Acoustic Emissions Sensor and Fuzzy C-mean Clustering Based Break Detection in Post-Tensioning Tendons

Abstract

Steel tendons have a vital role in prestressed concrete members and the stability of the members depends on tendons. The steel tendons corrode over time or when encased in grouts with excessive levels of chloride. The steel tendons are pre-tensioned or post-tensioned to 80 percent of their ultimate tensile strength, and significant corrosions can increase tensile stress leading to tendon breakage. For the purpose of safety and maintaining schedule, it is critical to develop a wire break monitoring methods for post-tensioning bridges. In this work, Fuzzy C-means clustering technique was employed to detect acoustic emissions released from breaking wires of post-tensioning steel tendons bounded in grout. To collect acoustic emissions of wire break and grout cracks, a constant rate tensile loads were applied to tendons embedded in grouts until a wire broke, and piezoelectric transducers attached to the tendons were used pick up the released acoustic signals. To improve the robustness of the clustering method, environmental acoustic noises were collected from a bridge and added to the database of the tensile tests. Wire break detection using acoustic emissions and Fuzzy C-mean clustering achieved 100 percent accuracy in detecting wire breaking acoustic signals while the acoustic signals of grout cracks and environmental noises were not detected as a wire break.