Comprehensive Bayesian structural identification using temperature variation. (15th June 2017)

• Record Type: Journal Article
• Authors: Jesus, Andre; Brommer, Peter; Zhu, Yanjie; Laory, Irwanda
• Is Part Of: Engineering structures. Volume 141(2017:Jun. 15)
• Page Start: 75
• Page End: 82
• Publication Date: 2017-06-15
• Subjects: Bayesian inference -- Model calibration -- Structural-identification -- Identifiability -- Temperature variation; Structural engineering -- Periodicals; Structural analysis (Engineering) -- Periodicals; Construction, Technique de la -- Périodiques; Génie parasismique -- Périodiques; Pression du vent -- Périodiques; Earthquake engineering; Structural engineering; Wind-pressure; Periodicals; 624.105
• DOI: 10.1016/j.engstruct.2017.01.060
• Languages: English
• ISSNs: 0141-0296

Abstract

Highlights: A temperature based Bayesian structural identification is presented. Inference of structural parameters converges to its true physical value. Feasibility relies on responses with high dependency of the calibration parameters. Abstract: A modular Bayesian method is applied for structural identification of a reduced-scale aluminium bridge model subject to thermal loading. The deformation and temperature variations of the structure were measured using strain gauges and thermocouples. Feasibility of a practical, temperature-based, Bayesian structural identification is highlighted. This methodology used multiple responses to identify existent discrepancies of a model, calibrate the stiffness of the bridge support and establish uncertainty of a predicted response. Results show that the inference of a structural parameter is successful even in the presence of substantial modelling discrepancies, converging to its true physical value. However measurements should have a high dependency on the calibration parameters. Usage of temperature variations to perform structural identification is highlighted.