Multi-objective collaborative optimization of mechanical and shrinkage properties for engineered geopolymer composites based on orthogonal experimental design

To achieve the synergistic optimization of mechanical performance and shrinkage control in Engineered Geopolymer Composites (EGC), this study developed a multi-objective optimization method based on orthogonal experimental design. Four key factors, including fly ash to slag (FA/GGBS) ratio, alkali content, water-to-binder (W/B) ratio, and fiber volume fraction were selected to design 16 mix proportions. Their 28-day compressive strength, tensile strength, ultimate tensile strain, as well as drying shrinkage and autogenous shrinkage strains at different ages (7, 28, and 56 days) were measured. Range analysis indicated that the FA/GGBS ratio was the dominant factor across all properties, with the range values at 28 days being compressive strength 18.98 MPa, tensile strength 1.07 MPa, ultimate tensile strain 4.41%, drying shrinkage 6149.6 με, and autogenous shrinkage 371.6 με. Analysis of variance further revealed that the FA/GGBS ratio had highly significant effects on shrinkage at all ages, while fiber content exhibited significant influence in reducing shrinkage strains during mid to late ages. Subsequently, quadratic regression models for the five performance indicators were established based on response surface methodology. Weighted composite objective functions were constructed by incorporating weights assigned through the fuzzy analytic hierarchy process, followed by multi-objective optimization using the NSGA-II genetic algorithm to identify the optimal mix. Under the optimal mix, the predicted values of the five performance indicators were 71.15 MPa, 3.00 MPa, 7.37%, 6911.76 με, and 4228.24 με, demonstrating excellent strength, toughness, and shrinkage control capabilities. Finally, phase composition and interfacial structural characteristics of the optimal EGC mix were analyzed using X-ray diffraction and scanning electron microscopy. The findings provide an effective approach and theoretical support for performance regulation and design optimization of EGC materials. • A multi-objective optimization framework integrating orthogonal design, Fuzzy-AHP, RSM, and NSGA-II was developed for EGCs. • Range and ANOVA revealed FA/GGBS ratio as the dominant factor, with fiber content crucial for shrinkage control. • The optimal mix achieved compressive strength of 71.15 MPa, tensile strength of 3.00 MPa, tensile strain of 7.37%, and low shrinkage values. • XRD and SEM analyses confirmed dense C-(A)-S-H gels and strong fiber–matrix bonding as the microstructural basis.