Prediction and Optimisation of Traffic Noise. A Study on Noise Barriers and Shielded Areas

Doctoral thesis, 2003

The main contribution of noise levels in cities is today road traffic. The strongest source for all road vehicles is most often the tyre-road contact. This source position close to the ground is fortunate with respect to secondary noise abatement measures, such as absorbing road surfaces or noise barriers. The first part of this thesis deals with noise reduction possibilitiesusing optimised surface treatments on the ground and on noise barriers, with special attention to low-height barriers. The insertion loss can be significantly improved through optimisation. The amount of improvement depends on barrier geometry, the impedance of the surrounding ground and on the frequency. The largest improvements can be found at low frequencies, but siginificant improvements can also be found at 1 kHz, where the traffic noise spectrum is strongest. A specialized surface on the road and a 1 m high barrier gives 7 dB better insertion loss than a rigid barrier of equal geometry when weighted with a representative traffic noise spectrum. The second part of this thesis studies noise in cities, where a many sources distributed over a large area are responsible for the total noise level. The noise levels on the silent side is in focus, which due to shielding has lower noise levels than the directly exposed side. Two new methods which are suitable for silent side calculations over large areas are presented.The main contribution to the noise level on the silent side is found to propagate over the rooftops of surrounding buildings and not around them. The noise abatement possibilities in cities are then studied using these models with respect to two features: 1) An increased absorption, and 2) A modified vehicle flow distribution. Both features show a promising noise reduction potential. A reduction of 10 dB may be possible on the silent side through increasing absorption. A traffic concentration to few streets gives larger areas with low levels than an evenly distributed vehicle flow. The amount of reduction is however dependent on the street grid geometry and the existing vehicle flow.