How can I check if my model makes sense?

[bojoka4052]

I have made the structure below with beam elements; the transverse beams (In the Y-direction) does not have all the properties defined and is only supposed to carry loads between the longitudinal beams (In the X-direction). The structure is supported by rollers, and I want to find out if the structure can withstand the loads it has to carry and if there is enough reinforcement (its a concrete bridge):

I can for example check the deformation with only dead weight applied and it seems to deform in a right way:

How else could I check if the model works? Should I make load combinations and see if the required reinforcement makes sense?

 

[FEA way]

Do some rough analytical estimates of the deflection of this beam structure (you will have to further simplify it for this purpose). Make sure that not only the scaled deformed shape makes sense but also stress and displacement values. The model is simple so there shouldn't be any serious issues with it but once you will add complexity to it (for example, model the structure using solid or shell elements with embedded beams as a reinforcement), issues may arise. Make sure that the mesh is sufficiently dense - check if the results converge with mesh refinements.

 

[ThomasH]

bojoka4052:
The structure you are analyzing is statically indeterminate. That means that the load distribution depends on the stiffness in the members. So you need to define all valid properties to get a reasonable result.

After that you have a 2-span beam with a cantilever in the x-direction. The "beam" consists of 5 beam elements that are parallel but a reasonable hand calculation should be possible.
In the y-direction you also have a 2-span beam but with two cantilevers. Again, reasonable results with hand calculation should be possible for comparison.

If the results does not convince you, simplify the model.

The evaluation of reinforcement should come after you have checked that the model behaves reasonable.

Thomas