Hi Tor,
Could you tell me how you modeled connection of the rectangle pipes, and plates? Its like a lap joint, and simply connection of plate and pipe surfaces by tie conection I think it is not accurate, becouse it will join whole surface, not edges.
Will making a tie coupling between surface of pipe and edges of plete will be ok?
The model was initially designed in Solidworks (not by me), but im assuming you mean the mesh model? I actually never finished it, as it was more of an example. In the past, i have approached such a problem with a glued/tied connection between surfaces, and be done with it. I’m well below yield, and the model is proven in the field, so ive felt confident about it, but ive recently been experimenting with adding welds to my 3D model, by simply adding connecting chamfers. I think that is more precise. I should say im a novice FEA analyst at the time of writing, so this may not be the right thing to do.
Im going through a lot of tutorials on using Salome Meca currently, and hope to bring some meshing video tuorials in the future to help out the FOSS FEA community.
If this both plates will be connected by tie connection (surface to surface) the result there will be not correct, i think, as welds could be possibly on around plate ( there will be a fillet weld only)
The file is really an awsome exercise/challenge of contacts.
My recomendation;
-Do not run all the contacts at the same time. Add them one by one and solve before going to the next. Like welding the pieces and doing the NDT. Each one has it’s own particularities and should work before going to the next.
-Take a second to think who is slave and who is master. Slave should have projection onto the master.
I have made some beam alignements for this reason . Some of them where not fully facing.
EDIT: The attached new mesh version has an additional shell adjustment on the plates area. It provides a cleaner result and Stress distribution. Probably less challenging too.