Slab design - volumetric vs. shell

[Anonymous]

Dear Support,

I need your help with explaining why below don't work. 😞

As a simple example I am trying to compare 2 simply supported slab models. One is build as a shell model and one as volumetric.

Both have the same dimensions: 4m span, 2m width and 300mm thick.

There are linear pinned supports applied to shorter sides.

I know that shell model is the center line model therefore I replicated that in volumetric model by setting the support lines at mid height to get equal and opposite stresses.

I have applied 10kN/m2 load on the whole slab surface.

The one thing that I don't understand is why it gives me different values of stresses in the same points.

Shell gives Sxx of 1.33MPa against 0.94 MPa of the volumetric moment in the mid span of the slab.

Please find attached models.

Looking forward to hear from you.

Best Regards

Jan

[Artur.Kosakowski]

Solid elements are less 'accurate' and such model requires denser meshing. I'm attaching the models with 0.1 m mesh size and doubled number of divisions along the height of the solid.

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

Artur,

Thank you for your answer.

I have an additional question if you don't mind.

I am designing piled bund tank area (see picture below) and in this moment in time all the elements I have modeled as shell elements.

The only thing that I am worry about is if this model is accurate specially for the plinths as they are deep and in shell model everything is centered as in real life slab and plinth are flush through the bottom surface.

Could you please state what would be your recommendation of modeling something like this in Robot.

Thank you in advance.

Kind Regards

Jan

[Artur.Kosakowski]

What is the scope of the project (what kind of output / results are needed) ? 

[Anonymous]

Hi Artur,

I need to do reinforcement design for all the elements of that structure and also define the pile loads for the contractor.

Kind Regards

Jan

[Artur.Kosakowski]

In Robot reinforcement can be only calculated for shells (not solids) and the assumption is that the governing effect is bending. If for pile caps you assume that design of reinforcement should be by the strut and tie method (shear) then you may decide to model them as rigid links for load transfer only as this is not supported in Robot. 

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Anonymous]

Hi Artur,

Thank you for your response.

Just one more thing.

On example would you be able to explain the principle of that interface?

Do you mean of leaving the hole in the slab and fill that with rigid links between the piles?

Looking forward to hear from you.

Kind Regards

Jan

[Artur.Kosakowski · Accepted solution]

Yes, something like that. E.g. a master node above the piles being the slave nodes if to want to apply horizontal forces due to wind etc. 

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.