[gmx-users] Question on pressure coupling

[Mohsen Ramezanpour]

Dear Gromacs users,

Inverted hexagonal phases are usually simulated (and should be) using a
triclinic box with either semi-isotropic or anisotropic pressure couplings.

I am not sure what a semi-isotropic and anisotropic means in a "Triclinic"
box, exactly?
The .log files look the same for the Triclinic box and rectangular box.


Here is an example:
Semi-isotropic (.mdp file):

compressibility         = 4.5e-5  4.6e-5
ref_p                   = 1.5     1.6

The .log file looks like:
compressibility (3x3):
      compressibility[    0]={ 4.50000e-05,  0.00000e+00,  0.00000e+00}
      compressibility[    1]={ 0.00000e+00,  4.50000e-05,  0.00000e+00}
      compressibility[    2]={ 0.00000e+00,  0.00000e+00,  4.60000e-05}
   ref-p (3x3):
      ref-p[    0]={ 1.50000e+00,  0.00000e+00,  0.00000e+00}
      ref-p[    1]={ 0.00000e+00,  1.50000e+00,  0.00000e+00}
      ref-p[    2]={ 0.00000e+00,  0.00000e+00,  1.60000e+00}

############################################

and for Anisotropic (.mdp file)

compressibility         = 4.5e-5  4.6e-5  4.7e-5  4.8e-5   4.9e-5   4.1e-5
ref_p                   = 1.5     1.6     1.7     1.8      1.9      1.1

The .log file looks like:

compressibility (3x3):
      compressibility[    0]={ 4.50000e-05,  4.80000e-05,  4.90000e-05}
      compressibility[    1]={ 4.80000e-05,  4.60000e-05,  4.10000e-05}
      compressibility[    2]={ 4.90000e-05,  4.10000e-05,  4.70000e-05}
   ref-p (3x3):
      ref-p[    0]={ 1.50000e+00,  1.80000e+00,  1.90000e+00}
      ref-p[    1]={ 1.80000e+00,  1.60000e+00,  1.10000e+00}
      ref-p[    2]={ 1.90000e+00,  1.10000e+00,  1.70000e+00}

How the scaling is done differently for Triclinic and rectangular boxes?
(The question is also valid for isotropic coupling, but for semi and
aniso is more confusing)
Or, is there any difference in the implementation of pressure coupling for
different PBC types at all?

Thanks for your comments in advance,
Cheers,
Mohsen
-- 
*Rewards work better than punishment ...*