Numerical investigation of non-linear equivalent-frame models for regular masonry walls

• Published in: Engineering structures 2018-10, Vol.173, p.512-529
• Main Authors: Siano, Rossella, Roca, Pere, Camata, Guido, Pelà, Luca, Sepe, Vincenzo, Spacone, Enrico, Petracca, Massimo
• Format: Article
• Language: English
• Subjects: Aseismic buildings; Boundary conditions; Buildings; Comparative studies; Computer programs; Computer simulation; Configurations; Earthquake damage; Earthquake effects; Enginyeria civil; Equivalence; Equivalent-frame models; Estructures de murs; Finite element analysis; Finite element method; Masonry; Masonry structures; Masonry structures URM walls Equivalent-frame models Seismic vulnerability Non-linear static analysis; Materials i estructures; Mathematical models; Non-linear static analysis; Nonlinear analysis; Numerical analysis; Open source software; Piers; Plastic properties; Plasticity; Seismic engineering; Seismic response; Seismic vulnerability; Sismologia; URM walls; Walls; Àrees temàtiques de la UPC
• ISSN: 0141-0296; 1873-7323; 1873-7323
• DOI: 10.1016/j.engstruct.2018.07.006

•Evaluation of EFM accuracy by testing URM regular walls in comparison with FE models.•The comparative study assumed as reference a micro-modelling finite element approach.•Two modelling criteria for defining piers effective height has been accounted for.•The influence of walls geometry on EFM accuracy is studied by non-linear analyses. The accuracy of the Equivalent Frame Method (EFM) in modelling the seismic non-linear behaviour of unreinforced masonry (URM) buildings is investigated for regular walls (i.e. walls with regular openings’ distribution) with different pier-to-spandrel geometrical relations. The developed EFM is composed of pier and spandrel elements with spread plasticity to simulate the flexural behaviour and lumped plasticity to simulate the shear behaviour. The investigation focuses on checking, by means of comparison with Finite Element Model (FEM) assumed as reference, the applicability of EFM to existing buildings. These structures are often characterized by geometrical schemes difficult to be represented by ideal frames. To point out the role of the geometrical configuration, the numerical results provided by the two modelling approaches are compared for different representative cases of regular walls characterized by pier-spandrel configurations rather typical in existing URM buildings. In addition to the innovative EFM approach, based on a fiber discretized beam element, also a more traditional approach, based on beam elements with lumped plasticity, is included in the comparative study. The two different EFM approaches were implemented in the software Midas GEN © [44], while an open source software was used to implement the FEM (Kratos Multiphysics [59–60]). All the models were used to perform static non-linear analyses under equivalent loading and boundary conditions. The evaluation of EFM and FEM is derived from a comparative simulation of a two-storey URM wall experimentally tested by other researchers. Two alternative approaches are assumed for the definition of piers’ effective heights in the EFM, i.e. the models proposed by Dolce [1] and Augenti [2]. The results demonstrate that remarkable differences may be detected in EFM and FEM predictions of the shear capacity and damage mechanisms as a function of pier-spandrel geometrical configurations. This result highlights the need for a cautious application of EFM to existing URM structures.