Full-scale tests of two-storey precast reinforced concrete shear walls: Investigation of strength and deformation capacity

Precast concrete structures behave differently than equivalent in-situ cast solutions mainly because of the narrow connections between the precast elements. In this context, the keyed shear connections are of particular interest, as they have to ensure structural integrity in order to transfer shear- and normal stresses between adjacent elements. Extensive experimental studies (based on push-off tests) of this type of connection have been conducted in the past. These tests typically showed significant softening in the post-peak regime of the shear-slip relationship. It is for this reason important to study how effectively such connections can be utilised on the level of structural systems, where redistribution of internal forces is often required to reach the global load carrying capacity. This paper presents results from full-scale testing of two precast shear wall structures. The structures were two stories high and simulated a typical segment of precast buildings in practice. Each test structure was composed of 12 precast elements (decks and walls) and tested to failure under a combination of vertical and horizontal loads. The main varying parameter between the two tests is the design shear capacity of the keyed connections, which was doubled from test T2 to test T3 by increasing the diameter and the yield strength of the U-bar loops. Both structures displayed a ductile failure with significant deformation capacity. However, despite the higher strength in the keyed connections, the load carrying capacity of test T3 was not increased compared to test T2. This behaviour can be explained by a comprehensive analysis of the measured deformation field obtained from 2D Digital Image Correlation. By comparing the deformations with information obtained from independent component push-off tests with equivalent connections, it is possible to pinpoint the different stress regimes that the critical connection experienced, when the test structure reached its load carrying capacity. The results show that at that point in time, large parts of the critical connection were still in the pre-peak shear-slip regime, while another part located between two window openings was already at the end of the post-peak regime with practically no residual strength left. The findings indicate that the effective design strength of connections should be chosen with caution when modern numerical rigid-plastic methods are used to calculate the load carrying capacity of precast structural s