Optimization of concrete mix design using genetic algorithms

Abstract

Most of the engineering construction projects of bridges, dams consume huge quantities of concrete and other materials. Economical design can be better achieved by optimising the quantities of materials without affecting functionalities of structures. Minimising the proportions or quantities of materials without affecting its functional characteristics in a concrete mix results in most cost efficient design process. This study presents the technique of optimisation of concrete mix design applying Genetic Algorithms using a developed MATLAB program. Concrete mix design was performed for different grades of concrete and water cement ratio. It is observed that the use of Genetic Algorithms resulted in economical mix by minimizing the cement content keeping the strength of concrete unaffected. The study results indicated that quantities of cement have been reduced by about 25-40 kg per cubic meter through mix design using GA technique. This resulted in about 6-10 percent reduction in quantities of cement for various cases of mix design considered in the study. Mix design performed using optimization techniques like GA proved to be efficient when compared to mix design using manual approach. Further, the models for predicting compressive strength under different proportions of materials can also be analyzed using GA approach presented in this study.