camber in steel beams

huangyhg

camber in steel beams
i would like to start a thread about cambering of steel members.  following are my general guidelines in cambering but if anyone has a different opinion or have other considerations, please feel free to contribute.
1. typically use 75% of precomposite dead load on a floor beam to calculate camber.
2. minimum camber is 3/4" and in 1/4" increment.  ignore any calculated camber that is less than or equal to 1/2".
3. no camber on beams at sloped roof if slope exceeds 1/2" per foot.
some engineers choose not to provide any camber on floor beams.  instead, they provide for an additional 1/2" of leveling concrete (which is poured together with the deck slab) in the dead load.  is there see any benefit in doing this?
thanks for y'all's input.
for steel bridges, i adopt (100% dl + 75% ll) to compute the camber. the idea is to keep it straight under most likely loading. these factors are recommended by our steel bridge code.
the camber is just for the purposes of 'finish and appearance' under a percieved load and is without any structural value (unless you are a fan of fe-nl analysis). the camber may be important for larger structures like bridges but can be more easily managed by leveling concrete if spans are smaller. i would also resort to leveling layer of concrete, wherever i could manage because it is so cheap as compared to cambering exercise which no fabricator loves to do. cambering requires bending, reaming of holes to correct tolerances and so on at the cost of large man-and-machine power.
aisc has some cambering guidelines in the manual (check the index ... i'm away from my desk so cannot cite the page number.   these depend a bit on the depth of the   
thanks flame and jae.
my question was intended for buildings so bridge criteria is not the one to follow.  for building framing   
whyun - yes, that is the section i was thinking of.  
we got into a camber issue recently with two large w21 beams spanning over a roof where a large condenser unit was positioned on top of the steel.  the manufacturer's of the unit showed up and tried to set the double condenser on the beams.  they of course flexed downward upon the application of the load and the light gage metal on the sides buckled a bit due to the deflection.
we had specified camber in the beams so that with the operating dead weight of the unit the beams would be close to flat.  we knew that the variables in getting this beam perfectly flat were the actual weight of the unit, the true e of the steel, and the potential for some degree of fixity in the simple bolted connections at either end.
the unit manufacturer complained that our beams didn't meet their spec.  we had used some fairly long spans (don't re  
a couple of rules we work by,
1.  don't camber any beam less than a w14x22 - just increast the size.
2.  don't camber a beam less than 20 feet long.  we have been told multiple times that it is near impossible to camber a short beam because the rigs don't work for that short of a beam.  
-doug