Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

RE: Curved Steel Beams

[Subject Prev][Subject Next][Thread Prev][Thread Next]
According to AASHTO (Section 10.15) and Fabrication Aids for
Continuously Heat-curved Girders (USS by Roger L. Brockenbrough), the
minimum radius of curvature shall not be less than 150 ft or the larger
of:

R = 14*b*D/(Fy^0.5 *Y*tw)

R = 7500*b/(Fy*Y)

where

b = widest flange width, in.
D = clear distance between flanges, in.
Fy = ksi (not to exceed 50 ksi)
Y = ratio of the total cross-sectional area to the cross-sectional area
of both flanges
tw = web thickness, in.
R = radius, in.

I'm not sure about the residual stresses, but if heat-curved, they
should be lesss significant than if cold-rolled.

Tony Powers
HDR Engineering

> ----------
> From: 	Tom Castle
> Sent: 	Wednesday, February 18, 1998 5:09 PM
> To: 	'seaoc(--nospam--at)seaoc.org'
> Subject: 	Curved Steel Beams
> 
> I am working on a structural steel framed roof with a curved "turtle
> shell" surface.  Of course the architect wants the volume ceiling to
> reflect the top side of the roof, so no trick flat beams below.  I
> have
> a couple options as to how to frame it, but was wondering what the
> practical limits are as to how tight a radius a wide flange beam can
> be
> curved.  The beam size would be about a W16x26.  I suppose that the
> beam
> can always be fabricated out of plates, but I if think I want to avoid
> this.  I know how to analyze the beam once it is made, but can it be
> made and are there residual stresses that must be taken into account?
> The radius at some points is as little as 12 feet.  Any thoughts?  It
> would be appreciated.  Thanks.
> 
> Thomas Castle, S.E.
> 
> 
>