A spectral fatigue damage prediction model for offshore structures under Gaussian random processes based on bimodal spectra

Abstract

Offshore structures frequently confront complex and variable marine environments, and the frequency-domain approach is often used in engineering to predict the fatigue damage of structures. To bridge the gap between stress spectra and stress range distributions, this paper proposes a novel stress range probability density distribution model comprising two stress range distribution functions, which can be well adapted to the shapes of the stress power spectral density (PSD) in various cases, and provides a more accurate representation of the rain-flow range probability density function (PDF). This study initially analyzes the stress range distributions of bimodal spectra using the rain-flow counting RFC method. Subsequently, and the unknown parameters of the stress range PDF are determined using the 1st-order moment of rain-flow stress amplitudes and polynomial relationships derived from normalized mean frequency and shape parameters. The proposed model is validated through the use of stress time-series data generated from random combinations of normal distribution functions.