Premium
Remaining stiffness estimation of buildings using incomplete measurements
Author(s) -
Eskew Edward L.,
Jang Shinae
Publication year - 2017
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.1899
Subject(s) - stiffness , benchmark (surveying) , relevance (law) , frame (networking) , structural health monitoring , least squares function approximation , computer science , engineering , event (particle physics) , phase (matter) , structural engineering , algorithm , mathematics , statistics , mechanical engineering , physics , geodesy , quantum mechanics , estimator , political science , law , geography , chemistry , organic chemistry
Summary After hazardous events, it is important to be able to quickly identify the remaining stiffness of affected structures for condition evaluation. Model updating can be used to update structural models to reflect current conditions based upon experimental measurements. Direct model updating is a simple and quick method of damage detection, but does not guarantee physical relevance. Least‐squares optimization can be used to accurately identify damage with physical relevance, but needs more measurements then updating parameters in order to produce an accurate solution. However, after an extreme event sensors on the structure may be damaged, creating a scenario with limited measurements which can render optimization techniques incapable of assessing the remaining stiffness. To address this issue, this paper proposes a two‐phase method to localize and then quantify the remaining stiffness of the structure. Direct model updating with limited measurements is used to localize potential damage to a subset of parameters, and a least‐squares optimization using the localized parameters is used to quantify the remaining stiffness in the structure. Numerical simulations using a simplified model based upon the phase I IASCE‐ASCE structural health monitoring benchmark problem with missing first floor sensors have been employed to demonstrate, and experiments using a five‐story steel frame structure are conducted to validate the methodology. Copyright © 2016 John Wiley & Sons, Ltd.