An efficient signal processing tool for impedance-based structural health monitoring

Various experimental studies have demonstrated that an impedance-based approach to structural health monitoring can be an effective means of damage detection. Using the self-sensing and active-sensing capabilities of piezoelectric materials, the electromechanical impedance response can be monitored to provide a qualitative indication of the overall health of a structure. Although impedance analyzers are commonly used to collect such data, they are bulky and impractical for long-term field implementation, so a smaller and more portable device is desired. However, impedance measurements often contain a sizeable number of data points, and a smaller device may not possess enough memory to store the required information, particularly for real-time analysis. Therefore, the amount of data used to assess the integrity of a structure must be significantly reduced. A new type of cross correlation analysis, for which impedance data is instantaneously correlated between different sensor sets and different frequency ranges, as opposed to be correlated to pre-stored baseline data, is proposed to drastically reduce the amount of data to a single correlation coefficient and provide a quantitative means of detecting damage relative to the sensor positions. The proposed analysis is carried out on a 3-story representative structure and its efficiency is demonstrated.