Numerical Modelling of Timber Building Components to Prevent Disproportionate Collapse

Abstract: An increasing number of multi-storey buildings are constructed with components made of engineered wood products, such as glulam or cross-laminated timber (CLT). Our understanding of the mechanical performance of mass timber buildings is well developed for foreseeable exposures and loads, i.e. under the commonly specified limit states in building codes. However, our knowledge is fairly limited about the ability of multi-storey timber buildings to survive the consequences of unforeseeable exposures, e.g. accidents, natural catastrophes or terrorism.Tall buildings with many occupants are required to resist the effects of unexpected exposures, such that a disproportionate collapse can be prevented. There exist three lines of defence to decrease the probability of a disproportionate collapse; I) decrease the probability of the exposure itself, II) decrease the vulnerability of the structure and III) increase the structural robustness. From an engineering perspective, it is most effective to address line II and III, and therefore they are the primary focus of this thesis.Structural robustness guarantees the mobilisation of alternative load paths (ALPs) after an initial damage, i.e. the removal of a bearing component. For timber, the development of ALPs relies primarily on the behaviour of the connections between building components. The activation of ALPs in a confined building part is often accompanied by non-linear responses and considerably large displacements whose effects on the surrounding structure need to be accounted for, which makes physical tests expensive. To gain knowledge about ALPs requires both, testing and modelling, however, given the abundance of different removal scenarios and the high cost, the possibilities of physical tests remain bounded. Numerical models are thus required, both, to augment the test results, e.g. by parameter variations or changed boundary conditions, and to build an understanding of the underlying mechanisms affecting the ALPs. In this thesis, non-linear quasi-static finite element (FE) models have been developed to investigate and quantify the ALPs at a  component level in CLT buildings. Furthermore, simplified non-linear dynamic models were used to study the effects of parameter variations on the collapse behaviour of a building bay.The vulnerability of a structure depends on the ability of its individual components to withstand loads above their intended design loads. The design of overly strong components, i.e. key elements, requires accurate knowledge about their mechanical performance. Timber is a material with a large inherent uncertainty adhering to its mechanical properties. Automated strength grading of timber boards can narrow this uncertainty, however, even with the current technologies, the variation in the produced material remains relatively large. The advent of computed tomography (CT) scanning equipment in sawmills has provided a new, ample source of information which can be used to derive mechanical models of logs and boards. In this thesis, a new method is presented to generate FE models from CT scanned boards, in which the knots, pith and local fibre orientations around knots were reconstructed from 3D image data. The method provides the first step towards predicting the mechanical properties of yet unsawn boards already at the log stage using continuum mechanical instead of statistical models, and planning the specific use of single boards in a structure during sawing. The results indicate that models from CT scans can improve future strength grading strategies, which would facilitate the design of key elements.The main objective of this thesis is to provide numerical modelling approaches for an improved understanding of the behaviour of timber components in multi-storey constructions regarding disproportionate collapse. The goals are to i) investigate and quantify ALPs, to increase structural robustness, and ii) develop mechanical models from CT scans of timber, to improve strength grading and thus decrease the vulnerability of future timber buildings.