Truss Tension Chord Bracing - Really Necessary? 4

[WillisV]

A general question that has long been a topic of disagreement within my design office - do trusses (in particular long span steel trusses) need to be laterally braced at the bottom tension chord in order to provide adequate bracing for the vertical web member "columns" to behave as the pin-pin members (K=1.0) that we typically assume in web member design.

James Fisher wrote an excellent paper on this very topic a while back (Engineering Journal - Third Quarter 1983). His article was, however, followed by many dissenting opinions.

Input anyone?

 

[structural01]

Tension must be laterally braced to meet slenderness ratio.

 

[UcfSE]

Is there a link that you know of where we could read that article? That sounds like a good read to me.

 

[WillisV]

http://www.aisc.org/Template.cfm?Se...e=/Ecommerce/ProductDisplay.cfm&ProductID=974

Free for AISC members.

 

[UcfSE]

Thanks, now if only I were a member or knew someone who is.

 

[VirtualEngineer]

AISC Memberships for 2 to 6 individuals in one firm costs only $160.

If you split it up that comes to $26.67 each.

You get free downloads of all 19 of their Design Guides valued at $60 each.

When the new 2005(?)Unified Design Manual comes out you will get half price on the manual.

You get free downloads of all the recent Structural Engineering articles and many of the articles from years gone by.

This price is a pretty good deal.

In this electronic age, membership is getting to be a necessity for staying up to date with the industry.

I feel that our employers should to pay the membership dues (of course you employers out there may disagree). The benefit of a better-informed employee does accrue to the employer; however if your employer won't pay the dues, try banding together with other professionals to split the cost.

Regards to all,

JPJ

 

[valleyboy]

I don't think there is a requirement to provide lateral bracing to the tension members of trusses. In BS 5950, the tensile capcity of a tie is a factor of the material yield stress x (Gross area of section - area of holes for fixings).

Slenderless limits dont apply to tension members, but do apply to compression members - hence sometimes restraints are necessary to limit the effectve length of the bottom boom which becomes a compression member under wind reversal.

VB

 

[WillisV]

The argument in Fisher's paper is that the vertical web members of a typical truss are in compression. These web members are typically designed as column elements with an effective length factor of 1 (their actual length) - i.e. pin-pinned. For this assumption to be true the bottom tension chord must provide adequate out of plane stiffness to cause the vertical web members to act as if "pinned" at the tension chord rather than "free." Otherwise the vertical web members would need to be designed as upside down cantilever columns with an effective length factor of 2. In order for the tension chord to provide this out of plane stiffness it sometimes becomes necessary to brace the tension chord out of plane at certain points.

 

[miecz]

Read the article and it makes sense to me. Many years ago, a brilliant engineer told me that beams (in this case, trusses) need to be braced at the tension flange as well as the compression flange. When I challenged his assertion, he convinced me with this argument: When you load a beam in the plane of its major axis, it wants to relax, that is, turn its minor axis to the load. It doesn't matter if the compression chord moves laterally to the tension flange, or the tension flange moves laterally to the compression flange.

 

[flamby]

Like a beam, truss is also susceptible to Lateral Torsional Buckling. It is a good idea to brace the bottom chord when the members are slender or there is a possibility of misalignment in the plane of truss.

Ciao.

 

[pappyirl]

In my experience it is unusual to laterally brace the bottom boom of a truss, unless stress reversal is a problem. In a previous thread entitled "Steel Truss Lateral Bracing", several posts mention how little stiffness is required to provide lateral bracing.

Steel Truss Lateral Bracing

Using this analogy, if the bottom chord of the truss is stiff enough to resist 2% of the compressive load in the vertical web member, then we can assume the pinned-pinned scenario to be applicable.

In relation to bracing the bottom flange of a beam; why would the bottom flange in tension want to move laterally if the top flange is restained? The lateral movement is due to the compression force in the flange resulting in buckling. Only in the unrestrained situation can I understand it being irrelevant which flange is moving.

 

[WillisV]

Pappyirl,

You analogy is correct, but only partly so. There are actually two distinct requirements for a member to act as a brace.

1. Strength - commonly referred to as the old 2% rule. The bottom chord does need to be strong enough to resist a certain percentage of the vertical web member load. Note that the 2% rule is out-dated and that the AISC LRFD V.3 has revised bracing strength requirements in chapter C.

2. Stiffness - this is a separate criteria INDEPENDANT of the force in the members that must also be met to be considered a brace. This criteria normally controls over the strength requirement. Again the stiffness criteria are provided in Chapter C of the AISC LRFD spec.

It is the stiffness criteria that is really pushing the tension chord bracing requirements as the stiffness of a long-span bottom chord is normally relatively small out-of-plane.

 

[JAE]

No one here has mentioned the fact that tension members always tend to snap back to a straight line, like a string on a guitar. Now this only applies to truss tension chords that are fixed at the ends of the truss (also like the guitar string). Flanges of beams, for instance, are in tension due to a continuous build-up of tension through the cumulation of horizontal shear along the length. The ends of a wide flange beam arent always fixed axially at the ends.

I once designed a large bowstring truss (horizontal top chord comprised of jumbo WF shapes and curved bottom chord made up of bridge strand). The truss spanned about 370 feet and was about 50 feet deep at the midspan point. We and another consulting engineer created some models to investigate the possibility that the truss would have a pre-disposition to snap out of plane. The models were somewhat inconclusive, but seemed to show the truss always trying to snap back to its original vertical position after we pre-deflected it sideways.

In any case, we opted to be "safe" and added diagonal struts from the adjacent roof framing to a point about 2/3 down the length of the verticals.

The seminar on stability that Yura (from the Univ of Texas) presents usually shows that shapes need to be either braced at the compression flange against translation, or braced against rotation of the entire section..either working fine. No mention of tension chord/flange bracing was brought up.

 

[miecz]

pappyirl,

I've never actually run or seen an experiment to prove this, so it's actually just a theory that goes like this:

Given it's druthers, a beam would rather present its weakest axis to the direction of the load. Imagine a beam tilted at 45 degrees about its longitudinal axis from the vertical. The compression flange is continually braced and the tension flange is not. As the load is increased the beam will tend to twist to present it's weakest axis to the load. You can see this with a paper model.

As the initial tilt angle decreases, the tendency to twist decreases, but is still there. For most rolled beams, the stiffness offered by the web and bottom flange is enough to resist the twisting effect. But a truss has very little web and flange stiffness.

 

[AggieYank]

for whoever mentioned that the tension flange of a beam needs to be braced as well as the compression flange, i completely disagree. LTB causes the compression flange to go out of plane, not the tension flange. As far as a truss and the tension bottom chord being stiff enough to act as a brace for the vertical members, thats another story.

 

[WillisV]

JAE,

I have also run numerous models of trusses with imposed out-of-plane deflections and have run into inconclusive results. I think in this case it is best to rest on the side of caution. This same line of thinking, by the way, is one of the reasons while steel joists are typically bridged at the first node from each end. The forces in the vertical web members are highest there and therefore in most need of bracing.

 

[Tomfh]

Willis,

How come the compression members are assumed to be cantilevers when not restrained at the tension flange?

Why would they be fixed against rotation at the compression flange?

 

[JStephen]

Consider a related situation: A vertical steel bar welded to, and supporting, thin steel plate (this is a tank roof). How deep can the bar be? The failure mode I can visualize is the bar just folding up flat against the plate (which is what happens on a paper model).

 

[miecz]

Thank you JStephen. Good to see that someone else makes paper models. How deep the bar can be depends on the thickness of the bar and the thickness of the roof plate. This is similar to a rectangular gas duct under positive internal pressure with thin bar stiffeners. Even a stocky bar will flop over if the roof (or duct)plate is very thin. I believe that even a wide flange will flop if the roof plate is very thin, even though the plate gives contiuous lateral support to the compression flange.

 

[vmirat]

I would think that the top chord of an OWSJ would fail before the bottom chord ever did. Therefore, if you design the top chord as the limiting factor, the bottom chord would never fail, hence no need for the lateral support.