Session 4 was chaired by R. G. de Jong. The session began with an invited paper by U. J. Kurze, who described present-day techniques for measuring, predicting and controlling the environmental impact of railway noise and vibrations. In his discussion of air-borne sound, Kurze mentioned that measurements are being made with single omnidirectional microphones as well as with directional arrays of microphones, and showed results for some of these measurements. He also compared several different analytical methods for predicting air-borne sound levels. After describing the available techniques for measuring structure-borne noise and the corresponding prediction schemes, he concluded that the latter are not as advanced as the analogous techniques for air-borne sound. Kurze briefly mentioned an analytical model, comprising a combination of dipole and monopole sound sources for computing time-histories of the sound pressure level generated during the passage of a train. In answer to a question asked by King, Kurze said that, for the front of the train, the model is composed mostly of dipole sources, whereas for the rear, dipoles form 80% of the sound sources with the rest being monopoles.
The first contributed paper, given by G. Janssen, outlined an acoustically based geographical information system being developed by the Dutch Railways. By using such input data as the predicted number of trains, the train types and speeds, the available sound-abatement measures on trains and along the track and the applicable noise regulations, the information system quantifies the acoustical consequences in terms of either the costs for appropriate sound-abatement measures or the size of the impacted areas or the number of people living within these regions. In response to Hanson's question as to whether the digital mappings of the railway lines, the topography, the population density and other information are commercially available, Janssen said that they are.
The next paper was given by H. Kuppelwieser, who described the three-part VIBRA computer program designed by the Swiss Federal Railways to predict vibrations and structure-borne sound in the vicinity of new or extended railway lines. In this talk, the sub-programs were described in detail: VIBRA-1 is the more simple model and VIBRA-2 is a more complex version. VIBRA-3 contains the database for the first two calculation models. According to Kuppelwieser, the VIBRA program provides a cost-effective and relatively simple method for obtaining predictions.
The next speaker, H. J. A. van Leeuwen, discussed a comparative study of 14 analytic models designed to predict the acoustical performance of noise barriers along railway lines. To a large extent, input parameters required by the various models are different. Perhaps not unexpectedly, the insertion loss of a barrier predicted by the various models differs from one another, in some cases quite considerably. In addition, predictions were made for absorbent and reflective barriers and the results were compared