Digital image correlation applied to in situ evaluation of surface cracks upon curing of MgO-containing refractory castables

•DIC highlights damage growth by means of increased crack density and crack openings for networks on monitored surfaces, whereas traditional techniques only monitor Young's modulus changes.•Tomography scans showed that cracks initiate on the surface, and then propagate toward the center of the specimen.•DIC results indicate that after the crack density reaches its maximum value, the crack opening displacements continue to increase as in-depth propagation is still active. MgO particles are added to high-alumina castables to provide in situ spinel formation at high temperatures. However, the MgO hydration upon curing may damage the material because of localized volumetric expansion. Usually, damage is evaluated by the ex situ overall measurement of Young's modulus changes during processing via Impulse Excitation Techniques. In this paper, an experimental setup was designed to use Digital Image Correlation (DIC) as an in situ alternative to evaluate damage. Tomographic scans highlighted that all cracks initiated on the sample surfaces, and propagated in the bulk in an intergranular mode. Crack initiation and growth were assessed, for different temperatures, via Surface Crack Density (SCD) measurements, and Mean Crack Opening Displacement (MCOD) fields. They provided important insights into heterogeneous expansion phenomena and crack network quantification; for example, the SCD flattened while the overall damage was still increasing. The results attested the usefulness of DIC for in-situ quantification of ceramics cracking during processing.