Dear Inaam Qazi

1- In longitudinal direction, you should model composite sections (one for each girder and the slab above it).

2- In transverse direction, you should model slab sections (and diaphragms if any).

I worked on this modeling many times.

If you have any questions, just ask.

Best wishes...

Dear Mohannad

For 2 D grid analysis I assign steel girder properties in the longitudinal direction. in the transverse direction i assign slabs , then i offset the girders so that slab lies on top of steel girders.But the problem arises that the deflection pattern is not intuitive and is giving wrong results.

You have talked about composite action modelling. I am not sure if it can be done in staad pro .

I think the problem is that girders in the longitudinal dir and slab in the transverse dir act seperately as they are at different levels and grid lines are not rigid to tranfer load from one longitudinal dir to transverse dir.

With Regards

Dear Inaam Qazi

Please read "Bridge Deck Behaviour" by E.C. Hambly

What type of finite elements are you using for the slab? If they are plate bending only, even if you offset them with rigid links, membrane forces will not develop to simulate the composite action of the girder (there is no shear transfer at the beam-slab interface because the relevant degree of freedom may not exist at 1-dof per node plate bending elements). In that case you would need 5-dof shell (bending plus membrane) elements. If those are not available, you can model the beam as a composite T beam (just use equivalent material area by elastic modulus ratio above steel to simulate concrete if need be) and place the plate bending elements at the beam's centroid.