Cambering Beams In a Non-Rectangular Bay 1

JJStructural said:

The steel weight is reduced about 30-40% with cambering.
...beam lengths vary significantly from about 15'-40' long.
Would it be better to beef up the member sizes in this situation or is it okay to have different cambers?
...I would like to go the most economical route.

Click to expand...

IMHO, using different member sizes is preferable on your project for several reasons:

1) "Reducing weight" of structural steel only to "save money" is false economy in the 21st century. Design to minimize labor and improve constructability.

2) Cambering is not a precise operation. If camber < 3/4" is needed... don't camber.

3) In fabrication, cambering beams < 24' long is not practical.

4) For beams < 14" deep the web is usually too thin for cambering.

5) For beams of any size, webs < 1/4" are too thin for cambering.

See the article: "Specifying Camber" in AISC Modern Steel Construction Magazine.

For bridge girders, we don't specify cambering of rolled beams anymore. We've found using plate girders, or accounting for deflection with a variable slab thickness, to be the more economical approaches where we require camber. Granted, ours are typically 24" deep or more for 60' or so spans. For smaller beams, the costs for cambering may be more favorable.

Rod Smith, P.E., The artist formerly known as HotRod10

Thank you for the responses. SlideRuleEra, if I were to camber the beams, most of them would be 14" deep or less so it's probably not a good option based on that alone. Hypothetically though, if my beams were larger, would it cause issues to have beams shorter than 24' NOT cambered adjacent to >25' beams that ARE cambered? The girder on one side of the bay is angled so all of the beams have different lengths.

JJStructural said:

..if my beams were larger, would it cause issues to have beams shorter than 24' NOT cambered adjacent to >25' beams that ARE cambered?

Click to expand...

Assuming each beam has simple supports and the framing supports a concrete slab, yes, that situation could very well be an issue. IMHO, it has to do with deflection under varying live loads:

Until the concrete slab is placed and has set, beam deflection (within reason) does not matter. Therefore dead load is not a problem for various beam deflections.

The two beams in your example probably have very different values for Moment of Inertia (I). As live load (assumed to be uniformly distributed on the entire slab) increases from zero to 100%, to two side-by-side beams deflect differently. This will stress the slab, which may or may not cause a problem... but is not desirable.

Here is what I would do. Since each beam is unique take the opportunity to design each beam so that it's relative deflection is the same under full live load... say L/360. The 40' beam will deflect much more than the 15' beam, but relative deflection for each is L/360 +.

To do this, consider the deflection equation from the appropriate beam diagram:



For each unique beam length, the only variable for which you have control is Moment of Inertia (I).
Select each beam based on it's value of "I", so that relative deflection of each beam is more or less constant (L/360).
There are two ways at your disposal to select "I".
Choose either a beam with a different depth or beams with a constant depth but differing weight / ft.
Probably will take a mixture of both to come up with a rational design.

This is not a "perfect" solution, but I believe it is a reasonable compromise.

Thank you, that is very helpful. I wasn't considering the relative deflection but it seems obvious now that it is important. Appreciate the help!

You are welcome. Cambering is good for making a slab "flat" for one specific loading condition, but does nothing to change a beam's deflection characteristics over a range of loads.