Comparison of Two Different Models Describing Railway Noise Generation and Radiation

Noise from railway lines causes human suffering as well as the need to erect high obtrusive noise barriers. Land too near the railways cannot be used for dwellings. The cost for society is high. Consequently, there is a need for tools enabling accurate noise predictions. To optimize the track, we need a prediction tool, a computer program, which tells how much noise a specific track radiates under certain circumstances. The most widely used prediction tool is TWINS; PRINT is a Korean equivalent, which is being used throughout this paper. Noise prediction results are compared to calculations with another, new, tool, RWN (Rail Wheel Noise). However, the two models yield different results, although the same input parameters (track properties) are used in the calculations. The reason is that PRINT is a frequency-domain model, and RWN a time -model. Performing the calculations in the time domain enables a more realistic description of the noise generation mechanism. The question to be answered is: are there situations when it is crucial to use a time-domain model in predicting noise and noise optimizing railway tracks? This paper tries to answer this question.