Newbie question on composite bridge modelling using SAP 1

[kww2008]

I am trying to model an existing bridge using the bridge modeller. It is a steel concrete composite bridge. The concrete slab is attached to the 4 steel beam top flange using shear studs.

The model uses a frame for the steel beam and shell elements for the deck. The results for a load case include the overall moment at a girder location and the corresponding component moment and axial force for the girder alone.

(1) Is this a wrong model to use as the neutral axis of the steel beam would be at mid depth of the steel beam and the neutral axis of the concrete slab would be at mid depth of the concrete slab because the neutral axis of the composite section would not be these (expected to be below and close to the top flange of the steel beam)?

(2) Are the component moment and axial forces for the girder provided so that stress and strain can be calculated using P/A +/- My/I based on the assumption that the steel section carries these forces?

(3) would a model using plate elements for the steel girder more appropriate for analysis steel concrete composite structures of the type decribed above?

 

[ishvaaag]

I have not (yet) faced bridge modeler, but it should be rare it placing (always) a compression head for some composite beams bridge at midheight of the web. If such is the case it seems to me your question should be a modelling issue addressable from the bridge module itself.

Other than that you may get some insight on the actual behaviour of the composite bridge from general SAP2000 modelling with just frame elements for the beams and shell elements for the slab through the use of some more or less rigid links (they could be even tuned in stiffness) between corresponding nodes.

This way your beam would get a semblance of closer values to those required for a steel check with the solicitations it will be seeing as a component of the composite bridge setup, for otherwise, as you describe, the model would be more some steel beams supporting some slab -at web's midheight.

Yet a bridge module should do better and have some built-in definition for the composite bridges addressing these issues, what I can't confirm (yet) for SAP2000.

 

[kww2008]

Thank you ishvaaag for your input.

When using bridge modeller, the steel beam is modelled using frame elements and the concrete slab using shell elements, and I think these two element types are offseted by an amount to represent the composite structure. Every node of the frame is body contrained with a node of the shell elements directly above it. After analysis the program gives the forces (BM. shear and axial force) acting on the steel girder. My question is whether I can at this stage treat the steel girder on its own and calculate the stresses at the top and bottom fibre using P/A + or - My/I. I am concern because if I do this I am assuming the neutral axis at mid-depth of the steel section. If I do not do this, how can I determine the top and bottom fibre stress of the steel girder. Hopefully this clarify my initial posting.

 

[ishvaaag]

You can ascertain if (as should be the case) the beam is really showing composite action, and hence, its own solicitations as a (separate) component of the composite structure by just looking to the axial forces readable for the beam. Assuming just an isostatic example, if you are getting significant tension, you are in your composite setup as you want, and the solicitations are steel-only, quite likely. If you are getting little axial tension, quite likely the beam is not seeing composite action, but this is scarcely probable out of your description, that, as you see, quite closely matches my proposal to model the composite action with just the general tools of SAP2000; where they have slave-linking between nodes I would be placing quite rigid (or tuned rigidity) links.