Experimental Versus Numerical Shear Modelling of Clay Brick Triplet Specimens

Numerical analysis of masonry walls has been predominantly undertaken using the simplified micro-model, developed in the 1990s, and has started to be used in engineering analysis and design of historic masonry structures. However, this involves a zero-thickness interface to and is not useful for determining where in the masonry joint the line of failure propagates. Building the detailed micro-model requires defining the binding parameters of brick and mortar. This requires the fracture energies in both tension and shear, dilation angle, friction coefficients, and shear stress limit of the system. Shear testing permits the determination of friction effects and shear stress limits experimentally. Shear triplets were tested experimentally for cement-lime mortars mixed at 1:1:6 and 1:2:9 cement to lime to sand ratios by volume. Pre-compression loads of 0.2 MPa, 0.6 MPa and 1.0 MPa were used in order to determine compression effects. Four specimens were tested at each individual loading level and with each mortar mix. Increased compressive loading would result in inelastic deformation prior to the shear test. The interface model allowed specification of the shear stress limit, crack opening stresses, as well as tensile and shear fracture energies, although challenges remain in precisely specifying these quantities based on experimental results. It was possible to reproduce the failure pattern at the interface location for brick triplet tests, while using calibrated elastic moduli determined from experimental testing in compression, and also while using the original elastic moduli as determined from material tests. This work enabled the expansion of the detailed micro-model to model the full behaviour of masonry shear walls, with the compression and shear effects of the brick, mortar and bonding interface working in tandem while also being numerically distinct.