Source: Lisa Kerle & Monika Steinbäck, TU Graz 2021

Noise is a sound that an individual perceives as disturbing. In 2019, a third of Austrians stated that they were disturbed by noise at home, with traffic being a significant source of noise at 48.5 %. Across Europe, around 140 million people are exposed to harmful noise, with road (113 million), rail (22 million) and air traffic (4 million) being the primary sources of noise. Around 1 million people are affected by industrial noise. In addition, high noise pollution also threatens wildlife.

The professorship Acoustics and Environmental Noise is an endowed professorship and receives five years of funding from the Austrian Research Promotion Agency (Österreichische Forschungsförderungsgesellschaft mbH) and our industry partners. Together with ÖBB-Infrastruktur AG, Linz AG Linien, Wiener Linien GmbH & Co KG, AVL List GmbH, ASFINAG, KTM F&E GmbH, and The Austrian Airports Association, we focus on transportation noise and its effect on humans. In the long term, we open up the application field of urban noise, particularly noise in residential areas, with the goal of a quiet city.

Our research starts with theoretical principles of noise generation at the source, radiation, and propagation, considering topography and atmosphere, such as meteorology, shielding, and ground attenuation. We develop fundamental physical knowledge to understand better how technical sound sources can be mastered from source to perception. This necessitates collaborations with other disciplines, from electrical engineering, mechanical engineering, civil engineering, and medicine. We apply our knowledge to design sound sources so that only desired sounds will be radiated while unwanted noise is mitigated. Our main competencies are

• Modeling and simulation of sound sources, sound radiation and propagation
• Smart and reliable measurement data acquisition, transmission, processing, and analysis
• Physical and psychoacoustic assessment of noise emission and immission
• Technical and economic assessment of noise protection measures and research and development of new approaches to noise protection

Team

Group lead

Christian Adams
 

Team

Julia Donnerer

Florian Kraxberger

Projects

Efficient Computation of Acoustic Wave Propagation in Large Domains by the Discontinuous Galerkin Method

Conventional methods for outdoor noise propagation commonly rely on geometrical acoustic methods such as ray tracing or mirror sources. While these methods generate results quickly compared to other  numerical methods, they lack physical plausibility. E.g., diffraction effects are not resolved, which increases the error with  increasing wavelength. This means that the results are becoming invalid for  low-frequency noise. On the other hand, transportation technology such as road and rail vehicles often radiate low-frequency noise. Hence, the need arises to develop physically plausible and computationally more efficient methods for outdoor noise propagation . We use a wave-based method to overcome these shortcomings of conventional outdoor noise propagation methods in the low-frequency regime. Our approach relies on the higher-order Discontinuous Galerkin (DG) method for the acoustic conservation equations. A matrix-free implementation of the software framework developed at IGTE utilizes large-scale computational resources, such as supercomputers, in an optimal manner by providing a high degree of parallelization. In this project, the workflow will be applied to large geometries, such that physically valid noise propagation results can be achieved also in the low-frequency regime.

Numerically computed noise levels in a valley for a frequency of f=20Hz. The computational results have been achieved with the software openCFS co-developed at IGTE

Similitude of vibrating plates with damping
Structure-borne and airborne sound can be measured on a scaled model of an original mechanical structure and transferred to the original using scaling laws. Experimental effort can thus be reduced. In addition, as much information as possible can be obtained from one experiment, e.g., by transferring the results of the model experiment to several size levels of a structure. Principles of similitude theory are therefore assumed. This allows vibroacoustic quantities (natural frequencies, vibration velocities, sound pressures, etc.) to be adequately transferred from the scaled model to the original (or vice versa). Previous work has assumed that damping characteristics of the model and original structure are identical. However, damping properties of model and original can differ significantly in a statistical manner. This reduces the prediction accuracy of model tests; their practical use is limited to the few cases in which equal damping between model and original can be ensured. A changing damping in scaled model tests is caused by the facts that (1) the boundary conditions of the original structure can only be represented approximately in the model and (2) variations in the material properties scatter among model and original. This leads to distorted similitude conditions. In order to clarify the exact relationships, existing similitude methods first need to be further developed in the project so that damping properties of vibrating mechanical structures can be scaled. A higher order similitude theory (finite similitude theory) has been proven to overcome current limitations of existing scaling methods on geometrically distorted plates and beams. This approach will be extended in this project to perform model tests on vibrating structures, where the damping of original and model are different. Finally, with this knowledge, the prediction accuracy of vibroacoustic model tests can be improved by transferring the damping properties as well as the vibration properties from model to original. Thus, the project will push the applicability of model tests in vibroacoustics significantly beyond current possibilities. More information