Creep Deflections of Composite Girders

I have a book called "Composite Construction Design for Buildings" by Viest, Colaco, Furlong, Griffis, Leon, Wyllie. In the version I have, Chapter 3 is on Composite beams, 3.5 addresses serviceability and 3.5.1 addresses deflections w/ sub-sections on instantaneous and long term deflections. They have an extensive reference list.

The reference that looks the most useful to me (though I haven't read it) is:
D15: IM Viest, RS Fountain, RC Singleton, "Composite Construction in Steel and Concrete for Bridge and Buildings" McGraw-Hill New York 1958.

I believe it gives some simplified methods for estimating long term deflection in a composite beam. And, compares them to more complex methods. It's sort of summarized in the book I have, but I don't think it's fully treated. At least not to my satisfaction.


My general thoughts if Long Term Deflection is a major concern:
1) Don't use light weight concrete which tends to have more creep / shrinkage issues. At least not without being very explicit about the need to avoid creep with guy who's designing the concrete mix.
2) You can increase the amount of reinforcement in the slab. Not only does this reduce deflection (even if you're not counting on it), it also helps create some continuity / restraint between beams. So, the beam would start to act as more like a fixed and less like a pinned support at interior ends of the beam. Again, not something you generally count on, but something to think about.
3) Obviously, try to camber out the self weight deflection. Then, if possible, wait a little longer to apply the cladding, so that the beam is fully set.

AASHTO suggests the use of transformed section properties with a value of 3n (where n is the modular ratio = Es/Ec) when transforming concrete to equivalent steel for permanent loads applied to the composite section of bridge girders. We use that value for stresses and deflections.

Rod Smith, P.E., The artist formerly known as HotRod10

Creep won't affect load applied before the concrete becomes composite as the concrete has zero stress at that point. You would have to watch out for shrinkage warping though.

Edit: this assumes the formwork is supported by the steel beams so the beams carry the wet concrete.

It's a fair jump in design effort to do properly and take account of potential variation in the concrete parameters. How close are you to your limits? If you give some details of geometry, concrete properties and load, I can give you my estimate fairly quickly.