Premium
Probabilistic seismic demand assessment accounting for finite element model class uncertainty: Application to a code‐designed URM infilled reinforced concrete frame building
Author(s) -
Alam Mohammad S.,
Barbosa Andre R.
Publication year - 2018
Publication title -
earthquake engineering and structural dynamics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.218
H-Index - 127
eISSN - 1096-9845
pISSN - 0098-8847
DOI - 10.1002/eqe.3113
Subject(s) - latin hypercube sampling , masonry , probabilistic logic , structural engineering , uncertainty quantification , unreinforced masonry building , opensees , uncertainty analysis , frame (networking) , nonlinear system , fragility , incremental dynamic analysis , engineering , sensitivity analysis , seismic hazard , building model , seismic analysis , computer science , finite element method , monte carlo method , mathematics , civil engineering , statistics , simulation , machine learning , artificial intelligence , chemistry , telecommunications , quantum mechanics , physics
Summary Reliable and robust probabilistic assessment of structures requires explicit consideration of all relevant sources of uncertainty, both aleatory and epistemic. This paper proposes a formulation to incorporate model class uncertainty in probabilistic seismic demand assessment (PSDA) of structures, where model class uncertainty relates to the use of different structural analysis models used to predict the physical response of structural systems. The application of the proposed formulation is illustrated through the assessment of a recent code‐designed reinforced concrete (RC) frame building with unreinforced masonry (URM) infills, which is a prevalent form of construction worldwide. The model class uncertainty analyzed in this paper is related to the potential selection of one of three state‐of‐the‐art masonry infill strut models. In the application example, nonlinear static pushover (NSP) analyses and tornado sensitivity analyses are performed to identify the important parameters of infill strut models affecting the seismic response of the infilled RC frame. A hybrid stripe analysis (HSA) approach is adopted to perform nonlinear response history analyses (NRHAs) of the example RC frame. To account for relevant sources of uncertainty, latin hypercube sampling (LHS) is used to quantify the minimum number of realizations of model parameters necessary to achieve convergence of the coefficient of variation (COV) of the engineering response parameters assessed. Results of the NRHAs indicate that incorporating model class uncertainty significantly affects the estimation of uncertainties of the drift hazard demand.