Análise de fadiga estocástica pela simulação Monte Carlo com variáveis de controle

Abstract

Offshore structures are subjected to time-varying load effects during their service life, which may result in fatigue damage to them. Hence, fatigue analysis is of utmost importance in the design of offshore structures in order to meet its safety standards. However, fatigue analysis may result in a high computational cost, since it requires a huge amount of structural dynamic analyses to consider all possible sea states that the structure may be exposed to. In this regard, the aim of this study is to investigate the Monte Carlo Simulation using Control Variates as a possible efficient method in fatigue analysis of offshore structures. Thus, different control functions are studied and tested, for instance, a quadratic polynomial function, a model that uses an interpolation function constructed by the Kriging method and a simpler function proposed by Low [1]. Importance Sampling is another technique investigated as an alternative to enhance the efficiency of fatigue damage assessment. Therefore, two fatigue analyses are presented, where one consists of a theoretical simplified model and the other consists of a FPSO mooring lines system subjected to several sea states composed by two combined waves conditions (wind-sea and swell waves). The results obtained point out that the application of appropriate control functions as well as adaptive approaches of Importance Sampling technique are likely to drastically reduce the number of simulations required to achieve the same level of accuracy in the assessment of fatigue damage when compared to the classical Monte Carlo Simulation.