Experimental study of grout defect identification in precast column based on wavelet packet analysis

Grout defects always exist in sleeves of precast structures, but studies on grout defect identification are rarely performed. This article proposes a combination method of dynamic excitation technique and wavelet packet analysis for sleeve defect identification in the precast structure. Hammer excitation on a 1/2-scaled two-floor precast concrete frame structure with column rebar splicing by grout sleeves is conducted to collect column acceleration responses. Moreover, the corresponding energy spectrum is obtained by the wavelet packet analysis. Furthermore, three defect identification indices, that is, percentage of energy transfer, energy ratio variation deviation, and energy spectrum average deviation, are calculated and compared. Robustness analysis of the energy ratio variation deviation is carried out by adding white noise in the original acceleration response signals. The results show that (1) the percentage of energy transfer, the energy ratio variation deviation, and the energy spectrum average deviation are positively correlated with the grout defect degree where the energy ratio variation deviation is more sensitive in the identification of defects; (2) the energy ratio variation deviation robustness of the original signal with the inputted multiplicative white Gaussian noises is better than that with the inputted additive white Gaussian noise; and (3) the proposed defect identification method can characterize the sleeve grout defect degree in column.