Prediction of sulfate resistance of cements produced with GBFS and SS additives using artificial neural network

Abstract

Concrete structures built on sulfate rich soil or wetland, or directly exposed to seawater are subjected to sulfate attack, which might be critical, as the durability of concrete is highly dependent on its resistance against sulfate compounds. The objective of this study is to develop a methodology for the prediction sulfate resistance capabilities of sulfate resistance of mortars prepared with cements incorporating granulated blast-furnace slag (GBFS) and steel slag (SS) as partial replacement of Portland cement clinker in different ratios. Three different combinations of GBFS and SS were utilised to partially replace Portland cement clinker at various proportions from 20% to 80%. Parameters such as specific surface, specific gravity, volumetric expansion, Vicat setting time, compressive strength, sulfate resistance and durability against high temperature were investigated on the produced cement samples. Furthermore, experimental results were also obtained by building models in accordance with the artificial neural network (ANN) technique to predict the sulfate resistance of cements. The results showed that ANNs can be successfully used to model the relationship between the sulfate resistance and each of the observed parameters. Keywords: granulated blast-furnace slag; GBFS; steel slag; durability; sulfate resistance; artificial neural network; ANN.