Is it possbile and how to apply an directional hydrostatic pressure by the pull code?

User discussions
1

GROMACS version: 2018-2021

  • List item

GROMACS modification: Yes/No

1
1869×422 263 KB

As shown in the figure，some researchers have explored the transport of water molecules through the pore of graphene oxide membrane under the applied pressure( Impact of Surface Ionization on Water Transport and Salt Leakage through Graphene Oxide Membranes | The Journal of Physical Chemistry C (acs.org)). Some of them were implemented by GROMACS. They said that constant forces（ Constant acceleration NEMD, acc-grps）, were applied on the graphene piston wall to produce the desired pressure.
But I learned that there is a bug with this method and it will be deprecated（https://gromacs.bioexcel.eu/t/constant-acceleration-md-anti-bug/1614）. It has been proposed to use the pull code instead. I have a few questions here.
1. Can the pull code (pull-coord1-type=constant-force) produce constant hydrostatic pressure? Someone told me that the real pressure generated by this method is variable and can not be known. Is this true？
2. If the pull code could work，how to set pull-coord1-k? Is it directly equal to the force that needs to be applied to the piston? Forgive me, this question may be silly.
3. If acc-grps and pull can not reach the expected hydrostatic pressure, is there any way to implement this kinds of research using Gromacs？
Thank you very much！

2
  1. It might. You will have to restrain or pull the middle membrane. The pressure reported by Gromacs will of course be of a different value (whole system), but I see no reason, why a fluid between a solid wall and a piston with some force should not have a defined hydrostatic pressure. It probably needs to be pretty large though.
  2. I guess yes.
  3. I don’t think there is another way. The barostats can not couple parts of the systems.