An Application of Stochastic Simulation to the Study of the Variability of Road Induced Fatigue Loads

Abstract

To adapt their products to the durability requirements, car manufacturers must possess a suitable methodology to quantify the large variety of road induced fatigue loads that vehicles' components undergo during their life. The load data acquisition process is generally based on long and costly measurement campaigns. Here an alternative load characterization methodology is proposed. It is based on stochastic modelling and simulation rather than purely on a collection of load measurements. Stochastic models, in the form of random processes, are proposed for vehicle- independent influential factors, namely road roughness and vehicle's speed. Random realizations are generated and used as inputs for multi-body simulations, describing the dynamics of any given vehicle. An arbitrarily large set of fatigue loads can therefore be obtained inexpensively, for any given elementary life situation, and subsequently, for the total life of the vehicle, using customer usage information. A study of the influence of a change in market region is performed as an illustration of the stochastic simulation methodology. For this study, existing measurements carried out by Renault within the considered market regions, are used as a source of statistical information.