Hello everyone, I hope you are doing well. I have a question regarding the use of the NPT barostat with controlled shear pressure in a system containing water. I tried to apply it along the xz plane, but I noticed a very aggressive change in the tilt factors. I was wondering: is this procedure problematic for systems containing water? My system consists of two graphene sheets separated by water.
Thank you very much for your help
Depends on what one expects to measure as the “shear modulus” of a liquid.
1 Like
Hi srtee,
Thanks for your reply. I’m indeed trying to measure the “shear modulus” of the entire system. According to the LAMMPS manual, when you do npt with controled xz , LAMMPS will control the pxz of the total pressure tensor by adjusting the box tilt (the xz tilt factor). Since my system contains water between two graphene sheets, I observe that the tilt factor grows very large and never stabilizes. so i’m wondring if the npt is suitable for a homogeneous system (crystals) and not suitable for system containing water ?
To be explicit: as far as it makes sense to me, a liquid cannot sustain shear strain (except maybe at very low shear rates at a suitable interface, certainly inaccessible to MD simulations), so it doesn’t make sense to talk about the “shear modulus” of a liquid.
A liquid can undergo shear stress that induces shear flow, for example Couette flow, but you can’t expect to barostat such a system (you’d impose a constant wall-to-wall length), especially a liquid as un-viscous as water.
1 Like
If your system contains water then I’d hazard to guess that - if it is solid - it is probably quite soft and viscoelastic. So, it might yield and flow easily like @srtee suggests. So, imposing constant shear stress with a barostat will perform a creep measurement.
If you do have a solid, but it is weak, its almost always better to do a strain-controlled simulation. Apply a small step-strain with a short fix deform run, then turn off fix deform and measure the shear stress over time. If the system has a finite shear modulus, the average value will be finite. If the system is a liquid, then the shear stress will decay to zero.
If your system is a fluid, then you may not want to measure a viscosity or relaxation modulus G(t) instead of a shear modulus.
1 Like