Longitudinal shear flow and torsion in box girders

[HighPanda]

I am doing design of a composite girder which is formed of a steel “U” trough with concrete slab at top.

Because of torsional and distortional warping stresses, do I have to design shear studs for these load effects? If so and warping stress being, say A (N/mm2) at the junction between the web and concrete slab, is the additional longitudinal shear flow equal to A x steel web thickness?

For the shear flow around the box section, is the transverse shear force resisted by shear studs equal to “shear stress near web/slab junction” x “web thickness” x “spacing of shear studs in longitudinal direction”?

 

[ishvaaag]

I think you need to care of these effects; if not they might provide detrimental enough to be a liabilty to some degree in the life of the bridge.

Respect how and how much. The principle of composite design that is worthy youmay try to respect is the elastical conjoint behaviour at service level. It then will need to also have the required safety factors for ultimate loads not as respectful of integrity as what at service load a correct behaviour for deformation and fatigue of the connectors demands.

Having the composite behaviour fully ensured at service level in plus of it having the necessary reserve to limit (even in worse state of interface integrity) will ensure a sound life for the bridge.

So in my view, first and foremost you need to determine with precision the shear forces at the concrete-U_steel interface, what you can do by some elastic analysis both for the service level and as well for the limit state. If you are as well required to provide full steel yield or overstrength capacity for the connectors is more a matter of a wish or statement of the code than of actual physical requirement once all structural matters are dealt with.

 

[ishvaaag]

And now a feasible manner to get the interface requirements with some accuracy. 1st you plan where your studs to make the thing composite are to be. According to the nodes yo then get, I would decompose my concrete beam in a matrix of rectangular plates of concrete, to the nodes of which I would tie with rigid links to corresponded symmetricalle placed at position steel plates, and then joined at bottom by another array of steel plates, also with rigid links to nodes in the central set of concrete plates. The forces at the links will represent with enough accuracy the tributary force to the corresponding stud. Using the actual studs for links wouldn't do better because they would take flexure to the central concrete set of plates, hence better use rigid links.