Research on dynamic indicators for void defect detection in rectangular cross-section CFST columns

This study aims to address potential internal void defects in rectangular cross-section concrete-filled steel tube (CFST) columns. A CFST internal void defect detection method based on vibration testing and modal analysis is proposed, with three dynamic indicators constructed from modal parameters for CFST void detection, including absolute difference of the diagonal gaussian curvature of the flexibility matrix, the gaussian curvature of the absolute diagonal flexibility difference, and the Bayesian fusion indicator. Indoor experiments were conducted using CFST specimens with and without internal voids, applying excitation with a shaker and capturing vibration response signals using a laser Doppler vibrometer. Modal parameters were obtained through modal analysis to validate indicators effectiveness. Further on-site experiments were conducted, and modal parameters under various void conditions and loading scenarios were obtained via finite element simulation, verifying the effectiveness of the indicators. Experimental results demonstrate that these indicators can effectively locate and identify cm-level internal void defects in CFST structures, with the Bayesian fusion indicator performing most effectively. This study validates the feasibility of dynamic indicators in CFST void defect detection, providing theoretical support and experimental data for this field.