One of the sources of sustainable energy with great, still untapped potential is wind power. One way to harness such potential is to develop technology for offshore use, more specifically at high depths with floating turbines. It is always critical that their structural designs guarantee that their natural frequencies of vibration do not match the frequencies of the most important oscillatory loads to which they will be subjected. This avoids resonance and its excessive undesired oscillatory responses. Based on that, a 3D finite element model of a 15 MW semi-submersible floating offshore wind turbine was developed in the commercial software ANSYS Mechanical ^® to study its dynamic behavior and contribute to the in-depth analysis of structural modeling of FOWTs. A tower and floating platform were individually modeled and coupled together. The natural frequencies and modes of vibration of the coupled system and of its components were obtained by modal analysis, not only to verify the resonance, but also to investigate the determinant factors affecting such behaviors, which are not extensively discussed in literature. It was found that there is strong coupling between the components and that the tower affects the system as a result of its stiffness, and the floater as a result of its rotational inertia. The platform’s inertia comes mainly from the ballast and the effects of added mass, which was considered to be a literal increase in mass and was modeled in two manners: first, it was approximately calculated and distributed along the submerged flexible platform members and then as a nodal inertial element with the floater being considered as a rigid body. The second approach allowed an iterative analysis for non-zero frequencies of vibration, which showed that a first approximation with an infinite period is sufficiently accurate. Furthermore, the effects of the mooring lines was studied based on a linear model, which showed that they do not affect the boundary conditions at the bottom of the tower in a significant way.

This article presents a probabilistic methodology for assessing corrosion initiation in reinforced concrete structures exposed to chloride ingress. The approach addresses key limitations of conventional analytical models by accounting for non-diffusive barriers and incorporating a rigorous statistical treatment of censored data to mitigate biases introduced by limited simulation durations. A combination of analytical solutions for diffusion from opposite sides with time-dependent boundary conditions is also proposed and validated. The probabilistic study includes the depassivation assessment of a hollow pier section. The blocking effect caused by rebars is statistically characterised through correction factors derived from finite element simulations. These factors are used to adjust analytical solutions, which are computationally inexpensive. Results show that neglecting the rebar blocking effect can overestimate the mean corrosion initiation time by up to 42%, while the use of censored data reduces bias in lifetime estimates. The observed frequency of censored events reached up to 20% when simulations were truncated at 100 years. The corrected analytical models closely match the finite element results, statistically validating their application. The case study indicates premature corrosion initiation (less than 10 years to achieve target reliability), underscoring the need to better reconcile the desired levels of reliability with realistic input parameters for depassivation.