Prediction of durability of the concrete using electro-mechanical impendence technique: an experimental and machine learning approaches

Abstract

The durability of concrete is a critical factor in ensuring the structural integrity and longevity of infrastructure, with a focus on sustainable development. Traditional durability assessment methods, such as laboratory-based tests, are time-intensive and impractical for real-world applications. This study proposes the use of the Electromechanical Impedance (EMI) technique as a non-destructive, real-time method to assess and predict concrete durability. By embedding Lead Zirconate Titanate (PZT) sensors in concrete specimens, the study monitors changes in admittance signatures-specifically conductance and susceptance signatures over time as the specimens undergo deterioration under exposure to sodium sulfate solutions. The measured signatures are used to derive equivalent structural parameters such as stiffness, damping, and mass, which serve as indicators of deterioration. The results demonstrate that the EMI technique provides a more sensitive, accurate, and efficient means of predicting concrete durability compared to traditional method, further machine learning approach viz. Support Vector Machine was applied to predict the durability. The data trained an Support Vector Regression Model (SVR), which found suitable for predicting deterioration levels. This research contributes to the advancement of Structural Health Monitoring (SHM) and sustainable building practices by offering a scalable and sustainable approach for real-time durability assessment, ultimately improving the long-term safety and performance of concrete structures, contributing to a more sustainable development.