Individual Bolt Bearing vs. Group Bolt Bearing

[ahypek]

Good morning folks,

Typically, when I design connections I check individual bolts for bearing and tearout capacity. On occasion, I'll end up with a single bolt failing in tearout while the summation of bolt bearing strength exceeds my requirement. For years, I've been making sure each individual bolt capacity exceeds the required strength but recently, I've been questioning whether this is too conservative.

For example:

AISC Table 10-1 shows that a beam web (coped at top flange w/ (3) 3/4" bolts) with an edge distance of 1-1/4" has an available strength of 133 kips/in (ASD).

If Lc = 1.25 - 0.8125/2,

Lower limit should be: (1.2*(1.25-0.8125/2)*(65ksi))/2.00 = 32.9 k/in for a single bolt in tearout
Upper limit should be: (2.4*3/4*(65ksi))/2 = 58.5 k/in

If Lc = 3-0.8125,

Lower limit should be: (1.2*(3-0.8125)*(65ksi))/2 = 85.3 k/in > 58.5 k/in

Therefore, assuming our beam web is 1" thick the total bearing strength is 149.9 kip.

But now what if the allowable reaction is 140 kip. Our bolt group exceeds the requirement but if we assume 140 kip/3 bolts -> 46.7 kip/bolt which is greater than 32.9 kip/bolt of tearout strength.

Do I assume that significant deformation has occurred at this bolt and now I only have 2 bolts resisting the reaction through bearing?

Would love to hear what you think.

 

[bootlegend]

I have thought about this also. The AISC commentary on chapter J has some notes about it on page 16.1-444 of the 15th edition. My interpretation is that for typical connections it is acceptable to sum each individual bolt strength and determine the total group strength and compare to total load on the group. If you have some sort of unique connection then you may want to be more conservative and use the individual strength compared to the load per bolt.

 

[ahypek]

I guess after rereading 16.1-410 (14th), AISC claims that "...the connection [very general/unclear] has enough ductility to allow all of the bolts to reach their individual effective strengths."

Thanks bootlegend.

 

[canwesteng]

I'm not sure about your question... All the bolts should be equally bearing on the steel before tearout occurs. Tearout should be checked only against the loads on that bolt and not used to derate its bearing strength.

 

[ahypek]

canwesteng,

Allow me to further clarify. All bolts are equally bearing on the steel but the bolt nearest to the edge does not have enough strength individually to resist tearout. I have always designed my connections so that my individual fasteners have enough strength but according to the AISC, this is not necessary and the summation of individual strengths is the total capacity even if one bolt is inadequate. I'm wondering if realistically, the bearing surface undergoes significant deformation and the remaining reaction will now be forced onto the other bolts.

 

[canwesteng]

This probably warrants an email to the AISC if I'm understanding correctly. Say you have 6 bolts, the bolt nearest to the edge will tearout with 1/6 the load? In that case I would say the connection is not strong enough.

 

[ahypek]

Exactly! The commentary states, in a very general way, that the summation of the individual fasteners is the actual capacity due to ductility of the connection even if one is inadequate.

Was hoping to discuss this with other engineers on here.

 

[dhengr]

Ahypek:
I believe that answering your question was basically the origin of the need for the block shear calc. for any given joint/connection. At the time, I think the AISC increased the single bolt values in shear & bearing so the possibility of what you are suggesting became more likely, and thus, the possibility of the bolted connection unzipping. In your case, you have the connection loading finally split btwn. six bolts, and we basically assume that each bolt takes one 6th of the load, but this requires that there may be some considerable deformations in bearing at some of the bolts, before all of the bolts pickup their share of the total load. A well designed and proportioned connection, proper edge distances, etc., can tolerate these deformations to arrive at the finally bolt cap’y. You have tolerancing problems with exact bolt dia., bolt hole size, hole location and matching from plate to plate, etc., so only one or two bolts may take most of the load until their holes deform enough to start to bring the other bolts into play. And, as in your case, if the edge bolt was the first bolt to start bearing/taking load, and the forces were in the right direction you could initiate tearout or plate splitting in several different ways, and at a initial bolt loading higher than the final average bolt loading. Then, as the connection starts to unzip, you have the total loading transferring to only five remaining bolts, etc.

 

[Settingsun]

For what it's worth, Australian practice is to check tearout for a single (critical) bolt. Except for very thin plates, it's usually not an issue.