effective length of elements in partial side-supported trusses 2

[majmaz]

Hi
It seems in planar trusses that side bracing of compressed chord is provided at all joints but tension chord is not restrained completely against lateral movement (side-supported at some joints), the effective length of diagonal and vertical elements for out-of-plane buckling in non-restrained panels is greater than L (K>1)!

Though the "K" factor can be calculated by analytical or numerical modeling of this kind of elements, I'm looking for any recommendation of codes or previous researches for effective length of diagonal and vertical elements in above mentioned trusses.
Can anybody give more information?

 

[Lion06]

This has been covered before on this forum. I'll look for the actual thread.

Many on this forum, and in my office, do not agree on this subject. I believe that this is more of a global buckling issue, and not a local buckling (buckling of a single diagonal) issue. Here is why - If the bottom chord displaces sideways and the digaonal goes along for the ride, its own buckling length is still its own actual length. You can think about this by imagining a vertical, slender pipe pin supported at its base with no lateral support at the top other than a cable that is slung over the top and attached to the ground on either side of the vertical pipe. If you tension the pipe, it may become unstable and flop down and rest horizontally on the ground, but its own buckling length (kl) is stil its actual length.

In my opinion, the bigger concern is the global stability (the pipe flopping over).

That being said, I've gone through a couple examples of energy calculations to check stability for such a case and have tentatively convinced myself that it's not a problem. See the attached paper for the energy method stability check I used.

 

[Lion06]

Here is another paper.

 

[Lion06]

Here's one more.

 

[Stillerz]

Structural EIT-
Isn't covered somewhere in the code that truss members may be taken to have k=1.0 for all members?
I'm brain farting here pretty bad....

 

[Lion06]

I don't think so. If it is and you find out where, please pass it along.

 

[Stillerz]

I must be thinking of pg 5-39 of the 89 code under "braced frames".

 

[ishvaaag]

The importance of "bracing" the lower chords is associated more than anything with avoiding a "house of cards" failure with all the trusses rotating around their support points due to eccentrical application of the loads atop. This kind of failure everyone has slipped off the side of a rolling trunk or seen the japanese humour program here called "Yellow humor" with all those giant rolling cylinders know. For trusses hanging from points over the lower chord, at any situation where such eccentricity can't be forfeited, or where some lever arm of the same may make rotate the trusses on their supports, lower chord or transverse bracing or stabilization against rotation is warranted.

 

[msquared48]

As ishvaaag said, the top and bottom chords need to be properly braced to prevent rotational failure.

Mike McCann
MMC Engineering

 

[Lion06]

I agree with the top and bottom chords being braced to prevent a rotational failure, but I believe this is more related to the idea that beams need to be rotationally restrained, at a minimum, at the ends, and that this isn't a local stability issue for the diagonals.

 

[BAretired]

I could be wrong, but I think you are all saying the same thing. A compression member braced at one end only is in unstable equilibrium. If the bottom chord is adequately braced, all of the compression members in the web have an effective length of their actual length, i.e. k = 1.0.

The King post shown in Fig. 11 of the Nair paper posted by SEIT requires a brace at the bottom of the king post. Without it, the structure is unstable and will fail under load.

BA

 

[ishvaaag]

I just paraphrase the last sentence in last BAretired post and say, without (bottom bracing) the structure may lose, maybe, who knows, its structural depth upon rotation on the supports, and then a structure calculated with a proper K, could still fail.

 

[majmaz]

Thanks for your replies. according to the replies and introduced papers by SEIT, I concluded:

1- Effective length of diagonals for this kinds of trusses is L(K=1).
2-Minimum stiffens of side supports are required for above assumption.
3-Diagonals can be designed as an ordinary column by formulas of AISC or other codes.

majmaz

 

[BAretired]

In particular, steel joists without bottom chord extensions at the ends should be laterally braced at or near the crank point.

BA

 

[slickdeals]

I am too late to jump in on this thread. But read the commentary of the AISC 3rd edition manual (Page 16.1-189) suggesting that K=1 for truss members. I don't know why this has been removed in the 13th edition or I can't find it.

 

[dik]

I have an old paper from the AISC indicating the importance of bottom chord bracing... ie, top chord bracing doesn't work well without it... I'll try to dig it up...

Dik