Prediction of Noise Transmission in Lightweight Building Structures

Lightweight wood frame buildings are becoming popular due to low production price and a reputation of being friendly to the environment. Modern production techniques allow for factory assembly of panel modules—like walls and floors—or even complete rooms preassembled as volume elements. The high degree of off-site assembly allows for a high quality production and helps minimize both production costs and the risk of unforeseen events that could otherwise lead to costly delays.

The standard EN 12354, describing a simplified statistical energy analysis (SEA) subsystem approach, provides a valuable tool to predict the flanking transmission of air-borne and structure borne sound already at the design stage. However, lightweight building structures typically do not meet the requirements for ideal SEA subsystems and, therefore, applying the EN 12354 standard to lightweight building structures may result in imprecise predictions. Furthermore, lightweight buildings have low-frequency problems due to the low mass of the structure and often unpredictability (variation) between the building acoustic performance of supposedly identical dwellings can be observed. Therefore, better prediction methods are essential to the development of better designs and thus to the future of lightweight buildings. In order to improve both structures and prediction models, the sources—and magnitude—of variation must be determined. In the pursuit of understanding of the nature of this variation, simplified structures and junctions need to be considered, as the complexity of complete buildings can lead to conclusions which are possibly not true for similar buildings due to the non-deterministic nature of lightweight buildings.

The present doctoral thesis covers various topics related to the issues mentioned above. First, a paper about the modal density of ribbed plates at mid/high frequencies is presented. The modal density in such plates is not a uniform distribution, but instead it undergoes an undulating behavior with corresponding pass bands and stop bands. It is demonstrated, how the modes can be divided in two groups, where one group shows pass band/stop band behavior, while the other has a nearly uniform distribution of modes. The suggested approach for SEA adaptation is to consider a ribbed plate as two SEA subsystems: One that contains modes related to waves traveling in the direction orthogonal to the ribs, while the other subsystem contains modes related to waves traveling parallel to the rib stiffeners. The investigations utilize Monte Carlo simulations to examine the behavior of nominally identical plates.

Next, two papers that utilize the Finite-Element Method, focusing on aspects of modeling ribbed wall panels in the low-frequency range, are discussed. One paper investigates the effect of including an acoustic medium (air) inside the bays of a single-stud double-plate wall panel, while the other paper considers different approaches for modeling plate/frame couplings in FE models, which utilize full three-dimensional, solid continuum finite-elements. When using structural elements such as beams and shells, couplings may be modeled as either line or point coupling. However, when utilizing full three-dimensional, solid finite-elements, the scenario is not that simple. The investigations of both papers are carried out as parametric studies in the commercial FE package ABAQUS.

Finally, an experimental part, that focuses on the uncertainty and variation in wooden junctions, is included. Ten nominally identical plate/beam T-junctions are tested using experimental modal analysis, and the results are discussed. Furthermore, a numerical study of the robustness of two higher-order modal parameter estimation methods is conducted in order to assess the applicability of modal analysis on such junctions. The results demonstrate significant variation in the estimated modal damping at modes with strong coupling between different parts of the structure.