Behrooz Ahmadi-Nedushan - Assistant professor, Department of civil engineering , yazd university

This article deals with use of advanced multivariate statistical techniques for prediction of compressive strength of concrete using ingredients proportions. The parameters representing the concrete are normally highly correlated. Different multivariate statistical techniques appropriate for analysis of multicollinear data are discussed, and efficiency of stepwise regression in analyzing multicollinear data set was investigated. Principal component regression was used to analyse the multicollinear data and to build a predictive model of 28-day compressive strength of concrete. The efficiency of the proposed method is verified using an actual concrete mix proportioning dataset. Results indicated that principal component regression in very efficient in prediction of concrete strength using mix proportion data.

Concrete, Compressive strength, Estimation, Mixing, Multivariate