Hybrid Neuro-Particle Swarm Model For Carbonation Depth Forecasting In Recycled Aggregate Concrete

Forecasting of carbonation depth in the use of Recycled Concrete Aggregate (RCA) as partial replacement to Natural Coarse Aggregate (NCA) is investigated in this study. Carbonation depth in Recycled Aggregate Concrete (RAC) is predicted by using Neural Network and optimizing the results at the same time. Neural Networking is of growing interest that utilize multiple layers considering various input parameters in modeling. Existing mathematical models that predicts carbonation depth only uses a limited number of influencing factors and several simplified ideal assumptions. With this, a hybrid Neural Network – Particle Swarm Optimization model using MATLAB is utilized to consider various important parameters in predicting the carbonation depth in RAC. Previous experiments exhibited that reaching 50 percentage of RCA replacement shows a decreasing compressive strength, thus, this study is confirmed by a carbonation experiment utilizing only 30, 40 and 50 percent replacements of RCA by weight.