A model for the Statistical Distribution of Road Vehicle Vibrations

This paper presents the initial results of a study\ud aimed at improving the method by which the vibrations\ud produced by transport vehicles are characterised and\ud simulated. More specifically, this paper focuses on the rigid\ud body vibrations generated by road transport vehicles in the\ud context of distribution of packaged goods and produce. The\ud research uses a variety of vibration data, collected from various\ud vehicle types and routes in Spain and Australia with\ud high-capacity vibration recorders. Vehicles used range from\ud small transport vehicles to large truck-trailers with both\ud airbags and steel spring suspensions while the routes travelled\ud include suburban streets, main roads and motorways. The\ud paper discusses the significance and limitations of the average\ud power spectral density (PSD) and explains why the average\ud PSD is not always adequate as the sole descriptor of road\ud vehicle vibrations as the process generally tends to be\ud non-stationary and non-Gaussian. The paper adopts an\ud alternative analysis method, based on the statistical\ud distribution of the moving root-mean-square (RMS) vibrations,\ud as a supplementary indicator of overall ride quality. The\ud measured data was used to compute the statistical distribution\ud of each vibration record, the shape of which was compared for\ud the entire set of records. The suitability of various\ud mathematical models, based on the Weibull and Rayleigh\ud distributions were investigated for describing the probability\ud distribution function (PDF) of road vehicle vibration RMS time\ud history. The paper proposes a single mathematical model that\ud can accurately describe the statistical character of the random\ud vibrations generated by road vehicles in general. It shows that\ud the model can also effectively describe the statistical\ud parameters of the process namely the mean, median, standard\ud deviation, skewness and kurtosis