We are surrounded by noise sources such as engines, household appliances, heat pumps and computers. A common treatment is to build an enclosure around the noise source that partly reduces the radiated sound. Typical examples are the engine bay of a car or the enclosure around a heat pump. The enclosures must have relatively large apertures to allow for cooling airflow. The room that is created by the enclosure has often a complicated geometry with numerous densely packed objects such as cables, pipes, containers, screens and sound absorbers. The enclosures of today are rudimentary and sub-optimal due to lack of fast yet accurate prediction tools for noise radiated through the complex geometry.
Classical finite element or boundary element methods are no realistic alternative in the mid- and high-frequency ranges, where computational effort is immense and the results are hypersensitive to small variations in geometry, boundary conditions, and environmental conditions. Therefore, methods based on spatial, frequency and ensemble averages of energetic variables have been developed. The goal of the project is to develop a numerical prediction tool that allows for optimisation of apertures, screens and absorbers in enclosures around noise sources in general.
The project is carried out at the Division of Applied Acoustics in the research group Vibroacoustics.
Keywords: high frequency methods, sound intensity, numerical modelling