Beam end restraint

[graybeach]

Hello all,

I'm in a discussion with an engineer I work with about whether and where stiffeners are required for the assembly shown on the attached sketch. The W18 beams will be installed beneath badly deteriorating concrete beams to prevent collapse should the concrete beams fail. The top flange of the W18 is unsupported as there will be a gap between it and the concrete beams. The steel beams will be installed in discrete locations so there is no way to connect them together.

This engineer will seal the drawings and is responsible for the design, but I was assigned to review it. That we have this review process with no protocol to solve disagreements is one of the annyoing things about my company.

The disagreement here is that I believe some stiffeners are required to restrain rotation of the beam ends. The design engineer disagrees because he says that the W18 is compact, and the W14 bracket is not the same as a column that could become unstable.

I'm trying to decide how much of a fuss to make about this. What do you think? Is there any way that stiffeners are not required? If stiffeners are required, where would you put them? Should any stiffeners be added to the bottom of the W14 bracket?

 

[nutte]

You need to restrain the end of the beam against rotation. This is similar to a seated beam connection. For these, you could use an angle on the top flange of the beam, or one on the web near the top flange, to provide stability. A pair of stiffeners would also do this.

If you're looking for a reference to bolster your argument, look in the Steel Construction Manual, starting on page 10-84 in the 13th edition. There is also a paper referenced that might be helpful, "The Results of Experiments on Seated Beam Connections" by Roeder and Dailey.

 

[dik]

Oh boy... depending on the reaction, you may not need a plate stiffener... it's a matter of web crippling at the intersection of the two beams... it may be a small area, but if the load is small, it should not be a problem. The stiffener can be used to increase the web crippling area.

You may want to 'dry pack' the space between the beam and the concrete beam.

I would, normally, as a matter of practice use a stiffener myself.

Dik

 

[rday]

I would stiffen.

I woud refer to AISC 13th Edition Section J10.7 on unframed ends of beams and girders.

 

[JAE]

I'd check for web buckling, crippling, and also I would definitely put a rotational restraint at the top of the beam.

Vertical stiffners would be another choice. I would do something here...not leave it as is that is for sure.

 

[graybeach]

Thanks everyone. I read the Roeder and Daily paper and realized that our connection is a lot like a stiffened seated beam connection. Maybe stiffeners aren't really needed, but top restraint most likely is. Now to deal with the politics...

 

[Lion06]

Compactness has nothing to do with whether a stiffener is required or not. It may be ok, but AISC requires beams to be rotationally restrained at the ends. The stiffener is a no-brainer, no stiffener, in my mind, means you need to proved it's rotationally restrained.

 

[a2mfk]

I agree that a top restraint is required, this is a seated beam connection.

A web stiffener would depend on the web buckling strength, a completely separate check that is usually not a problem except in short spans with high shears (in my experience).

Vertical stiffeners out-of-plane with the web could kill two birds with one stone, but I am guessing you only have one bird to kill. Murder that bird with a clip angle at the top of the beam.

 

[BAretired]

a2mfk,

Did you mean "Murder that bird with a clip to the top of the skull"?

BA

 

[tumbleleaves]

Pushing this post back to the top, because I found this topic of immense interest a long while back.

My understanding is that AISC requires ALL unframed beam ends to be stiffened because that is how the equations for lateral-torsion buckling are derived (how the ends of the beams are restrained during the tests that are used to develop the equations).

Most (nearly all) of the popular steel texts do not address connections with unframed unstiffened beams, even though it has to be done in practice all the time: although not necessarily by structural engineers, since they typically design only buildings and strictly to a code such as AISC. Just my 2 cents.

 

[graybeach]

Thanks tumbleleaves. It is amazing how this is subject is rarely mentioned in steel texts. I'm working on getting the stiffeners added. Can't do a top angle because there's nothing to bolt it to.