Parameter optimization and regression analysis for multi-index of hybrid fiber-reinforced recycled coarse aggregate concrete using orthogonal experimental design

• Published in: Construction & building materials Vol. 267; p. 121013
• Main Authors: Feng, Junjie, Yin, Guansheng, Tuo, Hongliang, Niu, Zhiqiang
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 18.01.2021
• Subjects: Hybrid fiber reinforced recycled coarse aggregate concrete; Modified multi-index synthetic weighted scoring method; Optimal combination; Orthogonal experiment; Reduced multivariate regression model; Orthogonal experiment; Optimal combination; Modified multi-index synthetic weighted scoring method; Hybrid fiber reinforced recycled coarse aggregate concrete; Reduced multivariate regression model
• ISSN: 0950-0618; 1879-0526
• DOI: 10.1016/j.conbuildmat.2020.121013

•rg of RCA is the crucial influence factor of mechanical properties of HFRRCAC.•Hybrid effect between HES and MPP fibers have little impact on SPI.•New evaluation method to assess the parameter combination was developed.•Reduced multivariate regression models have higher predictive ability. The optimum combination of parameters and the prediction models for four property indexes (PIs) of hook-end steel and macro-polypropylene hybrid fiber reinforced recycled coarse aggregate concrete through an orthogonal experimental design is presented in this paper. First, the analysis of variance was applied to study the influence of fifteen factors on all PIs. Then, the modified multi-index synthetic weighted scoring method in evaluating the parameter combination was developed and used to obtain the optimum performance for PIs. Finally, the multiple regression analysis was employed to propose multivariate regression models for the prediction of all PIs. The results show that the replacement ratio of recycled coarse aggregate is the prominent influence factor of PIs. Yet, the interactions between recycled coarse aggregate and fibers and the hybrid effect between fibers have little impact. The parameter combination with the highest score was obtained through the verified evaluation method. The mixed regression models are more accurate prediction and smaller residual standard deviation, and the reduced mixed regression models can optimize the predictive ability and well agree with experimental results. The evaluation method and prediction models can be extensively applied to the parameter optimization and property index prediction of multi-index system.