Quantification of Reinforcement Debonding in Damaged Mortar via Digital Volume Correlation

Background. Debonding between a cementitious material and a reinforcement is a mechanical phenomenon of great interest. It cannot be quantified directly through standard tests since it occurs within the material bulk. Objective. The goal is to develop an experimental method for quantifying debonding during in-situ pull-out tests that also induce damage in the mortar matrix. Method. A 1/50 scale foundation model is subjected to a pull-out test in an X-ray tomograph. A finite-element-based Digital Volume Correlation analysis with mechanical regularization is conducted based on a three-dimensional mesh constructed to reproduce the geometry of the foundation and reinforcement. Results. Heterogeneous regularization with a single-node mesh has little effect on the correlation residuals. Using split nodes to describe the interface drastically reduces the correlation residuals in the reinforcement. If cracking occurs in addition to debonding, introducing a heterogeneous regularization based on damaged elements improves the quantification of debonding. Conclusion. By splitting the nodes at the interface and localizing regularization in damaged elements, the reinforcement and mortar kinematics is better captured and thus debonding as well.