[lammps-users] Poiseuille Flow Between Infinite Parallel Plates

Hi Lammps users and developers,

I want to conduct a "Poiseuille Flow Between Infinite Parallel Plates"
simulation and I notice that there is an example for this. However, the
example uses fix enforce2D and is only 2 dimensional. I want to perform
this in three dimensions and any suggestions into how to go about this is

How do you impose a Pressure gradient? I see in the example that it uses
an addforce to the group of atoms that flow.However the top boundary has a
fix addforce also orthogonal to the flow direction. Why is this?

I plan to you use fix nvt hence I need only to thermostat the 2 axes
orthogonal to the flow so I will use compute temp/partial for this.Is this
the only thing I need to consider?

Jan-Michael Carrillo

It should be no different than any other 3d LAMMPS simulation
(starting from a 2d example). You would want to thermostat
only the non-flow directions. The P gradient can be done
in the flow direction via fix addforce. The other one in the example
was likely simply to press down on the top plate. Fix aveforce (not
addforce) might be more typical for that.