Multi-objective optimization of ambient-cured one-part geopolymer roller-compacted concrete incorporating 100% pavement-derived recycled concrete aggregate

This study developed and multi-objectively optimized an ambient-cured, one-part geopolymer roller-compacted concrete incorporating pavement-derived recycled concrete aggregate as a complete replacement for natural coarse and fine aggregates, with the intended use as a candidate material for further development toward pavement base and concrete pavement applications. Fly ash and ground granulated blast-furnace slag (GGBS) were used as precursors with a dry alkaline activator (DAA) to produce a dry-mixed “just-add-water” binder system. SEM, XRD, and XRF characterization showed that the RCA comprised natural aggregate and adhered old cement mortar, with microcracks, quartz-dominated crystalline phases, and a SiO₂-rich composition. A three-factor Box-Behnken design evaluated the effects of GGBS content, DAA content, and water-to-binder ratio (W/B) on Vebe time, 28-day compressive strength, and 28-day flexural strength. The reduced quadratic models were significant for all responses, showed non-significant lack of fit, and confirmed nonlinear effects within the investigated ranges. Desirability-based optimization identified 39.73% GGBS, 14.78% DAA, and a W/B ratio of 0.34, with an overall desirability of 0.94. The predicted Vebe time, compressive strength, and flexural strength were 40.06 s, 56.51 MPa, and 5.47 MPa, respectively. Confirmation tests yielded 40.67 s, 56.37 MPa, and 5.54 MPa, with prediction errors below 3%. Compared with a binder-equivalent natural-aggregate reference mixture, the optimized RCA mixture showed a higher Vebe time and moderately lower strengths. The results indicate that one-part geopolymer RCC containing 100% pavement-derived RCA can achieve a favorable balance between compactability and 28-day mechanical performance under controlled laboratory conditions.