Characterization of Metallic Corrosion Through Electrochemical Modeling and Electrical Impedance Imagin

Corrosion of metallic components affects the long-term performance of equipment and structures since corrosion can cause distributed and localized loss of materials. To estimate the rate of corrosion, experimental methods such as linear polarization based methods are commonly used. However, predicting the future rate of corrosion remains a major challenge. Although different simulation methods have been developed for predicting the rate of corrosion, these models lack sufficient experimental validation under controlled conditions. Also the prediction of corrosion models is a function of input parameters such as polarization behavior of electrodes. While experimental investigations have shown changes in the polarization behavior of electrodes when corrosion exists, the effect of uncertainty of input parameters on the model predictions has not been studied. Therefore, the primary goal of this thesis is to model corrosion, experimentally corroborate the simulation approach, and investigate the uncertainty propagation in the simulation data due to uncertainty of modeling inputs. To this end, a two-dimensional (2D) finite element method is developed to solve the coupled system of Poisson-Nernst-Planck (PNP) equations, the results of which are compared to the 1D and 2D galvanic corrosion experiments under the controlled conditions. The polarization behavior of electrodes are measured experimentally for 7 days in 1 day intervals, which are used to define boundary conditions of electrodes boundaries. Monte Carlo (MC) simulations are then used to quantify uncertainty propagation caused by uncertainty in the polarization behavior of the electrodes. The simulation data show good agreement with the results of 1D and 2D galvanic corrosion experiments.;Corrosion of reinforcing steel in concrete can result in structural safety and serviceability issues. A secondary goal of this study is to investigate the feasibility of the developed modeling approach to the corrosion of steel inside concrete pore solution. To this end, two different synthetic concrete pore solutions are used as electrolytes. The coupled system of PNP equations is solved using a finite element method to model corrosion of mild steel after depassivation. The effect of uncertainty of the input parameters is investigated on the model predictions using MC simulations. The simulation data support experimental results of corrosion current densities measurements of mild steel in the synthetic concrete pore solution.;Many corrosion rate estimates are based on an assumed linear relationship between a small electrical polarization of the metal and the corresponding current density. To develop such a polarization, an electrical connection to the metal is essential. However, securing such an electrical connection to the corroding metal is not feasible in many cases, such as when measuring the corrosion rate of metals in soil or in concrete. Another difficulty with the current methods is the accurate estimation of the polarized surface area of the material which affects the calculated rate of corrosion. The third goal of this thesis is to investigate whether Electrical Impedance Tomography (EIT) may be used to detect the corrosion rate of metals in electrolyte. EIT measurements are collected at different levels of corrosion of steel bar. Absolute imaging scheme is used in the conductivity reconstructions. The results of imaging indicate that the conductivity of the medium changes due to the migration of the corrosion products to the electrolytes. The reconstructions show the high concentration of the corrosion products around the steel bar. However direct observation of corrosion was not feasible with the imaging approach used in this study. Therefore, further development of models and experimental techniques is required.