Modelling Variability in Vehicle Service Loading Histories
by
Kyung Hyun Park
Abstract
Though multiple service loading histories may have the same general character for any given route, the magnitudes encountered in each history will vary from driver to driver. This variability in loading histories will be a function of the driving styles of the individual drivers, which in themselves have a significant level of variability. As the load history from any one driver will be variable and not a simple linear scaling of the history of any other driver, a simple linear scaling of any history will not characterize the variability of usage over a wide range of service conditions. In this thesis, the effect on fatigue of the variation of service loads produced by different drivers operating a vehicle over a constant route will be presented. This will be accomplished through a statistical technique of measuring and extrapolating cumulative exceedance diagrams to quantify the distribution of loading histories. Necessary in this analysis is the fitting of a single three parameter Wiebull curve to the exceedance diagram. Also discussed is an automated procedure to perform such a fit. Monte Carlo simulations along with the local stress strain approach will be used to simulate a distribution for the varying service conditions imposed by the different drivers. These varying service conditions are the result of a novel approach in which loading histories were obtained through the use of a Four-Wheel Drive All Terrain Vehicle.
Download (2 MB)