AnsweredAssumed Answered

Mixed meshing question

Question asked by Matt Cullin on May 29, 2016

I am trying to duplicate an example from a mixed meshing presentation to get a better understanding of what I can and cannot do in Simulation.  If you see the attached assembly file, I have a base plate, a custom profile square tube weldment, and a thin surface handle.  Obviously if I model the beam component with solid or shell elements, the connection between the handle and the square tube is taken care of automatically by the Global bonded default.  I want to use beam elements for the beam member. Solidworks is generally pretty clever, so I first tried a single weldment member.  The Global defaults took care of the contact between the beam end and the base (a solid), but was not able to deal with the handle connections midway between the joints of the weldment.  In the contact visualization it shows these colocated surfaces as being bonded, but on running fails due to instability.


I figured Simulation enforced the bond by forcing the local shell/solid and the beam's sub-elements to have the same position and curvature (i.e. the beam element section is normal to the solid profile).  In messing around with a different model, I noticed that a stiffening beam and solid meshes seem to converge at joints.  This makes me think that maybe the bonded condition is only enforced at the joints. If someone can give me insight into this, I'd appreciate it.


Anyway, I figured I'd just break the long beam weldment into smaller segments such that both joints of a beam segment would lie on the contacting shell faces.  I thought I was being clever.  My problem is that when I make a structural member length shorter than, or equal to, the section width, the joint placement algorithm places the joints on the sides rather than the ends.  Is there a way to "manually override this" and get the joints back where they belong?  I realize it is kind of weird to have a structural member whose length dimension is shorter than the width of the profile, but I wonder if it can be done.  Also, I realize that I could work around this by treating the short sections as solids or shells, but I am really trying to learn about the program's capabilities.  The actual model/results are inconsequential.


Thanks in advance for your help,