How to maintain atmospheric pressure in isothermal-isobaric ensemble

LAMMPS LAMMPS General Discussion
1

I want to obtain the stress-strain curve for a two dimensional material by applying uniaxial tension. I am relaxing the system to 300K temperature using isothermal-isobaric ensemble before introducing deformation to the simulation cell. I suppose I have to use this fix for relaxation (since the pressure unit is in Bars and temperature unit is Kelvin) -

fix NPT all npt temp 300.0 300.0 5.0 iso 1.0 1.0 5.0

However, I have found examples of this fix being used for atmospheric pressure -

fix NPT all npt temp 300.0 300.0 5.0 iso 0.0 0.0 5.0

Is it a mistake they have made? Or is the difference so negligible that the obtained output stress-strain data is identical?

Thanks in advance.

2

Not sure what the dollar signs are there for, both of your input lines as shown are invalid LAMMPS input.

Yes, a simulation at 1 bar pressure is formally different from a simulation at 0 bar pressure, and yes, many systems can show negligible change between those two pressures due to their low isothermal compressibility. I personally don’t understand why people do 0 bar simulations.

3

Thanks for your reply Dr. Shern Tee. My bad. It’s just 1 and 0 without the dollar signs. Perhaps a reason for doing Zero bar simulation is to reduce computational cost ? Or maybe simulations with one bar converges rarely ? Just a thought. Would be glad to hear content reasons.