Influence on the load‐displacement behaviour of steel‐to‐concrete connections with post‐installed adhesive anchors

Abstract

Designing frame structures requires knowing the behaviour of each member of the frame (beam, column, joint, etc.) regarding the axial forces, shear forces, and bending moments. Furthermore, the ductility of the structure under seismic loading is essential in earthquake regions. Nowadays, structural frames are built from different materials such as concrete and steel, to achieve better performance. Therefore, the behaviour of the connections between steel and concrete is essential. The steel‐to‐concrete joints were the focus of the INFASO project, where joint solutions with easy fabrication, quick assembly, applicability in old structures, sufficient ductility, and high loading capacity were developed. They proposed the use of anchor plates with welded studs or post‐installed fasteners such as adhesive anchors to connect the steel and concrete members. This paper focuses on the performance of post‐installed adhesive anchors. During their service life, post‐installed anchors are subjected to monotonic, constant, and seismic loading. Each of these loading approaches is described in the current standards. Seismic and constant loading tests are of importance to the long‐term behaviour of the anchors. Seismic loading tests are carried out using a predefined cycle pattern. According to TR049, seismic tests of category C1 (tension and shear) are performed with 140 load cycles, where the load amplitude decreases after 10, 30, and 100 cycles. On the contrary, C2 category tests increase the amplitude within 75 or 59 cycles, depending on the test. Constant loading tests apply when the anchors are installed in cracked concrete. In contrast to the mentioned patterns, this study observes the behaviour of the adhesive anchors when the amplitude of each cycle is increased by 5 % until the anchor fails. Standard short‐term pull‐out tests are carried out to determine the load increments. Various parameters such as the embedment depth, bond line thickness, hole cleaning, wet concrete, and elevated temperatures are studied. Confined and unconfined tests are performed. Overall, the reference short‐term failure loads are higher than those in the tests with incremental and cyclic loading. The anchors installed with reduced hole cleaning, in wet concrete or subjected to elevated temperatures have a lower failure load compared to the reference series. Two additional bond line thicknesses are used for comparison. The increase in the thickness influenced the failure load differently for confined and unconfined test setups.