Abstract
Calcium aluminate cement-based castables were developed in the early 1990s for the metallurgical and petrochemical industries, exhibiting exceptional mechanical resistance when heated over 1000 °C. In typical operation conditions, they withstand compressive stresses due to high temperatures and mechanical loads. The extraordinary material performance has led to interest in using these materials for developing building protection systems against fires and explosions. This application requires structural reinforcement to resist tensile stresses in the concrete caused by accidental loads, making the bonding of reinforcement crucial. The different temperature expansion properties of the castables and reinforcement steel further complicate the bonding mechanisms. This manuscript belongs to a research project on developing refractory composites for civil infrastructure protection. In previous studies, extensive pull-out tests evaluated various combinations of refractories and reinforcement types to determine the most efficient candidates for refractory composite development. Thus, this study employs ribbed stainless Type 304 steel bars and a conventional castable, modified with 2.5 wt% microsilica for a 100 MPa cold compressive strength. It uses the previous pull-out test results to create a numerical model to predict the bonding resistance of the selected material combination. Following the composite development concept, this experimentally verified model defines a reference for further developing refractory composites: the test outcome of a new material must outperform the numerical prediction to be efficient. This study also delivers an empirical relationship between the castable deformation modulus and treatment temperature to model the reinforcement pull-out deformation in the composite heated up to 1000 °C.