Experimental Study and Artificial Intelligence Modeling of Single-Component and Multi-Component Fiber-Reinforced Self-Compacting Geopolymer Concrete Based on Ground Granulated Blast Furnace Slag, with Evaluation of Fresh, Mechanical, and Shrinkage Properties

Abstract

The limitations of raw materials for producing Portland cement concrete in construction processes, greenhouse gas emissions from its production, and the high costs of alternative materials have become an intriguing challenge for engineers and researchers. In this regard, fiber-reinforced self-compacting geopolymer concrete based on slag and fly ash has been developed as an eco-friendly structural mix by replacing aluminosilicate-based pozzolans with cement. This study designed and produced a stable, self-compacting mix with structural properties, incorporating polypropylene and steel fibers. The developed mixtures were prepared by varying the alkaline concentration, pozzolan content, and fiber volume and were evaluated for fresh, mechanical, water absorption, shrinkage, and microstructural properties. The study was conducted at two alkaline molarities of 8 and 12. Results indicated that increasing the molarity from 8 to 12 improves the microstructure of slag-based self-compacting geopolymer concrete, enhancing its density, uniformity, and mechanical properties. These changes lead to increased overall strength and durability of the concrete, although they may raise the cost and consumption of the alkaline solution. Additionally, steel fibers contribute to reinforcing the concrete structure by forming internal resistant networks and stress distribution, effectively enhancing the tensile and flexural strength. These fibers are particularly effective in resisting deep cracks and heavy loads. Furthermore, the study employed the Decision Tree Model for compressive strength modeling, achieving prediction accuracy of over 88% using the developed relationships.