split surface singularities

[ozzkoz]

I'm having a problem with singularities at the boundary of split surfaces when a portion of the surface has a fixed BC applied to it. As a simple example, if I have cantilevered arm which is supported by a wall over half the arm, I've split the surface in two, and applied a zero displacement in the y direction constraint to one half. When I solve the model I find that there is a singularity along the split line, the stress goes up and up with each mesh refinement (in this case p-order). Is there a better way to constrain the model? Do I need to explore using some more advanced element, like a contact element or something similar?

As an alternative I can compute the reaction load ahead of time and apply it here, and then apply a displacement constraint to another portion of the model just to stabilize it and then check to be sure that any loads at that reaction are insignificant but I don't think this really models the problem as well.

 

[flash3780]

So, your model is basically a diving board?

Well, if you're interested in the stress at the edge of the support, I'd suggest modeling the support. It's a good rule of thumb to apply your boundary conditions sufficiently far from your area of interest.

Honestly, though, if you're modeling a simple beam you can just look in Roarks and solve it by hand, can't you?

//signed//
Christopher K. Hubley
Mechanical Engineer
Sunpower Incorporated
Athens, Ohio
--

http://engineeringliberty.wordpress.com

 

[ozzkoz]

This is a simplified example, my model is much more complex. It's actually a keyed assembly and I'm modeling the stresses in the keys. I guess I can model the assembly and use contact elements, this would be the closest thing to reality but I'm trying to keep the size of my model down. I was looking for some way I might model it using only the one part.