Improving half-cell potential survey through computational inverse analysis for quantitative corrosion profiling

Abstract

Corrosion of reinforcing steel is the primary cause of reinforced concrete (RC) premature failure. Half-cell potential survey based on ASTM C876 is a widely-used method to evaluate RC corrosion. However, the quality of the survey depends on the amount of data acquired, and data acquisition translates into labor time. Low quality surveys may lead to poor evaluation and misleading assessment. This study demonstrates that field survey augmented with computational methods is a promising approach to an improved diagnosis of RC corrosion. The approach augments field data from half-cell potential survey with computational method to diagnose the location and size of corrosion in the reinforcing steel of RC. The computational method integrates Particle Swarm Optimization into Boundary Element Method to carry out inverse analysis. The effectiveness of the approach was tested for an RC column of a local building affected by the 2004 Indian Ocean tsunami. The half-cell potential survey was used to find the general location of the corrosion in the RC. The computational inverse analysis was then used to diagnose the location and size of the corrosion based on the acquired field data. The actual corrosion location and size were then confirmed by breaking up the RC column. It was found that the approach was able to diagnose the location and size of corrosion up to 95.65% of accuracy. This study concludes that computational inverse analysis is a promising venture to improve the quality of and extend the usability of the half-cell potential survey for quantitative corrosion profiling.