Multi-objective optimisation of glass powder-modified carbon fibre cement-based sensors for enhanced mechanical and electrical performance

The development of sustainable carbon fibre cement-based sensors (CF-CBS) is constrained by the need to balance mechanical properties with sensing performance. This study develops a novel glass powder (GP)-modified CF-CBS and employs Response Surface Methodology (RSM) to optimise workability, mechanical performance, and electrical conductivity (EC) while enhancing sustainability. The face-centred central composite design within the RSM framework was used to evaluate the effect of GP and its interactions with the sand-to-cement ratio, carbon fibre (CF) concentration, and methylcellulose (MC) on the workability, mechanical performance, and EC. Scanning Electron Microscopy (SEM) images were used to analyse the microstructure, with a focus on morphology and CF dispersion. The GP-modified CF-CBS models exhibited high coefficients of determination (R²) of 91.5%, 95.8%, 94.3%, and 87.9% for workability, 28-day compressive strength, 28-day flexural strength and electrical conductivity (EC), respectively. ANOVA confirmed the reliability of all responses through significant p-values (< 0.05) and a non-significant lack-of-fit (p > 0.05). The result showed that the contents of S/C, GP, CF, and MC at 0.99– 1 .05,5–10%,0.35–0.45%, and 0.15–0.25%, respectively, led to optimal workability and mechanical performance near the percolation threshold. Furthermore, SEM analysis revealed that GP improved the microstructure by promoting the formation of C-S-H gel. Thus, GP reduced the percolation threshold of the CF-CBS by 38.6% at optimum MC% (0.21) and at all levels of S/C (0.5 to 1.5). The optimised GP-modified CF-CBS achieved stable, sensitive, and highly repeatable cyclic piezoresistivity. The developed models provide practical guidance for designing sustainable GP-modified CF-CBS for structural health monitoring.