Post your mesh. What kind of constraints do you have between these elements? Do you have any hand verifications of what you expect to find in these areas?

All parts are bonded and symmetric for contacts since there is no good way to include Weldments. My constraints are located at where the receivers would be attached to the component and are fixed. I don't have hand calculations to show what is expected. But from my knowledge, if you have a symmetric design where force is acting in the midpoint that forces/weak spots would be in the same location as it would be evenly distributed on both sides. If that's not correct, I'd love to know so I can attack this right, this just seems so odd to have one side be this uneven in SF when everywhere else the stresses are equal on both sides.

If everything in your model is symmetric (including boundary conditions and loads) then you could analyze half of this model and apply symmetry boundary condition to the faces on the plane of the cut.

Awesome, I'll give that a try. Everything is symmetric which is why I am so confused. I appreciate the help!

Your model, contacts and loads might be symmetric perhaps there is some parts of the mesh that are not symmetric. the FEA model will find even the most minute differences if they are modeled. I think using a symmetric model makes alot of sense and it will save you time.

In my opinion the likelihood of introducing BC errors is far greater than the time saving. Also, it doesn't really explain the problem does it? Forcing an asymmetric result to be symmetric by mirror imaging is a bodge.

Also welds can be modeled. How on earth do you think fatigue analysis of welded structures is done otherwise?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376: Eng-Tips.com Forum Policies http://eng-tips.com/market.cfm?

It's not about forcing this behavior, the issue should be resolved anyway. But good practice is to make use of symmetry if possible. Another thing is that those pins should be removed or at least simplified (especially when it comes to the wires hanging from them.

GregLocock, I unfortunately don't have a full version of Fusion at my use... I understand weldments can be implemented but I have no access or don't have the knowledge on how to implement them, hence me stating that there is no good way to include them. If you have an instructional video on how to do that, that'd be great instead of asking how on earth something can be modeled... Sorry for the confusion.

And FEA way, I appreciate the recommendation and the results are more consistent with the IRL failures we are seeing.

Cheers,
LR

Just an educated guess, but - take a look at the mesh of the triangular plate on the left hand side of the last picture you posted. That mesh does not appear to be symmetric about the obvious line of symmettry. Whether this is because that part is actually not modelled with symmetry, or because the mesher just did it that way, but that kind of mesh variation will create the kind of outputs you are seeing, with the caveat below.

The caveat - you are looking at a plot of "SF", which is kind of an arbitrary, binary yes/no plot. Small, rather insignificant variations in stress can cause the yellow/red "SF" boundary to shift quite a bit. Try looking at the Von Mises stress contours instead, you may see less apparent asymmetry.