curved beam

huangyhg

curved beam
can someone please shed some light on the design procedure for a curved beam. i assume it is slightly different than a traditional design. i don't have a particular problem that i am trying to solve, just something i'm curious about. thanks.
find a job or post a job opening
one thing to be aware of is that the neutral axis of a curved beam may not necessarily pass through the centroid of the section. treatment of curved beams is therefore separate from "straight" beams. see any strength of materials textbook for a full treatment, where you will probably see the analysis of a lifting hook, which is a classical example of the analysis of curved beams.
-- drej --
torsion must be considerred.
the calculations are more rigorous than normal concentric bending or axial loading.
good design will attempt to minimize torsion effects.
here are all the references that i have on curve beam design, all out of past aisc engineering journals.
  horizontally curves steel girders-fabricated & designed
  by w.m. thatcher, july 1967.
  analysis of curved girder bridges by charles culver,  
  darryl brogan & david bednar, january 1970.
  straight-element grid analysis of horizontal curved beam
  systems by herbert a. weissman, april 1970.
  approximate torsional analysis of curved box girders by
  the m/r-method by david tung & richard fountain, july
  1970.
  box girder bridge design-state of the art by c.p. heins,
  fourth quarter 1978.
  the application of flexural methods to torsional
  analysis of thin-walled open sections by thomas boothby,
  fourth quarted 1984.
i've read these papers, but, fortunately, so far, i never had to apply them to a practical problem.
i've found the easiest way is probably to discretize the beam into sufficient number of straight segments, and run it on the computer. avoids complicated equations that you don't understand and can't rely on. make sure you get the support conditions right.
are you referring to a curve in plane or out of plane? i suppose an in - plane curve would be more of an arch.
my original idea was a beam curved in a horizontal plane. although, an arch would also be considered a curved beam but in a vertical plane.
there is a good paper on the subject:
bradford, m and pi, y-l
design of steel i-beams curved in plan to as4100
australian journal of structural engineering, vol 3 nos 1-2 pages 85-97, 2000
as4100 is the australian limit state steel structures code.  the bending-torsion interaction equations proposed in the paper incorporate the effects of the included angle, secondary bending about the minor axis and torsion actions for several lateral bracing configurations.
all the above responses are correct and appropriate.  for a single curved beam, the lifting beam is the best design example.  for a system of curved beams (like a curved girder bridge deck), it is usually required to analyze/design the curved girders as a system (i.e. 3-d) because the lateral bracing components are considered as primary structural   
helical stair is also an example of curved beam. this can be solved by inputting the co-ordinartes and load magnitude in computer program. i use multiframe 3d program.
my question is: is the result from the computer program acceptable as regards its output on tosional stresses
use aisc torsion analysis for steel.
aci ch 11.6 for concrete.
i think most programs don't automatically consider torsion. some have particular switches you must flip before the torsions analysis is done if at all. i use hand calcs to check the computer output. then refine the computer models.