Tube Joint Clip Analysis

[jetmakerbrit]

Could somebody please suggest how to analyse a clip joining tubes together in a rack I'm trying to substantiate. The clip joins two tubes perpendicular to each other. One tube slides through the end of the clip and the other slides inside the clip. The attached image will hopefully explain more clearly.

Basically there's a moment being applied to one tube causing the clip to act like a socket to transfer the load to the other tube. I've looked through the text books I have for a socket burst or tear out type of calculation but so far nothing appropriate.

Another engineer performed a bending calculation across a section at the centre line of the leftmost fastener hole but this seems to ignore the mechanism by which the load is transfered between the two tubes.

Many thanks for your consideration and assistance in advance.

 

[GBor]

Which tube has the moment? I'm assuming that the two fastener holes on the right of the top view go all the way through the tube that slides in to the clip and the other fastener hole goes through the tube that slides through the end. I'm also assuming that the tube sliding into the clip is the one that has a moment applied some along some plane.

I think the other engineer was on the right track, but this may not be all that needs to be considered. Did you check shear on the fastener and bearing on the tube sliding through? Also, is it a pure moment so you can check torsion on the through-tube?

Don't forget to consider stress concentrations, and DT if this is a fatigue load.

You may want to resort to FEA for stress concentration factors, but I would hand-calc it for all the different potential failure modes since it is difficult to see which one is the most critical.

 

[jetmakerbrit]

Thanks GBor your assumptions are correct. Looking at this structure on A/C, I realised this clip pivots on the single fastener on the left and would therefore not carry a moment into the other tube. This rack was modeled using beam elements in Nastran so looks as though the moment DOF need to be released and the model rerun.

 

[rb1957]

"socket" ... not likely, it looks a flexible as an olypmic gynmast (sp??) ...

the tube attached with the two rivets "can" transfer moment (i'd fix the model so that the moment is released, or just ignore it if it's small), the loadpath out of this (IMHO, cheap piece-o-crap) clip would be a couple between the rivet in the other tube and the edge of the clip (as it bears up against the tube) ... neither loadpath nor the clip are very effective in bending, but i guess it is the worst case scenario.

if you wanted to make it more effective, add a tube over the top of the two rivets, which would make this part of the loadpath more effective, but the loadpath out into the other tube is still pretty wishy-washy.

i take it all the rivets are blinds ? (again, not great of moment)

 

[GBor]

I would say your Nastran engineer needs some training unless this joint was part of a much bigger model. If that is the case, then you should use the loads from your global model and look at this locally.

Have to agree with rb, this connection looks like it belongs on a hang glider, not an airplane.

I don't think I understand the source of the load, where and how it is applied. Think you can either add that to the sketch or give a description?

As far as substantiating this, it may be time to either 1) get a DER involved, or 2) tell whoever you are working for that this connection is horrible and needs to be redesigned. I know I would prefer to avoid it and the only reason I would put it on a glider of any kind is because people who are willing to jump off a high clip under a piece of celophane (sp?) would probably enjoy the thrill of expecting their wings to collapse at any moment.

Garland E. Borowski, PE
Engineering Manager
Star Aviation

 

[jetmakerbrit]

This clip is from an electronic supply rack on the C-130. The loads emanate from the weight of the LRUs and crash worthiness load factors. This joint is one of many similar joints in the rack. The fasteners holding this thing together are NAS603 bolts. I'm sure the moments are due to erroneous modelling - which is no surprise looking at the rest of the analysis. Thank you GBor and rb1957 for your thoughts.

 

[GBor]

That actually changes things a bit. Is the through-tube vertical? And the bolts are horizontal and penetrate through the entire tube?

I do think I understand the moment in this situation having designed a rack or two or three or more. I've seen racks substantiated this way before and have even substantiated one or two that way myself. You still need to check the shear through the cross-section, make sure your loads are conservative unless you know the cg of the equipment (probably not centered so that both rails get equal weight distribution), and work through the statics. I'm still not fond of the connection...would like to see another bolt or two in case there is an unknown crack in one, but they probably can handle a shear load.

I will retract (most of) my negative comments and say bending and shear on the clip...possibly FEA it since it really doesn't fall in to the category of "beam theory".