Fracture and brittle behavior of lightweight self-consolidating alkali activated concretes: a Taguchi-ANOVA approach

The increasing use of conventional cement-based concrete has raised serious concerns regarding high carbon emissions and resource consumption, necessitating the development of sustainable structural materials. In lightweight concrete systems, limited understanding of fracture behavior and brittleness, particularly in alkali-activated matrices, restricts broader structural application. This study aims to evaluate the fracture resistance and brittleness characteristics of Self-Consolidating Lightweight Alkali-Activated Concrete (SLAC) incorporating pre-treated expanded clay aggregate. The effects of fly ash-to-binder ratio, maximum aggregate size, and notch-to-depth ratio on fracture behavior were systematically investigated. A Taguchi-based experimental design coupled with regression modeling and validated through Analysis of Variance (ANOVA) was employed to identify the most influential factors governing fracture performance. Weight-compensated fracture energy method, two-parameter fracture method, and MATLAB-assisted fractal dimension analysis were used to quantify fracture energy, stress intensity factor, and brittleness indices, respectively. A closed-form predictive model was developed to estimate the fracture toughness from fracture load, improving prediction accuracy and reducing experimental effort. Results revealed that optimal mix F0-M16-N0.2, with 0 Fly ash-to-binder ratio, 16 mm Maximum aggregate size, and N/D ratio of 0.2, exhibited the highest fracture energy (117.5 N/m) and stress intensity factor (34.2 kN/mm). While mix F0.5-M16-N0.35 with increasing fly ash-to-binder ratio enhanced ductility (CTODc-0.029 mm, fractal dimension-1.124). Analysis of variance identified aggregate size as dominant factor influencing fracture parameters. Microstructural analysis revealed a refined and compact matrix, which collectively influences fracture toughness through enhanced stress transfer and crack path tortuosity. Regression models demonstrated reliability in optimizing eco-efficient mix designs for sustainable infrastructure applications.