do we ignore inelastic behavior of cold rolled beams

huangyhg

do we ignore inelastic behavior of cold rolled beams?
example:  we ask for a hss 6x6 rolled to a radius of 25'.  the fabrication shop takes a straight hss 6x6 and jacks it by brute force at third points until the beam has a radius of 25'.  to be bent to the radius, the beam yields and goes into inelastic behavior.
when the beam is designed, we assume elastic behavior, and use an allowable stress that is less than the yield stress (asd).
is there a justification for ignoring this seeming discrepancy?  is the beam heat treated to restore elastic behavior?
thanks
i asked this question of an engineer in our office regarding some beams that were curved to a pretty tight radius for a project we did not too long ago.
i was told that the beams are heated before they are bent (or curved) and that it's not really changing the properties or the elastic response of the   
aisc engineering journal has some good articles on the subject:
cambering steel beams, ricker, 4th qtr. 1989
cold bending of wide flange shapes for construction, bjorhovde, 4th qtr. 2006
both are available as free downloads from aisc.
they probably don't jack it into a 25' radius.  that's likely done with an angle roll or other similar equipment.  if you jack it at third points, you would get uneven yielding along the length.  you'd get a curve, but not a circular curve.
and as far as i know, that is indeed ignored.  just like the residual stresses inherent in the manufacture of the beam, and the residual stresses from any type of welding or bolting or cutting operation.
beams that are hot rolled have residual stresses which often which yield when added to working stresses. beams that are cold cambered often have additional built in stresses that are typically ignored in computing stresses.