I have a part that, when installed, is preloaded in bending by being clamped down, but has been modeled as a flat part for manufacturing purposes (as well, I'm not completely sure about the final shape, because the clamp is compliant, and also bends under the clamp load).  I have been asked to simulate the part and am trying to avoid using the non-linear module, because someone else here is typically needing that licence. 

I have created a few models to test various arrangements and techniques for doing this analysis, and shrink fit seems to achieve what I need to do in certain geometric configurations, but not others.

So my real questions is:  What sorts of "real" contact conditions have people made to work using shrink fit contact and the linear solver?  I got a successful run using parallel planes that had initial interference, but things broke down when I tried non-parallel planes, possibly because in the deformed shape the contact is between a surface and a line.  I intend to shortly try a "shrink fit" between a flat plane and a round surface (which is, again, theoretically line contact, but might work because there is tangency, where the non-parallel planes have no tangency.  I'll post back here with a report regarding how this turns out.

Files are attached: 

In "...Step2" run the study "Static 2-sym" this one worked, though in can fail if the interference is increased too much.

In "...Step3" there is only one study, and it failed.  I have tried tweaking a few things, and I either get virtually zero displacement/strain/stress or something insanely high, with the interference remaining in the deformed shape

If someone has made something like "...Step3" work, would you please share how you managed?

Thanks,

Tim