CDP model- How possible Tension region turns out to be compression under bending?

[Matte StrEng]

Dear all,

Using the CDP model, I modelled an RC beam for a 4-point bending test. I followed all the instructions as explained in an example in the literature. The beam was loaded for 10 sec. I observed that the tensile region in the bottom of the mid-span turned out to be a compressive region as crack develops and propagates.

How could it be possible even though the beam is still under loading? Please find the attached images showing the stress distribution (S33 in a longitudinal direction perpendicular to the X transverse - Y vertical surface) at 0.6 sec and 10 sec. As of the fourth second of the analysis, the compressive stress comes out and starts increasing despite loading as can be seen from the stress-time curve.

I really appreciate any help you can provide.

 

[FEA way]

The deformation of the bottom part looks really strange. Make sure that there’s no hourglassing in the model.

Try visualizing crack directions with maximum principal plastic strain, as described in the documentation for the CDP material model. Other output variables and animation of the deformed shape of the model may show you what exactly happens in this case.