Numerical and Experimental Investigation of Punching Shear of Slab-Column Connections Retrofitted with Ultra High Performance Fibre Reinforced Concrete (UHPFRC)

Abstract

Punching shear is described as a brittle failure mode that can occur in reinforced concrete buildings at the slab-column connection area. Many times, such a failure mode can initiate a potential progressive collapse of the building. An in-depth investigation was performed to examine possible structural failures of flat slab construction, as well as research punching shear both experimentally and numerically. This included validated Finite Element Analysis (FEA) models and a planned proposed retrofit method to strengthen the slab, promoting improved serviceability and resilience. A proposed retrofit method of using Ultra High Performance Fibre Reinforced Concrete (UHPFRC) as a tensile overlay was suggested for a square slab with side lengths of 1.8m and a thickness of 120mm, having extended square column stubs with side lengths of 200mm. Two slabs were constructed, including a control specimen (un-retrofitted) and a 30mm full tensile overlay retrofitted specimen with UHPFRC. Two more slabs are planned to be constructed, with all four being tested in future research. These additional two slabs will have full overlay thicknesses of 15mm and 20mm. FEA models were established and analyzed for all proposed slabs. All FEA were conducted in ABAQUS software using the concrete damaged plasticity model. Another FEA software was also considered (ATENA), but the obtained results were not in good agreement with test results. ABAQUS results showed that an increase of thickness provided greater punching shear strength. The control specimen was compared to code provisions to validate accuracy. It was found that ACI-318-19 proved to be most accurate, whereas CSA A23.3-19 was one of the least accurate. To model the constitutive behaviour of UHPFRC, emphasis was given to the tensile properties using the inverse analysis equations proposed in the recent Canadian codes (CSA A23.1-19 and CSA S6-19). A series of parametric studies were performed, varying model parameters using stress-strain or stress-crack width properties and the inclusion of damage. It was concluded that the stress-crack width provided better results for the post-peak softening along with the damage parameters compared to prism test results; however, the peak load was found to be greatly overestimated.