, because
melting in this case requires breaking of bonds. The only way you could in
principle simulate that is by using Morse potentials, but then you still
have the angle terms left which should be replaced by a three body
nonbonded potential. Interesting but non trivial, and not implemented in
GROMACS.

Groeten, David.
________________________________________________________________________
Dr. David van der Spoel, 	Biomedical center, Dept. of Biochemistry
Husargatan 3, Box 576,  	75123 Uppsala, Sweden
phone:	46 18 471 4205		fax: 46 18 511 755
spoel at xray.bmc.uu.se	spoel at gromacs.org   http://zorn.bmc.uu.se/~spoel
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++