Abstract
We present an approach, termed electrochemical tomography (ECT), for the in-situ study of corrosion phenomena in general, and for the quantification of the instantaneous rate of localized corrosion in particular. Traditional electrochemical techniques have limited accuracy in determining the corrosion rate when applied to localized corrosion, especially for metals embedded in opaque, porous media. One major limitation is the generally unknown anodic surface area. ECT overcomes these limitations by combining a numerical forward model, describing the electrical potential field in the porous medium, with electrochemical measurements taken at the surface, and using a stochastic inverse method to determine the corrosion rate, and the location and size of the anodic site. Additionally, ECT yields insight into parameters such as the exchange current densities, and it enables the quantification of the uncertainty of the obtained solution. We illustrate the application of ECT for the example of localized corrosion of steel in concrete.
Highlights
Localized corrosion is a common phenomenon in a wide range of materials and environments, including corrosion of underground metallic structures[1,2,3,4], reinforced and prestressed steel in concrete[5,6,7,8,9], and medical implants such as vascular stents in the human body[10]
The described approach, electrochemical tomography (ECT), is a promising method for the instantaneous determination of local corrosion rates of metals embedded in porous media
As opposed to traditional electrochemical methods, such as linear polarization resistance (LPR), its accuracy is not limited by the lack of information about the surface area of the anodic part of the corroding steel
Summary
Localized corrosion is a common phenomenon in a wide range of materials and environments, including corrosion of underground metallic structures (e.g., pipes, storage tanks, borehole casings, geotechnical, etc.)[1,2,3,4], reinforced and prestressed steel in concrete[5,6,7,8,9], and medical implants such as vascular stents in the human body[10]. The information gain, in our case, is determined to be the difference between the sampled posterior (the MCMC inversion results) and the prior probability densities, which describe the a priori belief about the model parameters. We conclude that we have reached convergence and sufficiently sampled the posterior in around 30,000 samples
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
