Innovation Approach for Modelling Compressive Strength of Fiber Reinforced Concrete Using Gene Expression Programming

Abstract

Recent advances in the field of construction materials have led to development of a variety of high performance concretes like steel fiber reinforced one (SFRC). It has been proved by many researches that the addition of steel fibers can improve various properties of concrete. The compressive strength of concrete (fc) is the main mechanical property in design of reinforced concrete structures. This paper deals with estimation of compressive strength of SFRC using gene expression programming (GEP) approach. In this regard, fine aggregate to cement ratio (FA/C), coarse aggregate to cement ratio (CA/C), water to cement ratio (W/C), fiber percentage (FP), superplastizer to cement percentage (SP/C) and fiber length to diameter ratio (L/D) were considered as the most important factors affecting the compressive strength of SFRC. To extract an accurate mathematical relationship from GEP approach, a comprehensive database was collected from literature with 115 mix design of SFRC. About 80% of the gathered database was used for training the model, while the rest was utilized for testing the model. The results indicate the acceptable performance of the developed GEP-based model, as the viewpoint of statistical parameters. The absolute fraction of variances for both training and testing datasets are more than 0.98 which approve a high correlation between the predicted values of the proposed model and the experimental results. At the end, a parametric study was carried out to investigate the efficiency of the developed model in predicting the tendency of compressive strength by changing the effective input variables.