Hurricane loss estimation in wood-frame buildings using Bayesian model updating: Assessing uncertainty in fragility and reliability analyses. (15th March 2017)

• Record Type: Journal Article
• Authors: Mishra, Spandan; Vanli, O. Arda; Alduse, Bejoy P.; Jung, Sungmoon
• Is Part Of: Engineering structures. Volume 135(2017:Mar. 15)
• Page Start: 81
• Page End: 94
• Publication Date: 2017-03-15
• Subjects: Bayesian model updating -- Hurricane loss -- Fragility -- Monte Carlo simulation; 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.2016.12.063
• Languages: English
• ISSNs: 0141-0296

Abstract

Highlights: Bayesian model updating method is proposed to quantify uncertainty in fragility of wooden structures. Bayesian regression model with conjugate priors is employed to find the posterior capacity. Importance sampling is used for efficient computation of failure probability through domain of wind speed. Total expected loss is calculated using fragility functions while considering model uncertainty. Abstract: This paper proposes a Bayesian model updating method for quantification of uncertainty in the analytical capacity predictions of roof-to-wall connections of wooden structures. The ultimate goals are to construct fragility models for various mitigation options accounting for uncertainties due to model discrepancy and experimental error and to determine the most cost-effective option under a potential hurricane in terms of aggregate loss estimates based on total probability of failure. The proposed approach is applied to both hurricane-clips (mitigated specimen) and toenails (unmitigated specimen) and the exceedance probability of different aggregated loss scenarios are found using both the proposed method and the existing lognormal approach. The advantage of the two-stage Bayesian model is that it provides the uncertainty bound on the exceedance probability which gives a measure of the factors unaccounted for in the capacity estimation process.