A probablistic framework using a discrete Fe-based homogenized model for the in- And out-of-plane analysis of masonry structures

Probabilistic-based type of analysis seems to constitute a state-of-art approach for earthquake engineers. Such strategy is, in the analysis of potential retrofitting measures for a building (or typology), needed for the development of vulnerability or fragility curves. This information is recognize...

Full description

Bibliographic Details
Main Author: Silva, Luís C. (author)
Other Authors: Milani, Gabriele (author), Lourenço, Paulo B. (author)
Format: conferencePaper
Language:eng
Published: 2019
Subjects:
Online Access:http://hdl.handle.net/1822/67865
Country:Portugal
Oai:oai:repositorium.sdum.uminho.pt:1822/67865
Description
Summary:Probabilistic-based type of analysis seems to constitute a state-of-art approach for earthquake engineers. Such strategy is, in the analysis of potential retrofitting measures for a building (or typology), needed for the development of vulnerability or fragility curves. This information is recognized to be extremely valuable within the decision process, since it allows the evaluation of the collapse probability (or other defined limit state) of a given structure under a given intensity measure. The computational development integrated with the advent of proper working methodologies led to the increasing of research in the field. Nevertheless, the current numerical FE-based modelling approaches are still cumbersome for the dynamic analysis of the masonry behavior. In this scope, a novel stochastic methodology, based on the Latin Hypercube Sampling, using homogenization concepts for the masonry characterization and employing a discrete-FE based model is herein presented. Both running- and English-bond masonry textures are suitable to be studied, being the system uncertainty considered on parameters affecting the material, mechanical, geometrical and loading conditions. The approach has been implemented on an English-bond masonry mockup tested on a shaking table in LNEC. An incremental dynamic analysis has been performed and fragility curves derived. It has been shown that the strategy is effective, accurate and very attractive from a computation point of view.