Prediction Model for the Compressive Strength of Green Concrete using Cement Kiln Dust and Fly Ash

Abstract

Abstract Integrating artificial intelligence in construction industry is a challenge that can help to move towards sustainable construction. Therefore, Artificial Neural Network (ANN), which is a computing system that simulates the human brain processes, can be helpful tool for prediction of the compressive strength of green concrete. Green concrete can be made using waste materials as a replacement portion of cement (supplementary cementitious materials) or aggregate that can benefit in the reduction of negative impacts on the environment and improve its compressive strength. This research aims to predict the compressive strength of green concrete that includes a ratio of cement kiln dust (CKD) and fly ash (FA), as industrial by-products, using artificial neural network technique and MATLAB software. The developed ANN model is based on the collected necessary information about the concrete components and compressive strengths from literature. Two models have been trained and tested. The first includes CKD in the concrete mix using 35 training samples with 3 hidden layers. While the second includes CKD and FA in the concrete mix using 42 training samples with 7 hidden layers. The results of both models showed a good prediction of the compressive strength of green concrete with error less than 10%. The benefits of this nondestructive approach may include preservation of natural resources, reduction of greenhouse gasses emissions, cost, time, and waste to landfill as well as saving energy.