Does Minimum Edge Distance Apply to Nutplate Rivets? 7

[DrRocketSurgeon]

I've talked to multiple DERs who have adamantly held that rivets responsible for holding nutplates in place DON'T qualify as fasteners (their only job is to hold the nutplate against torqueing during assembly), therefore standard edge distance requirements (say, 2*D in the AC 43.13 or the Bell SRM) don't apply. It's accepted as common wisdom among engineers around here, and the logic makes sense.

However, I can't find a single document or manual that spells this out. As is custom in aviation, everything we do must be backed up. Can anyone out there show me where it's written that nutplate rivets are exceptions to edge distance rules? Or, if this isn't true, can anyone show me where it's spelled out that the opposite is true?

TL;DR Does 2D edge distance apply to the rivets on nutplates?

Thanks in advance for your help!

- Pat

 

[WKTaylor]

DrRS...

"...It's accepted as common wisdom among engineers around here, and the logic makes sense."

Dear God say this isn't so. This 'lesson' was taught the hard way... loss-of-life and/or major structural cracking damage... decades ago.

This is a classic stress concentration buggar, that any liaison, design or stress engineer worth a pay check should be able to spot and fight tooth and nail to prevent.

Peterson's Stress Concentration Factors is a place to start.

Assume a 2-lug NP is clocked at 45-deg to an edge, so that 1-NP rivet hole has a 2D edge-margin, the bolt hole is next in-line and the last NP rivet farther away still. Here are the stress concentration factors to consider.

Nut plate rivets [usually a solid 3AD or 4AD] will swell-fill hole and provide a benign stress field... but not very effectively. Usually these holes are rough drilled, quick-pass in-out of the hole. Or worse Yet these NP rivets are both the tiny blind-rivets that have a pull-thru stem and barely [if-at-all] fill the hole.

Then 45-deg away from the edge NP rivet is the larger load-carrying bolt-hole, usually class 1 or class 2 clearance-fit.

Then 45-deg away from the bolt-hole is the last NP rivet install.

OH have I asked the loaded question???: what is Your NP rivet spacing 'regular' or 'reduced'??? Regular is 'tight spacing'. Reduced is 'god-awful-tight spacing'.

Oh have I mentioned that when holes align at a 45-deg or greater angle to the edge [in an along-the-edge-tension-field] that they effectively come into in-alignment with each other for area-out calculations... not to mention Kt increases due to alignment?

What about when the bolt is repeatedly installed/removed and the core hole bore is scored axially… or by rifling due to bolt threads??

Have I mentioned that there are structures rules in place for DoD aircraft forbid primary/flight-critical load carrying members [in current designs] to have any non-permanent bolts/nutplates fastened to it [allowing component installation/removal] for any purpose. IF this is necessary, then a secondary component [plate/sheet/etc] must be permanently fastened to the primary member… and that the secondary member is then allowed to have the bolt-and-NP holes in that member [carrying only panel loads to the primary airframe]. IF this member fails by cracking, it will not [should-not] transfer that cracking damage to the critical protected member.

NOTE. There are exceptions to this DoD rule due to the emergence of rivetless nutplates installed by cold expansion of a core bushing-like-element into the bolt-hole, ONLy... such as the FTI TukLoc or the ForceTec rivetless nutplates. The bushing element is cold expanded into place, permanently, protecting the hole from any damage caused by bolt insertion/removal; and eliminates the Kt of adjacent NP rivet-holes. IF the hole in the rivetless NP is damaged then it has to be carefully drill-out [press-out would be virtually impossible].

The reason XXXXXXXX F-Xs are no longer in the USAF inventory is simple: this nightmare situation occurred on the upper-cockpit longeron, aft-end of the cockpit, near the pressure bulkhead [as I recall]. Clocked NPs were inadvertently installed in production [by the shop]… in-lieu-of longitudinally clocked nutplates as intended by the designers… on this flight/load critical member… just to fasten an access panel in position. Fatigue cracking through one of the bolt-holes progressed to both 45-deg adjacent NP rivet-holes, which then progressed to edge break-out and then catastrophic section-failure of the longeron. This occurred in low-level high speed maneuvering flight: the cockpit with the student and instructor departed [fell away from] the aircraft on a ballistic trajectory onto the ground [desert] and the crew was killed [no time to react and tumbling "G" forces prevented reaching the ejection handles to escape]; and the aft airframe [what-was-left-of-the-fuselage, the wings, running engines, the stabilizers, etc] then tumbled/twisted in an a bizarre aero/ballistic descent to the ground.

When the cracking was evaluated, the cause was immediately evident to investigators… fatigue cracking thru the clocked NPs, on critical structure, as noted. Hence a fleet-wide F-5 inspection of the cockpit upper-longeron was accomplished. Almost every F-X [two-seat ‘tub’] was affected [found cracked]… and USAF/pilots lost confidence in the design. FMS acft have had many inspections and major mod/beef-up to mitgate this problem in their fleets.

The end-message here is as follows.

IF these NPs are installed [clocked as noted] on a primary/flight-critical member, in a tension/tension-fatigue environment... then fatigue cracking-failure is virtually guaranteed. This could be worse if the primary member has SCC or poor corrosion resistance issues to begin with.

IF the primary flight-critical member is isolated from this ugly NP-clocking situation... then no sweat... the secondary member may/maynot fail, but the failure won't be catastropic, IE: transfer of cracking damage to the primary member won't occur.

CAUTION: IF the secondary member is considered as additive to the primary member for strength and durability... then the design is grossly flawed and failure could still occur by damage transfer.

Your sense of the magnitude of the problem represented by the "group-think-common-wisdom" where Your work is probably accurate accurate.

Regards, Wil Taylor

o Trust - But Verify!
o We believe to be true what we prefer to be true.
o For those who believe, no proof is required; for those who cannot believe, no proof is possible.
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion"]
o Learn the rules like a pro, so you can break them like an artist. [Picasso]

 

[tbuelna]

Nice post wktaylor.

Just to help illustrate the situation I believe you were describing, below are the dimensions for a common floating shear-type (counterbored) two-lug nut plate used on aircraft. If you subtract 1/2K from C and divide by D, you'll note that the distance from the nut center to the edge of the rivet hole is less than 2D, and the panel would have the same size/configuration of holes. These nut plates are attached using flush rivets that use a c'sink on the far side of the panel.

Most aerospace OEMs I have worked at had their own design and documentation standards for various types of nut plates, including things like lug orientation, spacing, edge margin, panel hole tolerances, etc. Typically, as a minimum, we would provide dimensions to the fastener center of the nut plate, call out the p/n of the nutplate and rivets, show a pair of "plus" marks to indicate orientation of the lug rivet holes, and provide a flag note stating "install per company process X". The manufacturing procedure would provide details on how to prepare the panel (drilling, countersinking, deburring, corrosion protection, riveting, inspecting, etc). Since the rivet holes in the panel were often produced manually using a drill jig piloted at the center hole, the manufacturing procedure also included a tolerance for alignment of the nut plate lugs, usually something like +/-10deg of what was shown on the drawing.

Lastly, we should also consider whether the (lugged) nut plate is a floating or fixed type. Most of the nut plates I recall seeing used on aircraft were floating. But with a fixed nut plate, if there is a condition like loss of axial preload that results in a radial force being transferred from the bolt to the nut plate body, the rivets will be subject to that force until clearance between the bolt and panel hole is taken out.

You might want to read thru NASA-STD-5020. It gives a detailed discussion of the analysis approach used for critical mechanically fastened joints on spacecraft.

 

[BigBoyBob]

DrRocketSurgeon, it seems that there may have been some sort of game of "telephone" going on where you work. I can see how it might have started, since it is true that the nutplate rivets only function to hold the plate against bolt torque (in optimal operation, that is; tbuelna brings up a good point with a possible malfunction).

But Wil's analysis is (as usual) spot on. The area out and stress concentration problems are potentially huge. How many of those DERs would be willing to plug mislocated holes that close to a fastener and an edge and not worry about the edge margin?

Your instinct to "find it in writing" somewhere is a good one. One good reason to have ready access to specifications is exemplified in your post. Human memory is notoriously unreliable. Like you said, it's standard practice to back everything up. That isn't just to CYA, it's because we do what we do by standing on the shoulders of those who came before us. We don't have to reinvent the wheel every time we go to work, but we do have to pay attention to what the wheel-inventors did and didn't prove.

Good luck!

 

[crackman]

DrRocketSurgeon

I will add my two cents for what it is worth. I whole heartedly agree with Will and have had many similar experiences during my career. The only thing I can add is as a DER I have never taken the 2D edge distance as NOT applying to nutplates and I am particularly sensitive to this since I am a DTA DER. In fact, whenever working on old aircraft which have poor details such as the "mickey mouse" nutplates (45 degree) and those oriented towards the short edge, I always try (as best as I can when I am allowed) to have the nutplate holes either trimmed out and repaired if too short or plugged and a gang channel added. I my own opinion there is no reason to have nut plates ever riveted directly to primary structure which is critical. In fact, there are numerous options today including the rivetless nutplates.

Anyways, good luck.