Prediction of Compressive Strength and Workability Characteristics of Self-compacting Concrete Containing Fly Ash Using Artificial Neural Network

Abstract

This study aims to propose an artificial neural network (ANN) model for predicting the properties of self-compacting concrete (SCC). SCC has enhanced properties such as very high workability and it can go through very tight spaces between reinforcements without any application of vibration. To get the desired strength and workability, it is necessary to understand the parameters determining the nature and properties of SCC and the relationships involved among those parameters. In this study binder content, water to binder ratio, fly ash percentage, coarse aggregate, fine aggregate, and superplasticizer content are chosen as input parameters, and output results from the model are slump flow value, L-box ratio, V-funnel time, and compressive strength. An ANN model is constructed and its architecture is selected by evaluating the performance of a network with a different number of neurons for the optimum results. Then this model is trained, tested, and validated through a database of experimental test results gathered from various literature. The accuracy of this model is evaluated by evaluation matrices such as R and MSE. To check the efficiency, the current model comparison was made with an existing data envelopment analysis model (DEA). © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.