Progressive Collapse Resistance of Post-and-Beam Mass Timber Buildings: Experimental and Numerical Investigations on 2D and 3D Substructures

Abstract

Mid-rise to tall timber buildings are internationally gaining popularity. Reaching heights greater than 5 to 6 storeys has been made possible by the availability of engineered wood products, such as Laminated Veneer Lumber (LVL), Glued laminated timber (Glulam) and Cross Laminated Timber (CLT). These buildings are referred to as “mass timber buildings”. As the height of timber buildings increases, so do their potential risks of progressive collapse. Progressive collapse is characterised by a local failure of a load-bearing structural element which may propagate through the whole building, and ultimately causes its partial or entire collapse. While progressive collapse mechanisms of reinforced concrete and steel buildings have been widely researched, limited studies have been carried out on mass timber buildings. Their ability to resist progressive collapse and their load transfer mechanisms after the loss of a load-bearing element are currently unclear. First, to gain an initial understanding of the progressive collapse behaviour of post-and-beam mass timber buildings, a series of scaled-down 1×2-bay (2D) timber frame substructures were tested under a middle column removal scenario. The behaviour of the frames and the ability of three types of commercially used beam-to-column connections and a proposed novel connection, to develop catenary action under large deformations was measured. The system capacity in terms of the Uniformly Distributed Pressure (UDP) was also quantified. The test results showed that only the proposed novel connector was able to sustain the design pressure in international design specifications if no dynamic increase factor was considered, and therefore presented a potential solution to improve the robustness of post-and-beam mass timber buildings. Furthermore, progressive collapse of post-and-beam mass timber buildings cannot be resisted by the frame alone using the investigated currently used connections and alternative load paths must be found. Second to further explore the mechanisms of post-and-beam mass timber buildings against progressive collapse, four scaled-down 2×2-bay (3D) substructures, with CLT panels, were constructed and tested in the laboratory. Three substructures were tested under an edge column removal scenario, with substructures manufactured from two different types of beam-to-column connections. Namely, two tests were performed with a connection type commonly used in Australia, and one test with the proposed novel connection investigated earlier. The last substructure was subjected to two different corner column removal scenarios, with the substructure tested twice under different CLT panels configurations. The substructure was assembled from the commonly used in Australia beam-to-column connection. In all tests, two Uniformly Distributed Pressures (UDP) were applied to the floors in two stages: (i) a constant UDP of 4.8 kPa was first applied to the bays not adjacent to the removed column and (ii) an idealised UDP was then increasingly applied to the remaining bay(s) through a hydraulic jack connected to a six-point loading tree. The load redistribution mechanisms (alternative load paths), the structural response and failure modes were recorded. In general, experimental test results showed that the applied load was principally transferred to the three columns the closest to the removed column and that the CLT panels spanning over two bays were efficient in resisting the load. The layout of the CLT panels plays a critical role in resisting progressive collapse. A simplified analytical model, consistent with the current industry design practice and pre-defined alternative load paths, was used to predict the ultimate resistance capacity of the tested specimens and compared to the experimental capacities. Overall, the simplified methodology was found to be conservative. Third, finite element (FE) models were developed using the component model and validated against the 2D and 3D experimental results. The properties of the springs to be used in the component model were obtained from additional experimental component tests. CLT panels were simulated using layered shell elements while beam elements were used for the beams and columns. In the 2D numerical model, the ultimate load was accurately predicted and the development of compressive arch and catenary actions were well reproduced. The validated 2D model was then used to build the 3D model. For all tests, the 3D numerical models accurately predicted the overall load-displacement responses, load redistribution mechanisms, failure modes and strain developments in the beams and CLT panels. The validated numerical models were used to conduct a series of parametric studies to further examine the structural responses of the post-and-beam mass timber buildings. The results indicated that the structural capacity would be reduced when only using onebay long CLT panels, compared to using either staggered or all two-bay long CLT panels. Also, beam-to-column connections of the frames connected to the removed column could locally support the CLT panels above, providing an additional alternative load path for the structure in the context of progressive collapse, which is normally neglected in industry design practice.