Restrained shrinkage of masonry walls

State of the art computational rnechanics, in combination with experimental programmes have a lot to offer in providing insight, characterization of total behaviour and predictive ability of structural masonry. Here numerical research towards rationalizing masonry wall movement joint positioning and spacing is reported. Despite restrained shrinkage cracking being the dominant cause of damage in masonry buildings in The Netherlands [1], movement joint design remains an art, guided by thumb rules.
The behaviour of cementitious materials and specifically creep and shrinkage, depends strongly on their moisture content. For realistic analysis of these phenomena it is therefore imperative to know the moisture distribution in masonry at any time. There is evidence of weak coupling between moisture diffusion. the driving force behind masonry hygral shrinking, and mechanical stress and strain. except when cracks arise. The drying acceleration caused by cracks is ignored here and one-way coupling is assumed, namely the shrinking strain field dependence on pore humidity. The subsequent mechanical analysis is assumed to have no effect on the moisture distribution and thus the shrinkage strain, even in the event of cracking.
Greep is accounted for by an aging Maxwell chain in a novel combination of visco-elasticity and plasticity. This approach is chosen for the necessity to account for the orthotropic failure characteristics of masonry. being much stronger in tension parallel to the bed joints than in the orthogonal direction. An existing masonry material model, based in multi-surface compu-tational plasticity and accounting for orthotropic elastic and plastic behaviour, was enhanced to include visco-elastic behaviour