Structural response predictions compared to material property estimates for structural integrity assessment under operational uncertainty. (February 2018)

• Record Type: Journal Article
• Authors: Lee, Kyunghoon
• Is Part Of: Computers & structures. Volume 196(2018)
• Page Start: 49
• Page End: 62
• Publication Date: 2018-02
• Subjects: Structural integrity assessment -- Operational uncertainties -- Bayesian inference -- Certified reduced basis methods; Structural engineering -- Data processing -- Periodicals; Electronic data processing -- Structures, Theory of -- Periodicals; 624.171
• DOI: 10.1016/j.compstruc.2017.10.012
• Languages: English
• ISSNs: 0045-7949

Abstract

Highlights: Structural integrity assessment was carried out under operational uncertainties. Certified reduced basis methods were utilized for rapid yet rigorous simulation. Bayesian MCMC inference was performed with forward model error and measurement noise. Response predictions were found more effective than material properties. Abstract: To examine structural integrity in consideration of operational uncertainties, we utilized structural response predictions obtained by transmitting estimates of structural system parameters into a simulation representing the actual system. For this purpose, we resolved inverse parameter estimation by Bayesian inference and constructed a forward model using certified reduced basis methods. For demonstration, we applied the proposed assessment strategy to two isotropic structural systems under static, linear elastic deformation: a beam and a plate subject to bending and shearing dominant loadings, respectively. Numerical investigation with emulated damage cases showed that structural response predictions were superior to material property estimates for inspecting structural integrity, particularly when operational uncertainties were not ignorable.