Even though a building meets the regulations with respect to impact sound, structures with wooden frames may receive complaints from the inhabitants due to structure-borne sound in the lower frequencies. In this thesis measuring methods that are more precise than the traditional measuring methods have been used to see if they could give a better knowledge about flanking transmission in the low frequencies in timber-framed structures. The measuring results have been compared to FE models of a complete room. By comparing the non-traditional measurement results with the results from the FE model a straight forward comparison could be made.

Two field measurements have been made within the scope of this thesis. In the first field measurement a two-storey house with four apartments was studied. A traditional tapping machine was used as a source, and an accelerometer measured the responses in the apartment below. This study showed differences in acceleration levels on the surrounding surfaces in the room below the source. The non-bearing apartment-separating wall was the surface that had the largest acceleration levels. In the second field measurement several accelerometers were used simultaneously to measure the responses. With the help of experimental modal analysis and by measuring at several points at the same time the operational deflection shapes of all surrounding surfaces in the reception room could be captured. In this measurement a shaker was used as a source, so that the applied load could be measured the whole time. The load had constant level containing all frequencies within the studied frequency range. Simultaneously, a microphone was used and the acceleration levels in the surrounding surfaces could be compared to the sound pressure in the room. The measurements showed, not surprisingly, that at the frequency where the largest sound pressure was found all surrounding surfaces had acceleration peaks.

An FE analysis of the room from the second field measurement has been made. The analysis shows that by using a load similar to the load used in the experiments the velocity level of the responses are in the same range both in the analysis and in the experiments. It also shows that damping has a large influence and that the damping needs to be studied further to be able to use the model as a prediction tool for flanking transmission.

However, even though the results from the measurements and the analysis have some deviations, a great deal can be learned about the phenomenon of flanking transmission of timber-framed structures by using the proposed methods.