An EMI-Based Clustering for Structural Health Monitoring of NSM FRP Strengthening Systems

Abstract

The use of fiber-reinforced polymers (FRP) in civil construction applications with the near-surface mounted (NSM) method has gained considerable popularity worldwide and can produce confident strengthening and repairing systems for existing concrete structures. By using this technique, the FRP reinforcement is installed into slits cut into the concrete cover using cement mortar or epoxy as bonding materials, yielding an attractive method to strengthen concrete structures as an advantageous alternative to the external bonding of FRP sheets. However, in addition to the two conventional failure modes of concrete beams, sudden and brittle debonding failures are still likely to happen. Due to this, a damage identification technology able to identify anomalies at early stages is needed. In this work, some relevant cluster-based methods and their adaptation to electromechanical impedance (EMI)-based damage detection in NSM-FRP strengthened structures are developed and validated with experimental tests. The performance of the proposed clustering approaches and their evaluation in comparison with the experimental observations have shown a strong potential of these techniques as damage identification methodology in an especially complex problem such as NSM-FRP strengthened concrete structures.

Keywords

• Structural health monitoring
• PZT sensors
• electro-mechanical impedance
• hierarchical clustering
• k-means clustering
• NSM-FRP strengthening