Strain, being superiorly sensitive towards damage, has been widely explored in modal testing based damage monitoring. Piezo sensors have been successfully evaluated for their sensitivity toward incipient damage detection in experimental modal analysis under forced excitations based experimental modal analysis. However, the efficacy of piezo sensors for damage monitoring under operational/ambient conditions is yet to be investigated in depth. This study addresses this gap by evaluating a scaled-down bridge model under pedestrian motions for damage monitoring by piezo sensors under operational modal analysis (OMA). Artificial damage is created by introducing a vertical cut at four locations with two different intensities. Displacement and strain responses are measured using accelerometers and piezo sensors, respectively, and corresponding displacement and strain based modal parameters are extracted using impact hammer and under operational pedestrian motions. Modal flexibility based curvature and coordinated modal assurance criteria (CoMAC) are employed for damage monitoring. Variation in modal flexibility based curvature is adopted as damage sensitive feature (DSF). rComac is defined as the difference of CoMAC from unity and is also adopted as DSF. Mahalanobis distance based outlier analysis proves the strain-based DSF’s superiority to the displacement-based DSFs for damage monitoring, especially for the incipient damage states. CoMAC is observed to possess higher sensitivity than the curvature for incipient damage. Neighbouring sensor locations exhibit potential damage in higher damage states with Mahalanobis distances higher than the critical value. It is found that the CoMAC-based damage strategy is better approach toward damage monitoring combined with curvature as DSF. In overall, the piezo sensors are successfully established as effective strain sensors for damage monitoring in OMA under output-only identifications.
Read full abstract