Modelling clearance holes MSC Nastran/patran sol 400

[freeder3]

Hi, I have a need to model a group of fasteners which have clearance. I was modelling the fasteners using RBE3 with a CBEAM connected to my plates which have a solid mesh including the holes.i have contact between the plates. This method worked well hower was overly conservative for shear loads.

This approach did not account for the clearance and therefore always had shear transfer. Can anyone offer any advice as to how to accurately simulate the effect of clearance?

I have tried modelling the collar, tying that to the beam and specifying contact with the beam and the hole. This works great for the shear loads , shear loads are only being transferred once the gap is closed but axial forces have increased nearly 3 fold!

Can anyone offer advice? I can of course incorporate a solid mesh of the bolt but this would increase the amount of contact interactions considerably and run time.

Many thanks all!

 

[rb1957]

when you say "overly conservative for shear loads" was the shear component causing you to upsize the fasteners ? Or is it "just" the shear loads are bigger than you'd expect (with the clearance holes) ?

it is "easy" to reduce the stiffness in the in-plane directions ... with a CBUSH and GAP elements. But, how will these loads get across the joint ? How do these in-plane loads travel in the real world ?? Are these loads likely to be transferred by friction (we use some sealant to fill the space between plates (to keep out moisture) and this is a really good adhesive !) ? Or are you thinking of a "monte carlo" simulation ... N different gaps at N different fasteners ?? You may "wheel out the pope" ...


... and declare some fasteners shear effective, and others not. You may design this in, with smaller clearances at the nominated fasteners.

But I suspect this is another example of overdoing FEA (because we can) and not apply hand calcs to simple problems.

 

[SWComposites]

why in tarnation do you want to model the fastener holes? you are not going to get a correct local stress distribution with RBE3 or RBE2 connections to the holes, nor even if you try to incorporate contact between the fastener and hole. just eliminate the holes in the model, connect with springs/bars/cbushes at the fasteners, get the fastener loads, and compare to basic bearing and fastener shear allowables.