[gmx-developers] Pull code and virial

[Berk Hess]

Justin MacCallum wrote:

> Hi,
>
> I think I've discovered a problem with the way the pull code interacts 
> with pressure coupling. As far as I can tell the extra forces added by 
> umbrella sampling and AFM pulling are not included in the virial, 
> leading to unexpected volume changes in the system when the 
> thermodynamic mean force is high.
>
> I'm not totally sure how to fix this. Is moving the pull code force 
> calculation to before calc_virial enough? I'm not totally clear on how 
> the virial is calculated in GROMACS.
>
> I should point out that even when the man force is very high, the 
> volume change is still quite small, on the order of one percent. This 
> does however make a large difference in potential energy due to large 
> changes in water-water interactions.
>
> I don't think this a big problem for most simulations using the pull 
> code, but if you want to look at volume changes, or potential energies 
> there may be artifacts.

I guess the best way to fix this would be to calculate the pull virial 
(also for constraint pulling)
in the pull code, store it in the forcerec and add it in calc_virial, 
just as is done
for the grid part of the electrostatics virial.

Berk.