Tension Only Bracing - Single Angles 2

[Everynameistaken]

Hi All,

I see there have been several threads on this already but do not seem to answer our specific question.

We have a low rise building < 15m using a system that has low ductility. As such we are permitted to use tension only bracing and no strict capacity design (CSA S16 conventional construction)

The most efficient is to use cross braces angle, we use two bolts at each end and connect in the middle with a single bolt.

I have usually considered the central connection point as the length when checking the KL/r limit for the tension brace.

Ie. 10m node to node, with a cross so the length is now halved so my kL = 1.0 x (0.5 x 10m). I know there is quite a bit of conflicting literature on this point, but since we are not relying on this brace point for a compression member I am on with this.

My question is more about which r should be used when checking the kl/r limit. For a system with very low ductility (R < 2.0) the limit for kL/r < 300.

Since the angles are held square at each and by the gusset and at the centre we have typically used the r about the geometric axis. These are the higher r values from the code. The principal axes has a very low r in one direction and could impact the angle size due only to the slenderness limit.

My understanding , for tension only systems with low ductility, this kl/r < 300 limit is more about sag and if the sections are held in place and connected together, the geometric axes seem to make sense.

Let me know your thoughts

 

[human909]

Sag and vibration are the only big concerns if you are on the slender side. Sag can be an issue for horizontal cross bracing but much less of an issue with vertical cross bracing. Vibration can be an issue if there is vibrating equipment or it is exposed to wind.

The code I use doesn't have slenderness limits. But I've been burnt by pushing slenderness limits in situations involving vortex shedding. The actual member was easily strong enough, but with laminar winds at the right speed there was severe resonance. (L/r in my case was around 200)

 

[dik]

Is 300 the limit, or is it just recommended? I seem to recall there is an exclusion for 300. Most PEMBs use rod with a KL/r greater than 300.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[XR250]

I think just about any rod bracing will exceed 300

 

[OldDawgNewTricks]

My comments are based on AISC, not CSA, so take them as you will.
1. 300 is a recommendation, not a hard limit. It is advisable to comply with the recommendation because experience has shown that it helps to avoid problems. But you may exceed it if you have fully considered the ramifications (sag, vibration, etc.).
2. The recommended limit of 300 is not applicable to rod bracing.
3. The limit also applies to the fabricated length, not just the unbraced length in the completed structure. Very slender members are also susceptible to damage during fabrication, transportation, and erection.
4. I would check the Z-Z axis, not just the X-X and Y-Y axes.

 

[KootK]

1) As others have noted, there are plenty of cases where I wouldn't sweat the 300 limit.

2) Were I to enforce the limit, I would use the geometric weak axis. Regardless of the rotational restraint at the ends, I feel the geometric axis will be the most flexible with respect to movement and, therefore the most critical.

3) The benefit of the rotational restraint points will be a function of how long the member is between those points and how torsionally stiff the cross section is. A very short piece probably would be forced into its geometric axis patter. A long piece with crap torsional stiffness (single angle) probably would not.

 

[Agbsh]

The effective length (k[sub]z[/sub]) is 0.85 times the half diagonal length considering the radius of gyration in the z-axis, r[sub]z[/sub] and (k[sub]x[/sub] = k[sub]y[/sub] = 1.0) times half length in r[sub]x[/sub] and r[sub]y[/sub] (see screenshot from El-Tayem and Goel). All the three limits (0.85 * (L/2) /r[sub]z[/sub]), (0.85 * (L/2) /r[sub]x[/sub]) and (0.85 * (L/2) /r[sub]y[/sub]) to be in range of 200 to 300 (recommendation as mentioned above).

The paper I mentioned is available free of charge in AISC:

https://www.aisc.org/Effective-Length-Factor-for-the-Design-of-X-bracing-Systems

 

[lexpatrie]

If nobody minds I'll slap that into an FAQ for the AISC forum.

[link

https://www.eng-tips.com/faqs.cfm?fid=2919

]1986 - Tension Only Bracing (effective Length Factor for the Design of X-bracing Systems)[/url]

If you're looking to cite the article, here's the official version of that, from AISC.org.

El-Tayem, Adel A.; Goel, Subhash C. (1986). "Effective Length Factor for the Design of X-bracing Systems," Engineering Journal, American Institute of Steel Construction, Vol. 23, pp. 41-45


If there's anything else or anybody knows those two articles I remember reading on the subject, if you'd like you can use the feedback feature on the FAQ.

 

[lexpatrie]

I can't seem to edit FAQs, so I'll dump the information here for the moment.

Picard, A.; Beaulieu, D. (1987). "Design of Cross-bracings Part 1: Theoretical Study," Engineering Journal, American Institute of Steel Construction, Vol. 24, pp. 122-126

Design of Cross-bracings Part 2: Experimental Study - Picard and Beaulieu, 4Q1988

Discussion and Closure, Design of Cross Bracings Part 1, 2 - Picard and Beaulieu / Stoman

Those are all engineering journal, and while there are no links, if you need them you should be able to find them via aisc.org. I can't find part two or the closure currently.

 

[canwesteng]

I believe that is for tension compression braced systems and relies on the tension in the opposite direction bracing the compression brace as well as restraint from the gusset. I'm not sure it is the same rationale for bracing tension only members against sag.

 

[lexpatrie]

If you mean the articles I mentioned, I just pulled the names off my index, I didn't reread them. As I recall, it's tension and compression bracing used together, meaning the idea was the compression brace worked and the tension brace tended to be used to brace the compression leg, and take forde as well. That was a design vogue back when for a whole...

As I recall this was kind of the death knell for that approach because it doesn't work that way in testing.

The other article... I'm not sure, the time frame for the 1988 articles was before the vogue of using HSS (né Tube Steel) so the angles didn't have much compressive strength. I don't really follow using a k factor in what seems to be tension bracing, so I suppose it's trying to use a "solo" compression brace versus X bracing, i haven't read that one.

Sorry, maybe I should be quiet until I've read the new article and reread the older ones. I'll have to pester AISC about the part 2 and the closure.

 

[dik]

That 0.85 times half the diagonal length is consistent with the 0.45 times the diagonal length that I've used as a guide. The 0.45 came from an early publication published by one of the Aluminum suppliers.

 

[canwesteng]

I don't have a problem with the article, I just don't think it applies to tension only bracing as described by OP. The braces might not slap in the wind though, but they will steel seem to vibrate.

 

[dik]

That's why with round BAR, you attach together with a 'U' bolt.

 

[human909]

On a tangent topic....

I've pretty much given up on utilising tension angle and cross bracing in general for most of the steel structures I design. (Though my peers certainly haven't) There are practical issues like sag, vibration and less flexibility penetrations through the braced plane. There are also structural issue of non-linear stiffness behaviour and unpredictable load path behaviour with discontinuities. (eg when the compression brace buckles then load path can completely change possibly in a very short space of time.)

If I do use cross bracing I prefer tensioned rod. As it many but not all the above issues and has good economy if D-nuts can be used, it loses some economy if turnbuckles are used.

I prefer single load path bracing. Which would normally be HSS but can also be angle, H-section or channel depending on the circumstances.


My above preferences probably won't win me the cheaper bids in the PEMB game. But my clients generally aren't after the VERY cheapest price and I believe I am generally quite economical in my steel design because I have spent extensive time talking to steel fabricators and erection crews.

 

[canwesteng]

I lean on double angle T+C bracing a lot. Connections tend to be cheaper than HSS or W braces. Sometimes it doesn't really add much tonnage over single angle T bracing, though for long spans I'm still using single angle tension braces.