Prediction of Blast Furnace Slag Concrete Compressive Strength Using Artificial Neural Networks and Multiple Regression Analysis

High performance concrete compressive strength modeling is a complex process. This study investigates the relationship between compressive strength of blast furnace slag concrete with its constituents and to predict blast furnace slag concrete compressive strength using two methods: (1) multiple regression analysis and (2) artificial neural networks. Results from study showed that the use of artificial neural networks in compressive strength modeling provides higher accuracy in predicting compressive strength of a given mix proportion. However, the multiple regression model is able to give an equation representing the relationship between the compressive strength of concrete with its inputs. Both compressive strength prediction models can be used as additional tools in the decision making of a blast furnace slag concrete mix design.