End Fastener Loads in Combined Splice/Doubler Cases

[SAITAETGrad]

Continuing from thread2-95161, I have been trying to find a way to use Swift or other resources for the determination of repairs to structural members. For example, say you have a heavy, axially longeron member with a crack or a cut-out on a free end. You place in a repair angle that transfers load as a splice over the lost strength section but also has a length where it unloads the longeron.

Swift is primarily intended as a resource for fatigue analysis of skin repairs and is extremely helpful for pressurized aircraft. I'm a little lost however when it comes to analyzing other structural element repairs.

 

[rb1957]

as you say, swift is intended for fatigue analysis.

for static analysis i'd consider the ultimate strength of the rivets/bolts. if some of the original fastneres are ineffective ('cause maybe you've shortened the splice to remove damage) then you have to provide another way to get load out of the terminating structure and into the splice. you can replace the lost piece of the splice (essentially now a packer) and add a repair angle. this'll pick up the now ineffective fasteners and transfer the load into the splice, so it'll have to overlap the splice. it'll put these overlapped fasteners into double shear (the splice is picking up load from both sides) so their allowable is higher and everything should be ok.

from a damage tolerance perspective, this'll probably be transparent. the terminating member is still seeing the load leave in much the same manner.

issues ... the fasteners into the repair angle are going to be longer than the original ones, and so fastener bending is something to be considered. if you have to up-size the replaced fasteners, this'll change (slightly) the load distribution between the fasteners; something else to watch. also (if you up-size the fasteners) check the edge distance.

and if you're getting lost, find a compass and a point of reference

good luck

 

[SAITAETGrad]

Sorry if I wasn't clear. My concern is with regarding fastener end loads (i.e. DTA). Thanks for your help. Any further comments would be helpful.

 

[planedr]

Set up a small matrix and do the math! Come on do Simon proud.

Niu has a very good example on page 234 of Airframe Structural Design. Depending on how deep you want to go or need to go, you can use Niu example calcs and be done with it. He covers 3 to 5 fasteners which would be typical for splicing most members.

Again though what year did you grad?

 

[rb1957]

your comment "Swift is primarily intended as a resource for fatigue analysis of skin repairs ..." lead me to think that you were also concerned about the static strength of the repaired joint. my post was saying that for static analysis you don't Need to consider the fastener stiffness; from the ultimate strength point of view the outer fastners will yield and load will redistribute untill all the fastners are working at the same level.

from the DT perspective, unless you're upsizing the fasteners, or extending the splice, there shouldn't be any significant change in the distribution of load transfer out of the terminating member; you'll have the same number and size of fasteners. One thing you Could consider is that you've changed the inspection of the less critical fasteners on the splice (by covering them with the repair angle) ... the structural inspection program of the splice is going to be based on the most critical end fasteners, but should require inspection of all the fasteners. now that you've added the repair angle (on top of the splice) you'll need a different inspection procedure to inspect the splice.

if you are changing these (the size and/or number of fasteners), then the job is quite complex. yes, as planedr notes, there is a procedure in Niu and you can build a compliance model ... it is going to be complex, and take some time, and the biggest factor affecting the analysis will be the bolt stiffness model you assume (Swift, aka Douglas Aircraft, and Niu present very different models). Niu presents quite a nice fatigue analysis (determining an effective Kt including the effects of load transfer) which you may use as substantiating the threshold inspection (some certifying people like this, some don't) For the repeat interval, didn't depend on how big the detectable damage is. if it is large, then you could rationalise that the situation is unchanged by the repair (that the crack would make the end fastener ineffective). if it's small then you'll need to do a crack growth calc as well, which if you're new to DT and repairs isn't something to be expained in these forums ... it'll take too long to go over all the considerations.

i'd suggest that there are many ways to rationalise your solution without a detailled analysis, and if you're in the repair world you're probably already late. i sympathise, you're in a difficult position; repairs need a lot of experience to understand the significant features and you never time for analysis you don't Need. it sounds like you're fresh out of school and doing the best you can with the resources available ... your company should be helping more.

good luck

 

[Sparweb]

SAITAETGrad:
You could also try sitting down with an engineer from the regional office. It's what I did a few years ago. The ones I know are very knowledgeable about these cases, and without doing the work for you, they'll walk you thru the steps you need to take, plus hand you half a library to leaf thru on the subject. It's a good thing to figure out before the repair job comes up. As RB1957 pointed out: there's no time to figure it out while the plane's AOG.
If you don't recognize my name, I'm a fellow alumni (1999). Call me; I'll buy you lunch.

Steven Fahey, CET

 

[planedr]

Steve,
Me too 1997. I'm back in Canada (Calgary) next week, buy me lunch?

I worked on a frame analysis with a 12 bolt stepped fitting. We used a matrix to chew through the numbers not to bad really except for the size, it was similiar to what Niu did.

But if SAITAETGRAD needs a DTA, he need's that and then some. TCCA will give you 12 months if they buy into the basic method of the repair / design. Do you really need a DTA though? or can it be shown to be equivalent to something similiar?

 

[planedr]

Sorry I meant to add

Al Feys CET,EIT

 

[Sparweb]

Sorry Al, I haven't been keeping up with my forums! If you're still in town call me at:
two-five-zero-eight-zero-two-seven (the folks who run the forum don't like obvious phone numbers or e-mail addresses, and for good reasons).

One more comment for SAITAETGRAD: Your original post asks about an axially loaded longeron with a crack or a notch near a free end. I've just had my head inside an abused Piper. There's a notch in a frame flange right where it gets spliced to the other half of the frame. The net conclusion is that the flange of the frame is unloaded in that area - so no worries about the notch. The splice is doing all the work at that section. By standing back and noticing there's very little load in the flange in that area, I've saved myself the trouble of analyzing anything. Depending on where your notch/crack problem is from the end of the longeron, you may be able to draw some conclusions that contract the scope of your analysis.

On the other hand, the way you pose the question, I think you have a longer-term project in mind - perhaps just preparing yourself for the inevitable. Having some of these problems worked out in advance can be a big advantage when the AOG situation falls on your lap.

If you're still working on this one, I think I could dig up an old Boeing document with end-load charts.

Steven Fahey, CET