Load Modelling for Fatigue Assessment of Vehicles - a Statistical Approach

Doctoral thesis, 2007

A way of modelling loads in customer usage for fatigue life considerations of vehicles is discussed. The starting point is lateral vehicle loads acting on trucks. Customer usage is described by using road classification. Within each road class the loads are assumed to be similar and a particular customer can be modelled on the basis of the distance the driver drives in each road class. A parametric model is built up to describe the lateral loads found in measurements. From this model, an explicit expression for the expected damage is derived based on the rain flow cycle count method, the Basquin equation and the Palmgren-Miner rule. This model aims to capture the external loads coming into the vehicle in the lateral direction. In particular, it focuses on the operational environment and customer behaviour. From measurements, the parameters in the model are estimated. The variation of the parameters over the measurements is studied using the analysis of variance method. In the study, the road class as well as more local factors such as market and driver are considered in order to make sure that the road class can be regarded as global. The loads in the road classes can then be modelled using the results from the analysis of variance. Practical considerations in the design of experiments to gather data for the models are taken into account. The analysis technique is applied to the load model using data from a field study made in three different markets. Conclusions are drawn on the possibilities of classifying road loads and the need for design of experiments to use field measurements more efficiently.