Couette flow: appropriate ensemble and calculation of shear stress

Dear lammps users,

I am trying to simulate the Couette flow where a shear flow is applied to water confined between two moving walls.

I have done both literature review and search in LAMMPS mailing list in order to find the appropriate ensemble to integrate the equations of motion and correct thermostat. I have found different approaches in the literature and mailing list. Should I apply nve to both water and walls? Or should I thermostat the wall using nvt and apply nve to water? Can anyone give some insights to this matter? Or should apply nve to whole system and just thermostat the flow?

Second, how can I calculate the shear stress exerted by water to the wall?



The initial Q is what is you model for the wall. If it is just rigidly
moving atoms, you don’t need to thermostat them. You must
thermostat the fluid (water) because the dragging wall will
add energy to it. Langevin is a typical good choice. You can
look at the examples/VISCOSITY wall example for how
shear stress is measured - in that case in order to derive
a viscosity.