Fatigue life evaluation of automotive mechanical components by using smart design algorithm

The aim of this work is to develop a so-called "smart design" in MATLAB (TM) environment. The design process consists of an innovative and automatic algorithm aimed at rapidly evaluating the fatigue damage of automotive mechanical components, starting from the load cycle that the component is subjected. In this work, the target component is the front halfshaft of a sport utility vehicle. The algorithm works as follows. A driving cycle has been acquired through an experimental setup consisting of a control unit, a GPS and an OBD port. A mathematical model of the vehicle, based on longitudinal vehicle dynamics, has been developed in Simulink (TM) environment to evaluate the traction torque at the wheel. A torsional fatigue tests campaign has been carried out on nine front halfshafts of the vehicle to evaluate the torque amplitude-number of cycles curve of the component. The output histories, computed with the rainflow method, have been referred to a load ratio R = -1, according to the experimental fatigue tests, by the application of three different mean stress fatigue criteria: Goodman, Gerber, and Morrow. At the end, cumulative fatigue damage on the component has been evaluated by the application of the Palmgren-Miner rules. The results show that the Goodman criteria, with a fatigue damage value of around 20%, represents the most precautionary criteria for the fatigue design of an automotive component.