[gmx-users] surface tension coupling

[Erik Lindahl]

David L. Bostick wrote:
> 
> Hi,
> 
> For surface tension coupling, the first ref_p value in the mdp file is the
> reference surface tension times the number of surfaces in [bar nm]
> according to the manual.  So does this mean that if I wish to simulate a
> bilayer with constant surface tension where z is the bilayer normal, there
> are 2 surfaces?  So the ref_p value should be 2 times the target surface
> tension?
> 
> David

Hi David,

Since there is no way the program can determine the number of surfaces,
the surface tension you give in the parameters is the TOTAL surface 
tension in the system - just as the value reported in the energy file
(The cryptic description is "surface tension * number of surfaces").

Surface tension for bilayer simulations can be tricky, since some 
authors report the tension for the entire membrane while others quote 
the tension per monolayer.

Assuming you know that you want the tension 'x' in each monolayer, you
should use '2x' in the scaling if you have a single membrane, and for
instance '4x' if you would simulate two stacked bilayers.


Cheers,

Erik