Vibration performance of hybrid steel-CLT floors

Sammanfattning: In the light of today’s effort to achieve sustainable future of the planet, timber as building material makes a comeback on the construction market. Since the requirements on the buildings and the internal comfort increase, there is a need for finding new solutions and products; one of them is cross-laminated timber (CLT), which has the potential to be used for high-rise buildings due to its mechanical properties. The aim of this work was to study the vibration performance of CLT floors as it is often the governing factor in design of CLT structures unlike for other common building materials. The orthotropic mechanical properties of CLT were determined by the shear analogy method and verified with a finite element (FE) model of a simply supported beam compared to hand calculations of shear forces, bending moments and deflections. The properties based on Timoshenko’s approach were evaluated as less precise regarding the deflection. The non-composite structural behaviour of a steel-CLT hybrid floor structure was predicted for FE dynamic analysis based on a comparison between modelling exercise and hand calculations. Two different methods, the Concrete Society (SC) and Steel Construction Institution (SCI) methods, both seemed to be applicable for determination of the response factor first since the mechanical properties are not used as input in the calculations. These two methods differ in certain aspects, and based on FE analysis of simply supported slab even the resulting response factor for the CLT differs significantly. Moreover, the hand calculation results were similar to those of the FE analysis for the CS method, but in less agreement for the SCI method. Nevertheless, it is not recommended to reject the latter method based on this study and further studies should be performed on real structures with response factor known from on-site measurements. A part of the first floor of Canary Wharf College was modelled and analysed, and previous measurements of the frequency and response factors enabled a validation of some assumptions. The SCI approach showed to be inadequate for this type of structure and therefore only the CS method was applied further. Analysis of the floor structures supported by walls demonstrated similar results from both the measurements and the dynamic analysis. However, if the floor slab was supported by beams, the response factor was significantly overestimated although on the conservative side. This difference suggests that the modelling of such conditions are not satisfactory. The CS method appears to assess correctly the magnitude of the response factor for CLT floors supported by walls but overestimates it in case of beam supports. The first finding shall be confirmed through analysis of other structures and a more extensive research should focus on the latter one to determine more exact behaviour of the model under different conditions.