Signal Response-Based POD for Reliability of Damage Detection in Reinforced Concrete Structures

Structural concrete damage evolution can be made by non-destructive testing (NDT) methods. These methods are divided into two subcategories as active and passive methods due to their applications. While defect characteristics are analyzed by measuring the changes in the material caused by externally provided energy in the active methods, passive methods use the energy emitted from the internal processes of the material without using any external source. Even though the test conditions of the passive NDT methods can be controlled, the randomness of the obtained data affects the probability of detection (POD) of the damage inspection. Particularly in concrete, an excessively heterogeneous material, sensor-damage distance and damage characteristics influence the reliability of the inspection procedure. This study aims to investigate the reliability of the passive NDT method for monitoring damage in concrete by focusing on effects of the sensor placement and damage severity. For this, with Acoustic Emission (AE) technique, detection reliability was comprehended on structural bending test of a reinforced concrete beam by signal response-based POD. As the AE parameters such as amplitude and energy are related with the damage severity, both of them were utilized to generate the POD curves with respect to the beam’s deflection. Furthermore, compared to prior studies on concrete damage inspection, Berens model and confidence interval analysis were applied by providing quantitative framework to prove the robustness and sensitivity of the method for detecting even low-load level damage.