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

 

[hokie66]

Columns take all the gravity load. Bracing is for lateral loading.

 

[TLycan]

Hokie66, how you modeled in SAP2000 or what ever program so that the whole gravity force will transfer to the columns?

 

[hokie66]

I don't know about SAP, but just leave out the braces, and the load will go to the columns.

 

[TLycan]

This means I have to make two models one for the gravity load with no vertical bracing and one with the vertical bracing for lateral loads. and even in the last case with the load combination ( D + L +lateral) the vertical bracing will take again part of the gravity load

 

[hokie66]

Oh, the problems of using computer models.

 

[msquared48]

Pin the braces and make them tension only.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[TLycan]

Thank You All. I think, yes, the idea is in making the braces to carry tension only.

But I have two other Questions:
do people really do this when modeling a multi-story building?
The other question why not design the columns and bracing to their share of forces especially this is what really should happen in reality; that is the brace will take part of the gravity loads

 

[hokie66]

How many storeys are we talking about? What type structural system? What are the typical column and brace sections?

And to answer your question...structural engineers use a wide variety of modelling techniques, including the good old pad and pencil, especially in the preliminary design phase. When it gets to the final design, many of us design the floor system first, run the loads down for the columns, design the columns, and only then refine/design the lateral system. Sometimes there have to be changes to the gravity system to accommodate loads imposed by the lateral forces, but we don't revise the bracing/lateral system to take gravity loads.

 

[TLycan]

I am talking about a six floors refrigerator building with total height of 25meters. the building dimensions is 7meter width X 22meters length. the system consists of secondary beams simply supported on the main frames which spans the short directions. the lateral stability is achieved through main frames in the short directions and vertical bracing in the long direction between the columns. the vertical bracing configuration is X-Bracing , V-bracing and sometimes a single diagonal; this is because piping is going in and out and the bracing have to avoid them.

most of the columns sections are HEB500. the Vertical bracing are 2channels UPN240 Forming box section
thanks for your quick response

 

[hokie66]

In that case, I doubt that any practicing engineer would assign gravity load to the diagonal braces. When we model building structures, we don't try to figure out exactly what path the forces take, we just make sure there is a reliable path. Often, the structure is smarter than us.

 

[msquared48]

Not smarter Hokie, just more stubborn.

Mike McCann
MMC Engineering

http://mmcengineering.tripod.com

 

[koohi]

Because sap2000 method of analysis it is inevitable. Your structure is undetermined and the loads be shared between all member in a joint. It is a real event. But in analysis and in vertical loading case, if you don’t like axial force create in bracing if your bracing system is X you must design structure (columns) without bracing elements in vertical loads case and then design all of structure for vertical and lateral load cases.

 

[JAE]

I agree that developing or purposely defining a competent load path can work for some structures but if you have heavy X-braces in a multi-story structure, the X-braces DO take axial load - in reality.

If you imagine a pair of columns with a floor beam below and above, with an X-brace - if the upper two joints move relative to the lower two joints then the X-braces MUST take axial load. There is no getting around this. The question is whether you must consider this axial load in your design (i.e. define a competent load path per hokie66 above).

I think the concern I would have is that if you have a multi-story building with fairly rigid braces which soak up some axial load, perhaps significant axial load, then under a seismic event or wind load condition your brace would have additional axial force where you would have somewhat less capacity due to earlier buckling.

 

[SteelPE]

If there is a load path for the load to go into the brace then I design the brace accordingly. I have designed plenty of bracing like this. I am currently looking at designing some connections for a fabricator (US procedures) where I suspect the engineer designed the bracing for gravity load (greater compression forces than tension forces). The EOR even specified that no welding is done on the bracing until the slab is poured (good luck with that one when pouring a suspended slab in the winter) to try and reduce the loading on the connections.

You can always adjust your columns (by hand) to carry the full load w/o the bracing. How difficult is that?

 

[abusementpark]

I see no problem with designing the structure how it will behave, which should be captured by a properly built FEM. As SteelPE said, you can always overdo the columns if you want to sleep a little easier.

 

[WillisV]

I design the braces to take their share of the gravity load under gravity+lateral combinations because they will. I also design the columns (and beams if you are in a chevron or x type configuration) to be able to handle 1.2D+1.6L gravity only load cases as if the columns aren't there as a point of personal preference.

As far as having to create two different computer models in order to do the above, boo hoo, it takes 10 extra minutes, try doing it with moment distribution...

 

[WillisV]

err...replace columns with braces in second sentence above.

 

[TLycan]

Thanks Everyone

 

[paddingtongreen]

This is one of those strange cases where both are correct. Designing for vertical loads going into the bracing is clearly correct, but experience tells us that designing for only lateral load works. The answer is perhaps in the different member sizes produced by the designs. I tend to think that if a bracing designed for lateral only is ever hit by full lateral load plus its theoretical vertical load, something yields, bolts slip etc., and so adjustment is made.

That's the only explanation I have for the fact that both designs work in practise.

It would bother me to put in columns that were light because some of the D & L loads were in the bracing.

Michael.
Timing has a lot to do with the outcome of a rain dance.