Modeling and analysis of fiber-reinforced high-performance concrete strength prediction based on nonlinear programming

Abstract

In this study, the nonlinear programming software Design-Expert and Lingo were used to establish a mathematical model between the fiber volume fraction and the strength of high-performance concrete for steel and polypropylene fibers in single or mixed concrete. The interaction of the two kinds of fiber in the matrix was considered. The results show that the fiber-reinforced high-performance concrete strength model conformed to the quadratic and cubic functional relation. The significance analysis of the model regression variance was performed, showing that the amount of polypropylene fiber had no significant effect on splitting strength and flexural strength of concrete while, steel fiber and hybrid fiber affected the compressive strength significantly. Additionally, polypropylene fiber helped increase the compressive strength of concret; and steel fiber had a greater effect on improving splitting tensile and flexural strength. The correlation between the predicted and tested strength of both models was demonstrated by SPSS statistical analysis, with greater significance when compensating for the discrete points of the predicted strength. A comparative analysis of existing strength prediction models verified the model's applicability for fiber-reinforced fiber-performance concrete. By modifying the discrete points, the correlation between the experimental values of compressive, splitting and flexural strength of concrete and the predicted values of the Pearson coefficient increase by 14.19%, 32.78% and 9.58%, respectively. The ratio value of predicted strength to tested strength is close to 1, while a large error appeared when the existing models were used for fiber-reinforced high performance concrete.