Framework for Optimization of Multi-source Railway High Speed Noise Models Through Hybrid Methods Combining Acoustic Simulations and Close Perimetric Noise Measurements

The prediction of future noise levels of railway lines prior to their construction has evolved with the years. This evolution has reflected the fact that aerodynamic noise sources at high speed are not only placed at low height, but also in the upper parts of the train (Gautier et al. in High speed trains external noise: a review of measurements and source models for the TGV case up to 360 km/h, Seoul, 2008 [1]; Noh in J Rail Rapid Transit 228:307–322, 2014 [2]). It is common practice in environmental noise modelling to represent a train by multiple sources with speed relationships (Marshall et al in Derivation of sound emission source terms for high speed trains running at speeds in excess of 300 km/h. Springer, Berlin, Heidelberg, pp. 497–504, 2015 [5]). Careful estimation of the emission laws of these sources is critical for estimating the noise impact in the surroundings, particularly in presence of barriers or where buildings overlook the railway. Building on the results and conclusions previously reported by the authors in (Sica et al. in Pass-by noise assessment of high-speed units by means of acoustic measurements in a perimeter close to the train, Ghent, 2019 [1]), this work presents a framework for optimizing a multi-source railway noise model by combining accurate simulation with test data obtained with microphones close to the perimeter of the railway. Acoustic models to determine the screening of noise sources by the train body and other complex acoustic phenomena (i.e. directivity, diffusion and diffraction) have been coupled to numerical optimization tools, in order to improve accuracy of noise estimations of the multi-source railway noise model.