LAUSR

Abstract The present study considers the prediction of extreme values of the second-order hydrodynamic parameters related to offshore structures in waves, where the application of Gaussian distribution is not valid. Particularly, this study focuses on a characteristic function approach in the frequency domain to estimate the probability distribution of the second-order quantities, and the results are compared with direct simulations in the time domain. The stochastic behaviors of the second-order hydrodynamic quantities are investigated with the characteristic function approach, which involves eigenvalue analyses of Hermitian kernels constructed with quadratic transfer functions. Three different second-order responses are considered: the springing responses of TLP tendons representative of the sum-frequency problem, the slow-drift motions of a semi-submersible platform moored in waves as a representative of the difference-frequency problem, and the wave run-up around a vertical column for regular and irregular waves. The applicability of the present approach in predicting extreme values is assessed by comparing the results with the values obtained from time-domain signals.