Dimensionless analysis of pulse-like effects on the seismic behavior of a dam based on wavelet-decomposed near-fault ground motions

Abstract

In this paper, the inherent self-similarity between seismic characteristics and the dynamic behavior of a concrete gravity dam under decomposed near-fault ground motions is investigated. The representative low- and high-frequency components were decomposed using the continuous wavelet transform method to explore their effects on the dam response. The Koyna concrete gravity dam system was used as the basis for a numerical model. A series of dam dynamic responses under near-fault motion with and without pulse properties were obtained by numerical analyses. The relationship between the proposed dimensionless indices (intensity measure ratio (ΦIM), pulse period ratio (T1/Tp)), and response ratio (ΨEDP, m) was established to study the inherent self-similarity in the dynamic behavior of dams, which has an explicit physical significance with smaller dispersion. The response proportions of two main frequency components in the original near-fault record were revealed. Furthermore, a series of new threshold values were calculated based on the intersection of the pulse and high-frequency ratios, and the range of influence of the low- and high-frequency components was quantitatively divided. Finally, a predictive model was proposed for the dam responses to near-fault ground motions containing the proposed dimensionless indices that exhibited reasonable patterns. The pulse-like seismic characteristics and seismic responses to near-fault ground motions were analyzed and selected in the context of sensitivity, inherent self-similarity, threshold point, and regression error. Results show that the self-similarity between ΦSa(T1), ΦASI and ΨDVR,pulse,ΨDVRn,pulse is superior to that of other parameter ratios, indicating that more attention should be given to the role of intensity and pulse parameters in practical concrete dam seismic design.