Numerical Study on the Behavior of Natural and Recycled Aggregate Concretes

This thesis investigates the mechanical behavior of natural and recycled aggregate concretes (NAC and RAC) using a multi-phase numerical approach to predict performance. Growing interest in sustainable materials has led to using recycled concrete aggregates (RCA) to replace natural aggregates, creating Recycled Aggregate Concrete (RAC). Due to their different structural components, RAC and NAC exhibit varying mechanical properties, particularly in strength and durability. This study validated tension and bending (flexural) test models by comparing numerical results with similar models from the literature. Following validation, a multi-phase approach modeled NAC and RAC for the bending (flexural) and tension tests, providing a detailed analysis of each component’s role in concrete behavior. The NAC model used a three-phase system of natural aggregates, mortar, and interfacial transition zones (ITZ). In contrast, the RAC model included five phases by introducing old mortar and ITZ as additional RCA elements. Calibration procedures determined the mechanical properties of each phase in NAC and RAC, using a combination of empirical data and assumptions where direct measurements were not feasible. Parametric studies examined how variations in phase characteristics and aggregate properties affect concrete behavior. Results indicate that RAC exhibits lower tensile and compressive strength than NAC, attributed to the inherent weakness of RCA components. The calibrated model enables the prediction of RAC behavior using basic mechanical properties of RCA, and it also provides estimates for the mechanical properties of phases, like ITZ and old mortar, which are challenging to measure experimentally. The multi-phase modeling approach effectively captured the unique behavior of RAC, especially highlighting the impact of old ITZ and mortar phases on overall strength. Key findings underscore the importance of these phases in determining concrete strength and point to limitations due to assumptions about RCA properties, indicating a need for further experimental validation. Advisor: Seunghee Kim