What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

strength of riveted joint test

jcarne

jcarne

Well Known Member
Patron
I don't think I have seen this article mentioned on the forum yet. Very interesting read on tests conducted on rivet joints. I know these were tested to failure and yielding would certainly occur before this but really impressive numbers for only 6 rivets. I guess this is why Van's and almost everyone else seems to agree that a less than perfect rivet here and there is no big deal. What do you guys think of the results?

https://glasair-owners.com/resources/strength-of-riveted-joints/

Highly suggest reading the whole thing but results are in two tables at the bottom.
 
These results have been reported many times, since the NACA studies during WW-II. That being said, I compliment you on the effort! Unusual and laudatory! Good Man!

I do have concerns with the slightly "less than perfect" query.

It's subjective and could possibly lead to a slippery-slope condition.

A little bit of "warm up" or practice before each riveting session will get things "in tune" and acceptable rivets become an easy result!

As the Army slogan says: "Be all that you can be!"

Onward and upward,
 
Marc Bourget said:
I do have concerns with the slightly "less than perfect" query.

It's subjective and could possibly lead to a slippery-slope condition.


Onward and upward,
Click to expand...

I fully agree with this, not saying do a crappy job, just pointing out the unnecessary risk of drilling one in a thousand because the results you end up with may be worse...
 
Last edited:
Rivit

I work with rivit joints on an almost daily basis.
Good article, but Boeing and other manufacturers
Will sometimes use a ?shear? rivet hole.
Like drilling a #28 hole then putting in a #30 rivit.
That makes the shank swell slightly, witch in turn
Would make for a stronger shear fastener. (In theory)
 
Interesting real world test, the handbook design strength of a double dimpled .032 sheet with single shear is 217 lb. The lowest test result was 26% higher than this, meaning that the design strength accounts for some variance in the installation. While we should not be sloppy, there is some margin assuming Vans used the handbook values in the design process.
 
Rivet

Actually, that is true of setting a regular rivet as well. The standard hole for a -4 rivet is a size #30 which is .1285 or about .003 larger than the rivet. When the rivet is driven, it does indeed expand into the hole, creating an incrementally larger shear strength than the standard .125 rivet...

It was fun pulling rivet joints apart in school...
 
I'm glad you all liked the read, I found it quite informative. We can look at design criteria all day long but it sure is nice to see some real world data.

Sue said:
I work with rivit joints on an almost daily basis.
Good article, but Boeing and other manufacturers
Will sometimes use a ?shear? rivet hole.
Like drilling a #28 hole then putting in a #30 rivit.
That makes the shank swell slightly, witch in turn
Would make for a stronger shear fastener. (In theory)
Click to expand...

rocketman1988 said:
Actually, that is true of setting a regular rivet as well. The standard hole for a -4 rivet is a size #30 which is .1285 or about .003 larger than the rivet. When the rivet is driven, it does indeed expand into the hole, creating an incrementally larger shear strength than the standard .125 rivet...

It was fun pulling rivet joints apart in school...
Click to expand...

Very interesting observation/experience guys, I never thought about rivet swelling could lead to higher shear strength but it makes sense.

BillL said:
Interesting real world test, the handbook design strength of a double dimpled .032 sheet with single shear is 217 lb. The lowest test result was 26% higher than this, meaning that the design strength accounts for some variance in the installation. While we should not be sloppy, there is some margin assuming Vans used the handbook values in the design process.
Click to expand...

Bill this was my thought exactly! I saw the design strength was roughly 217lbs unless some other special conditions were there; this test showed properly set rivets at 337.5 lbs!
 
I did a similar test about 30(!) years ago when building a 4.
I found that freshly solution heat treated rivets (driven ?soft? and then aged for a few days if I recall correctly) sheared almost exactly at book value.
Aged shop rivets from the kit which were driven in received state tested higher. I assumed because they were a bit harder.
 
Many thanks for this.

This is a brilliant article, and it is also very reassuring for a first time builder like myself.

Never having squeezed rivets before and having just starting a 10 SB kit is a daunting process. Initially I was checking every rivet with the gauge but now I can just look and be satisfied most of the time without getting the gauge out.

I have had to drill a few rivets out that were obviously under or over squeezed. But I think the majority are well within acceptable limits.

I think all builders should read this article when they start. I am sure that millions of rivets have been drilled out and replaced unnecessarily and are probably not as strong the second time around, even though they look better on the outside.
 
amerkarim said:
Initially I was checking every rivet with the gauge but now I can just look and be satisfied most of the time without getting the gauge out.
Click to expand...

When my dad is bucking he still uses the gauge for anything he can get it on, it just makes me laugh but I need the time to grab another rivet so I don't mind the constant feedback (even though you can hear it most of the time).
 
This is good information and it illustrates the point that there is generally more strength in a joint than the published values. But it is a mistake to assume a strength more than that.

When the government or a big company like Boeing or Airbus develop their joint strength allowables, they create hundreds if not thousands of test coupons that cover a variety of holes sizes, sheet thicknesses, fastener types, etc. They run the results through a rigorous statistical analysis with different assumptions for probability and confidence (ex. 95/99 or 90/95). The result is a minimum value that is all but guaranteed for the application.

Here's a link to MMPDS, Chapter 8 that contains public domain data for rivet strength allowables. I highly recommend reading it when it's late at night and you really need to sleep :)
 
StressedOut said:
This is good information and it illustrates the point that there is generally more strength in a joint than the published values. But it is a mistake to assume a strength more than that.

When the government or a big company like Boeing or Airbus develop their joint strength allowables, they create hundreds if not thousands of test coupons that cover a variety of holes sizes, sheet thicknesses, fastener types, etc. They run the results through a rigorous statistical analysis with different assumptions for probability and confidence (ex. 95/99 or 90/95). The result is a minimum value that is all but guaranteed for the application.

Here's a link to MMPDS, Chapter 8 that contains public domain data for rivet strength allowables. I highly recommend reading it when it's late at night and you really need to sleep :)
Click to expand...

Thanks for the link, looks like a great late night read. Or where I live, a "winter read". :D
 
In the photo of the AN470 rivet, the rivet pictured has no dimple in the head. I think this means it is a soft AN470A3 rivet, rather than the higher strength AN470AD3 rivet, which would be identified by a small dimple in the head.

I don't know what rivets he used for the actual test, but if he used the same rivets as photographed, it explains the significant short-fall in strength of the 470 rivets compared to the 426 rivets, and the fact that they all failed in shear, except for the extremely overdriven ones. And that also makes me suspicious. I think if he had used AD (hard) rivets, he would have seen more head tipping and sheet tearing at the holes.
 
scsmith said:
In the photo of the AN470 rivet, the rivet pictured has no dimple in the head. I think this means it is a soft AN470A3 rivet, rather than the higher strength AN470AD3 rivet, which would be identified by a small dimple in the head.

I don't know what rivets he used for the actual test, but if he used the same rivets as photographed, it explains the significant short-fall in strength of the 470 rivets compared to the 426 rivets, and the fact that they all failed in shear, except for the extremely overdriven ones. And that also makes me suspicious. I think if he had used AD (hard) rivets, he would have seen more head tipping and sheet tearing at the holes.
Click to expand...

Interesting observation and good catch. However, I still think he used structural for a couple of reasons. The picture showing the soft rivet looks like it is pulled from some other source. Also, the strength of a soft rivet is less than half it's AD counterpart. Therefore, I think if soft rivets were used (AN470A) the results would be substantially less (I'm estimating 50% ish less). Anyone else with more experience able to verify?

Edit: after looking at the pictures again on the test coupons (the bottom two pictures of the test coupons) it looks to me that the one remaining rivet on the top left has the structural dimple.
 
Last edited:
jcarne said:
Interesting observation and good catch. However, I still think he used structural for a couple of reasons. The picture showing the soft rivet looks like it is pulled from some other source. Also, the strength of a soft rivet is less than half it's AD counterpart. Therefore, I think if soft rivets were used (AN470A) the results would be substantially less (I'm estimating 50% ish less). Anyone else with more experience able to verify?

Edit: after looking at the pictures again on the test coupons (the bottom two pictures of the test coupons) it looks to me that the one remaining rivet on the top left has the structural dimple.
Click to expand...

I'm inclined to agree. The one test I did with a soft "A" rivet showed a shear strength about 60--65% of a "AD" rivet.
 
You must log in or register to reply here.
Back
Top