Hybrid timber-concrete floor panel systems with a novel hollow core construction

Abstract

Hybrid timber-concrete (HTC) floor systems have gained wide acceptance within the structural engineering community as an alternative for reinforced concrete (RC) floor systems. They consist of a concrete layer and a timber layer acting predominantly in compression and tension, respectively. HTC floor systems are lightweight and with less embodied energy compared to RC floor systems. However, perfect composite action is difficult to achieve in HTC floor system resulting in part of the concrete layer acting in tension, thus reducing the material efficiency.In this thesis, an HTC floor system with a novel hollow core (called “HTCHC floor system” hereafter) is proposed. This novel floor system consists of a top concrete layer, bottom timber layer, and a fibre reinforced polymer (FRP)-timber hollow core sandwiched between them. Thickness of each layer is designed so that the bending neutral axis of the HTCHC floor panels is within the core, resulting in concrete and timber acting predominantly in compression and tension respectively. Use of the hollow core near the section neutral axis to replace the solid timber or concrete in the HTC section is the key innovation in this study, as it significantly reduces the self-weight and increases the efficiency in material utilization without sacrificing the structural performance. Two types of FRP-timber cores are designed and tested in this study: (a) a corrugated-shaped core, and (b) a waffle-shaped core. The fabrication method and flexural behaviour of the HTCHC floor panels with each type of core are presented and discussed.First a corrugated core system (called “HTCCC” hereafter) was proposed. Utilising the corrugated core helps to reduce the concrete volume by at least 40% compared with a conventional HTC section with the same total height and timber thickness. Corrugated core also provides spacing for building services to be run through the floor. Two core configurations were developed: with core orientates parallel to the span for maximum one-way spanning capacity and with core orientation transverse to the span for generation of additional transverse spanning capacity without compromising the longitudinal span capacity. In total, eight HTCCC floor panels were prepared and tested, with the flexural capacities and critical failure modes analysed for each. Effects of different core geometries, shear force transfer methods, and manifested composite action were also closely studied. Longitudinal specimens achieved the best composite action and correspondingly the highest panel performance, with a 73% ultimate moment carrying capacity and an 85% stiffness efficiency at serviceability limit state (SLS), compared to an idealised HTC section with full composite action. In the transverse pattern specimens, shear bolt reinforcement was found to be essential for maintaining the high panel performance. It was also shown that with adequate enhancement of interfacial shear transfer and proper geometry, transverse patterned panels could closely match the one-way spanning capacity of the longitudinal core HTCCC panels.Based on the findings of the HTCCC floor panels, a new waffle core system (called “HTCWC” hereafter) consisting of a rectangular shape core with waffle-grids was designed. This core design allowed a uniform concrete layer, which can eliminate the undesirable stress concentrations within concrete which were observed in the HTCCC panels, and reduced concrete volume by 67% compared with a conventional HTC section. Employing digital fabrication techniques and computer-numerical-controlled (CNC) machines, integral connections were used in this prototype for interlayer shear transfer. The highly accurate and efficient manufacture process gave excellent quality control of the integral connectors, which simplified the manufacturing process, and improved the composite action and the structural performance of the system. Four specimens consisting of two different types of interfacial connections were fabricated and tested under four-point bending. Test results showed that the HTCWC floor system achieved better interlayer behaviour and significantly higher load capacity than the HTCCC system. They had 80-86% ultimate moment capacity and 90-93% stiffness efficiency at SLS, compared to an idealised HTC section. They also had a 44% higher weight-specific load capacity than equivalent RC floor system, and 11%-29% higher capacity than the HTC floors.Finally, a simple numerical model capable of considering the slip between concrete and timber layers was proposed to predict the behaviour of HTCWC floor panels. In this model, a new two-layer element was proposed, with the concrete and tensile timber parts modelled as the two sublayers, and the core was modelled using interlayer springs. A stiffness matrix was proposed for this new element, and then nonlinearities of the core and sublayer materials were introduced by explicitly updating the stiffness matrix at each step. Results showed that this model provides very similar prediction as the finite element software ABAQUS for elastic composite beams while using only about 10% of the elements. It also showed an accurate prediction of the HTCWC specimens regarding the overall stiffness at the linear range, the nonlinear response, and the ultimate capacity.