PSI - Issue 57

Baran Yeter et al. / Procedia Structural Integrity 57 (2024) 133–143 Baran Yeter & Feargal Brennan/ Structural Integrity Procedia 00 (2023) 000 – 000

134

2

1. Introduction In accordance with the UN’s sustainability goal regarding affordable, reliable, sustainable, and modern energy for all, offshore wind has emerged as one of the fastest-growing sectors in renewable energy. Encouraged by the recent substantial success of the offshore wind sector, the public and private investors seek innovative solutions to maximise ocean renewable energy. In this regard, the floating offshore wind turbine is paramount to take advantage of high wind potential over deeper seas resulting in higher capacity factors. However, there is a danger of creating dependence on intermittent renewable energy, such as offshore wind, which points out the significance of energy storage by non electrical energy vectors to ensure a resilient energy system. The green hydrogen production from offshore wind can be directly linked to such developments by transforming ocean renewable energy into new energy vectors such as energy storage, renewable heating, and the decarbonisation of transport (road, rail, marine and potentially aviation). The semisubmersible is a prevalent support structure type for floating offshore wind turbines (FOWT) due to its stability, ease of manufacture and deployment. Moreover, it offers more space for hydrogen production and storage. However, the semisubmersibles consist of welded-tubular joints, which are potential sources of hotspots, leading to crack initiation. Thus, fatigue is a critical design concern for such dynamically sensitive structures (DNV, 2014). FOWTs can also be towed back to shore for inspection, maintenance, and repair, which opens the door for a damage tolerant (DTD) approach to be adopted for structural integrity management. This is particularly crucial because the DTD approach for structural integrity management combined with risk-based methodology allows for customised solutions for optimal structural dimensioning, material choice and adaptable intervention planning, which are unavailable for fixed-bottom offshore wind turbines. As the probabilistic structural integrity assessment is the backbone of this overarching risk-based damage-tolerant framework, the ultimate objective of the present study is to assess the structural integrity of a fatigue-critical structural detail of a semisubmersible FOWT under variable amplitude loading (VAL), which can affect the remaining life estimation due to the load sequence and overload-induced retardation effect. The uncertainty regarding fatigue crack growth (FCG) and its propagation are vital for a DTD as the time-dependent structural reliability estimate determines the action plan to mitigate the risks (e.g., structural health monitoring, non-destructive inspection and maintenance). Traditionally, the stress-life (S-N) approach has been recommended practice to deal with high-cycle fatigue by the classification societies. The reason why is that the S-N approach can be pretty suitable for structures containing small defects under low load levels, i.e., elastic stress levels (Besten, 2018). However, according to the study reported by Moan (2018), most fatigue failures in offshore structures (oil and gas platforms) were due to abnormal initial defect size due to the welding process and corrosion fatigue. Fracture mechanics is a local approach dealing with the very nature of fatigue cracking under cyclic load. It offers a better model including some of the essential issues related to FCG for offshore wind structures, such as the effect of corrosion (Adedipe et al., 2016; Jacob & Mehmanparast, 2021), load sequence effect and overload-induced retardation (Brennan, 2014; Yeter et al., 2015), material properties of high-strength steel (Igwemezie et al., 2019). However, it is worth noting that as the fatigue damage model get more sophisticated to account for all the influencing factor, the uncertainty related to modelling and remaining crack life estimation increases as well, which makes the use of probabilistic assessment account for all relevant uncertainties to make a better-informed decision concerning asset integrity management. In this regard, Maljaars et al. (2015) demonstrated that different load cycle sequences could yield very different fatigue lives, especially which could stem from significant retardation effects as a result of overloads. However, such retardation effects are less observable for random load history compared to the VAL with single overloads. One of the reasons that can cause such scatter in crack life is the characterisation of the plastic zone and size, as also argued by Sumi (2014), correspondingly the model used to estimate the stress intensity factor (SIF) in the plastic zone. The study reported Micone and De Waele (2019) confirmed that there was still a substantial ambiguity in the effect of VAL when there was plasticity-induced acceleration and deceleration of crack growth, including the models to predict crack behaviour under VAL, which points out the importance of quantification of modelling uncertainty regarding the remaining crack life estimation. However, for the fatigue design of ship and offshore, there has been a considerable amount of effort to produce a standardised load-time history to account for the load sequence effect for structures (Cui et al., 2011; Li & Cui, 2015; Li et al., 2020). In this regard, Brennan (2014) emphasised the importance designing time histories that are representative of the stochastic service loading of offshore wind turbines under both wind- and wave-