[gmx-developers] Why is not the thermostat used in the first step?

[hessb at mpip-mainz.mpg.de]

Quoting ramon at juguete.quim.ucm.es:

> Hello,
> 
> I see that in the first step the thermostat is not used.
> 

This has been changed in the CVS code.

> This is a problem when one wants to equilibrate a system where
> molecules overlap. Using a Berendsen thermostat with a time constant
> of timestep/10 would allow to remove overlaps for these systems.

This would not help as v(dt/2) is scaled with T(-dt/2) which
has no meaning at the start of a run. T(-dt/2) is just given
by the input velocities.

Berk.


-------------------------------------------------
This mail sent through IMP: http://horde.org/imp/