A Numerical Investigation on the Pitting Corrosion in Offshore Wind Turbine Substructures

Pitting corrosion is a common cause of concern for steel structures in an offshore environment. As geometric stress concentrating features, corrosion pits can potentially act as fatigue crack initiation sites. The current study is a part of the MAXWind project which, amongst others, aims to develop numerical tools for a more accurate estimate of the remaining lifetime of in-service wind turbines. UGent is responsible for developing an advanced corrosion-fatigue model which will be used to build “smart S-N curves”. The smart S-N curve is a novel concept that takes the level of corrosion into account. To this end, the entire evolution of corrosion fatigue is divided into three major phases including pitting corrosion, short fatigue crack propagation, and long fatigue crack propagation, see Figure 1. The main focus of this work is on pitting corrosion and its transition to short fatigue crack propagation. A phase-field modelling approach [1],[2] is used to simulate the autonomous growth of a corrosion pit. The corrosion phenomenon - pitting corrosion in particular – is a complex electrochemical process that is influenced by various environmental factors such as temperature, dissolved oxygen, pH, salinity, etc. [3]. Phase-field modelling is a robust technique that is capable to incorporate a vast range of influential parameters. In essence, in a phase-field model, each phase (here, metal and electrolyte) possesses a constant value in the bulk (0 for Pitting corrosion model transition transition Integrated model Experimental validation Short fatigue crack model Long fatigue crack model Phase-field modeling approach NR model X-FEM 18 th EAWE PhD Seminar on Wind Energy 2 – 4 November 2022 Bruges, Belgium electrolyte and 1 for steel), with a continuous interpolation between the bulk values across the interface between phases. The evolution of the system is a result of constrained minimization of free energy for which the advective Cahn-Hilliard equation is used. The Nernst-Planck equation is used to describe the diffusion of ions within the electrolyte, and the Butler-Volmer-type kinetic expression is used to calculate the reaction current density throughout the process. For more information and formulations see [1]. First, an electrochemical characterization was performed for structural steel grade S355 in an environment that is representative of the North Sea. This study is crucial to evaluate the electrochemical behaviour of this steel grade and will support further studies towards predicting pit dimensions in offshore wind turbine support structures in the North Sea. To this end, potentiodynamic polarization tests were implemented for S355 steel. An Ag/AgCl electrode was used as reference electrode in the tests and potential values are obtained against this electrode. The corresponding corrosion potential and current density were obtained as -711 mV vs. Ag/AgCl and 0.1534 A/m2 , respectively. For a metal, the more negative the corrosion potential is, the more susceptible it will be to corrosion [4]. In practice, corrosion protection systems and coatings will be applied to the metal structure, which will increase the value of corrosion potential [5]. Using the output of the experiments as input to the phase-field model, a parametric study was performed to assess the effect of the applied potential on geometrical parameters (pit width and depth) and electrochemical parameters associated with pit growth rate (metal cation concentration and reaction current density), see Figure 2. It was found that for an applied potential of -600 mV vs. Ag/AgCl, the corrosion process stays in the activation-controlled regime throughout the simulation time (150 seconds). Applied potentials of -550 to -500 mV vs. Ag/AgCl take the system to the current-resistance-controlled regime where the metal cation concentration does not reach the saturation. The higher the applied potential is, the more pitting corrosion is accelerated until it reaches to the point where any additional increase in applied potential will not have any 18 th EAWE PhD Seminar on Wind Energy 2 – 4 November 2022 Bruges, Belgium additional influence on pit growth rate. As it is easier for the metal ions to diffuse into the bulk electrolyte near the pit mouth in bare steel, pit width increases with a higher velocity in comparison to the pit depth. All numerical results will be validated with dedicated experiments. Any changes in temperature will cause changes in electrochemical parameters such as ionic diffusivity, corrosion potential and corrosion current density. As future work, a parametric study will be conducted on the effect of temperature on the corrosion pit growth rate. Ultimately, pit dimensions extracted at every time step will serve as input to a short fatigue crack propagation model [7]. Once a corrosion pit nucleates, the local stress in the material increases at the discontinuity. Therefore, in parallel to the pitting corrosion study, a finite element analysis was performed to assess the stress concentration factor (SCF) of corrosion pits and to identify the effect of different normalized geometrical parameters on SCF, see Figure 3. According to [8], a crack will initiate from a pit under the following conditions: (a) the pit size must pass a critical value, and (b) the crack growth rate must be higher than the pit growth rate. The concept of SCF can assist to address the first criterion. A pit becomes susceptible to transit to a crack as soon as it causes a considerable increase in the local stress. Results show that an increase in a/2c or b/c leads to enhanced SCF when the loading is oriented parallel to the pit’s major axis (i.e., = 0), see Figure 4 (a) and (b)