Movement Without External Forces

[luc46552]

Hello, I am making a model that consists of 3D deformable "beams" (not beam elements) that are held together by strain energy. By the end of the simulation, the goal is to get the final stressed state of the assembly shown below.

This is done in multiple steps. First, all beams are fixed in place through a boundary condition, and surface contact is defined between the beams and the "ground", which is modeled as an analytical rigid body. Next, surface contact is defined between the surfaces of the beams, and the boundary conditions that fix the beams are released to achieve the final stressed state. In this step, the beams are only held together by surface contact with pressure overclosure and friction, while they are in no way tied to the floor. The problem is, the beam assembly moves in space at an increasing speed as if there was some external force acting on it, which eventually causes the simulation to abort.

I do not understand why this occurs, since no loads or external boundary conditions are defined, and all forces between the beams should be equal and opposite. I have tried fixing a single beam or node in space while this happens, but this just reduces the reliability of the simulation further since the assembly is still trying to move as it is fixed. So my question is, if anyone has encountered this before, what could possibly be causing the beam assembly to move without an external load, and how can I prevent this to complete the simulation?

 

[FEA way]

Make sure there is no initial velocity or prescribed displacement/velocity boundary condition left in this step. But probably the reason is in beams-floor contact definition. Examine it carefully, increase friction, maybe use general contact instead of contact pairs. Check contact status variables (COPEN, CPRESS, CSLIP, CSTATUS). Also try using inertia relief instead of modeling the floor.

Is the floor fully fixed at its reference point ? If not, it should be.