A mode shape sensitivity-based wavelet feature extraction method for interface debonding detection in concrete-filled steel tubes
Mohtasham Khanahmadi, Borhan Mirzaei, Gholamreza Ghodrati Amiri, Majid Gholhaki, Omid Rezaifar
Abstract
Abstract The use of concrete-filled steel tube (CFST) composite columns is increasingly prevalent in the construction industry, particularly in high-rise structures. A common issue in CFST columns is interface debonding between the concrete core and the steel tube. If this debonding progresses both superficially and deeply, it can lead to instability and buckling of the column, posing a serious threat to the overall structural integrity. This study presents an innovative and effective method for extracting damage-sensitive features using horizontal, vertical, and diagonal detail coefficients derived from the wavelet analysis of corrected modal signals. The study introduces the total normalized irregularity detection index (NIDI T ) as a damage detection metric. The results indicate that NIDI T is highly effective in identifying and detecting debonding regions. NIDI T quantifies the accumulation of irregularities and disturbances in the affected areas, allowing for the detection of concrete surface debonding from the steel tube. The findings show that NIDI T can accurately and efficiently detect damage in middle and end-edge regions, addressing a significant challenge in structural health monitoring with high precision.