Forecasting the impact strength of the pozzolan concrete with adaptive fuzzy logic models

Abstract

This objective research to study the possibility of used properties pozzolan material for concrete cement the ternary blend by fly ash, metakaolin and silica powder by replacement cement partially on the strength for Flexural and impact by replace Portland cement. Consideration of concrete by varying the ratio there are three values are (w/d) of 0.38, 0.55 and 0.8 follow on standards ACI 211.1-91 and variable amounts of replacement cement some. The ternary blend of the fly ash, metakaolin and silica powder of percentage 0, 15, 20 and 25 and by weight of cement. and the control specimens (w/b) without contain fly ash, kaolin and silica powder, the slump test in the range of 3-5 cm and the use super plasticizer to control the slump test of concrete. By casting specimens of cylindrical diameter 10x20 cm cube 10x10x10 cm and the beam size 10x10x35 cm, after curing 28 days and test compressive, tensile and flexural strength of concrete, all specimens The results showed that The increase will help improve the quality of the concrete physical properties 35 percent up on the cement with clay, 15 percent will have the strength to most of the compression is reduced. The increased of compressive and Tension strength as the percentage decrease when replacing cement with fly ash, 35 percent up on replacing cement with clay to pull the maximum percentage of 10 bending when replacing cement with clay at a ratio of 10 percent is paid bending. The maximum value of fly ash, the flexural increased replace cement in the ratio of 50 per cent are increased the concrete with the most and the last place of cement. As a result of the porosity of the concrete with lower amounts of tightness concrete has increased. Fuzzy logic model is presented for predicting the strength for Flexural and impact by replace Portland cement. The result of the study from experimental case, the graph of membership function are present (low, medium and high). The results showed that the model Fuzzy Logic. Can be used fuzzy logics to predict as well