Electromechanical impedance-based prestress force prediction method using resonant frequency shifts and finite element modelling

Abstract

The piezoelectric-based smart strand technique could effectively monitor the prestress (PS) force in post-tensioned systems. In the previous study, the PS force was predicted by supervised regression models where the relationship between the EMI feature and the PS force was pre-determined through real-scale experiments. Nonetheless, obtaining the EMI response from an operating structure under damaged states is challenging and costly. To solve this shortcoming, this study proposes a simple alternative method to predict the PS force in the smart strand without conducting preliminary tests on a real-scale structure. The idea is to utilize a reliable numerical model corresponding to the target structure to develop a regression-based PS force prediction model. The experimental validation shows that the proposed methodology can accurately predict the PS force in the smart strand. Since the proposed method does not require experimental EMI data under damaged states, it is promising for practical applications on existing structures.