Premium
Arch dam system identification using vibration test data
Author(s) -
Mau S. T.,
Wang S.
Publication year - 1989
Publication title -
earthquake engineering and structural dynamics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.218
H-Index - 127
eISSN - 1096-9845
pISSN - 0098-8847
DOI - 10.1002/eqe.4290180404
Subject(s) - superposition principle , modal , vibration , system identification , identification (biology) , test data , normal mode , arch , mode (computer interface) , frequency domain , engineering , control theory (sociology) , mathematics , structural engineering , computer science , mathematical analysis , data modeling , physics , chemistry , botany , software engineering , control (management) , artificial intelligence , polymer chemistry , biology , operating system , quantum mechanics
In an effort to study the dynamic characteristics of an arch dam system from the vibration test results, a systematic method of frequency‐domain system identification is developed. The governing equations for system identification are based on a non‐classical modal superposition method. The non‐classical model is shown to be derivable from a general matrix formulation of the dam system. The conventional classical modal formulation becomes a special case of the general non‐classical formulation. The modal parameters of the non‐classical and the classical formulation are to be identified. The system identification method includes a single‐mode sweep procedure for initial parameter estimation and a progressive multiple‐mode parameter identification scheme that contains an information criterion for the determination of the optimal number of modes to be included in the identification process. The method is applicable to data measured at more than one point on the dam and to data that include both the amplitude response and the phase response. The method is applied to the vibration test data of two dams. Based on the results of these applications, the adequacy of the classical model and the non‐classical model is compared and the effect of the phase data on the parameter determination is discussed.