Corrosion level estimation in reinforced concrete beams by acoustic emission sensing and selective crack measurements

Abstract Accurate corrosion assessment plays an important role in ensuring structural safety of reinforced concrete (RC) structures. However, on‐site assessment of existing concrete structures presents many challenges, including high costs, limited inspection timeframes, and difficult accessibility. To facilitate inspection‐based corrosion assessment, this paper presents a novel approach by combining short‐term acoustic emission (AE) monitoring with selective crack width measurements for corrosion level (CL) assessment. AE sensing is a monitoring technique which can detect ongoing internal degradation mechanisms by analyzing ultrasonic waves emitted by the damage process. Yet, two major challenges arise on site: (1) practical limitations prevent continuous AE monitoring over the structure's entire lifetime and (2) AE can only generate relative results in the absence of reference measurements. This paper addresses both challenges to bridge the gap between laboratory experiments and on‐site monitoring of corroding RC structures. First, short periods of AE monitoring data are analyzed to investigate the potential of AE sensing over limited timeframes. Second, AE data of corroding RC beams are combined with sparse crack width measurements in order to obtain absolute CLs. The proposed methodology is experimentally validated by corroding eight beams with varying dimensions, corrosion zones, and reinforcement layouts. The experimental results prove that the estimated CLs closely match the rebar mass losses, with a mean absolute error of 1.53% CL for beams reaching up to 14% CL, confirming the potential of the combined AE and crack width measurement technique as an efficient and accurate condition assessment approach. Moreover, AE sensing provides detailed spatial variability of the rebar corrosion in the monitored zone, which is challenging to obtain with conventional techniques. By using this dual‐technique approach, shorter monitoring periods prove nearly as effective as continuous AE monitoring in accurately estimating CLs.