Improving the Accuracy of Electrochemical Experiment Data for Artificial Intelligence–Based Carbon Steel Corrosion Analysis

Abstract

Carbon steel is an essential material for constructing pipelines in different industries, but serious casualties and significant economic loss could occur due to the corrosion of the pipeline steel. The artificial intelligence–based method has been shown successful in evaluating the corrosion problem in pipelines. The accuracy of the corrosion data used in the artificial intelligence–based method is crucial. The data can be collected from field tests and laboratory experiments. Compared with field tests, the corrosion data obtained through laboratory experiments, such as electrochemical measurements, are more comprehensive, but the results are often inaccurate due to the nonuniform surface state. This study aims to propose a method to improve the accuracy of electrochemical experiment data by modified the surface morphology. To achieve the objective, a technique was designed to prepare the uniform morphology. Moreover, different electrochemical measurements, texture analysis, micromorphological characteristics, and chemical composition analysis were performed under the nonuniform and uniform surface conditions in this work. It was found that the addition of 0.5×10−3 M Fe2+ was effective in preparing a homogenous electrode surface while almost having no additional impact on the corrosion process. The results proved that the surface morphology of the carbon steel electrode significantly influenced the results of the open circuit potential and potentiodynamic polarization. A uniform surface could improve the accuracy of electrochemical measurements.