Vibration and structure-borne sound from railways are complex issues that, currently, are not being investigated to the same extent as air-borne sound. For those of us working with these problems, it was surprising how much attention was given to vibration-related topics in this workshop on ''railway noise''. Characteristic of the situation today is an increase in the construction of new railway lines and the upgrading of old ones through urban areas. With the cramped space available, the lines get close to structures or they are put underground. This leads to more complaints about vibrations and structure-borne sound. Successful technologies to reduce or screen off the air-borne sound draws more attention to the remaining disturbances transmitted through the ground and structures. The handling of these problems today and in the future, are demanding tasks, requiring new methods and technologies.
In his invited presentation, Manfred Heckl (Technical University of Berlin) opened the session by giving a sound theoretical background of the various creation mechanisms, the transmission and the reception of structure/ground-borne sound and vibration. The theory was visualized by illustrative examples of measured data. He pointed out how hard it is to predict vibration transmission through the ground and into buildings, and the need for large amounts of empirical data to gain increased understanding and quantification of the uncertainties. The lack of efficient countermeasures, particularly for low frequency disturbance, was pointed out.
The four following papers presented tools for predicting vibrations and structure-borne sound. Two of them were theoretically based. The model of Hugh Hunt (University of Cambridge) was based on a systematic use of transfer and frequency response functions. Approaches to determine the model parameters still remain. Chris Jones (British Rail Research) presented a model based on a fruitful substructuring, where the creation was first treated as a detailed model of vehicle-rail-embankment system, with a crude model of the ground, and then coupled to a more complete model of the transmission medium. Comparison with measurements demonstrated the ability of the model to predict vibration at the sleeper, and further how problematic the prediction of the transmission through ground and buildings is. The two other papers presented empirical tools. The one by Christian Madshus (Norwegian Geotechnical Institute) focused on vibrations and the effect of soft ground. The prediction model covered inherently the uncertainties and variability in the underlying empirical data, by giving confidence levels for the predicted values. He also presented a database for empirical data and model parameter estimation. Michael Breslin (Ashdown Environmental Limited, U.K.) presented a model used for predicting vibrations and structure-borne sound in buildings above bored tunnels. The accuracy of the model prediction was demonstrated. He also addressed criterion values for structure-borne sound and perceptible vibration.