EMI-based damage prediction of blended concrete under sulphuric acid exposure using ML and ANN models

Concrete structures are exposed to a wide range of aggressive chemical exposure conditions during their service life that can impact their durability and performance over time. Sulphuric acid exposure is a primary cause of rapid degradation in concrete, leading to rapid mass and strength loss. In such cases, embedded sensor-based structural health monitoring techniques are required for a more comprehensive assessment rather than surface-based non-destructive testing methods. This study utilizes an embedded piezo sensor for non-destructive degradation damage monitoring of blended concrete exposed to a 5% concentration of sulphuric acid over 50 days using EMI technique. The degradation damage is qualitatively identified by comparing the baseline and exposure state extracted impedance data. Furthermore, the extracted impedance data are used for degradation damage (mass loss) prediction of concrete using ML and ANN models. The findings reveal that the bagged trees ensemble and wide neural network models demonstrated the highest predictive accuracy, achieving R<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> values of 0.96 and 0.95 respectively.