No abstract

Multi-storey timber buildings up to 6 and more floors are increasingly built in many European countries. The challenge with these buildings can be that with traditional intermediate floor constructions in timber it can be difficult to fulfill the standard requirements and even when they are met, low frequency transmission can still cause complaints. Additionally it is difficult to develop appropriate light weight floor constructions since it is well known that the correlation between the standardized evaluation methods using the tapping machine and the human perception of impact noise can be poor, especially in buildings with light weight structures. In the AcuWood project, measurements and recordings on different intermediate timber floor constructions in the laboratory and the field were performed covering a wide range of modern intermediate timber floor constructions. Additionally, one intermediate concrete floor with different floor coverings was included in the study. Besides the standardized tapping machine, the modified tapping machine and the Japanese rubber ball and "real" sources were employed. Subjective ratings from listening tests were correlated to many technical single number descriptors including the standardized descriptors and non-standardized proposals. It was found that the Japanese rubber ball represents walking noise in its characteristics and spectrum best, taking into account the practical requirement of a strong enough excitation for building measurements. The standardized tapping machine, with an appropriate single number descriptor, L′ nT,w + C I,50-2500 or slightly better, L′ nT,w Hagberg 03 , leads also to an acceptably high determination coefficient between the descriptor and the subjective ratings. Additionally, the study delivered data, from which proposals for requirements for the suggested single number ratings are deduced, based on the subjective ratings.

Analyzing data from an earlier investigation revealed uncertainties concerning data evaluation. Therefore, data points comprising 1/3-octave impact sound data and interview data from the earlier investigation are partly replaced by new data points. The data removed pertain to horizontal measurements. The investigation is then completed with data for 10 new housing units pertaining to vertical impact sound measurements. Including the new floor structures, a total number of 22 vertical data points are included. It was found that the current international standard method is unsatisfactory, and that the single-number value proposed in the earlier investigation suffers from shortcomings: when analyzing the new data sample, the correlation coefficient fell from 87% to 83%. The best choice proved to be a reference curve with a steep positive slope of 5.5 dB/octave between 50-100 Hz and the curve becomes a straight line. This reference curve resulted in a correlation coefficient again equal to 87%.

Wooden building systems are becoming more common. Still, there are a huge variety of floor assemblies in the market. The floor assemblies normally become the weakest part due to impact load from walking persons. So far, there are no reliable standardized calculation models available regarding prediction of impact sound in the entire frequency range. Therefore, the design is always based upon previous experiences and available measurements. For the development of prediction models, the first approach is to carry out a grouping of various available floor assemblies. From that, the aim is to trace similarities and carry out simplifications. Correlation is found between the single number ′ + L C nT,w I,50-2500 and the mass per unit area. It is also found that the ceiling system is useful in order to optimize the construction. The data will be further processed and used in the model development and to propose optimization of wooden floor assemblies.