Shell solid contact with localized stiffness

[lukerickert]

Hi all, I am going crazy with what looks like a very simple problem.

It is a deck structure (the only challenge should be dealing with the thousands of parts)
but when I tried establish contact between a solid part (normal or rigid) and a shell model where there is a beam or stringer directly under the object I cannot get proper convergence, even with a very simple test model.

I would like to have quad mesh on the solid part but I haven’t done that as I need to use Netgen or something externally. Anyway I don’t think that is the real issue here.

I am doing everything I can think of to make my test easy to converge. The edge has a fillet, the part is slightly tilted so the contact will be gradual, the solid is constrained and there is only a small displacement with tiny initial steps. I have been working with exponential contact to see if that will work and while it will run it never manages to actually establish contact. (This is a version of the pmx file). I also tired converting the solid to be rigid and that didn’t seem to make any difference.

If anyone has any ideas I would be very appreciative. I am completely confused why this isn’t straight forward but the extra stiffness from the stringer must somehow really mess up the contact.

Thanks

Luke

[FEAnalyst]

There are certain combinations that often lead to non-convergence in CalculiX. Especially shells + rigid body + Nlgeom but it may also happen with shells and contact. In such cases, the best way might be to replace shells with solids using the Thicken Shell Mesh tool in PrePoMax.

Generally speaking, it’s best to start with simple models and then add complexity gradually. I would increase the prescribed displacement to make sure contact is properly established though.

[synt]

probably the problems due to knot exist at shell intersection, using separated parts of them by offsetting and connect with tied contact type may eliminate these conditions.