LAUSR

Abstract The degree of bending (DoB) characterizing the through-the-thickness stress distribution has a profound effect on the fatigue behavior of tubular joints commonly found in steel offshore structures and the determination of DoB values is essential for improving the accuracy of fatigue life estimation. Probability density functions of the involved random variables are necessary for the fatigue reliability analysis of offshore structures. The objective of present research was the derivation of probability density function (PDF) for the DoB in tubular KT-joints commonly found in jacket-type offshore platforms. A total of 162 finite element (FE) analyses were carried out on 81 FE models of KT-joints subjected to two types of axial loading. Generated FE models were validated using experimental data, previous FE results, and available parametric equations. Based on the results of parametric FE study, a sample database was prepared for the DoB values and density histograms were generated for respective samples based on the Freedman-Diaconis rule. Thirteen theoretical PDFs were fitted to the developed histograms and the maximum likelihood (ML) method was applied to evaluate the parameters of fitted PDFs. In each case, the Kolmogorov-Smirnov test was used to evaluate the goodness of fit. Finally, the Generalized Extreme Value model was proposed as the governing probability distribution function for the DoB. After substituting the values of estimated parameters, six fully defined PDFs were presented for the DoB in tubular KT-joints subjected to two types of axial loading.