The ONIX model: a parameter-free multiscale framework for the prediction of self-desiccation in concrete

The traditional approach for predicting self-desiccation is to simulate hygro-mechanics directly at the macroscale and to provide hydration-related inputs via phenomenological constitutive models. This manuscript presents instead a novel method that consists of obtaining inputs to such constitutive relations from direct simulations of cement hydration at the microscale, using a state-of-the-art simulator, namely the Cement Hydration in Three Dimensions (CEMHYD3D). This allows avoiding lengthy calibrations from experimental data. The prediction capabilities of the proposed model are demonstrated using experimental data of self-desiccation relevant to about 50 different mix designs of concrete, mortar and cement paste, with water to cement ratios ranging from 0.20 to 0.68 and silica fume to cement ratios from 0.0 to 0.39. The mixes are characterized by various cement chemical compositions, particle size distributions and Blaine finenesses, and the experiments span numerous time scales, from one week up to two years.