When I had to use curved beams to support concrete to get circular
parameters it has always been my position that the top and bottom flanges must
be supported. I use a small angle brace from the bottom flange to the deck at
about the third points. The torsion caused by the difference between the cord
of the circle and a straight line between supports is analyzed for torsion and
the warping stress treated like weak axis bending stress.
The best option is to keep the beams straight and cantilever the deck.
From: Sandip Guha [SMTP:sandip.guha(--nospam--at)cshqa.com]
Sent: Wednesday, March 15, 2000 11:07 AM
Subject: composite and non-composite steel beams-curved in plan
I am in the process of designing a roof and a floor framing system with
steel beams/girders that form the perimeter of a circle of diameter 106 ft.
Four columns, spaced evenly along the circle, supports the metal roof deck
and the composite floor slabs (4-1/4" lightweight concrete over a 1-1/2"
metal formdeck- total thickness is 5-3/4"). My questions are:
1) In roof diaphragm, can the metal deck be considered to
resist the torsional stresses resulting from the steel beams that
are curved in plan (radius of curvature 53 ft., arc length 83 ft.,
eccentricity of curved girder is 6.74 ft at center of beam
and nil at supports (48" dia concrete columns).
2) In floor diaphragm, can the composite floor be considered to
be able to resist the torsional stresses resulting from curved
Thank you for your prompt attention. Sincerely,