Connections and Fatigue Behaviour of Precast Concrete Insulated Sandwich Panels
fatigue , BFRP , connections , shear connector , model , sandwich panel
This study investigates two aspects of precast concrete insulated sandwich wall panels, namely mechanical connections and fatigue behaviour. In the first part, flexural tests were performed on panels with various end support conditions, loading orientations, and reinforcement and shear connector materials. Bolted angle connections were used to simulate practical support conditions, while loads were applied in a manner to simulate windward pressure as well as suction. Panels with steel and basalt fibre-reinforced polymer (BFRP) longitudinal reinforcement were tested and compared. Discrete steel and BFRP shear connectors were also used and evaluated. The bolted angle connections provided partial end fixity, thereby increasing the overall strength and stiffness relative to identical panels simply supported by rollers during testing. In all cases the bolted connections succeeded in developing the full strength of the sandwich panels. Panels with steel reinforcement failed due to rupturing of flexural reinforcement, while a panel with BFRP reinforcement failed due to rupturing of shear connectors and crushing of concrete in one wythe. Panels loaded in the direction of wind pressure achieved higher peak loads than identical panels loaded to simulate suction. An analytical model accounting for material nonlinearity, end support conditions and partial composite action from the shear transfer system was developed. The model accurately predicted flexural stiffness, while the peak load was underestimated in most cases. In the second part of the study, seven fatigue tests were performed on four panels with either steel or BFRP flexural reinforcement and shear connectors. Cyclic bending was conducted at two loading amplitudes: a high (Pdef) and a low (Pstr) load, representing serviceability limits for deflection and stress, respectively; both considerably higher than the maximum national wind load. The effect of a moderate axial load, as in loadbearing walls, was examined. The panels initially had a Degree of Composite Action (DCA) of 76-84%. The axially-loaded steel-reinforced panel achieved 1M cycles under Pstr, then another 1M under Pdef. Its DCA reduced to 73 then 65%. Without axial load, 1M and 0.24M cycles were achieved under Pstr and Pdef, and DCA reduced to 69 and 22%, respectively. The BFRP-panel failed at 0.07M cycles at Pstr. Its DCA reduced from 76 to 69%. It was then axially loaded and retested successfully to 1M cycles. Stiffness degradations of 12-50% consistent with DCA reductions were observed.