Numerical homogenization-based seismic assessment of an english-bond masonry prototype: Structural level application

A multiscale strategy is evaluated at a structural level for the analysis of unreinforced masonry structures. The mechanical characterization of the masonry is deduced from homogenization-based micro-scale finite element (FE) models. The derived data are here employed at a structural level via a discrete FE model. The discrete FE model is composed of quadrilateral rigid plates interconnected through vertical and horizontal interfaces. On the interfaces, between adjoining discrete elements, a model that accounts for the in- and out-of-plane behavior of masonry, with damage and plasticity, is adopted. Such interfaces represent the material pre- and post-peak regimes, its orthotropy, and, depending on the micro-model assumed, account by three-dimensional shear effects that are especially important for multi-leaf walls and complex regular textures. The discrete model has been implemented in an advanced structural analysis software where powerful built-in features as the arc-length method, line-search algorithm, and implicit or explicit solver schemes are available.The multi-scale model is applied for the dynamic study of a small English-bond masonry house prototype subjected to a series of consecutive earthquake records. Detailed comparisons between the experimental and numerical data are presented, including the results obtained through a continuous total strain rotating crack model. Quasi-static and dynamic analyses are conducted. Results demonstrate that when enough experimental information is available on the masonry components under tension, shear, and compression regimes, the approach predicts well the seismic structural response in terms of time-history displacements, seismic capacity, and damage patterns. The required computational cost (CPU time) is very attractive.This work was supported by FCT (Portuguese Foundation for Science and Technology), within ISISE, scholarship SFRH/BD/95086/2013. This work was also partly financed by FEDER funds through the Competitivity Factors Operational Programme - COMPETE and by national funds through FCT – Foundation for Science and Technology within the scope of the project POCI-01-0145-FEDER-007633