simply supported deep beam

Hi,

I have a problem with a simply supported deep beam which has a uniformly distributed load. The beam is simply supported at both ends but I have to consider the results for the following 3 scenarios
- where the support is at the bottom of each end of the beam
- where the support is at the mid point at each end of the beam
- where support is at top of each end of the beam

I have run this using FEA and I'm finding significantly different deflections and stresses when the beam is supported at the midpoint at its end and the other two scenarios...can anyone explain the exact reasoning for this?

Sincere thanks

Daniel