Investigation of prediction methods for low height noise barrier implementation

Abstract

In areas where the more common tall noise screens produce an unreasonably large reduction on the quality of the urban environment, absorbing low height noise screens, in short LHNSs, can be introduced. Due to the screens relatively low height and the fact that they benefit from a placement as close to the source as possible, the visual impact on the environment is small. Furthermore studies have shown them to be cost effective and produce equal, or in some cases, higher attenuation than their more traditional counterparts. However, due to the scale and cost of projects regarding railway infrastructure there is small room for error. Because of this, the current uncertainties in the prediction methods are often enough to warrant the usage of other noise reducing devices even though the LHNS could be viable solution. To give the Swedish Transport Administration a solid basis on which to perform calculated estimates an investigation is launched. A literature study shows that to make good estimations of the sound field from a train screened with a LHNS, a combination of more traditional methods and numerical simulations are required. It also shows the importance of an accurate source model including the relative energy distribution between the sub-sources of a train. Based on the findings a model utilising the 2.5D boundary element method is implemented to simulate the pressure field with and without screen for different train shapes. Geometrical estimations of a X60 commuter train, an empty industrial wagon and a case without a wagon are studied. The low height noise screen was modelled after the dimensions of an S-block 250. The simulations show that a reasonably good fit to measurement data from other projects can be achieved in third-octave bands. However, in some cases, large deviations can be noted. Because of this further validations with a designated test setup is advised before drawing any definitive conclusions. Apart from the identified large deviations the results calculated here consistently show that a considerable reduction can be achieved by the implementation of LHNSs and that they warrant further consideration in future infrastructure projects.