Limiting the braces to resist smaller forces 1

[BAGW]

Hi,

I am posting this just for concept understanding, new way of looking at the building design that I have never done before. If there’s a building with a moment frame bays and braced frame bays, obviously all lateral forces will be resisted by braced frame bays as they are very rigid as compared to moment frame bays.
Say the building is very low seismic area and with R=3.

If total base shear is 1000kips, braced frame will resist 900kips and moment frame 100kips. Can I limit the brace frame to resist only 500kips and design the moment frame to resist the rest 500kips? Is that an allowable design procedure? Choosing force magnitudes to be resisted by a particular system?

 

[structSU10]

The load follows stiffness. if you want the moment frame to resist half the load, it has to be the same stiffness as the braced frame.

 

[Ron247]

Off the top of my head I would think it would be the same stiffness required in order to get equal drift. Since bracing is so much more efficient, I would think it would be very large moment frame.

 

[KootK]

Is that an allowable design procedure?

I suppose it would be possible if you could guarantee that the brace frame failed in a ductile manner. That's pretty tough to do with low seismic detailing though.

Are you considering multiple braced / MF bays within a common framing line? Or spread out to different framing lines.

 

[BAGW]

Spreading out in different framing lines.

 

[JAE]

I'd be very careful that I don't create a condition where the brace would fail and then ALL the force shifted to the moment frames - the structure would "unzip" rather fast in that case.

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

 

[MIStructE_IRE]

In reality the braced frame will take the lion’s share. It will always be stiffer than a moment frame - they’re called sway frames for a reason.

I would design using braced bays only if it were me. As above, you don’t want to end up with a case where the bracing fails before the sway frame moves enough to become active.

 

[KootK]

Can I limit the brace frame to resist only 500kips and design the moment frame to resist the rest 500kips?

Spreading out in different framing lines.

Any chance this might be the byproduct of a flexible diaphragm load distribution? Post a plan sketch if you're able.

 

[JLNJ]

Echoing JAE above, you might want an enveloped solution which allows for some healthy wrong-ness in your guess as to what percent goes to which frame line.

 

[CANPRO]

This approach might be a little questionable, but I think it is fundamentally valid:

[ul]
[li]Use bolted x-bracing for your tension bracing. Provide long slots in line with the tension.[/li]
[li]Each tension brace will have (4) bolted connections with long slots - one at each end of the brace and one on each side of the splice plate where the x-braces cross[/li]
[li]With a 3/4" bolt in a standard long slot, say we get about 1/2" play each way x 4 locations = 2" play in the brace[/li]
[li]If your brace is at 45 degrees, that gives you about 1.5" free play laterally.[/li]
[li]This gives your moment frame 1.5" of movement to pick up as much load as possible before the tension bracing starts to take over.[/li]
[li]The rest is as structSU10 says, just a matter of getting the stiffness right.[/li]
[li]Counting on the slotted connections as a fuse is probably ill-advised.[/li]
[li]Just thought it was a neat idea. Woo hoo, Friday night [/li]
[/ul]

 

[milkshakelake]

JAE and JLNJ: What kind of situation would occur where the braces fail and then all the forces get transferred to moment frames? OP said that the analysis shows that the (theoretical) 500k goes to braced frames and 500k goes to moment frames, or 900k-100k or whatever. So it should distribute that way. If we assume that braces are so stiff that they'll take the forces, and then potentially fail, and then the moment frames will take 1000k, then "structural analysis" goes out the window. Braces might carry 100% of the load initially, but once they start failing, they start redistributing to the moment frames.

To draw a parallel, we can distribute forces to shear walls based on their stiffness. We don't assume that the long, stiff shear walls will fail, and then the short ones will take up all the slack. This is good and conservative, but will result in a lot of reinforcement for short shear walls, and change the stiffness anyway.

Another parallel: I don't remember exactly but you can do intermediate moment frames combined with CMU or concrete shear walls. The shear walls will be stiffer, but the code requires moment frames to take 25% of the lateral forces, not 100%.

Sorry if I'm missing something fundamental here. I thought about ductility, but if you use R=3 per ASCE, the seismic forces are large enough that you should be okay.

 

[JAE]

milkshakelake - I would think that if you had a combination of very stiff braces, and more flexible moment frames, with a more rigid-than-flexible diaphragm, then it is conceivable that your loads would drive into the braces initially. Depending on the design, and if the brace connections or anchors to foundations aren't up to strength or ductility then you might get an abrupt let-go of stiffness and loads in a brace - shedding all the moment frame.

The OP did not say 500k and 500k.

They said 900k to braces and 100k to moment frames from their initial analysis and then suggested specifically designing the braces to only be able to take 500k.
That is under-designing on a [red]strength[/red] basis to shed load to the moment frames....instead of the correct way of trying to design the [red]stiffness[/red] down to shed load.

CANPRO was on target in my view of trying to limit load (vs. limiting stiffness) by using slots, etc. I'm just not sure I trust slotted holes to actually work correctly and as intended....i.e. you could get binding, bad alignment in the slot initially, etc.





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

 

[dhengr]

BAGW:
It slips my mind at the moment, but what is the name of the proprietary beam/col. connection detail where an hourglass shaped cutout is applied to the outer beam flg. edges/tips, within a foot or so of the column connection; with the intention being that a fairly well defined and predictable moment hinge can/will form at this cutout? Why not use this same kind of thinking on your braces in your braced frame? Make the braces, tension only, know their yield strength accurately, and design the cutout shape to start to yield at your predetermined value of tension. I’m not sure how accurately/confidently you can model your whole bldg., so I would want the two systems to overlap in their strengths, so that neither sloughed too much load off on the other, overloading it cap’y. Of course, you still have to pay very close attention to the relative stiffnesses of the two systems.

 

[JAE]

I might add that my last point was a wordy summary of what structSU10 succinctly posted in the second post above.

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

 

[KootK]

It slips my mind at the moment, but what is the name of the proprietary beam/col. connection detail where an hourglass shaped cutout is applied to the outer beam flg. edges/tips, within a foot or so of the column connection; with the intention being that a fairly well defined and predictable moment hinge can/will form at this cutout?

Informal = Dogbone
Formal = RBS = Reduced Beam Section

 

[dhengr]

Thanks KootK, that sounds about right to me.

 

[BAGW]

How do one detail braces to resist only the amount of force engineer wants to? It all goes back to stiffness again as everyone is pointing out here. Its easy to do a RBS connection as its well tested and the detailing requirement is well documented. Here its a different scenario. I liked the concept of slotted hole like suggested (atleast makes sense on paper), but what happens when their is multiple cycles of loading? We wont know the location of the bolt after first cycle of load and unload. Seems like there is no research on limiting members to design for loads engineer likes.

 

[dhengr]

BAGW:
This is one of the difficulties, or confusions, of the way the codes and our methods are written these days. We are designing for the max. or extreme condition, but we rarely really see that. In the extreme, we expect some joints and members to start to fail, yield, have large deflections, cause moment hinges, etc., well beyond serviceability conditions, as long as they don’t completely fail (falling apart) and killing people. Some members, connections and the like will need to be replaced or repaired after this extreme event, but 99.99% of the time we are operating at a lower load and stress level, and then the relative stiffness of the two systems governs. My dogbone idea on the max. loaded tension braces is a ‘fuse element’, a pretty clean and clear cut/simple P/A situation, which relieves the load on that frame at some point, changing the/its relative stiffness, and causes the load to start to shift to the other system. Again, you better be darn sure of your bldg. modeling so you have a good handle on these various loading proportions. Slotted holes have proven not to be a particularly reliable detail for these types of conditions. Those joints tend to rust and seize up and the slots tend to fill up with rust and solid crap and then not allowing the hoped-for movement. We really don’t know how to design a slip critical joint to slip at a specified load.