This study focuses on variable amplitude loadings applied to automotive chassis parts experiencing carmaker’s specific proving grounds. They are measured with respect to time at the wheel centres and composed of the six forces and torques at each wheel, within the standard vehicle reference frame. In the scope of high cycle fatigue, the loadings considered are supposedly acting under the structure yield stress. Among the loadings encountered during the vehicle lifetime, two classes stand out:
To separate both effects, a frequency decomposition method is proposed before applying any lifetime assessment methods. The usual Rainflow counting method is applied to the Driven Road signal. These loadings, depending on the vehicle dynamics, are time-correlated. Thus, the load spectra is set only thanks to the vehicle accelerations time-measurement. Then a damage summation rule is applied for life assessment, in a very computationally efficient way and carrying a strong physical meaning.
Giving some requirements regarding the signal (stationarity, ergodicity, Gaussian distribution), less time-consuming methods such as spectral ones can be applied to the Random Road signal. As the initial random road signal does not meet these requirements, (road surface variations and shocks) an algorithm performing offline detection of multiple change points is implemented to deliver a series of suitable random signals. This overall method enables to distinguish the loading type impact on fatigue life, based on a customer measurement or a manufacturer's proving ground.