Parametric interpolation method for long-term extreme response and probabilistic fatigue prediction of offshore structures

Abstract

This work presents a semi-analytical methodology to efficiently evaluate the multi-dimensional integrals found in the analyses of the long-term extreme response and probabilistic fatigue of offshore structures. Some traditional methodologies for longterm extreme response and probabilistic fatigue analysis of marine structures described in literature are discussed in this study. Besides these methods, a new procedure developed in this work for optimizing these calculations is explained in details. The proposed method is based on the parametric interpolation of parameters which describe the short-term time-series responses, e.g., statistical parameters of peaks probability distribution and short-term fatigue damage. This interpolation method is then used to efficiently evaluate the long-term convolution integrals. Three comprehensive examples are presented showing the efficiency and accuracy of the proposed method when compared with the complete long-term integration methods (“brute force” methodology). In the first example, an analytic model of a single degree of freedom is analyzed in the frequency domain. In the second one, a Steel Catenary Riser (SCR) is evaluated based on time-domain simulations. In the sequence, a similar analysis is performed for a Steel Lazy Wave Riser (SLWR).