Reinforced Concrete Corbel Testing

[PreciseLogic]

Dear STKO

I hope I am posting this in the right location.. I am having great trouble matching two STKO models with real-life testing as reported in literature. The literature, attached, tested numerous precast corbels. For representative samples A3 and B1, one with stirrups and one without, I just cannot get my STKO models to get close to tested values.

For specimen A3, the tested value was 28.0 kips - STKO gives 41.4 k The real specimen failed in diagonal tension (rebar did not even yield during test, but STKO takes the flexural steel into strain hardening)
For specimen B1, the tested value was 47.0 kips -STKO gives 43.6 k The real specimen failed in flexural bending

So these two models, using exactly the same modeling techniques, grossly over-predict in one case and under predict in the other. I have attempted to model from simple to more complex (including strain hardening)

Please help me figure out why my models are so far from tested values. P.S. I am modeling these in 3D, as opposed to 2D, because once I validate my modeling materials and techniques, the goal is to predict capacities for corbels with eccentric load (in the global y direction) and that cannot be done with a 2D model (which is also why I am modeling the corbel plates with a zero-length-contact element).

Sincerely

PreciseLogic

[STKO Team]

Please extract from that document the information related to one of your model (materials, setup, etc..) so that it's easier for us to check We cannot go through a 100 pages document

[PreciseLogic]

Dear STKO

Yes, of course, my apologies. I took the pdf article and reduced it to just the 9 pages that describe the "A" and "B" series of corbels. I tried to highlight in the pdf the pertinent details.

Sincerely