Hi,
I am running 2D rolling simulations in Abaqus/Standard using quite a large roller (R=257.45mm) and a thin sheet (h0=2mm). I am using this as a comparison for an analytical model that predicts the through-thickness stress/strain distributions in the metal sheet during the steady state part of the process. Therefore, I need to be content that my rolling simulations are in a steady state. Since I care about through-thickness variations, roll force/torque are not necessarily brilliant measures of a steady state for me. I would prefer to check if the strain is changing from time to time for example.
So this is what I am trying to do. I am comparing different strain components at different times during the rolling simulation. The way I do this is I take a vertical line at some fixed x position and I interpolate the strain values onto this line so that I can plot strain vs z which is the vertical spatial coordinate. If a steady state has been reached, then these curves should not change from time to time. I have two results of this attached. One picture is from a simulation with 10 elements through the thickness of the sheet and the other is from a simulation with 30 elements through the thickness. In each plot, the interpolated curves are from the last 3 frames of the rolling simulation. The step in which each of these pictures pulls data from is 0.1 seconds long and there are 20 frames in each simulation, so the frames are 0.005s apart. I have chosen 0.1s as the length of the step as this represents the real time taken to roll the length of sheet I have defined. The roller is rotating at 5rad/s.
My question is that the results from the simulation with 10 elements through the thickness seem closer to a steady state than the 30 through-thickness elements one. Does anyone have any idea why this is? I would've thought the opposite would occur?
