A spectral method for fatigue analysis based on nonlinear damage model

For estimating structural fatigue damage or fatigue lives under random loads, time-domain methods are accurate and able to combine with nonlinear damage evolution models but too inefficient. Spectral methods, in contrast, are highly efficient but mostly based on Miner’s linear damage accumulation rule, which cannot reflect the nonlinear process of real fatigue damage. In this paper, by combining the block cycle jump technique which is frequently used in time-domain methods, with Dirlik’s spectral method for fatigue life estimation, a new spectral method for fatigue damage estimation that can be based on nonlinear damage models is proposed. Based on experimental data and numerical simulations, the ability of estimation nonlinear fatigue damage and computational accuracy and efficiency of this method is verified. Results show that this method inherits the advantage of high efficiency of classical spectral methods while its accuracy is also sufficiently high.