Interfacial Adhesion–Strength Detection of Structural Silicone Sealant for Hidden Frame–Supported Glass Curtain Wall Based on Nonlinear Ultrasonic Lamb Wave

Abstract

Degradation of interfacial adhesive strength is the main reason for failure of the hidden frame–supported glass curtain wall that is now widely used in the construction field. In this paper, a nonlinear ultrasonic Lamb wave detection method was developed for hidden frame–supported glass curtain walls. Both theoretical and experimental studies were carried out to verify the feasibility of the method. First, an excitation signal modulated by the Hanning window was generated through piezoceramic transducers. The artificial intervention of thermal aging was applied to accelerate the degradation of the interfacial adhesion strength, and the reflection signal of different aging periods was obtained. Then, two methods of discrete Fourier transform (DFT) and wavelet packet decomposition were used to feature extraction, and normalization and regression analysis were adopted to compare two methods more reasonably. The experiment results show that the amplitude integral method using the second-order relative nonlinear coefficient can reflect the interfacial adhesion strength sensitively, whereas the variation curve of the third-order relative nonlinear coefficient was not obvious. The result of the wavelet packet energy method was close to the amplitude integral method and consistent with reality. The second-order relative nonlinear coefficient fluctuated over a small range at first; then there was a decline, and it finally increased with the thermal aging time. However, the mean squared error (MSE) of the wavelet energy method was smaller. The wavelet packet energy method is more accurate and has a higher allowable error ability in characterizing the trends.