Vibroacoustic performance of wooden buildings : Prediction and Perception

Författare

Juan Negreira

Summary, in English

When the Swedish building regulations in 1994 allowed wooden multi-storey buildings to be built, this type of lightweight construction became popular due to its low cost and ease of construction, as well as wood being a plentiful resource in Sweden. However, complaints amid inhabitants are often reported due to nuisances caused by disturbing vibrations and noise propagating in the building. Still in 2016, no vibration limits are given in any international standard due the complexity involved. Certain guidelines and guide values have simply been suggested instead. The vibrational response of wooden buildings has therefore become an issue to be tackled during their design phase. Accordingly, the aims of the research presented in this thesis can be divided into two basic categories: (i) development of prediction tools for the verification of vibratory and acoustic performance before a building is constructed and (ii) the development of indicators of human exposure to floor vibrations.

As of today, there still exist no accurate and reliable methods for predicting the vibroacoustic performance of wooden buildings. Product development is carried out on an empirical basis, involving both observations and the experience of engineers. Time and costs could be reduced by addressing issues of vibration during the design phase, for instance by using numerical methods (e.g. finite element simulations) as prediction tools; since experiments on prototypes and existing buildings are both time consuming and expensive. Development of such accurate finite element prediction tools is the major objective of the research dealt with in this work. In line with this, finite element models of a prefabricated timber volume element based building were created in the investigations presented, and specifically the flanking transmission occurring was analysed. On the basis of the conclusions drawn in that study, other investigations aiming at improving the accuracy of numerical prediction tools were performed. Thus, the question of whether or not air and insulation in cavities of multi-storey wooden buildings affect the transmission of structural vibrations was investigated. The conclusions showed that acoustic media must be considered when predicting low frequency vibroacoustic behaviour of such buildings by use of numerical models. Likewise, a method for extracting the properties of elastomers, frequently used in timber buildings at the junctions as a vibration-reduction measure, was also developed in order to have reliable assessments of the material properties involved as input for the finite element models employed. This was done through performing analytical calculations, and carrying out finite element simulations and mechanical testing in a uni-axial testing machine. Moreover, modelling guidelines on how to model different types of beam-plate connections (both glued and unglued together with screws) when creating predictive models, are presented. The guidelines were drawn by comparing measurements to their calibrated numerical models. In addition, a time-efficient frequency domain method to implement the tapping machine into predictive models of impact sound insulation, is presented.

Regarding the perception part of the thesis, the investigations performed aimed at supplementing the lack of existing studies addressing human response to floor vibrations. In order to obtain a better estimate of an acceptable level of vibrations in dwellings, measurements on real floors while people walked on the them, as well as when they sat down while another person was walking, were performed. The accelerations, velocities and deflections they were exposed to during the test were measured. Indicators of human response to vibrations were extracted by determining relationships between people's answers to questionnaires about their perception and experience of the vibrations, and different parameters as determined by measurements. Several indicators were found to describe people's answers to questions both regarding vibration annoyance and vibration acceptability.

The measurements performed in the thesis were carried out using transducers developed and calibrated in-house within the frame of the project; all the construction processes, calibrations, accuracy and justification of their use to the applications dealt with here, are also thoroughly described in the thesis.

In summary, adequate knowledge of the vibrational performance of wooden buildings such as that obtained here by use of measurements and of finite element simulations is seen as paving the way for further development in this area. The conclusions drawn in the thesis will ultimately entail time and cost savings for the industry as well as help dwellers to feel more comfortable in their homes.

Avdelning/ar