Hi everyone,
I am modeling the attached connection using FEA.
Goal is to Test if the connection will support a given weight.
All calculation should be done in the elastic domain. No plasticity is allowed.

When looking at each of the part seperately, things went fine. But when modeling the assembly I doudbted the FE reults.

I want to double check the assembly resistance by analytical calculation.

Could you please suggest the easiest methodology to follow.
Do you have any tips to improve the FE modeling for a such problem.

Please note that many manufacturing details were removed in the 'sketch'.
Regrads

"All calculation should be done in the elastic domain. " ... ok, linear FEA should be good enough to solve this ...
"No plasticity is allowed." i'm 99% sure that you'll have plasticity ... the bolt is cantilevering the red piece, carrying the moment across it's threads ... it won't take much load before there's plasticity at the root of the thread ... it doesn't mean it's failed, it won't be noticeable, but it would happen.

if you're relying on clamping the faces (due to preload in the bolt) to react the moment, the same argument applies ... not much preload before the thread root yields.

you can take some moment out as "heel and toe", as a couple between the bolt and the lower corner of the red piece ... same arguement applies.

not sure how you "When looking at each of the part seperately, things went fine." ... the most real reaction would be to react the moment as "heel and toe" couple, which you can't do as individual pieces (ok, can't do easily). for FEA, you'll have a contact problem, and somehow you should model the preload in the bolt ... i guess you could say the preload is small, and the "toe" reaction is at the base of the red piece.