Rivets - simple misunderstanding or misled by liars?

[wktaylor]

OK, So, something about the AN470-AD4-2 Part number was totally bizarre to me... so I checked the last revision of AN470.

For 1/8 Dia rivets [#4... and for #3 and for #2 Dias also] the minimum length is [-3 = 3/16] ... so the actual MINIMUM size 1/8 Dia rivet by code-table = AN470[*]4-3

It is inconceivable to me how/why anyone would sell 'AN470[*]4-2' rivets that are completely 'off-spec'. Here is the embedded table in AN470 spec, thus...



NOTE1. AN470 rivets [Part Numbers] were actually superseded/replaced by MS20470 PN Rivets [part numbers] between 1959 and 1968
Free AN, MIL, MS. fed, etc specs are available at: https://quicksearch.dla.mil/qsSearch.aspx

In summary... an AN470AD4-2 is non-standard because it is a ridiculously short shank length... get longer rivets and use a rivet shank cutter as needed...

 

[Angusnofangus]

4-2 rivets can be used as a double flush rivet on a trailing edge.

426's, for sure. But the OP mentioned 470's. Not so good for double-flushing.

 

[Victor Bravo]

4-2 rivets can be used as a double flush rivet on a trailing edge.

^^^^^^^^^^^^^ And there you have it ^^^^^^^^^^^^^^

 

[cvairwerks]

The H rev for MS20470, in the requirements section, item 3 states that not all lengths are listed in the specification and for those not listed, length is in 1/16" increments. So a xxxxxxxxx4-2 is a viable length, but most likely not a normal stock item under NSN requirements.

 

[geraldmorrissey]

4-2 rivets can be used as a double flush rivet on a trailing edge.

That I would like to see. On thin skins you would have to dimple, that would eat up some of the rivet length on one side. It would be close.

 

[Chris.r.Ingram]

They wouldn’t make them unless someone bought them. Those folks wouldn’t buy them unless they had a reason to do so.

 

[cvairwerks]

We had the occasion to need a couple of AN426DD28 rivets for a repair task. Our special projects guys simply machined them up out of bar stock. Got lucky shooting them with a 9x gun and a 25+ pound bucking bar.

 

[wktaylor]

Cvaiwerks... the PN AN426DD28 [cited in post #27] is incomplete... a diameter code is required after DD... before 'god-awful -28' length'

Odd that You need DD [2024-T4 heat treat required] rivet in a 'god-awful -28 length'... that you hand-make. Why not just make/install a headless 'slug-rivet' made from 2024 rod-stock, SHT/quenched [ice-box] driven [bucked or squeezed*]... then mill-the tail/head-flush that is in the countersink???

NOTE1. Solid 'hard' rivets installed* in holes deeper than 4xDia tend to buckle and incompletely fill the holes. However, 'ice-box' driven in the 'dead-soft' [SHT/quenched => '-W'] condition, being the exception.

 

[cvairwerks]

WK- It's been probably 22 years since I did that repair. Diameter was 28 and length would have been somewhere around a -5. We had to plug 4 or 5, 7/8" holes that were miss drilled in a high stress bulkhead. We could only do a small 100 degree chamfer, instead of a normal countersink. Engineering called for making up the DD's and driving them double flush. Both sides were then shaved, sanded and lightly polished to match the bulkhead contour. When correctly drilled, the new hole would cut out about 1/4 or so of each plug. The repaired holes then had Ti sleeves bonded in them and ever so slightly reamed for fastener clearance. Nerve wracking, as screwing up the repair would set that a/c back about 3 weeks and several hundred manhours of tearing it down to drill a new pair of bulkheads to fit.

 

[wktaylor]

Cv... did I read this correctly: the DIAMETER was 28 and LENGTH was 5???? That is very odd for any rivet.

This sounds like You were plugging large Dia holes. In my experience these holes would normally filled with precision-Dia machined interference-fit plugs [with small/shallow 100-deg head], using 'freeze-fit' method, IE: shrink the plug using LN2, 'quickly-press-it-in-place' and simply allow it to warm to RT [expand] for a tight/permanent interference fit.

 

[cvairwerks]

Yep, plugged some 7/8" diameter fastener holes that were drilled off location due to an incorrectly located drill fixture. Freeze fit was deemed not a good repair as when the holes were correctly drilled, a portion of the plugs would be cut away...essentially cutting a crescent out of them. Frame was probably .125" thick, so the plugs were made to the AN426 dimensions and cut down after driving. We could only chamfer about .050" deep, so lots of head to removed after install.

Were we plugging an over sized hole, and the plug would be trapped on both sides, we would use freeze fit plugs., but where it's not that way, we do double flush plugs.

Biggest thing is to get people to do the work right and not make dumb mistakes that cause the rework.

 

[rdj]

I'm spending an inordinate amount of time pondering solid rivets, given that the plane in my garage is a Zenith and I rarely encounter one. Something about seeing an impossibly short rivet while shopping on Aircraft Spruce just sent me down the rabbit hole. Anyway, I'm with Billski--I think the short rivet must be used for filling useful lightweight holes. They also cost 2x to 3x as much per pound as the other longer rivets, just to make needing one that much more painful. They might make good double-flush rivets too, but I wouldn't know on a Zenith.

The Billski and F-16 examples of long-length assemblies were educational. My Bingelis books also noted their use in torque tube assemblies. Doh.

Anyway, I updated my Aircraft Spruce shopping tables with the MIL-R-47196A spec data, to see if those "Rules of Thumb" actually match the MIL specs. Turns out they do for the most part:




With the exception of the dimpled countersink min pitch and those crazy short rivets, the 1.5x exposed tail, shop head height/width, and min/max edge distance and pitch "rules of thumb" do indeed either meet or get close to the MIL specs. So not a lie after all, just a misunderstanding on my part.

Bob

 

[Turd Ferguson]

They also cost 2x to 3x as much per pound as the other longer rivets, just to make needing one that much more painful.

You need 1#? That's a lot of rivets. Or one rivet? Either way, just cut a longer rivet to length.

 

[Pops]

You need 1#? That's a lot of rivets. Or one rivet? Either way, just cut a longer rivet to length.

That is what my extra long rivets are used for. Just cut to fit the rare time you need one or two.

 

[BJC]

Am I misunderstanding something (probably), or is this 1.5X number just a bald-faced lie pulled out of thin air, and the truth is far more complicated? Or is Aircraft Spruce stocking rivets that no one can ever possibly use?

So not a lie after all, just a misunderstanding on my part.

Whom did you think was lying?


BJC

 

[rdj]

Whom did you think was lying?


BJC

"Rules of Thumb". I've been bitten by them before. It appears I can trust the rules of thumb I've read in Kitplanes, Sport Aviation et al., at least concerning solid rivets.

 

[agpilot24]

WK- It's been probably 22 years since I did that repair. Diameter was 28 and length would have been somewhere around a -5. We had to plug 4 or 5, 7/8" holes that were miss drilled in a high stress bulkhead. We could only do a small 100 degree chamfer, instead of a normal countersink. Engineering called for making up the DD's and driving them double flush. Both sides were then shaved, sanded and lightly polished to match the bulkhead contour. When correctly drilled, the new hole would cut out about 1/4 or so of each plug. The repaired holes then had Ti sleeves bonded in them and ever so slightly reamed for fastener clearance. Nerve wracking, as screwing up the repair would set that a/c back about 3 weeks and several hundred manhours of tearing it down to drill a new pair of bulkheads to fit.

Just curious, my 4x rivet gun does the little ones I have needed from glider to gulfstream…3-6 diameter. What kind of gun will drive a 28-5? I’m certainly not disputing it was done, but that one will take quite a punch to drive! And quite the bucking bar. I’d love to see the tools!

 

[BJC]

"Rules of Thumb". I've been bitten by them before. It appears I can trust the rules of thumb I've read in Kitplanes, Sport Aviation et al., at least concerning solid rivets.

I never thought of a “rule of thumb” as a person.


BJC

 

[OBwanBD5]

So, most of the solid rivets we use are AN426 or AN470. Here's one carried by Aircraft Spruce: AN470AD-4-2, specified as 1/8" diameter x 1/8" length which seems to match the AN decoder ring.

Now, pretty much every experimental aircraft site I visit states that the 'exposed tail' of the rivet (ie the amount sticking out the backside of two sheets of aluminum) should be 1.5 times the rivet diameter, which my calculator tells me is 0.188" for a 1/8" diameter rivet. My calculator also tells me that getting 0.188" of 'exposed tail' from a rivet only 0.125" long requires magical aluminum sheets with a thickness less than zero (-0.063" to be exact). Even more trustworthy sources than the EAA and HomebuiltHELP such as Hanson Rivet specify a 'clinch allowance' (which seems to be the same as 'exposed tail') of 0.125". That at least drops the magical sheets down to zero width sheets.

Am I misunderstanding something (probably), or is this 1.5X number just a bald-faced lie pulled out of thin air, and the truth is far more complicated? Or is Aircraft Spruce stocking rivets that no one can ever possibly use?

Bob

Gidday Bob,

I'm an A&P Tech from NZ with 40 years of Experience including 15 years of teaching my Trade. You can buy AN426 AD4-2 (MS20426AD 4-2 Rivets are the same) Solid Rivets, and they are 4/32" in Diameter (ie 1/8") and 2/16" in Length (ie also 1/8") of which around 0.040" is taken up in the Countersunk Head, but due to their very short shank length, they would mainly be used on double Countersunk Trailing Edge application as they wouldn't form a correctly sized Tail (or Shop Head) which is typically 1.5 D wide and 0.5 D high, which typically needs the 1.5 D protrusion of the unformed shank. 4-4s were typically the shortest rivets we formed, although a 4-3 was sometimes safer. As to Trailing Edges, even in larger Aircraft (eg C-130 Hercules) the Trailing Edge remains thin, and is often a Flight Control with skins as thin as 0.016" at times, so dimpled Countersinks are required, and due to the thin skins, there is no need to use 1/8" Rivets. I would recommend using 3/32" Rivets instead of 1/8", and if the skins are riveted together, consider using AN 470 AD3-3s, forming the bucked tail, then finishing off with double Universal snaps to round the formed tail. The amount of drag caused by the raised heads is negligible. If you want it to be completely flush then use the AN 426 AD 3-3s but put a strip of 0.040" Aluminium between the two skins to give rigidity, and allow the double Dimple with the tails milled flat. Some Trailing edges actually used a thin Double Arrow Extrusion which allowed the skins to be fitted flush, or a simple Triangular Extrusion. (something like this...)

 

[cvairwerks]

Ag: We used a very old Chicago Pneumatic 9X pistol grip gun and a bucking bar made from a piece of 2.5" bar stock, about 18" long. Same combination we used to drive a group of #5 dia. Monel rivets to install a titanium duct all the time. Main thing was to drive just enough to fill the chamfers and no more. I had a really good shooter teamed up with me that day and we made it happen without any failures or cracking. Took us longer to shave, profile and dress the repairs than it did to shoot them.