Vertical Bracings takes a considerable amount of the gravity load 1

[TLycan]

when modeled a multistory steel building In SAP2000 or any FEM Program the vertical bracing takes about 35% of the gravity load. I don't Like that; I like designing the columns for the whole gravity load. I want to know what do people do? Do they really design the brace for its share of the gravity load and the columns for the rest?. If not how they model it

 

[haynewp]

If you can define the connection so that it is welded after a significant part of the dead load has been placed (for example, instruct them to install the braces after the concrete has been poured on the floors) you can reduce some of the load into the braces. But they would still see part of the subsequent dead and live loads.

 

[JoshPlumSE]

There are two ways to do an analysis on these types of buildings and their are two types of programs.

One school of thought is to model everything together (i.e. the ETABs way) and solve it all at once. The other school of thought is to separate out the gravity solution from the lateral solution (i.e. the RISAFloor or RAM method).

Both methods will have their own pros and cons. But, I'd say this thread does a good example of explaining why this 2nd method is relatively popular. If you've got braces that would take significant load out of columns (or beams in a Chevron type frame) under pure gravity loading, then many engineers would prefer the 2nd method over the ETABs way.

 

[haynewp]

I think it would be unconservative to completely ignore the compression braces under gravity loads since this will lessen their capacity to take lateral.

 

[fancypants]

I don't think I have seen it mentioned in this thread and may have no bearing on the OP's problem although it is relevant. I will inject into the conversation that if there are seismic requirements when using R > 3, then for V and inverted V braces the code AISC 341-05 specifies the beam should be designed assuming the braces provide no support of the dead and live loads.

 

[ToadJones]

I only skimmed through this post, but I usually make all my vertical braces "inactive" for the gravity cases to be sure that the vertical braces don't rob too much gravity load, then, I run another set of combinations with the braces "active" and design for the worst case.

I would think that the only gravity loads you'd want the braces to be able to withstand would be from column shortening.

 

[JoshPlumSE]

Haynewp -

If you solve your gravity only load cases (or gravity + notional loads) in your lateral model then the braces will be designed for their share of the gravity loads. That

 

[haynewp]

I meant generally speaking. It may be that whatever software you are using can account for the additional compression in the braces from gravity in a lateral model and combine the results. But only solving gravity cases is not going to get you the combined effect on the compression braces.

 

[NS4U]

if the bracing takes 35% of the gravity load, then design the columns for 35% more load.

the analysis isnt wrong... load follows stiffness. so your bldg wont fail. your emgineering judgement is right. i'd have the same concerns as you. you just need to go model crazy. just upsize the columns by a reasonable amount and be done with it

 

[NS4U]

you DONT need to go model crazy. just add some extra load to the columns

 

[SteelPE]

Depending on what jurisdiction governs your design you may end up with special loads in the bracing/columns anyway. A jurisdiction that I do a bunch of work it requires the designer to comply with section 8.3 of the AISC Seismic Provisions when selecting a R=3 system. I had to be a little cunning to figure out how to check this inside of the software. Kind of a pain but that is the cost of doing business today.

 

[MJB315]

If you assume that the lateral braces are a load path for gravity forces, won't only the top level of columns be undersized? I know a bunch of assumptions come into play (^ bracing, uniformly distributed loads that are equal on all levels, etc.), but how much force really gets taken away from the columns?



"We shape our buildings, thereafter they shape us." -WSC

 

[JAE]

MJB315, A lot of it depends on the angle of the braces (wide, more horizontal vs. steep/more vertical braces).

 

[Michel60]

Good morning all,

This one seems to have brought lot of attention and I think points out some limits to consider in the total reliance on computer model results. Most programs in a conventional analysis mode complete the formation of the structural system in a zero gravity environment and then turns on the gravity switch once all the stiffness present (not the way a building is constructed in the real word). This approach will tend to be less accurate for the amount of gravity load present in braces, particularly for upper stories. In fact in high-rise structures it's not unusual to see computer model columns on upper floors go into net tension due to column shortening and load redistribution in the bracing system. Doing a sequential loading model or using a parallel model with reduced brace axial stiffness can be helpful to bracket the solution for the column/brace forces, as others have suggested.

Also if you chose to go with an design that rely on a large portion of the gravity load being carried in the bracing I think you'd have to question the potential ductility and performance of that system. Where does the gravity load go when the brace buckles?

regards,
Michel

 

[hetgen]

Hi,

I would design the X-bracing as a rope tension only and the V and single diagonal braces to take their share of the gravity load under gravity+lateral combinations.