What is the difference between RC slabs (one way and two way), shells and panels? Is the load transfer same?

[meetshah432]

I am trying to model a pedestrian bridge that is essentially a 2 span portal frame with 12m and 3m spans. The deck is placed symmetrically on the single beam(400mm x 700mm) (horizontal member of portal). The deck is 3.2m wide i.e. 1.6m either side of the beAM. Attached a snap shot for geometry reference:

Self weight, parapet SIDL of 1kN/m on long edges, 3kN/m2 SIDL, 5k/m2 of LL are the loads applied and I'm checking the results for 1.35DL + 1.35SIDL + 1.5LL.  
I modelled it with 2 way slabs and shells and the results vary greatly. 
Shell model:

Slab model: 

I checked the reactions and other basic checks. All loads are being considered and equilibrium is maintained. Can someone explain the load transfer mechanism and degrees of freedoms of the slab and shells? 

I cross checked the model with other softwares. If we model just the portal frame with equivalent UDL on horizontal top beam, the BMD results match with shell model, with moments around 850kNm. But if we model the slab as shell elements the moments are around 140kNm. 
Without slab: 

With slab:

I believe the stiffness and load transfer between shells and slabs have a lot to do with the changes in moments and will drastically affect my design.
Can someone kindly explain what is happening on the back end and what are the assumptions of the choices between shell and slab?

[Rafacascudo]

Without having the model ( zip and attach both),

If you set your meshed panel to have 1 or 2 way simplified distribution method , then this panel will act like a cladding and have no stiffness taken in account. The only benefit is having its self weight generated automatically.

And it doesn't matter being 1 or 2 way , because the only "load receiver" os your beam.

With a meshed panel model with pure shell calculation model( Load transfer=analytic) ,the slab/panel has stiffness and both, beam and slab will "share" the loads and divide the task of resisting them.

That's why the beam moment in this case is so smaller.

[meetshah432]

I will share the model shortly, but I think the results are opposite. The moments are large for shell and less for slab which is exactly opposite of what I would expect indicating that only weight is considered for shell but weight and stiffness is considered for slab. It did not agree with the description mentioned as well.