Local Corner Moments in RC Slab

[ashley_calleja0806]

I am currently modelling an irregularly shaped reinforced concrete slab and am experiencing high/localised bending moments at the sharp corners of the slab panels.

I am trying to model a one-way slab, with linear pinned supports along the supported perimeter edges and linear realeses between each panel. At several of the sharp/acute corners, Robot is producing local moment peaks.

I have tried several approaches to determine whether this is a modelling issue, including:

• Refining the finite element mesh around the affected areas;
• Checking the linear support conditions along the slab edges;
• Adding a column/support as corners nodal supports simulating column.

Despite these changes, the local corner moments remain. I have attached the Robot model (.rtd) for reference.

Could anyone advise whether these corner moments are likely to be FE singularities/local plate effects associated with the sharp geometry and intersecting support boundaries, or whether there is something in the modelling that should be changed?

Any feedback on the model would be greatly appreciated.

[Krzysztof_Wasik]

Hi @ashley_calleja0806 

Usually you can expect peaks of moments in sharp slab corners. In this particular case most of those (peak ) values are caused by uplift effect in sharp slab corners. Slab corners are trying to move up. Since UZ is blocked quite large negative reactions are produced in such sharp corners.

See regions with negative reactions (linear Fz reaction diagram is activated in screenshot below) 

You can reduce this effect by allowing UZ movement (uplift UZ+) in slab supports. 

Model will be calcualted in nonlinear way (calculations will be longer) but for such support type negative reactions will not be produced.

And peaks for bending moments will be reduced