Implementation of Soft Computing Techniques in Forecasting Compressive Strength and Permeability of Pervious Concrete Blended with Ground Granulated Blast-furnace Slag

Abstract

Urban expansion and infrastructure development have exacerbated environmental issues by creating impermeable layers on the earth's surface, resulting in flash floods and reduced groundwater levels. These problems can be alleviated by using pervious concrete to enhance pavement drainage capacities. However, pervious concrete has limited applications due to its lower strength properties, which are attributed to its mix proportions featuring minimal fine aggregate quantities and an open-graded mix. This study examines the impact of incorporating Ground Granulated Blast-furnace Slag (GGBS) as a supplementary cementitious material in pervious concrete on its strength, drainage capabilities, and water absorption. Further, Artificial Neural Networks (ANN) were used to predict the mechanical and permeability properties of pervious concrete mixes with varying GGBS proportions. The study's results indicate that using GGBS as a 35% partial cement replacement with 10 mm aggregates significantly increases compressive and flexural strength by 28% and 20%, respectively. While permeability values were slightly reduced, they remained within acceptable limits for drainage properties. The developed ANN models outperformed the traditional MLR model, serving as a viable substitute logical tool for forecasting strength as well as permeability. Ultimately, adding GGBS to pervious concrete not only enhances strength but also contributes to environmentally friendly construction practices.

Keywords

• artificial neural networks
• ground granulated blast-furnace slag (ggbs)
• pervious concrete
• supplementary cementitious material