Question about fastener and bolted joint shear failure modes 1

[JLBK56]

Hi everyone. I've got a bit of a philosophical question regarding shear failure of bolted joints. It is often common practice in my industry to check whether shear tearout, fastener shear, or bearing stress is of concern in a design. However, the general understanding of the behavior of bolted joints in shear indicates that until the joint slips, all the load is taken by friction. Say that the joint is predicted to slip, eventually the bolt(s) will engage their clearance holes in shear. However, given manufacturing tolerances, isn't it likely that only a limited number of fastener will engage in shear (bottoming out against its hole) before the others, depending on the direction of loading? In theory, a "zippering" effect could be observed where one fastener fails followed by others subsequently, since the load is taken mostly by one until it fails, then the next, etc.. This of course assumes that the joint design does not permit load redistribution to other fasteners before failure. In other words, using nominally predicted shear forces from a FEM or other hand calculation would not be appropriate for estimation of shear failure in this case, despite what is typically done in stress calcs. Curious what others think and how you all have handled such calculations in your designs/analysis. Thanks!

 

[KootK]

I agree. For this reason, bolt shear failure mechanisms need to be ductile and capable of redistribution. This is also why we don't use bolts and welds to resist a common force.

 

[SWComposites]

For this reason bolted joints should be designed to be clearly bearing critical, not fastener shear critical, so as to provide bearing deformation and fastener load redistribution.

 

[human909]

Both answers above are correct. But to more directly answer the question asked.

In normal bolted circumstances with normal tolerances, significant hole elongation will occur to allow multiple/all bolts to start sharing the load. A 20mm bolt in a 22mm hole will have a very small contact area and will deform by a few mm before the bolt gets anywhere near capacity. Meanwhile the connection has settled enough an other bolts are also taking up the load.

For this reason bolted joints should be designed to be clearly bearing critical, not fastener shear critical, so as to provide bearing deformation and fastener load redistribution.

In practice this is often not the case in simple structural shear connections. I don't dispute the logic, just telling it how it is.

 

[Stress_Eng]

I’ve read / heard of numerous approaches to carrying out a bolt group analysis, going from the typical assumed linear distribution (distributed via shear / bearing areas / combination of, etc), the limited bearing capacity approach (assumed bolt linearly increases to a limit and then remains constant), to the point where bolt joint displacement v force curves are used. And let’s not forget, we have 2D and 3D methods. I’m sure there’s many approaches that can be used! The subject can get quite involved!

 

[Italo01]

I agree with human909. According to DG17, a A325 bolt will deform 0,25in(6mm) before breaking, so there's enough room for redistribuion of forces.

I don't know about other codes but the brazilian code has a provision in which a connection greater than 1270mm on the direcion of the force must have the bolt capacity reduced due to reduced redistribution capacity and 1270mm is much greater than anything i've ever seen.

 

[SWComposites]

And the approaches to bolted joint design and analysis in aerospace/aircraft and civil structures is likely quite different in the details.

 

[rb1957]

I think there're several schools ... one says bolts must be bearing critical (which pushes the design to thinner sheets) and the other which doesn't mind so much.

A typical AN3 (2000 lbs shear) in Al (120ksi bearing) (these are round numbers, ok) is good for thickness less than 2000 = 120000*(0.2*t) ... t = 10/12in = 0.83in
but a good ol' BJ5 rivet would be 600 = 120000*0.15*t ... t = 0.03 ...

so large (steel) fasteners and particularly heavy lugs should be bearing critical, but rivets would typically be shear critical. And that's where aerospace probably differs from civil the most ... I doubt civil would "ever" use Al rivets.

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

 

[271828]

In AISC Land, there are two basic types of connections: slip-critical and bearing.

For bearing, the effective strength at each bolt is the minimum of bolt shear rupture, bearing, and tearout. Add the effective strength of each bolt to get the available strength for this "shear transfer" limit state. The underlying assumption is that there's enough ductility in all three of these limit states to allow redistribution.

This approach was mentioned in the 2010 Specification at the end of Section J3.10, but the approach wasn't used in the Manual tables in the 14th or 15th editions. In the new 16th ed. Manual, many tables have been reformulated to use this approach. See Table 10-1 for example.

 

[human909]

I think there're several schools ... one says bolts must be bearing critical (which pushes the design to thinner sheets) and the other which doesn't mind so much.

I'm not sure why there is a sudden influx of aerospace references in this thread. While I think it is useful to acknowledge the behaviours of various critical modes apply the design of plane fasteners to that of buildings and bridges seems problematic.

While bolt shearing might be the critical aspect in many basic structural connections (eg shear plates) they are normally significantly stronger than the member capacity due to bending or buckling.

 

[dik]

It's good to know how fasteners 'work'. It's good information

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[SWComposites]

Well, why aerospace comments here? a) we deal with a lot of complicated bolted joint issues, b) the OP field is Mechanical without saying what specialty, c) the OP didn’t mention anything about the questions pertaining to Civil structures, d) several of us aerospace structures types read this forum for education and amusement.

 

[dik]

About 40 years back, due to a missed x-bracing in a project, rushed for opening, I repaired the missing angle with sheet metal and pop-rivets... first time I've ever used them structrually. The partner in charge was surprised, but happy with the solution. Structure is still standing. I just like fastener stuff... many structural failures are a result of the fastener going south.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[ProgrammingPE]

I don't know about other codes but the brazilian code has a provision in which a connection greater than 1270mm on the direcion of the force must have the bolt capacity reduced due to reduced redistribution capacity and 1270mm is much greater than anything i've ever seen.

The American institute of Steel Construction's (AISC) Specification for Structural Steel Buildings also includes this provision (AISC 360-16). See the footnote for Table J3.2, Nominal Strength of Fasteners and Threaded Parts, which says "For end loaded connections with a fastener pattern length greater than 38 in. (950 mm), Fnv shall be reduced to 83.3% of the tabulated values." (AISC 360-05 used a length of 50 in. (1270 mm) with 80% of Fnv which matches your dimension.)

The purpose of this provision is because in long connections, the bolts that are closest to the applied load see a disproportionate part of the total load due to the strain in the connection plates. This behavior is predictable, and its impact is accounted for by a 0.9 length reduction factor already built into the Nominal Shear Strength, Fnv. Longer connections require the additional reduction factor referenced above as well.

As others have mentioned, the impact of the hole manufacturing tolerances are generally ignored when determining the shear strength of bolted connection due to their ability to redistribute the load to all of the bolts.

Structural Engineering Software:

http://www.structuralcentral.com/tools/steel-connection

Structural Engineering Videos:

http://www.youtube.com/structuralcentral

 

[rb1957]

I think it is up to the OP to filter the responses he's getting. Maybe he's in the "AISC" world, and takes that guidance.

I hope the comment about "aerospace" means that either "we're not worthy" or "you guys over-science this stuff". I know I've learnt quite a bit, and hope I've contributed positively too. Except when you guys go on and on about concrete ... geeze ! (c'mon smile, the only time I've needed concrete was to ballast a plane)

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