bypass stress and fastener load interaction

[sszuch]

I am wondering if there is an analytical method developed to include both bypass stress with fastener load for static and fatigue analyses. A simplified description of the problem is an axially loaded flat plate in tension in the X-direction with a perpendicular fastener load (bearing)applied in the Y-direction. I have made a Mechanica model, but it seems unclear which is the driving factor or how much influence one has on the other.

Currently I am attempting an equivalent strength approach comparing an oversize hole to the nominal hole size. Lug analyses (fastener load) and Anet / Agross (bypass) yield a positive margin, but there is clearly an increase in fatigue stress levels. The only methods I have available allow me to compare one case at a time. The bypass fatigue case shows an increase in maximum stress levels, and the fastener load case shows a decrease. Is there a method to find the net change in fatigue stress levels?

 

[40818]

You seem to know the question your trying to ask, but it doesn't come across. Please ask a specific question, as you dont give enough relevent information for a decent answer, your post jumps along with tidbits of relevance. I dont understand what your seeking guidance on to be able to help.

 

[sszuch]

Sorry, maybe I included too much information.

All of the anaylitcal methods I have come across so far deal with bypass stress and fastener loads that are in the same direction, i.e. two tension members fastened together.

Is there an interaction equation for an axially loaded member with a fastener load applied perpendicular to the direction of the bypass stress?

How can I predict fatigue life with this arrangement? I have not found any stress concentration factor curves for this configuration.

 

[rb1957]

since you have a model already, quick and easy 1st step would be th model the axially loaded faster (axially by-pass and axial pinload) and compare this to you loading (axial by-pass and transverse pinload) ... personally i don't think there'd be too much difference.

but i have a couple questions about the model, presumably it is non-linear (since a linear model would produce unrealistic local stresses associated with the load transfer). As such i think you've got a good estimation of what's going to happen, maybe run some models to verify that your results are real (apply only the axial load and even a linear model should show a stress concentration of 3 for your fatigue loads, a non-linear model would probably be needed for your ultimate loads). Apply the transverse load on its own to see what happens.

 

[sszuch]

The model is linear and is only used as a reference to verify that the stress distribution looks as anticipted. However, the prime contractor does not accept Mechanica models as an acceptable form of analysis, which makes this a little more difficult. The models were initially created to see if there was any appreciable difference in stress levels with the applied fastener load at a given bypass stress. This was to settle any concerns I had due to a short edge distance configuration.

Two seperate models were run with and without the fastener loads. The models show that the stress levels increase approximately 20% at the edge of the plate with the applied fastener load over the bypass only case.

Being that I cannot use these models for analysis, I was wondering if there were any other references out there that might cover this configuration. My library and experience is limited to Bruhn, Peterson, and Roark.

 

[rb1957]

could you use the models to verify that the maximum stress for the transverse pinload (your configuration) is much the same as an axial pinload (typical analysis configuration), then analyze it as such ?

trying to get sensible numbers out of a linear model is difficult since there will be localised yielding near the hole, which is permissible ('cause it isn't "significant").

i'm having trouble seeing where a short edge distance could arise ... maybe you have a large sheet with tranverse stiffener applying a transverse load; so you could have short eD on the stiffener where there isn't the longitudinal stress (in the sheet). In the panel i'd combine the longitudinal tension stress and the transverse shear stress to check the ultimate condition. i'd check the fatigue condition by assuming that this loading is equivalent to axial by-pass and axial pinload which i think is conservative.

 

[prost]

If the fastener load is perpendicular to the applied load, how does this stay in equilibrium? what force applied where is counteracting the X direction force? and what force coming from where is causing the Y direction fastener load?

 

[sszuch]

The actual part in question is a lower wing plank with integral risers. At the rib caps, an H-Clip attaches between the risers of the plank and to the lower rib cap tabs. The short edge distance condition is caused by misdrilling the fastener hole in the riser common to the H-Clip. I made the assumption that the skin resists shear in the panel and the risers resist bending of the plank. This particular riser location is a few bays away from the plank splice, and the shear was neglected based on these assumptions. The pin load comes from fuel pressure and aero loads on the plank. The only loads I have available are the axial stresses in the plank. I appreciate the help so far, and maybe this will clear up the confusion I've created.

Steve

 

[rb1957]

ah, so you've got axial stress in the stiffener, and an out-of-plane load (normal to the axial stress; just the other normal to the one i was thinking of, but explains the "lug" calc mentioned) due to pressure, rib crushing, etc; and you've got a short eD for this load.

Statically, i'd be surprised if there is a real static problem. i'd expect to see some localised plasticity and minor load redistribution.

Fatigue, you can use Petersen to find the Kt due to the short eD and factor the crack-at-a-hole geometry solution by (the increased Kt)/3. To include the rib cleat load, i think your model is the best tool to evaluate this; if your customer doesn't like this, ask him what will work, what was done for the cert analysis (static and fatigue), passed by test ??

for myself, i'd repair the short eD hole in the stiffener with an interference fit plug, and make a new rib cleat, picking up the stiffener at a more convenient point. but no doubt you'll say that the part has gone downstream and you're playing catch up. if the cleat is attached with a rivet, maybe using a Hi-Lite pin (with it's small amount of interference) would be a compromise ?