Multivariate regression models in estimating the behavior of FRP tube encased recycled aggregate concrete

This study applied newly developed multivariate statistical models to estimating the mechanical properties of recycled aggregate concrete cylinder encased by fiber reinforced polymer (FRP). Two different types of RFPs were applied, namely flax FRP and polyester FRP. Ten independent variables were predefined including the FRP type and cylinder size. It was found that several mixed models outperformed the traditional linear regression approach, based on the accuracy and residual value distribution. Individual factor analysis indicated that the fiber thickness and layer number had more significant impacts on the strength and strain of FRP-encased concrete's transitional point, compared to their impacts at the ultimate state.