Prediction of Flexural Strength of Portland–Composite Cement Mortars Substituting Metakaolin Using Fuzzy Logic

Abstract

In this study, Fuzzy Logic models have been introduced to predict flexural strength values of cement mortars. For this purpose, reference cement mortar containing only Portland–composite cement, and mixtures having metakaolin replacing 5, 10, 15 and 20% by weight of the Portland–composite cement were produced. The mortars’ flexural strength values were established at 2, 7, 28 and 56-day with standard cement test. In addition, Fuzzy Logic prediction models were created by using fuzzy triangular number coefficients and Gauss membership function to predict flexural strength of cement mortars. Subsequently, experimental with fuzzy results are compared. Accordingly, the correlation coefficient of flexural strength of cement mortars for fuzzy triangular number coefficients and Gauss membership function were found 0.84 and 0.87, respectively. These results show that between experimental and fuzzy results are a good harmony, and can be successfully applied in civil engineering applications.