Premium
Comparative study of vibration‐based parametric identification techniques for a three‐dimensional frame structure
Author(s) -
Antonacci Elena,
De Stefano Alessandro,
Gattulli Vincenzo,
Lepidi Marco,
Matta Emiliano
Publication year - 2012
Publication title -
structural control and health monitoring
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.587
H-Index - 62
eISSN - 1545-2263
pISSN - 1545-2255
DOI - 10.1002/stc.449
Subject(s) - modal , identification (biology) , parametric statistics , finite element method , vibration , frame (networking) , parametric model , computer science , minification , modal analysis , process (computing) , system identification , calibration , experimental data , inverse , algorithm , engineering , data mining , structural engineering , mathematics , statistics , acoustics , measure (data warehouse) , chemistry , biology , operating system , telecommunications , geometry , programming language , botany , physics , polymer chemistry
SUMMARY The entire structural identification process, starting from measurement of the experimental response to ambient vibration, and concluding with the assessment of a representative dynamic model, depends upon an informed selection of appropriate data treatment techniques. The paper compares different procedures for identifying both modal and physical models. Four approaches (EFFD, ERA, SSI, and TFIE) for output‐only modal identification are discussed. Their performance is evaluated using experimental data extracted from the ambient vibration response of a three‐dimensional frame. The discussion of the identification results throws light on the intrinsic characteristics of each procedure, while the particular features of the structure, presenting unexpected local modes, demonstrate the necessity of a hierarchical treatment of the available information, based on a solid engineering knowledge. The calibration of a physical model is approached using an exact inverse procedure, based on a parametric analytical model, or automatic techniques of error minimization, based on finite element multi‐models. The achievement of a good relative agreement, measured through performance‐based indexes, increases the general confidence on the results independently obtained from each modal identification and model updating method. Copyright © 2011 John Wiley & Sons, Ltd.