Experimental and Numerical Investigation of Emulative Connections in Precast Concrete Walls

Grouted connections are widely used in the precast concrete construction. For instance, in precast concrete walls, they are used to connect assemblies of vertically stacked panels. The connection is comprised of a grout cylinder bound by a corrugated metallic duct, which is used to house a large diameter reinforcing bar bridging the horizontal gap of stacked panels. The connection is used to provide vertical continuity to the assembly, and to help resist tensile demands from in-plane bending. Current design guidelines consider such connections through a bar-in-concrete treatment, disregarding its composite nature and the confinement effect of the duct. This has resulted in excessively long connections that could induce planes of reduced stiffness in precast wall panels. In this thesis, a research program was tailored to investigate the disparity between the real behaviour of grouted connections and their current design code idealization to offer alternative more realistic design provisions. The experimental program was divided into three phases. First, an exploratory study of the bond behaviour of grouted connections under monotonic loads was conducted. Second, the behaviour of grouted connections was compared to bar-in-concrete specimens under monotonic loading. Third, the cyclic behaviour of the connections at various embedment lengths was examined under quasi-static loading. Knowledge gained in the experimental program was used in analytical treatments to develop a novel model that can accurately depict the behaviour of these connections. Results from the various experimental phases reveal that the bond failures developed in grouted connections are not characterized by brittle tensile splitting modes, irrespective of the level of bond stress along the assembly at different embedded lengths. It was observed that the presence of the corrugated duct offers a continuous restraining field against radial expansion of the grout, causing the bars to be mobilized in much shorter anchored lengths than those suggested by current standards. A numerical model was developed to reproduce the behavior of grouted connections with reasonable accuracy. Its accuracy and computational efficiency should allow modelling full-scale precast wall assemblies. In this thesis, a research program was tailored to investigate the disparity between the real behaviour of grouted connections and their current design code idealization to offer alternative more realistic design provisions. The experimental program was divided into three phases. First, an exploratory study of the bond behaviour of grouted connections under monotonic loads was conducted. Second, the behaviour of grouted connections was compared to bar-in-concrete specimens under monotonic loading. Third, the cyclic behaviour of the connections at various embedment lengths was examined under quasi-static loading. Knowledge gained in the experimental program was used in analytical treatments to develop a novel model that can accurately depict the behaviour of these connections. Results from the various experimental phases reveal that the bond failures developed in grouted connections are not characterized by brittle tensile splitting modes, irrespective of the level of bond stress along the assembly at different embedded lengths. It was observed that the presence of the corrugated duct offers a continuous restraining field against radial expansion of the grout, causing the bars to be mobilized in much shorter anchored lengths than those suggested by current standards. A numerical model was developed to reproduce the behaviour of grouted connections with reasonable accuracy. Its accuracy and computational efficiency should allow modelling full-scale precast wall assemblies