Robust Inverse Design of Cementitious Materials for Biogenic Degradation Resistance

Cementitious materials in infrastructure undergo coupled chemical, mineralogical, transport and mechanical deterioration, yet development still largely follows a forward trial-and-test paradigm. This study formulates the problem as constrained robust inverse materials design over experimentally bounded design spaces. Published data on biological degradation, ye’elimite–aluminate–calcium (YAC) binders, modified CA2 and calcium aluminoferrites were reorganized without imputing missing observations. Mechanism–material relations were made explicit through an evidence-constrained compatibility matrix, while mechanical and processing requirements defined feasible regions evaluated by minimum-intervention, Pareto and robustness analyses. Fluorine-modified CA2 combined a synthesis temperature of 1200 °C with 67.76% hydration at 28 days and was the only non-dominated measured state in that subspace. Its guaranteed two-objective robust-dominance radius was ±4.0%; in 100,000 Monte Carlo realizations, it remained Pareto-efficient in 100% of cases for perturbation levels up to ±10%. In the mechanical subspace, YAC Variant II reached 32.1 MPa after one day and 60.2 MPa after 90 days, while CEM I 52.5R retained the highest mature strength; both were non-dominated. Calcium-aluminoferrite data demonstrated a reachable composition–processing phase field. The framework organizes degradation knowledge and phase controllability into candidate-selection decisions and suggests a basis for incorporating phase-, microstructure- and nanoscale descriptors when matched experimental data are available.