Font Size: a A A

Applications of statistics to minimize and quantify measurement error in finite element model updating

Posted on:2009-10-21Degree:M.SType:Thesis
University:Tufts UniversityCandidate:DiCarlo, ChristopherFull Text:PDF
GTID:2442390005951906Subject:Engineering
Abstract/Summary:
Finite element model updating is the process of modifying the parameters of a finite element model so that the model more accurately reflects the behavior exhibited during a non-destructive test. The test data can include displacements, tilts, and strains from static tests and mode shapes and natural frequencies from dynamic tests. The research utilizes a computer program developed at Tufts University called PARIS, short for PARameter Identification System. Finite element model updating can be a useful tool in bridge health monitoring.;This research applies statistical methods to the parameter estimation process. The formulation for the maximum likelihood estimator is developed and applied to PARIS. The Fisher information matrix is developed and used to aid in non-destructive test design. The inverse of the Fisher information matrix, the Cramer-Rao lower bound on variance, is also used to evaluate reliability of final estimates.;The methods developed in this work are applied to test data from a scale bridge model located at the University of Central Florida (UCF). The test data includes displacements, rotations, and strain measurements as well as mode shapes and natural frequencies. These data are used to update a finite element model of the structure.
Keywords/Search Tags:Finite element model, Mode shapes and natural frequencies, Fisher information matrix
Related items