Overstrength Connection at Grade

[driftLimiter]

I know this question has been beat to death But I have a particular arrangement That I am curious of others opinions.

The connection is a Gusset Plate connection to a Buckling Restrained Brace at the column baseplate.

AISC 341 - 16 Is clear that connection design forces should be based on the adjusted which I am just going to call Overstrength Load for the purpose of this discussion.
The connection approach we would like to take is to transfer tension into a grade beam via reinf. that is developed into the grade beam and welded to the gusset plate.

The grade beam spreads out the lateral forces to the elements we have designed to resist it. The foundation resistance is based not on the Overstrength loads, but instead the ductile seismic force Qe.

SOO, the approach I'd like to take is to proportion the reinforcement based on the force Qe. But since it is connected to the steel connection (and its providing the lateral resistance at the base of the frame).
It feels like I have a short circuit because the connection design (at least on the steel side) is going to be designed to resist the overstrength loads, but the elements supporting that connection are not.

I guess this is the same quandry that we run into anytime we are designing a connection requiring over strength, but a foundation that does not. I could simply add more bars to be able to show we have enough bars for overstrength, but then I wonder why I would do this if the foundation system doesn't have the overstrength capacity anyway.

 

[Greenalleycat]

I can't speak to your codes, but for us the overstrength force follows the whole way through

You designate one ductile element and design it for your reduced ductile loads (Qe or whatever)
Then you take the overstrength demands from that ductile element and follow it through the system until everything is shown to withstand those loads
So the grade beam should have reo to handle the overstrength AND the rest of the foundation system that it connects to, including the supporting soil, should be designed/checked to withstand overstrength forces

 

[YoungGunner]

I think it depends on what was designed for overstrength and the intent therein. Section 12.3.3.3 in ASCE 7-16 for "elements supporting discontinuous walls or frames" has some interesting insight in the matter. On one hand, they allow for beam connections that support discontinuous shear walls to avoid being designed for overstrength, but if the discontinuous element needed to be designed for overstrength per some other requirement outside of 12.3.3.3, then it sounds like overstrength needs to be transferred throughout.

 

[Greenalleycat]

To be honest, regardless of what the code clauses could be read to interpret, this is definitely one of those issues that you need to take a step back and evaluate

If you're undertaking a capacity design process then it needs to be followed throughout the whole structure, or else how do you know that you're actually achieving the mechanism you intend?
Your actual failure mechanism might be something random like foundation beam yielding or soil bearing failure - how does that influence the brace performance?

 

[YoungGunner]

Your actual failure mechanism might be something random like foundation beam yielding or soil bearing failure - how does that influence the brace performance?

I was under the impression with, at least from the notes associated with ordinary moment frame design, that at some point some other mechanism limits the force that is transferred through the system, which I will assume is because of additional ductility and/or movement. Which is ok. If the footing starts shifting, that dissipates the energy in the system in a new way and the maximum force the system needed to be designed for is actually reduced. So at some point, can't we say that overstrength doesn't need to track all the way through?

 

[driftLimiter]

Thanks for this feedback so far.


@Youngunner, the Overstrength force in this context is the Expected Strength of the Buckling Restrained Braced Frame, the provision is from AISC 341-16.

@Greenalleycat, I agree I need think about this from the system level. Luckily I am not using the grade beam for the vertical forces only as a tie beam for dragging the shear.

A little further reading seems to suggest that the foundation system receiving the brace should also be designed for the expected axial demand of the baseplate. It also acknowledges that the foundations may be designed for lower strength according to the code, but that it would require significant inelastic displacement of the foundation system. Which may not agree with the assumptions used by AISC 341 to get the R factor of the system. Which I believe is just what @Greenalleycat is saying.

Link

 

[driftLimiter]

So at some point, can't we say that overstrength doesn't need to track all the way through?

If these mechanism create additional displacement demand on the system then it could have less ductility than assumed for the design level force, there could also be more drift. This could indeed compromise the strength of the LFRS. But perhaps not all mechanisms would have such consequences. This aught to be investigated when you don't have a lot of redundancy especially.

 

[Greenalleycat]

@Younggunner, you may be right that, in practice, the consequences of yielding somewhere else (soil, foundation beam, whatever) may be beneficial
But that's more like luck than design intent and shouldn't be relied upon in this context (in my country anyway - I don't know conventions and codes in the US)

Here, the point of considering overstrength in a ductile system is that you are forcing the building to behave in a controlled way that you have designed for
If you suddenly start ignoring the overstrength load path halfway through your structure then you may as well have ignored the ductility because you have no control over that actually occurring