Modelling Compressive Strength of Concrete Containing Crushed Glass and Natural Gravel Using Artificial Neural Network

Abstract

ABSTRACT

This research modelled the 28-day compressive strength of concrete containing crushed glass and Bida Natural Gravel (BNG) based on deep learning algorithm using the MATLAB neural network toolbox. A total of 240 (150mm × 150mm × 150mm) cubes were cast from 80 mixes generated randomly using Scheffe’s simplex lattice approach. The compressive strength was the mean 28-day strength of three cubes for each of the experimental points. The resulting batch for each mix was used as input data while the laboratory results for compressive strength was the output data for the ANN-model. The developed model will be able to predict the 28-day compressive strength of concrete containing 0% - 25% crushed glass as partial replacement for fine aggregate, water- cement ratio ranging from 0.45 – 0.65 and concrete grade M15 – M25. The architecture of the network contained 6 input parameters: water to cement ratio, water, cement, sand, crushed glass and BNG, 20 neurons in the hidden layer and compressive strength in the outer layer. The performance of the developed model was examined using Mean Square Error (MSE) and Correlation Coefficient (R). Results showed that 6:20:1 model architecture for compressive strength had an MSE values for training, validation and testing are: 0.15, 4.14, 1.15, 0.86 respectively. Regression values for training, validation and testing are: 80%, 65%, 85% and 75%. The study concluded that a shallow Neural Network architecture with 20 neurons in the hidden layer is sufficient for predicting the 28-day compressive strength of concrete.