Nonlinear Finite Element Analysis of Reinforced Concrete Slab-Column Connections with Headed Stud/Bolt Shear Reinforcement Subjected to Punching Shear

Abstract

While many experimental tests have been conducted by various researchers on the punching behaviour of reinforced concrete flat slabs supported on columns with headed stud/bolt shear reinforcement, there are still many parameters which have not been adequately studied in the laboratory due to cost or time constraints. As a result much of the current code provisions for designing slab-column connections against punching shear are based on empirically derived formulations based on tests of partial scale isolated slab-column specimens. Researchers such as Genikomsou and Polak (2015), Navarro et al (2018), and Lapi et al (2020) have proven that these experimental tests can be supplemented using properly calibrated nonlinear finite element analysis (NLFEA) models in the commercial software ABAQUS.

In this thesis a three-dimensional NLFEA model is calibrated using interior slab-column connection specimens tested by Adetifa and Polak (2005) in concentric punching and exterior slab-column connection specimens tested by El-Salakawy et al (1998, 2000) in punching with unbalanced moment. This calibration uses the concrete damaged plasticity (CDP) model in ABAQUS, and includes discussion of the main parameters which influence the CDP model and values used in this calibration. The calibration also includes a study conducted to determine how to effectively model the shear reinforcement and shear reinforced area based on the “stem-star” method used by Genikomsou and Polak (2016). This includes a detailed analysis of the modelling of the shear stud star (S3) diameter to ensure enough rotational capability was provided in the shear reinforced region of the slab without significantly reducing the predicted capacity of the model. Through this study it was determined that a gap of 6-10mm should be provided between the column face and the first S3, and that the S3 diameter of subsequent rows of shear reinforcement had a negligible effect on connection behaviour. Additional effects such as changing the moment-shear ratio and the effect of adding openings near the column were also considered during calibration.

These calibrated models are then used to conduct several parametric studies on parameters related to the shear reinforcement in the specimens. Three parametric studies are presented in this thesis. The first study investigates the effect of changing the number of shear reinforcement rows and the spacing between adjacent rows of shear reinforcement. In this study it was determined that the total shear reinforced area has a larger effect on the connection than the number of reinforcement rows, and that when spaced between 0.75d and 1.5d no inter-stud punching could occur in any of the models.

The second study investigated the effect of changing the size and number of openings adjacent to the column, and the change in placement of shear reinforcement which must occur as a result of the opening changes. This study determined that 90% of the connection capacity was maintained when the opening width to column width ratio was 0.6 for two openings and 0.33 for four openings. This study also determined that providing stud rails outside of the openings in a double-cruciform arrangement has no significant effect on the behaviour of the shear reinforcement when compared to their usual placement.

Finally, the third study investigated the impact of anchorage-controlled shear reinforcement, which was proposed by Topuzi et al (2017) to increase the ductility of slab-column connections under cyclic loading without inducing higher lateral stresses in the connection, on the overall capacity of the connections. In this study it was determined that the inclusion of anchorage-controlled shear reinforcement results in less than a 10% drop in concentric punching connection capacity for all considered connections, and therefore could be suitable for inclusion in slab-column connections as proposed by Topuzi et al (2017).