steel column with the geometry of a greek cross

[cesaramorim]

Hi,

I've been suggested to use a steel column with the following geometry

This is a greek cross geometry and it was used by Mies Van Der Rohes in Barcelona Pavilion.

I thougt about doing it with 6 "L" shaped beams but I'm not sure the finish will polished between each shaped beam.

But the really relevant question is the safety design of it. The eurocode 3 shows how to determine the class of a single "L" beam and other singular elements, but that's as far as it goes.

Has anyone had this kind of trouble?

 

[TehMightyEngineer]

Looks like your message has been cut off.

This should be analyzed as a cruciform column, which used to be a fairly common shape as it has flanges on four sides allowing for easy connection of beams to the column. It has an unusual failure mode of flexural-torsional buckling. A google search of cruciform columns will likely turn up a lot of details and research. Not sure what the eurocode has regarding it but if it's like the AISC then there are likely provisions for it.

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[cesaramorim]

I think I've come up with a solution.
I tested my model with a HE 120B section from arcelor mittal and it was safe.

Therefore:

The dimensions fit perfectly. And like I've read before, if it's stupid but it works, it ain't stupid

What do you think?

 

[TehMightyEngineer]

Those angles will be incredibly flexible welded on only one edge like that, at face value I don't like it.

The more typical way I've seen cruciform columns created is welded tee shapes to the web:

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[cesaramorim]

But the H120B beam is enough to sustain the load by itself, what could go wrong with 4 aditional "L" shapes?

Also I've never seen that kind of cross section, I've just run trough the arcelor sales program, and couldn't find such kind of beam.

Thanks for your help

 

[Guest262653]

The following publications might be helpful:

https://sydney.edu.au/engineering/civil/publications/2011/r921.pdf

http://www.newsteelconstruction.com/wp/wp-content/uploads/TechPaper/NSCApril06_Tech.pdf

Check as well paragraph 6.3.1.4 of EN1993-1-1.

 

[structSU10]

Its not a piece you just 'buy' its a fabricated assembly. they take a two w shape members, cut one down the middle to make two WT sections, then weld the WT sections to the center of the web of the other W shape. The true cruciform shape is much less susceptible to local buckling issues than your angle solution, and has a more 'reliable' load path to get load to the built-up portions of the section.

 

[TehMightyEngineer]

Even if your original wide flange alone is acceptable, those flexible angles will likely exhibit global buckling. As you're relying on the original column for structure this isn't going to reduce your column capacity but will likely deform finishes placed over the angles. Plus, you're welding on a lot of metal. It would be better to have that metal actually do something rather than just be filler. Welding on WT sections as struct suggested will reduce you're required column size and likely give you easier connections to roof and floor beams (as applicable).

Professional and Structural Engineer (ME, NH)
American Concrete Industries

http://www.americanconcrete.com

 

[KootK]

It would help if we had a better sense of what you're trying to accomplish. What is the loading and column height? Are there particular cross section dimensions that you're gunning for? What's the application? There are a ton of clever folks here to help but they'll need some parameters to get started.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[cesaramorim]

The columns are to provide support to something close to a cantilever. I estimate their load as 100kN of compressive force and minimal (around 10kN.m^2) bending moments. The model confirms those loads.

I have a long slab with 13 m that has 4 columns on its extremity, and is partially supported by 2 walls on the extremity. (The concrete walls are below the blue marks)

The problem so far is finding a cross section with such geometry that when cut will be compatible with a "L" section. To create a perfectly simetrical cross, something like this would have to happen

This means the web would have to be 3x the flange, or maybe I'm doing something wrong, I'm not sure.
But if it's really like this, there's no profile with such geometry

Edit: I think I'll probably will have to cut 2 "I" sections in half and get 4 "T" sections that dont meet at the end. That way I could get 1/4 of the load on each T member, and analyse each one individually wich would be simple.

This is also what I understand from the image on the first post. It would make the analysis simpler, but I'm not sure if the fact they don't meet at the end makes my load assumption wrong. For example, the T element on the inner side of the slab, would surely have more load than the one in the extremity.

 

[structSU10]

I don't understand why you think you need this. Not knowing the spans you are supporting, there seems to be no need to provide such a member. A standard column sections should be able to handle this no problem, and is much cheaper than a fabricated assembly. Are there other reasons as to why you are trying to use this detail?

 

[KootK]

Thanks for the additional detail. If you've already got concrete construction above, perhaps a concrete cruciform column would be an attractive alternative.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[cesaramorim]

For architectural reasons that's really the only reason why I must get a greek cross there.

Thanks everyone for the tips

 

[racookpe1978]

If it looks prettier, and if the archt. can get the owner to pay for it.... go for it!

Now, be ready to work a little harder to paint that inside space.

Actually, the WF + doubled WT does look elegant, looks "clean" and does NOT look like a piece of "ugly, plain-old-structural steel" Plus, it can be welded with a simple jig and frame. The welds do not need to be continuous, but a staggered weld is a bit more difficult to calculate. Then again, you only need to calculate it once.

A WF can be split fairly easily into WT's, but find out if buying them separately as WT is actually going to be cheaper - Many places purchase WF, then cut them in half to sell you WT's that you could have cut yourself.

 

[structSU10]

Do they want exposed structure, or is the column going to be wrapped with finish? If the column is to be wrapped, I would be inclined to have the finish create the cross shape, and leave the column a standard column. If it is architecturally exposed, make sure they are paying attention to required fire rating - if this thing needs intumescent paint it will look ugly real quick. Otherwise, the comments above about how to build it still stand. This link may give some useful information:

http://eprints.utm.my/3658/1/JTDIS43B[05].pdf