Evaluation of approximative methods for rainflow damage of broadbanded non-Gaussian random loads

Two different methods for approximating the fatigue damage of broad banded non-Gaussian random loads using the rainflow cycle count method are evaluated using loads measured on a truck. Results for Gaussian loads are summarized, and transformations for non-Gaussian loads are discussed. One of the two methods is based on the spectral moments of the process, and the results are obtained as a linear combination of an upper and lower bound. The second method is based on the assumption that the sequence of turning points of the load can be considered a Markov chain, for which results can be obtained. Measurements performed on different markets are used to study the two methods. Results are presented in terms of expected damage and amplitude spectra. Problems and possible improvements of the models are discussed.