[lammps-users] A problem in Shear Triaxial Test for an RVE with the pressure control commands

I recently had a problem at my work; I would ask you to guide me if possible.

I intend to simulate a triaxial test immediately after the hydrostatic test. To do this, I fill a periodic box of the material I want in the form of several spherical particles. Then with the help of the fix npt command, I bring the pressure of the Box to a specific value and equality in all directions. It is also important to me that the temperature stays constant at 300 Kelvin. After the Hydrostatic step, using an unfix and fix, I perform the next step, i.e., triaxial, with the same fix npt command. In this way, I consider the pressure of lateral faces (for example, y and z directions) constant and increase the axial face pressure (x-direction) and also consider the temperature as constant at the same 300 Kelvin. But the problem is that the pressure in the axial direction eventually increases to 2 or 3 gigapascals and does not exceed this value, or the lateral faces’ pressure increases with it. These 2 or 3 gigapascals are a very small amount for me, and I want more difference. Also, I need the density value decrease in the triaxial part, while with this value of 2 or 3 gigapascals, it does not change significantly.

I also used the Berendsen barostat, and what happened was that it experienced a difference of more than 2 or 3 gigapascals while the lateral pressure remained constant, but the other outputs, such as stresses, etc., had problems. So It did not work well.

I also tried to do this using the fix deform command, but it is challenging to control the lateral faces’ movement so that their pressure has a constant value. Also, in this case, I couldn’t use a thermostat to keep the temperature constant at 300 Kelvin.

For more information, I’m using Mishin-Ni-Al-2009.eam.alloy for Nickel and Aluminium materials, and the whole simulation is done in 400 picoseconds.

Best regards,
M.Ashtari