Rheological Optimization of Self Compacting Concrete with Sodium Lignosulfate Based Accelerant Using Hybrid Neural Network-Genetic Algorithm

Abstract

One of the most useful innovations in concrete technology is Self Compacting Concrete that has the ability to flow efficiently and maintain material homogeneity. The rapid change in the behavior of concrete due to accelerating admixtures can significantly affect the workability properties of the mixture and reduce its ability to flow efficiently. To describe the influence of superplasticizers blended with accelerant on the rheological properties of SCC, several mixtures were tested for Slump Flow, L-Box, and Screen Stability tests. Artificial neural network was used to obtain a model describing the constitutive relationships between the material components and workability parameters of SCC and was optimized using Genetic Algorithm. Results showed that ANN was able to establish the relationship of rheology to the concrete material components and GA derived the optimum proportion for best rheological performance. Most of the design samples of SCC with blended superplasticizer and sodium lignosulfate accelerant were not able to perform well in the flowing ability due to inefficiency of the fresh SCC to flow. The increasing dosage of accelerant however rendered strong stability between the concrete particles allowing the SCC samples to resist segregation and maintain material homogeneity.