Deep Column Weak Axis Buckling 1

[DCBII]

I have some very deep columns (W40x249) that are very tall. I'm trying to reduce the unbraced length of them along the weak axis by adding vertical bracing between columns.

Normally, I'd say framing a braced frame strut into the center of the column web would be enough, but on a W40 column a W10 strut seems kind of... puny.

Do I need to brace both flanges? Add full depth stiffeners? Use two struts (one at each flange of the W40)? I'm curious how others would tackle this problem.

 

[peal724]

I agree with you about the scale and proportion. Consider this, instead of weak axis being centered how about two lines of bracing alligned with flanges.

 

[dmaughan25]

I have the same issue. We are designing 130 ft tall bent structures that are supported laterally at the top by a truss. I have wind load on my columns so they are now beam columns.

I am curios if there is any literature on double clip connections and there use to brace the top and bottom flange of the columns. My columns are W24-W30. So I am thinking aI cannto count on a Double clip grapping close enough to the flange to truly prevent twist, but I know it would prevent some twist.

 

[dmaughan25]

Here is an overall view of the structure

 

[DCBII]

dmaughan25

Our office has actually done a lot of conveyor bents with W24 bent legs. We've typically put a full depth stiffener on the opposite side of the connection and they've performed find (some of them have even been hurricane tested). A couple of guys in our office have been doing conveyor bents for 30+ years this way.

For some reason doing this in the W40 range makes me a little uneasy though. My application is a perimeter "wind column" spanning 140' (hence the deep column to control deflection).

 

[RFreund]

I will see if I can dig up something more concrete, possibly form yura.
However I will add this for now and hopefully someone can correct me if I'm wrong:
I think it depends on what unbraced length you are considering: Flexural Buckling or Torsional Buckling.

For flexural buckling, the concern is lateral movement, which to me seems like a member framing into the web would be able to meet this criteria. For the axial strength/stiffness requirements of the brace, I would look at AISC Appendix 6.

For torsional buckling, the concern (I believe) is rotation. In this case you would need to check if the brace member can restrain rotation. For the moment strength/stiffness requirements of the brace, I would look at AISC Appendix 6.


EIT

http://www.HowToEngineer.com

 

[dmaughan25]

Thanks a ton for the answer. I can see how the stiffener would take care of the issue. Is there any equation that shows the amount of lateral torsion induced by a certain moment in your column?

 

[dmaughan25]

Yeah I have read Yura's article on fundamentals of beam bracing. It basically comes down to if your connection can resist the twist from lateral torsional buckling, but I cannot find anywhere what this twist force is unless it is the required strength given by equation A-6-1 and A-6-5

 

[TLHS]

I'd probably put in stiffeners, but I'm not fully convinced that you actually need them assuming the connection point at the web doesn't punch out or flex so much that the bracing isn't effective.

A centreline brace would stop the lateral translation of the section, regardless of the connection to the flanges. The lack of connection to the flanges just leaves the chance that the brace causes some sort of torsional failure at the brace location. I assume it would end up happening at some load level just due to initial imperfections. However, this seems like a failure mechanism that requires a lot more energy.

I'm curious about this and if I have time I'll dig a bit.

 

[dmaughan25]

The problem is AISC says the connection must prevent twist (Appendix 6.3) and a gusset oriented parallel to the flanges in my opinion has no capacity to resist twist in the flanges. I think the double clips will have some capacity and with the web stiffer I think you could make it work.

In this scenario I think the brace has adequate strength, it is the connection to the brace that I am worried about

 

[DCBII]

Correct me if I'm wrong... but I think most wide flange columns fall into section E3 (flexural buckling) of the code (13th Ed.) rather than E4 (torsional and flexural torsional buckling). Does this imply that torsional buckling is not a consideration? In that case would I only have to restrain traslation and not rotation?

 

[RFreund]

I forgot since this is a beam-column.
The compression flange should be braced to prevent translation or the entire cross section should be braced so to prevent twist relative to the restrained end of the beam. But I think we are ahead of all that at this point.

Stiffener seems logical.
As for torsional strength/stiffness I was referring to A-6-9 and A-6-11.

EIT

http://www.HowToEngineer.com

 

[dmaughan25]

I believe you are right DCBII for column only action. I am looking at beam column action so torsional buckling is a consideration under F2.2. For beam-column column must be restrained against translation and twist. (see definition of Lb in section F2.2) Lb=length between point that are either braced against lateral displacement of the compression flange or braced against twist of the cross section

By the way how do I change my discipline to structural on this website. I set it up as civil/environmental but I am structural.

 

[hokie66]

dmaughan25,
At the bottom of the page, there is a "contact us" link. Suggest you try that and request the change.

 

[RFreund]

Also for columns, AISC notes that for some cases (doubly symmetric sections) where the torsional unbraced length exceeds the flexural unbraced length, then torsional-buckling may 'control'.

The paper "Fundementals of Beam Bracing" does given an example on how to find the rotational stiffness and does talk about stiffeners. The detail shown does not seem to provide much torsional restraint but at the same time that is a HUGE section.

http://www.aisc.org/store/p-994-fundamentals-of-beam-bracing.aspx

EIT

http://www.HowToEngineer.com

 

[WillisV]

Torsional buckling is always a consideration with columns, it just rarely controls over weak-axis flexural torsional buckling. If you only put a brace into the middle of a deep column you will prevent flexural torsional buckling about the weak axis at that point but will still need to check torsional buckling for the entire unbraced length.

 

[RFreund]

Wills - can you elaborate how at brace at the center can prevent LTB? Even if it was pin connection? It seems that this prevents lateral translation but the section can twist, no?

Thanks!

EIT

http://www.HowToEngineer.com

 

[WillisV]

RFreud - didn't say it prevented LTB, said it prevented Flexural-Torsional-Buckling (column weak-axis buckling). Entirely different limit states, LTB is a beam based (moment) limit state. I was speaking directly to the issue of having to check a column for compression loads for pure torsional buckling if the brace prevented weak-axis flexural torsional buckling but not twist. You would also need to check lateral torsional buckling for flexural limit states along the full length of the member unless the brace was adequate to prevent twist (and if it was, the torsional unbraced length for axial loading checks would also be reduced).

 

[DCBII]

I think WillisV has it right. The unbraced length for flexural buckling may be reduced by a simple pinned strut framing into the column, but the unbraced length for flexural torsional buckling would remain the full length of the member unless provision were made to prevent twist of the section. In other words, bracing the column for flexural buckling does not necessarily brace it for flexural torsional buckling. I think I need to restrain twist of the cross section.

 

[RFreund]

Yikes - sorry Willis, I miss read your post. Which is actually good cause that means my understanding was correct and is basically what I said before just with more confidence...

EIT

http://www.HowToEngineer.com