Optimization of graphene nanoplatelet incorporation in cementitious composites using Taguchi and response surface method

Graphene nanoplatelets (GNPs) have shown potential to improve the performance of cementitious composites, but their effectiveness strongly depends on dosage and dispersion quality. This study investigates the combined influence of GNP dosage and ultrasonication time on the mechanical, durability, and microstructural properties of cementitious composites. A Taguchi L9 orthogonal array was used to design mixtures containing 0.01%, 0.05%, and 0.10% GNP by mass of cement, dispersed by ultrasonication for 5, 10, and 20 min. A control mixture without GNPs was also prepared. The mixtures were evaluated using 7- and 28-day compressive strength, water absorption, permeable voids, scanning electron microscopy, signal-to-noise ratio analysis, analysis of variance, and response surface modelling. The results showed that the mixture containing 0.01% GNP and 20 min ultrasonication achieved the highest 28-day compressive strength of 76.8 MPa, compared with 67.4 MPa for the control mixture. The same mixture also produced the lowest 28-day water absorption of 7.47%, compared with 8.35% for the control. The lowest 28-day permeable void content was obtained at 0.05% GNP and 5 min ultrasonication, indicating that strength and pore-network refinement were related but not always governed by identical mechanisms. SEM observations showed that low GNP dosage combined with adequate ultrasonication promoted a denser hydration matrix, whereas higher dosages produced heterogeneous regions and possible agglomeration. Statistical analysis confirmed that GNP dosage was the dominant factor controlling performance, followed by ultrasonication time and their interaction. The findings demonstrate that effective use of GNPs in cementitious composites requires a balanced combination of low dosage and sufficient dispersion energy rather than increased GNP content alone.