[gmx-users] freezegrps changing coordinates in NPT simulations

[niexuechuan]

Hi Jon,
I note that when using "pcoupl = Parrinello-Rahman", the frozen group will be scaled. But when "pcoupl = berendsen", the frozen group will stay at its initial position, and will not be scaled.


Xuechuan


 








At 2015-05-15 01:57:10, jwillcox at andrew.cmu.edu wrote:
>Hello again,
>
>I'm running a simulation with graphene as a periodic molecule.  I've
>frozen the carbons in place in the mdp file using:
>
>freezegrps               =  CG_CE
>freezedim                = Y Y Y
>
>but when I run an NPT simulation, the entire system shrinks (including
>moving the carbons closer together) until it crashes.  Everything works
>fine for NVT and minimization.
>
>Are there any tricks to freezing groups in NPT simulations?
>
>I've also attached my mdp file.
>
>Thank you!
>
>Jon
>