I am about simulation of a box which should be confined vertically , and then sheared horizontally. I have 3 layers, the lower layer (which has 2 sub-layers) is frozen and the upper layer (which has 2 sub-layers as well) is rigid. They are meant to keep the lower layer fixed and move the upper layer (rigid layer) with constant velocity and looking at the micro-mechanical properties of middle layer. For confining vertically , I have used a "fix mesh/surface/stress/servo" to press it down and we know that we cannot use this mesh for move/mesh command to apply shear! Considering these facts, I have got 3 problems now:
1- When I move my stl plane downward and extract the forces on it, it never gets to target value, lot of oscillations, sometimes 0, negative or positive. On the other hand, although I only put force vertically, I see non zero values for horizontal components of forces acting on wall. I tried different "max vel " but still I can not get a permanent stable force on it!
2- The particles go inside each other under pressure! I do not understand this behavior! You can see the attached picture.
2- for next step, when I want to shear my box to see the behavior of middle layer in terms of macro friction coefficient (Fy/Fx, applied to this layer) and due to the problem I discussed in 1st paragraph, I had to make the upper layer rigid, move it horizontally with constant velocity and then calculate forces on the upper wall. It is the simplistic way and not the most precise way, but available one. I am wondering if anybody knows any trick to move a mesh with constant velocity and be able to measure forces on it (force controlled walls). Notice that, this mesh is used in previous stage of experiment (confining).
I have attached a picture, showing my model so far. the two upper layers are rigid and two lower layers are frozen.
Thank you in advance for your responses.