Two Bolt Column Shear Tabs Stiffness

[structengr23]

I am new to the commercial connection design field. At our company, we typically design the beam to girder and beam to column shear tabs with a single column of bolts and design to 1/2 Capacity of the beam. The shear tabs are typically 3/8"x4"xVarious Lengths, so they are not considered extended, just your typical shear tab. Our calculations show that all the shear tabs are good for the loads. The EOR of the primary steel wants two columns of bolts when our calculations don't call for it. He says that this adds to global stiffness of the overall structure and that if we don't use this detail (2 column rows of bolts), then he will have to add more bracing to the structure. I am skeptical that overall stiffness is enhanced by this request. He did not answer my question when I asked about his member end releases in the global model. For a shear tab, the member ends should be released for major and minor axis rotation. I am thinking of pushing back on this request for two columns of bolts. If he has considered them partially restrained against rotation in his global model, I could understand. But he did not say that is the case. He just said that the structures on the Texas gulf coast have to meet windstorm criteria and they've always used two column rows of bolts because it enhances the structures stiffness. He could not give me any technical support as to this assertion, i.e. FEA Analysis, Calculated Spring Rotational Stiffness, etc.

 

[hokie66]

This should depend on the contract for construction. If the EOR has specified or detailed the connections the way he wants them, that is what goes. If he has left the design to someone else (your firm?), then if he insists, you may have a claim. But it sounds like a contractual issue, not a technical one.

 

[structengr23]

Yes, you are exactly right. In the general notes, he specified the two bolt column detail. We asked the fabricator, since we are contracted to them, if we could use our normal detail. The fabricator said to use our normal detail. He should have consulted with the EOR first. We don't usually get direct access to the EOR. We have to go through the fabricator. Lesson learned. Ask for fabricator to get EOR confirmation first. We will need to rework the details. I'm okay with that. The fabricator realized he messed up and is going to have to pay us for the rework. I just from a conceptual standpoint want to get a feel for the stiffness benefit. I know it's not a moment connection, but it still has I'm assuming maybe a 30% moment contribution with a two bolt column. But, still, I don't see that taking the place of bracing. In order to get a true stiffness contribution, he would have to model some rotational springs for the member end releases to represent the shear tab as being more than a pinned connection.

 

[JAE]

The EOR here is a being a bit weird.

Two rows of bolts may induce some stiffness in the connections, but I'm not aware of other engineers typically doing this - or simply blanket-specifying connections like this to meet Texas gulf coast wind loads. Your framing needs to be analyzed for the applied "gulf-coast-wind" and the two rows don't magically keep the building standing vs. one row. Stability would depend more on the column-beam overall stiffness as the EOR is suggesting that they have a moment frame of some sort.

Also - with two rows and some implied fixity, you'd perhaps get negative moment in the beam...which the EOR maybe didn't design for the added bottom chord unbraced length...who knows....only the EOR so as hokie66 suggests you have to follow their lead.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[hokie66]

Just to modify my answer above, I'm not saying the EOR is right, and he should have given you a better explanation.

 

[skeletron]

Is the EOR concerned about rotational stiffness of the beam (ie. torsion)? With one column of bolts, the supporting beam is liable to rotate because the weld would take the full unbalanced moment. With two columns of bolts the rotation could occur at the bolt group.

I've only dealt with two columns of bolts in a shear tab designing industrial structures. But connections in commercial structures can be a lot different.

My first red flag reading this, by the way, was designing the connections to 1/2 the capacity of the beam. Designers need to stop doing that! Specify the loads and let the fabricator do the connections.

 

[CANPRO]

If he wants the connection to add additional stiffness to the structure, he should be specifying a moment value at the connection as well. Otherwise he's just blindly specifying this connection type hoping it will add stiffness to the structure.

Whats being used for the main lateral resisting system for the building? If he wants a slightly stiffer than typical shear connection to add to the lateral stiffness of the building, sounds like the building is going to have serious drift issues, i.e. the main lateral resisting system is going to have to deflect significantly before these softer connections start to contribute anything to the lateral resistance.

 

[CANPRO]

skeletron, I agree that 1/2 beam capacity is a bit dated (but dead simple from the perspective of the connection designer). I see a lot of EOR's specifying shear values as a result of a beam being loaded to its full moment resistance (V=4M/L) - makes for lighter loads in most cases, but now you have a different design for each beam/span combination. And short span beams can get into ridiculous shear values.

If they listed the shear value for each beam that would be great, but these days I'm lucky to get drawings with the grid dimensions, so I'm not holding my breath for more connection info.

 

[canwesteng]

1/2 beam capacity yields very reasonable connection designs for typical beam framing arrangements, and provides the next engineer to work on the structure some decent connections to work with. I know AISC loves to harp on adding beam end forces, but this is because they represent fabricators, not owners or engineers. The owner is much better served spending the money on connections instead of spending money on an exhaustive list of beam end forces and taking on the risk of an error on the drawings here leading to connection failure.

Although I would question how you are hitting 1/2 of beam shear capacity with shear tabs with a single row of bolts for every size of beam on the project, unless it's framed fairly lightly.

 

[BUGGAR]

Anybody concerned about the weld?

 

[Ron247]

As long as you are not restraining the flanges, the connection should be closer to a shear connection than a moment connection. Even a 1-column set of bolts resists some moment. It is apparent that the more columns of bolts you add, the more the connection migrates to a moment connection. The only problem I have with the requirement is modeling something as true-pinned and then constructing it as semi-fixed to fixed. If they want semi-rigid, specify the magnitude. Pinned versus fixed does not remove loads, it shifts some portion of effects of the loads to new locations.

I would definitely make a record of my conversation with the EOR. I understand a project requirement, but since they EOR makes the requirement, they should be the one that shoulders most of the burden should things go South due to the requirement. Other than the magnitude, wind in Texas doesn't act any different than wind in Ogden, Utah. Randomly changing rigidity tends to change the moments and shears in the structure. Take a simple rigid frame with 2 pinned based columns and a single fixed-end rafter. Design all components to a .995 CSR and then arbitrarily make the columns twice as stiff. See what that does to the rafter. That's an extreme example.

 

[structengr23]

Folks, I appreciate the responses. Lots of good points. At the end of the day, we are beholden to the EOR. I think the two column detail does pull the rotation restraint to the beam web and maybe takes some moment away from the weld at the girder, which is a benefit. I've found some studies that say a single column gives you 15-20% moment restraint. But, I haven't seen any studies that say a double column row gives you more moment restraint. Regardless, any assumed moment restraint would need to be modeled as a spring at the member ends to see the benefit in the model. I'm not sure this was done. And if it wasn't and the single column detail was used, the member end releases would still be pinned, so where would bracing need to be added. The model results would not change. But if he did use springs and he reduced them to pinned, he would see a stiffness reduction and he could justify his argument. I think it's one of those cases where "this is what we've always done". I agree that the 1/2 capacity is maybe suspect and they should specify member end forces to better corelate the joint design to the global model.

 

[Ron247]

I see 1-column could be more fixed than 2-column depending on the bolt layout. 1 column of 8 bolts spaced 3" apart versus 2 columns with 4 bolts each spaced 3" may have the same fixity. The eccentric load tables in the AISC give some ideas on that.

 

[MotorCity]

I would challenge the EOR on his request and ask for further technical clarification. At the end of the day, you will have to submit to the EOR's request if he stands his ground.

I don't give much credence to the "That's the way we've always done it" or "Because I said so" lines of reasoning. Not to mention you may be adding unnecessary cost to the project