| dc.description.abstract | Due to their many benefits including ease of construction, open floor plans and reduced cost reinforced concrete flat slabs are one of the most common structural systems. However, these systems, particularly flat slabs supported on columns alone (flat plates) are susceptible to punching shear failures. Punching shear failures are brittle and occur due to the complex state of stress at the intersection of the slab and column. Due to the brittleness of the failure, a punching shear failure of one slab-column connection can lead to the progressive collapse of an entire structure, which makes proper punching shear design essential.
Past research has led to the development of mechanical models and design provisions which can be used to estimate punching shear capacity. However, most of this past research has focused on square or circular slab-column connections, even though alternate support geometries are common. For example, elongated connections, such as rectangular slab-column or slab-wall connections, are commonly used due to lateral force considerations. Additionally, design provisions for special-shaped slab-column connections, such as L-, T-, and cruciform connections, have been included in major worldwide design codes for over 40 years. The existing research database for alternate column geometries is limited, particularly in the case of special-shaped slab-column connections, and thus, there is a need for a comprehensive study of the influence of support geometry on punching shear to allow code committees and practicing engineers to properly understand the behaviour of elongated and special-shaped connections.
In this thesis, a combination of linear elastic finite element analysis (FEA) and calibrated nonlinear FEA (NLFEA), implemented in Abaqus 6.13, was used to study the influence of support geometry on the punching shear behaviour of interior elongated and special-shaped connections without shear reinforcement. In addition to the support geometry, the influence of monotonic uniaxial unbalanced moments, and the presence of slab openings near the connections were also studied, as past research has demonstrated that both of these factors can have a significant negative influence on punching shear behaviour. Due to the overall size of the study, FEA was used in place of experiments due to the large cost and time required to conduct large-scale testing, though past experimental results from literature were used to calibrate the nonlinear finite element models (FEMs).
In the slab-wall study, the punching shear behaviour of interior slab-wall connections (SWCs) without shear reinforcement was studied using a nonlinear FEM implemented in Abaqus/Explicit. The nonlinear FEM used in the slab-wall study, which was based on the Concrete Damaged Plasticity model (as with all calibrated nonlinear FEMs used in this work), was an extension of a FEM previously developed by the author for an analysis of the punching shear behaviour of interior rectangular slab-column connections (R-SCCs) without shear reinforcement [1]. In this study, the influence of loading conditions (various methods of concentric gravity loading and combined concentric gravity loading and uniaxial monotonic unbalanced moment), wall size (thickness and length) and presence of symmetric slab opening layouts (nonsymmetric opening layouts were not considered due to limitations in available computational resources) on the punching shear behaviour of slab-wall connections were studied. While the FEA predicted that punching shear failures of slab-wall connections could occur before one-way shear failures, the concentric punching shear capacities of the connections, which increased as the wall size increased, were typically higher than the expected load demand in the prototype structure. The unbalanced moment capacities of slab-wall connections were also found to be higher than typical unbalanced moment demands. The punching shear behaviour of interior slab-wall and rectangular slab-column connections with and without slab openings was also compared and found to be similar, though differences do exist and are discussed.
The slab-wall study and the author’s previous analysis of rectangular slab-column connections [1] demonstrated that the ACI 318-19 [2] punching shear design provisions were inadequate for elongated connections. Therefore, alternate punching shear design methods were developed. These methods were developed based on the FEA results and verified through comparisons to available experimental results where possible. For interior square and rectangular slab-column connections, three alternative methods, which are intended to replace the existing ACI 318-19 provisions, were proposed. For interior slab-wall connections, which are defined as connections where the maximum support dimension exceeds 14d (where d is the effective slab depth), a simplified design method, in which the slab-wall connection is decomposed into a combination of two-way and one-way shear regions was proposed.
The study of the punching shear behaviour of interior special-shaped slab-column connections (SS-SCCs) is divided into two parts.
In the first part, the concentric punching shear behaviour of special-shaped slab-column connections was analyzed using a nonlinear FEM in Abaqus/Explicit, which was calibrated to experimental results of one square and eight rectangular slab-column connections tested by Hawkins et al. [3]. The influence of column shape (L-, T- or cruciform), column size and influence of slab openings at various locations around the column on punching capacity was studied. Based on the FEA results, the shear force distribution around special-shaped slab-column connections was found to be non-uniform, particularly for L and T connections, where a significant portion of the load is transferred along the outer edges of the L and base of the T, respectively. However, regardless of column shape, slab openings near special-shaped slab-column connections should be located near the diagonal portion of the critical perimeter assumed in ACI 318-19 [2] or in the region between the column flanges to minimize their influence on punching shear capacity.
In the second part of the special-shaped connection study, the punching shear behaviour of special-shaped slab-column connections subjected to concentric shear and monotonically increasing uniaxial unbalanced moment was analyzed. Since the Hawkins et al. [3] specimens only had flexural reinforcement on one side of the slab, a FEM calibrated to the experimental results of Drakatos et al. [4] (which were tested under combined concentric shear and monotonic or cyclic uniaxial unbalanced moment) was used. Due to limitations in available computational resources, only one column size, and one opening layout, which involved a single square opening located between the column flanges, were considered per column shape (L-, T- and cruciform-shapes). The FEA results demonstrated that slab openings located between the column flanges typically have a minor influence on punching capacity when the moment is applied away from the opening. However, when the moment is applied towards the opening, the openings reduce the punching capacity of the analyzed special-shaped slab-column connections by approximately 10%. Therefore, openings between the column flanges do have some influence on the punching shear capacity of special-shaped slab-column connections subjected to combined unbalanced moment and concentric loading (especially because the moment direction cannot be controlled in most, if not all, practical applications.
It should be noted that in both parts of the special-shaped connection study that rectangular slab-column connections were also analyzed. These additional analyses further expanded the database for elongated connections and helped quantify the benefits of using special-shaped connections in place of rectangular connections. Furthermore, the accuracy of the ACI 318-19 [2] and ACI 421.1R-20 [5] punching shear provisions, for predicting both punching capacity and the influence of slab openings, was evaluated through comparisons to the FEA results in both parts. Concentric punching capacities according to the ACI 421.1R-20 method were typically conservative compared to those from the FEA, though the method is time-consuming and not suitable for iterative design. Therefore, simplified modifications to the ACI 318-19 concentric provisions were proposed for the initial design of L and T slab-column connections (no modifications were required for cruciform connections). For specimens subjected to combined unbalanced moment and concentric loading, the ACI 421.1R-20 provisions result in very conservative capacity estimates, particularly for connections with slab openings. Additionally, the ACI 421.1R-20 predictions are very inconsistent compared to the NLFEA predictions, which suggests that the ACI 421 method is not accurately capturing the behaviour of slab-column connections subjected to combined unbalanced moment and concentric loading. With respect to the influence of openings, the existing provisions were found to typically overpredict the influence of openings located between the column flanges. Furthermore, the ACI 421.1R-20 provisions were found to not accurately predict the influence of moment direction on punching capacity. Unlike the FEM, ACI 421.1R-20 predicted higher unbalanced moment capacities when the moment was applied toward the opening. | en |
