Improvement Estimation Accuracy of Impact Detection Using Metal-Core Piezoelectric Fiber/Aluminum Composites

Metal-core piezoelectric fiber/aluminum composites are promising candidates for use as sensors for structural health monitoring (SHM) systems. These composites exhibit excellent mechanical properties compared with conventional piezoelectric ceramics because the fiber is embedded into the aluminum matrix using the interphase forming/bonding method. Furthermore, it has emerged that the output voltage characteristics of composites show distinct anisotropy. Herein, an impact detection system with more practical estimation accuracy is developed by performing wavelet transformation on the output voltage. To investigate the practicality of the proposed method, the estimation results are compared with the results using the conventional output voltage. By exploiting the anisotropic response, the new system is expected to reduce the number of sensors required compared with conventional systems, while retaining accuracy. The results show that the present system could be applied successfully using models of the output voltage and signals obtained from composites fixed in two different positions. Application of the wavelet transform method to the output voltage reduced the estimation errors in comparison with values estimated directly from the output voltage, with maximum estimation error of 4.38 mm. This accuracy cleared the minimum required accuracy for an SHM system; thus, confirming the validity of this approach.