Search

This thesis studies the behaviour of hybrid systems consisting of tubular steel columns and laminated glulam timber beams. The research includes experimental investigations at the material and structural system levels as well as several numerical and assessments. Firstly, an extensive experimental programme is conducted on the determination of the material properties of the glulam. This involve compressive, tensile, shear and bending material tests. In addition, one-dowel connection tests are carried out to examine the interaction between the steel fastener and the timber volume around the fastener hole with the aid of Digital Image Correlation (DIC) techniques. Subsequently, detailed three-dimensional detailed models are constructed and their results are compared with the experimental tests. To account for the material damage, the foundation zone approach was used. This model was shown to be able to correctly simulate the crushing response of wood in the embedding region. The implementation of this region into the model allows an accurate simulation of the damage accumulation process. To this end, new relationships were formulated as part of this thesis, that can be used to estimate the material characterisation as a function of the crushing volume. The proposed model allows for a relatively low dependence on the radius of the foundation volume adopted making it applicable to a wider range of varying geometrical configurations. The validity and accuracy of the proposed modified foundation models were examined against the experimental force-displacement curves, and good agreement was found between the experimental response and the numerical simulations The second experimental programme is concerned with the performance of timber beam-steel column assemblages. Two timber beam-to-steel column alternatives are examined: a) top and seat angle connection and b) slotted-in T-stub connection with bolts. The configuration of the connections and their set-up are presented, followed by detailed results in the form of figures and tables as well as observations from the tests. The main behavioural patterns are identified and key response characteristics such as stiffness, capacity and failure mechanisms are discussed. Detailed finite element models were created to simulate the experimental timber beam-to-steel column connection tests. The models included advance features such as contact phenomena, bolt pretension and orthotropic material definitions. The foundation zone approach developed in this thesis was implemented. The results of the models were validated against the experimental results and good agreement was found. Additionally, component-based models were also formulated for the prediction of the flexural response of the connection types tested. These expressions were developed to estimate the stiffness, capacity and ultimate loads of the connections and the results were validated against the experimental and numerical findings. The component models proposed were able to predict the response of the connection accurately including those cases where screws and bottom-wedge angles were involved. In the concluding part of the thesis, the developed finite element models are employed in a parametric assessment in order to highlight the influence of key geometric and material considerations. The dearth of information on timber-steel hybrid connection of this type, design methodologies must be developed to offer a preliminary assessment on the prediction and evaluation of the key characteristics and hence the results are also compared with simplified analytical expressions. Finally, the last section of the thesis summarised all the findings and numerous possible future research are identified.