An experimental validation of the fatigue damaging events extracted using the wavelet bump extraction (WBE) algorithm

Abstract

This paper describes an experimental validation of the fatigue damaging events that were identified and extracted using a wavelet-based fatigue data editing technique. This technique, known as the Wavelet Bump Extraction (WBE) algorithm, is specifically designed to summarise a long record of fatigue variable amplitude (VA) loading whilst preserving the original load cycle sequence. Using WBE the fatigue damaging events were identified and extracted in order to produce a mission signal. In order to validate the effectiveness of WBE in practical applications a VA road load time history that was measured on a road vehicle suspension arm was taken as a case study. Uniaxial fatigue tests were performed using the original signal, the WBE mission signal and the individual WBE extracted segments. A mirror polished specimen of SAE 1042 steel was tested using a servo-hydraulic machine. The fatigue lives measured for these VA loadings were then compared to the fatigue lives calculated from a VA strain loading fatigue damage model. The results show a good fatigue life correlation at the coefficient of 0.98 between the prediction and experiment. For the road load time history considered, the WBE mission signal was found to be only 40% the time duration of the original time history while maintaining 60% of the fatigue damage according to analytical calculation and 87% according to experimental testing.