Shear strength of 2117 aluminum rivets 2

[Abbyyyy]

Greetings, everyone!

This is my first time starting a thread, and I'm hoping to receive your help.
I've encountered a problem: commonly used solid rivets, like NAS1097/NASM20426, etc., all have rivets made from 2117 aluminum alloy, otherwise known as AD rivets. According to the technical specification NASM5674, the shear strength of AD rivets is required to be 26KSI. However, after researching and testing actual products, the real shear strength is essentially between 33KSI and 34KSI, exceeding the specification requirement by 23% to 26%. Does anyone know how the shear strength requirement for solid rivets in NASM5674 is established? Why is the specified value in NASM5674 so much lower than the actual ones? Thank you!

 

[SWComposites]

The minimum strength may be based on:
- value that the rivet manufacturer is willing to guarantee, and/or,
- old data from many years/decades ago, at which time actual strengths were lower.

Why is it a problem? Just use the spec min value for design.

 

[Abbyyyy]

Thank you for your reply!
I need this value for some calculations. If the specified value could be raised according to the actual product value, it would bring many benefits.
I find it somewhat strange for the actual product's shear value to be so much higher than the specified value. For other aluminum solid rivets, such as 7050 rivets, otherwise known as E/KE rivets, the difference between the actual product's shear value and the specified value is only about 10%.

 

[verymadmac]

I note MMPDS-0 gives Fsu =30 ksi for driven AD rivets. I seem to recall that establishing statistical design values rivets (typically a B value) requires a substantial number of tests to be carried out, the number of process taken to install a rivet and the low repeatability of hand driven rivets are going to hammer your standard deviation of the test results.

 

[rb1957]

and I think that's the difference. MMPDS table 8.1.2 ... AD rivets (driven, ie work hardened) have Fsu = 30ksi. Use that value, properly referenced, not the material spec.

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[rb1957]

what calc uses the shear strength of the rivet, rather than the allowable load ?

Remember this value is reduced by sheet thickness and/or CSK. There are various tables to look this up, but annoyingly you need to look at earlier versions of MMPDS (or Mil Hdbk 5)


"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[SWComposites]

Material and fastener suppliers often like low specification requirement values so they always pass and never have rejectable batches.

 

[Ng2020]

"If the specified value could be raised according to the actual product value, it would bring many benefits."
...And lots of additional testing to verify the higher design value. See EASA CS / 14CFR 2x.613.

 

[rb1957]

Sorry, no one should be talking about changing spec values. The spec is for "undriven" material. MMPDS has allowables for "driven" rivet material.

Forget the raw material spec. Use the allowable in MMPDS. Using "Fsu" is also misleading in calculating rivet allowables, as it doesn't include the factors that CSK and or sheet thickness require.

I don't know why yo Need Fsu. Perhaps you can explain some ??

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[WKTaylor]

OK some things need clarification.

NASM5674 26-KSI shear strength requirement is for UNDRIVEN shank in steel dies... spec use only. 'Driving' the 2117-T4 solid rivet actually induces compression cold-working of the shank as it crushes/swells to fill the [slightly larger] hole [and form the 'bucked-tail'].

MMPDS/MIL-HDBK-5 driven shear strength values are based on testing... with control test parameters for defined grip-length [affects swelling] and specific hole diameters... installed using 'good-practice hand-driven' [bucked] techniques. This sets a statistical baseline for very +99% confidence of all rivets meeting/exceeding the value listed [~30-KSI]. For materials these allowables would be considered 'A' basis... where strength is lowered for statistical safety purposes.

HOWEVER... your statement is a sketchy in many regards. 'Special cases' are excluded from consideration... unless the appropriate [proprietary] testing has been accomplished and is approved by whatever agency. Where strength is increased with 'special controls' for wire-alloy, dimensional processing, finishes, installation practices... usually at premium cost for satisfactory statistical safety purposes... this would be considered/classified as 'B' basis... premium quality.

NOTE1. OH yeah... and as mentioned by others... the MIL/MMPDS values were 'derived/codified' decades ago.

NOTE2. The apparent FSu values You cited for test rivet installs... 33-to-34-KSI... should actually be compared to 30-KSI... sooooo...
33/30=1.100
34/30=1.133
... not as dramatic as you stated.

NOTE3. Machine installation [drill-insert-squeeze, repeat]... of conventional rivets MS20470, MS20426, NAS1097, etc] could provide increased shear strength allowables due to high repeatability/consistency/confidence. And yep... there are other unique practices for flush rivets not discussed here.

NOTE4. OK, aluminum alloys and fabrication processes [heading/forming/heat-treatment/etc] have been improved over the decades... so... I suspect that MIL/MMPDS shear strength tables could stand changes... especially for some of the newer solid rivets [NAS9313, NAS9314, NAS1241-OS, NAS1242-OS, etc]. Please send bags-of-$$$ for testing

CAUTION1. I had recent bad-experience with production mechanics creating a 'new way' to drive solid [D] rivets for high volume production... that created consistent bad installs that were never corrected 'on-the-spot'. And no-one in engineering was ever consulted for advice RE this 'practice'. Then the shit-hit-the-fan. After MAJOR gnashing-of-teeth and soul searching... there was a safety inspection requirement and mandated replacement of hundreds-of-thousands of rivets, due to 'no confidence' in the installations... way-out-of-spec... what a nightmare.

I hope this makes sense.

Regards, Wil Taylor
o Trust - But Verify!
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation, Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", HBA forum]
o Only fools and charlatans know everything and understand everything." -Anton Chekhov

 

[SWComposites]

please send bags-of-$$$ for testing LOL! EVERYONE wants higher allowables; NO ONE wants to pay for them.

and re CAUTION1 - yeah, shops dreaming up their own ways to "improve" processes without any engineering oversight (and sometimes with) usually ends up in a giant cluster **** nightmare.

 

[Abbyyyy]

Thank you all for your responses. I'm appreciative of the enthusiasm shown towards my first post!

The reason for using the specification value was to consider the varied performance of products from different manufacturers. I used a reduction coefficient based on the ratio of the product's actual performance to the specification value, aiming to ensure the conservatism of the final results. However, it seems that this method might make the results too conservative. Therefore, I was considering whether it could increase the specification value to improve the reduction coefficient and hence enhance the final results. I realize this might be a problematic approach, and I will carefully reconsider my thoughts.

Thank you, everyone, for your reply!

 

[Abbyyyy]

send bags-of-$$$ for testing

Indeed, establishing and modifying a standard/specification requires a lot of resources and effort. Unfortunately, I didn't understand this process before, nor did I consider the difficulties involved.

 

[SWComposites]

Abby - it would be helpful if you would explain why you want to compare the performance of rivets from different rivet manufacturers.

 

[rb1957]

agreed, we know how strong a "spec" rivet is. it doesn't matter that some raw material may have a higher Fsu than the (raw material) spec. "no one" (well, no one I've seen) uses this approach.

as with SWC ... "what are you on about ?"

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[WKTaylor]

Aby...

For higher reliable shear strengths, there are higher strength alloys for solid aluminum rivets, such 'D' = 2017-T4 [inactive] or 'E'/'KE' = 7050-T73.

Regards, Wil Taylor
o Trust - But Verify!
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation, Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", HBA forum]
o Only fools and charlatans know everything and understand everything." -Anton Chekhov

 

[LiftDivergence]

Maybe I'm missing something in this thread, but you wouldn't use the MMPDS values as the joint allowable. At least I wouldn't think to.

It can easily be checked, the shear strength values listed in a table such as 8.1.2(a) are literally just the based on the listed Fsu and the cross sectional area of the shank.

That would only act as the allowable joint load in the case that the joint was shear-critical, no?

In reality, most joints experience a transitional failure somewhere between bearing of the layers and shear deformation of the fastener. At least if the joint is properly designed.

We don't want to have shear critical fasteners per design. Joint strength should be based on transition failure data.

This is a plot I made for another post:

You can see, the shank direct shear values are the upper limits of joint strength. Your actual value would be lower than that. So yeah, MMPDS can give you the theoretical upper limit of strength, if the joint is shear critical. But that's about it.

Basically, I would not recommend using direct shear allowables from the fastener spec or MMPDS to determine joint strength in a structural substantiation.

Keep em' Flying
//Fight Corrosion!

 

[rb1957]

really ... don't use MMPDS section 8 for fastener/joint strength ??

yes, table 8.1.2b is the upper limit.

plain shanks are adjusted by factors in table 8.1.2.1b.

CSK fasteners from tables with the specific fastener, like table 8.1.2.2f for BB (426AD) rivets.
which is a PITA for 1097 rivets (since they only have E material, table 8.1.2.2n)

shear critical fasteners ? why not ?? sure, some places have a preference (or a design rule) for bearing critical which pushes the design to either large diameters or small thickness ... but this is preference, not a "golden rule".

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[SWComposites]

Yes, LD is correct. Many riveted and bolted joints do not reach fastener shear strengths. MMPDS/Mil-5 used to publish "joint strength" allowables which covered specific combinations of sheet materials and thicknesses and fastener types/diameters/etc, with the values given in terms of load/fastener and represented the most critical failure modes. Not sure if that data is in the latest versions.

 

[rb1957]

are we all saying that using Fsu is not directly applicable.

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.