Abstract
In the presence of an electrolyte and an electrochemical potential difference, galvanic corrosion can occur, increasing the corrosion rate on the more-active material and decreasing the corrosion rate on the more-noble material. Although galvanic corrosion is generally considered an isolated mechanism on its own, it often induces other localized corrosion events. For example, micro-galvanic couples between active intermetallic precipitates and the more-noble matrix of a given alloy can initiate and accelerate pitting corrosion. Similarly, macro-galvanic couples between dissimilar materials in a crevice geometry can have a large enough driving force to induce crevice corrosion which would not have been present if the crevice former was inert.
In modern structures, neither scenario above is avoidable. Alloying elements are added within a material to achieve an optimization of mechanical properties, leaving the substrate susceptible to pitting via the electrochemical potential differences between the precipitates and matrix. Furthermore, precipitate-strengthened aluminum alloys require mechanical joining, rather than welding, which introduces dissimilar materials into electrical contact. A common form of mechanical joining uses fasteners, which can create an occluded region between the fastener shaft and fastener hole. The geometry of both a cylindrical fastener hole and a hemispherical surface pit can act as a stress concentrator, increasing the probability for crack initiation and propagation. Therefore, both macro- and micro-galvanic induced localized corrosion events are extremely relevant in terms of the structural stability and lifetime of a component. This work takes a mixed experimental and computational approach of validation, prediction, and finally mitigation of galvanic-induced localized corrosion in three sections: 1) galvanic-induced crevice corrosion, 2) pitting corrosion, 3) computational methodologies to accommodate complex systems. The effect of geometry, environmental factors (chloride concentration and water layer thickness), and material selection were investigated in each section, and limitations within each simulated scenario were discussed.
Validation of a finite element method (FEM) model assuming the Laplace equation was first conducted through comparison with literature and in-house experimental panels for a SS316/Ti-6Al-4V/AA7075 fastener/panel galvanic couple. The computational and experimental results determined that in fastener-in-panel systems, severe galvanic-induced crevice corrosion occurred within the fastener hole, independent of the visible surface corrosion damage. Therefore, two main strategies were determined to mitigate corrosion damage within the susceptible fastener hole region, 1) to lower the overall galvanic current, possible through the application of a sol-gel coating or less-noble fastener, 2) to concentrate the majority of current on the surface of the panel, rather than the occluded fastener hole, which may be achieved through controlled surface defects, bulk WL, and a raised fastener head. A machine learning algorithm was also created through the resulting FEM data to predict under which conditions, environmental and geometric, the majority of current would occur in a creviced region, with the goal to understand and prevent problematic scenarios.
The work on macro-galvanic couples was extended to account for, and mitigate, stress corrosion cracking. A materials selection framework was created, combining linear elastic fracture mechanics (LEFM) and FEM, to limit scenarios in which high crack growth rates or high corrosion rates may occur. The potential distributions predicted through the FEM, specifically in the highest-stress region, were used to optimize scenarios which fell in the base of the “U”-shaped crack growth rate dependence on potential. Current density distributions and the total current in the system were then used to further narrow down the material selection scenarios.
In terms of micro-galvanic coupling, pitting of SS316 was investigated with the same computational approach as above. The stability of a given pit was determined through the critical pit stability product (i⋅x)_crit and the repassivation potential (E_rp), for a variety of pit geometries, finite cathode sizes, and water layer thicknesses. In all tested conditions, E_rp predicted a higher stability of pits to continue growing than (i⋅x)_crit, leading to the consideration that the conventionally measured value of E_rp may be too conservative. Pulling upon two recent pitting framework developed in literature, an equation was proposed to calculate E_rp based on the transition potential (E_T), the critical percent saturation (f), and the anodic E-log(i) Tafel relationship of the pit base (b_a). Utilizing data from literature on SS316, consistent values of E_rp were determined which were approximately 70 to 120 mV more electropositive than convention.
An accumulation of the macro- and micro-galvanic coupling work was conducted by simulating a SS316/AA7050 couple which was experiencing localized pitting events on the inhomogeneous AA7050 surface. It was determined that to best correlate the simulated results with the experimental scanning vibrating electrode technique (SVET), the experimental boundary conditions needed to account for both anodic and cathodic deviations from generic bulk conditions. Although FEM was not able to account for individual pitting events, as the location, size, and distribution of the activated regions would be needed as input parameters, the average current density distribution and total current magnitude showed good comparison between the computational and experimental data, once the boundary conditions were modified.
