Which is stronger: AN 426 or AN 470?

[claycookiemonster]

Refresher: AN 426 are the flush head rivets, and AN 470 are button heads.

Not considering the appearance or drag coefficients of either. I could also believe that there may be a slight difference between dimpled versus countersunk 426's. Not talking about that either, for now.

But if strength is all that matters, which would you use? It looks to me like the button headed 470's may have a greater area against the SURFACE of the metal being riveted, so their strength would be more in compression. Dimpling for 426's would seem to sacrifice some direct compression strength, but perhaps make up for in in the shear resistance the dimples provide against each other.

Hey, football is over. I've got to think about something.

 

[Mel]

Hey, football is over.

Yeah, sure.....Football is over.....Just like road construction is over.....NOT!

 

[David Paule]

The shear strength of the shank is the same for the same diameter.

The AN470AD rivets are stronger in tension than their AN426AD counterpart. Generally, rivets are designed for shear loads, which usually dominate. But some assemblies do have tension loads. I don't know why you're referring to compressive loads since for those, the mating parts themselves carry that in direct contact.

For the flush rivets, a dimpled joint is stronger in shear than a countersunk joint.

For all rivets, the thickness of the pieces being riveted and their edge distance play a major role in determining the joint shear strengths.

Bottom line, follow the plans. And if you're anticipating building an RV-15, especially follow the plans!

Dave

 

[Auburntsts]

Not an engineer, but U.S.A.F./Navy T./O. 1-1A-1 Structural Hard-ware https://www.tinker.af.mil/Portals/106/Documents/Technical Orders/AFD-082416-1-1A-1.pdf

in table 5.3 seems to indicate, if I understand it correctly, that regardless of a rivets alloy, that for the same size rivet, AN426s in a dimple are strongest in shear, followed by the AN470 Universal head, with 426s in a CSK the weakest.

 

[claycookiemonster]

Yes, I stand corrected. I meant tension, but said compression.

Yes, this sprouted from a discussion over in the RV-15 waiting room about what appear to be significant 470's on many parts. The follow-on conversation where the cross over in aerodynamic drag of 470's versus the extra labor of 426's will be fun to watch when the -15 kit is debuted - if the button heads remain.

Aesthetically, there are places where each shine. If you haven't made a Van's Toolbox, it has both and is gorgeous.

 

[Marc Bourget]

I recall reading a report in an early Thorp T-18 Newsletter where a builder realized a 15 MPH gain (to 150 IAS) when he "recreated" his (O-290GPU conversion power) T-18 to flush rivets.

In a WW-2 Detail Design Drafting Textbook ,from Northrup, they reported the power to maintain a (IIRC) 180 MPHcruise speed in a DC-3 sized twin engine transport was cut more than 50%.

It would appear that the time you spend dimpling will be made up very quickly in the first few cross-country flights!

 

[lr172]

Not an engineer, but U.S.A.F./Navy T./O. 1-1A-1 Structural Hard-ware https://www.tinker.af.mil/Portals/106/Documents/Technical Orders/AFD-082416-1-1A-1.pdf

in table 5.3 seems to indicate, if I understand it correctly, that regardless of a rivets alloy, that for the same size rivet, AN426s in a dimple are strongest in shear,

and I believe it is significantly stronger. Because it is not really the rivet preventing the shear but the base metal inside of base metal that is preventing the shear. The rivet tension is required to keep the metal in metal joint from separating, as that is what has to happen to allow shear. I was always under the belief that with dimple to dimple connects, the base metal offsets were in shear and the rivet is in tension (shear force is trying to push one dimple out of the other. NOT an ME so just Larry's thinking, not fact.

 

[PilotjohnS]

Which is stronger?

To me it doesnt matter. Both will cause the metal to deform before breaking, so therefore whichever is stronger only matters in an crash.

I think the real question is which is more durable in service? Which will last longer and stay tight with repeated loads, flexing, etc? In this case I would guess the dimpled rivet would be best. I suspect the pop rivets are the worst in this respect, and is why I dont want to see the RV 15 with pop rivets; I cant see pop rivets holding up with thousands of back country landings. But I have no real data and trust that Vans does so I will believe Vans if they say the pop rivets will hold up. JMHO. YMMV

 

[JonJay]

I recently saw a publication with charts covering many types of rivet alloys, types, joints, etc…. Now I can’t find it!
Perhaps one of the engineers has it in a handbook or something.
It covered machine countersinking, dimpling, protruded head, etc…. In convenient and easy to read charts showing shear strength for each.
Anybody recognize or have it?

 

[HFS]

What I Have - by Gil Alexander in 1995

Here is a compilation of "Countersinking versus Dimpling" by Gil Alexander (an early RV "adopter") from 1995, as excerpts from MIL-HDBK-5F (of which I have one - thanks Dave Paule).

I have done hundreds of shear test to failure of different joint types over the years - just for my own edification. What became very clear is that dimpled joints fail in shear by the dimples moving, one relative to the other to create "separation" of the two services, causing a recess in the flush head, which in turn accelerates the failure of the joint - not yield (for design purposes), but failure to separation. The attached pic shows a close up of this phenomenon after it occurred, but before complete joint failure.

Sorry about the orientation of the first two - I tried everything I know to make these pics "upright" - but nothing I did worked. Maybe someone can tell how to fix this - it seems to happen a lot with this platform.

HFS

 

[David Paule]

As HFS mentioned, often joint failures include both tension and shear, even if the load is pure tension. Since shear usually includes some bending, the actual failure itself can get rather complicated. The approach successful designers use is to rely on the test data in MIL-HDBK-5 or it's later versions. That data is based upon actual tests.

Dave