Macro and micro-scale modelling of masonry structures using the Discrete Element Method

Masonry is a heterogeneous anisotropic material, which is composed of units (e.g. bricks, stones, blocks etc.), bonded together with or without mortar. It is probably the oldest building material that is commonly used today. Although masonry is easy to construct, its mechanical behavior is non-linear and thus complex to understand. The need to predict the in-service behavior and load carrying capacity of masonry structures has led researchers to develop several computational strategies and tools that are characterized by different levels of complexity. Such models range from considering masonry as an anisotropic continuum (e.g. macro-models) to the more detailed ones considering masonry as an assemblage of units and mortar joints (e.g. micro-models) (Lourenco 1996). Although, the macro-modelling approach based on the Finite Element Method (FEM) of analysis is an efficient approach for the purpose of design or the understanding of the global behavior of a structure, the approach is lacking success since progressive failure or collapse mechanism cannot be obtained; since failure is smeared out in the continuum. An alternative and promising approach which can represent the discrete nature of masonry is presented by the Discrete Element Methods (DEM). The approach was initially developed by Cundall (1971) to model sliding of rock masses in which failure occurs along their joints. It was later applied to model the mechanical behavior of masonry structures (Lemos 2007; Sarhosis et al. 2016). This paper aims to highlight the ability of DEM to simulate the mechanical behavior of masonry at different scales. Two examples are investigated including: a) the mechanical behavior of a masonry prism subjected to direct compression; and b) the mechanical behavior of a masonry arch bridge with backfill material. In both cases results are compared with experimental findings and comparisons are made. Reliable prediction of masonry strength can allow one to reduce the costly and timely experimental testing, gain a better insight into the structural capacity of the mechanisms which drive failure of material and structure and avoid the reliance on conservative empirical formulas.