Structural fire performance of bolted glulam beam-to-column concealed connections

Abstract

Recent increased interest in the application of engineered-wood products, such as glued-laminated (glulam) timber and cross-laminated timber (CLT), in the Canadian building construction market has prompted amendments to be made to the building codes of several provinces. Until recently, the National Building Code of Canada (NBCC) restricted the use of wood as the primary building structural systems’ material by limiting its application to buildings up to only four storeys. The current version of the National Building Code of Canada, NBCC 2015, increased the height restriction to six storeys. The experimental research study detailed in this thesis has been carried out to investigate the behaviour of concealed steel-glulam bolted connections subjected to monotonic loading at both ambient and elevated temperatures. At ambient temperature, sixteen full-size test assemblies were examined, representing a total of thirty-two tested beam-to-column glulam bolted connections. The connections in eight of the test assemblies were strengthened perpendicular to the wood grain with long self-tapping screws (STS). For both unstrengthened and strengthened connections, increasing the number of bolt rows, each row having two bolts, from two to three rows increased the connection’s moment capacity with increments more than that obtained by increasing the bolt’s end distance from four to five-times bolt diameters. STS-strengthened connections failed in a relatively ductile manner in contrast to the unstrengthened connections. Increasing the number of bolts to six bolts in three rows in strengthened connections reduced the occurrence of brittle failure modes compared to connections with four bolts in two rows. Critically, STS-strengthened connections experienced increased moment-resisting capacity between 1.3 and 2.4 folds. The results of the ambient tests were used to verify the calculated moment capacities of the test assemblies in order to determine the full-service design load of the connection exposed to CAN/ULC-S101 standard fire in the large-size fire testing furnace accommodated at Lakehead University’s Fire Testing and Research Laboratory (LUFTRL). Outcomes of the fire resistance tests revealed that increasing the number of bolt rows from two to three, each of two bolts, increased the glulam beam-to-column connection’s fire resistance time by greater increments than those achieved by increasing the bolt’s end distance from four to five-times bolt diameter. Strengthened connections were found to have an increased fire resistance time compared to unstrengthened connections. Also, the STS strengthening increased the fire resistance time of the glulam connections by greater increments than those obtained by increasing the number of bolt rows and the bolt’s end distance.