Seminars

Corrosion damage is an extremely widespread danger to the durability and safety of structures. Pitting corrosion can lead to accelerated failure of structural components by perforation, or by acting as an initiation site for cracking. Most models for the evolution of pitting corrosion consider that the corrosion reaction only affects the evolution of the metal surface and they cannot capture changes in the mechanical properties in the layer immediately below the solid/liquid interface. These changes, such as embrittlement induced by corrosion or stress-dependence of the diffusion processes in corrosion, are determining factors in explaining how Stress Corrosion Cracking is triggered and how it progresses in time. In this presentation, I will introduce a novel peridynamic model for the evolution of damage from pitting corrosion. We model the anodic reaction in corrosion processes as an effective diffusion process in the electrolyte/solid system combined with a phase change mechanism using peridynamics. This allows autonomous movement of the interface. No explicit interface conditions need to be enforced. In contrast to the classical models in which a Stefan condition controls the propagation speed of the solution/metal interface, in our peridynamic model the motion of the interface is part of the solution, not part of the problem. To capture subsurface degradation due to corrosion, we introduce a corrosion damage model based on a stochastic relationship that connects the concentration (carried by the diffusion bonds) in the metal to the corrosion damage (carried by the superposed mechanical bonds). We extend the 1D model to 2D and 3D, and formulate a coupled-model in which, in addition to the concentration-dependent damage, we allow material phase change based on damage. This coupled model accounts for broken mechanical bonds that enhance corrosion rate. We apply this model for activation-controlled and diffusion-controlled corrosion. We also investigate the effect induced on the pit shape evolution by the presence of a lacy cover. We predict that when the passive film is weak, corrosion in the horizontal direction is activation-controlled, while in the vertical direction, it is diffusion-controlled. The time-evolution of the pit profile (shape and size) computed by our model matches well those measured in experiments.