Moment frame bracing with steel joist

[ugandabob]

I have a steel moment frame. The beam in the frame supports steel joists connected to the top flange only. Can I rely on the joists to brace the beam for the axial force it carries, or is my unbraced length the entire length of the beam?

 

[connectegr]

Are the joist simply bearing on the top flange of the beam? If the connection does not resist an axial force, the flange in not braced.

http://www.FerrellEngineering.com

 

[BAretired]

The bottom flange must be braced in order to consider a reduction in unbraced length. This could be done with bottom chord extensions from joist to bottom flange.

BA

 

[a2mfk]

thread507-287322

recent, extensive thread on the subject, with a great accompanying technical article link...

Like BA and connect are alluding to, in a moment frame your bottom flange will also be in compression, so you in effect will have two unbraced length checks to do. Your bottom flange may very well control your design and like BA said you will need diagonal kickers to brace it.

Always consider up-sizing the steel member to avoid bracing, steel tonnage vs bracing and stiffeners is very often more economical.

 

[tenguy23]

As a side comment, the AISC 13th Edition has provisions for Nodal Bracing (Appendix 6 and Commentary Appendix 6).

If your joist connections can transfer axial (as connect pointed out), you can design your brace as a lateral stiff spring that must satisfy a minimum stiffness & a minimum strength to overcome 1% of the axial in the main member.

If the brace can satisfy these limits, then your main member under compression can be considered to have a smaller unbraced length.

Unfortunately, as BA & a2mfk pointed out, in a moment frame you will need to check both the top/bottom flanges...

 

[TTUengr51]

If the joists are adequately attached to take compressive brace forces, you could count on the top flange being braced in compression. However, the bottom flange could still buckle, so I would be hesitant to use the joist spacing as the unbraced length for compression. You could use something in between the two extremes, but I am unsure of any publications code wise that would back this up. Would be a judgment call.

This also would hold true with negative bending on the beam for lateral torsional buckling.

One thing to consider to brace the bottom flange would be a kicker member (i.e. angles) to brace the bottom flange back up to the diaphragm. This is typical in the pre-engineered metal building industry to brace the bottom flange of the bents back to the purlin framing.

Nick Deal, PE, SE
Michael Brady Inc.

http://www.michaelbradyinc.com

 

[Lion06]

Unless you have very long beams I don't think it will make much of a difference. The axial load that gets into a moment frame beam is typically very small. These beams, in my experience anyway, are typically governed by controlling lateral drift. That, along with the fact that there are typically a lot of moment frames in a building, means each one takes not a whole lot of load. It also makes sense when you think about the fact that the moment frame members are really working in bending. I'm not sure how you typically model moment frames - if it's a single frame ( two columns and a beam) I put the windward a leeward forces at their actual respective locations. If you do this, you'll find the axial load is een smaller (and will actually probably be tension). If your software calcs the lateral loads for you and assigns their location in the diaphragm (as RAM SS does) then you'll have to pick a node to disconnect or you'll get no axial force) - this is the most conservative approach, and will result in the highest compressive (or tension) force.

To address the OP, though, I don't think a joist connected to the top flange only braces the beam against compression. More importantly, it doesn't brace it for negative bending and that will be more of a problem than the axial bracing.

 

[gyung]

Effective length for axial compression, I wouldn't consider joists as lateral restraints.

If the member is subjected to large axial forces for some situations like conveyor belts with equipment loads, If I were the designer, I would rely on those secondary members in the bracing system as lateral restraints, and making sure a clear and simply load path of the bracing system from the major loads to the ground supports.

P.S. If beam subjected to large hogging bending moment, which bottom flange subject to compression bending stress, you could consider "fly-brace" to shorten your effective length in bending.

 

[csd72]

One other option is to check the top and bottom flanges separately as the bottom flange will have a much lower stress.

 

[rscassar]

No, the joists fixed to the top flange of the beam will not restrain the member for compression (pure axial load) buckling. The axial force is assumed uniform over the entire section so I would want to restrain more than just the top flange.

For lateral-torsional buckling (flexural), then yes, I would say that the joists provide restraint to the top flange. I believe I have had this debate with SEIT in the past regarding whether this detail provides lateral restraint or not.

 

[Lion06]

Kik-

Have I ever said that this detail provides lateral restraint for axial load in the past (with only one flange braced)? If so, I must have had too much eggnog before I started typing.

 

[WillisV]

Even if the joist detail could provide lateral restraint for axial load to resist flexural torsional buckling it certainly doesn't resist pure torsional buckling (see 13th Ed. E4) which if you run the numbers for most cases provides similar capacities to unbraced flexural torsional buckling about the weak axis of the member - so in reality it is a moot point.