In recent years, there has been an increasing trend in Australia and New Zealand towards the use of long-span timber and timber-concrete composite (TCC) flooring systems for the construction of multi-storey timber buildings. The popularity of these flooring systems is because of their low cost, easy construction and the use of environmentally sustainable materials. Due to their light-weight, such long-span floors are however highly susceptible to vibrations induced by service loads. Although longspan timber and TCC flooring systems can easily be designed to resist the static loads using currently available design guidelines, it is crucial to also investigate the dynamic behaviour of these floors as the occupant discomfort due to excessive vibration may govern the design. Moreover, many structural failures are caused by dynamic interactions due to resonances, which highlight the importance of investigating the dynamic behaviour of flooring systems. To date, there are very limited design guidelines to address the vibration in long-span floors, especially composite floors, due to a lack of sufficient investigation.
In 2009, a research consortium named Structural Timber Innovation Company (STIC) was founded, with the aim to address various issues encountered with structural timber buildings including timber and TCC flooring systems. STIC is conducting Research and Development (R & D) work in a number of key areas to provide a new competitive edge for commercial and industrial structural timber buildings. The R & D work is undertaken with three parallel objectives at three universities, namely, the University of Technology Sydney (UTS), the University of Canterbury (UC) and the University of Auckland (UA). The focus of UTS is the assessment of various performance issues of long-span timber only and TCC flooring systems for multi-storey timber buildings. The work presented in this thesis deals with the investigation of the dynamic performance of timber only and TCC flooring systems, which is one of the sub-objectives of the research focus at UTS.
In particular, the presented research assesses the dynamic performance of long-span timber and TCC flooring systems using different experimental und numerical test structures. For the experimental investigations, experimental modal testing and analysis is executed to determine the modal parameters (natural frequencies, damping ratios and mode shapes) of various flooring systems. For the numerical investigations, finite element models are calibrated against experimental results, and are utilised for parametric studies for flooring systems of different sizes. Span tables are generated for both timber and TCC flooring systems that can be used in the design of long-span flooring systems to satisfy the serviceability fundamental frequency requirement of 8 Hz or above. For floors where vibration is deemed to be critical, the dynamic assessment using the 8 Hz frequency requirement alone may not be sufficient and additional dynamic criteria such as response factor, peak acceleration and unit load deflection need to be satisfied. To predict the fundamental frequency of various TCC beams and timber floor modules (beams), five different analytical models are utilised and investigated.
To predict the cross-sectional characteristics of TCC systems and to identify the effective flexural stiffness of partially composite beams, the “Gamma method” is utilised. Essential input parameters for the “Gamma method” are the shear connection properties (strength, serviceability stiffness and ultimate stiffness) that must be identified. Therefore, a number of experimental tests are carried out using small scale specimens to identify strength and serviceability characteristics of four different types of shear connection systems and three of them were adopted in the TCC beams. The connections included two types of mechanical fasteners (normal wood screw and SFS screw) and two types of notched connectors (bird-mouth and trapezoidal shape) with coach screw.
Traditionally, the composite action of a system is determined from static load testing using deflection measurements. However, static load testing is expensive, time consuming and difficult to perform on existing flooring systems. Therefore, two novel methods are developed in this thesis that determines the degree of composite action of timber composite flooring systems using only measurements from non-destructive dynamic testing. The core of both methods is the use of an existing mode-shape-based damage detection technique, namely, the Damage Index (DI) method to derive the loss of composite action indices (LCAIs) named as LCAI1 and LCAI2. The DI method utilises modal strain energies derived from mode shape measurements of a flooring system before and after failure of shear connectors. The proposed methods are tested and validated on a numerical and experimental timber composite beam structure consisting of two LVL components (flange and web). To create different degrees of composite action, the beam is tested with different numbers of shear connectors to simulate the failure of connection screws. The results acquired from the proposed dynamic-based method are calibrated to make them comparable to traditional static-based composite action results. It is shown that the two proposed methods can successfully be used for timber composite structures to determine the composite action using only mode shapes measurements from dynamic testing.
This PhD research provides a detailed procedure for designing and investigating the short term static behaviour of a proposed long span timber floor system for non-residential applications that meets serviceability and ultimate limit design criteria, with the use of timber as the only structural load bearing part of the system. In this study the behaviour of two types of LVL are investigated through a number of experimental and analytical tests. As a result of the tension and compression tests, a suitable constitutive law is developed which can accurately capture the stress-strain relationship and the failure behaviour of LVL, and it can also be incorporated into FE analysis of any LVL beam with similar structural features to the tested specimens. Further, the results of the full scale four point bending tests on LVL sections are used to identify the behaviour of LVL up to the failure point and to develop a finite element model to capture the behaviour and failure of LVL. Moreover, after investigating the long span timber floors, one system is proposed to be fabricated for the extensive experimental and numerical investigation. The results of the full scale experimental tests together with the numerical investigation provide a robust model for predicting the performance of any timber beams with similar structural features to the proposed system while the dimensions and spans can be varied according to special requirements such as dynamic performance or fire resistance requirements.
The objectives and scope of this study are to conduct long-term experimental test on timber-concrete composite beams, analyse the results to determine the creep coefficient of the composite system and compare the experimental results with the analytical solutions in accordance with Eurocode 5, in which the effective modulus method is used to account the effect of creep. To achieve the aforementioned objectives, a long-term laboratory investigation was started in August 2010 on four 5.8m span TCC beams with four different connector types. The specimens have been under sustained loads of 1.7kPa and subjected to a cyclic humidity conditions whilst the temperature remains quasi constant (22 °C). During the test, the mid-span deflection, moisture content of the timber beams and relative humidity of the air are continuously monitored. The long-term test is still continuing, two TCC beams were unloaded and tested to failure after 550 days, while the other two TCC beams are still being monitored and this report included experimental results up to the first 1400 days only. The long-term investigation on the two timber only composite floor beams commenced on March 2013 and the results are reported for the first 800 days from their commencement.