Problem with MES prescribed displacement

[sstantz]

I have a whole host of problems with the simulation I am currently trying to run. But let's start here:

I attached a screenshot of the first time step in the simulation. From left to right, let's call these parts "outter tube," "wedge," and "cone." (the "cone" is actually two parts bonded together - a wedge and a piece of "inner tube").

I have a prescribed displacement pushing the cone downward 1.44 inches. There are symmetric boundary conditions appropriate for a quarter model and a bondary condition fixing the bottom of the outter tube. There is frictionless slide, point to surface contact between the cone and wedge. Also, There is a tied point to surface contact between the wedge and other tube (simulating friction keeping the wedge still -since frictional contact was seemingly not doing its job - but more on that later).

The model starts out with the top of the cone (top of the inner tube portion) flush with the top of the outter tube. The big problem here, is that in time-step one, you can see that the cone moved UPward - the opposite direction of the prescribed displacement. I cannot understand why it is doing this. After the first time-step, it begins moving downward, in the proper direction. But, first, it moves up; this has occured with several other attempts also with different boundary conditions. Prescribed displacement should do exactly as it is named (i.e force the top surface of the inner tube to move downward - not upward at any point).

Also, since I mentioned a problem I was having with friction, let me comment on that also:

Previously, I was trying to use frictional contact with the outer tube to keep the wedge in place as the prescribed displacement occured. The reason for this was that ultimately, I would have liked to have the wedge begin moving after the top of the wedge came in contact with the lip on the cone. Just to be sure I was not going crazy, I did a quick hand-calc to verify that the wedge should, at no point, move (despite the angle of the cone, until contact is made on that top surface). But, the friction was not keeping the wedge still, it was simply being pushed down. I have no idea why the friction is not working. The program should not have problems with such a simple and trivial calculation.

Please offer any suggestions with fixing this problem - (as an additional note: I am new to using this software before, although I do have a small amount of experience using ANSYS. I have yet to have a simulation complete succesfully. Convergence is a problem, sometimes early, sometime late in the time frame of the analysis, depending on the settings I tinker with).

[sstantz]

I got the Cone/Inner tube assembly to stop moving upward on time-step one. I did this by having the prescribed displacement start after time=0 rather than at time=0. I don't know why this helped.

The convergence is still awful. Worse than it was before, actually. Less than 10% into the simulation and the time step starts reducing drastically. I tried lowering the contact stiffness. It seemingly had no effect. I tried allowing some penetration, also did not help.

The friction issue is still unresolved. and I cannot continue working out the problems in this thing, until I get better convergence. I keep trying suggestions from the users guide, but nothing is helping.

[zhuangs]

About the convergence,  I would like to provide the following comments:

(1) Your mesh might be so coarse that there is large initial penetration between the contact parts.  For the curvature contact surfaces, even though there is no initial penetration, the coarse mesh (non-matched nodes) can result in penetration. Thus, try to use finer mesh.

(2) Try to use user-defined contact distance with a samll value, such as 1.e-5, and so on.

-Shoubing