Corrosion-induced cracking of reinforced-concrete (RC) covers is well known, yet key knowledge gaps persist. Most studies isolate uniform corrosion or a single non-uniform corrosion pattern and ignore the effects of boundary restraint and structural configurations, leading to inaccurate predictions of cracking thresholds and crack propagation patterns. This study systematically investigates the influence mechanisms of constraint conditions and structural configurations on corrosion-induced cracking behavior using the phase-field model. The results indicate that the non-uniformity of steel corrosion is a critical factor governing cover cracking. As the corrosion non-uniformity coefficient increases, the critical corrosion level exhibits a monotonic decreasing trend—from 0.95% to 0.15% under strong constraints and from 0.52% to 0.15% under weak constraints. Concurrently, the crack morphology evolves from a single radial crack to a wedge-shaped crack oriented toward the peak corrosion side. The influence of constraint conditions is dualistic, while strong constraints enhance the failure threshold, their mitigating effect diminishes markedly under highly non-uniform corrosion. The critical corrosion threshold for eccentrically arranged corner reinforcement is significantly lower than that for centrally arranged reinforcement; the corrosion angle only induces slight crack deflection and minor threshold fluctuations; and the curved top section, due to its weaker equivalent constraint, exhibits inferior crack resistance compared to the linear top section. Three-dimensional analysis reveals a pronounced longitudinal discreteness effect, which not only substantially elevates the critical corrosion threshold but also leads to diverse spatial failure modes. This work links rust-expansion eigen-displacement to crack propagation within a unified phase-field framework, providing materials-level criteria for evaluating corrosion tolerance and guiding the design of cover materials and reinforcement layouts to enhance durability.

Mix proportions for 3D-printed concrete must satisfy requirements for transport, extrusion, interlayer stability, and hardened performance. Heterogeneous raw materials, equipment, and test conditions often prevent direct use of published mixtures. This study analyzed mixture evidence from 1003 deduplicated publications at three levels: cross-study associations, within-study comparisons, and linked printability–mechanical evidence. Among mixtures with an explicit forming method and source-reported printability, 476 mixture-level aggregate units from 157 studies showed a negative association between the water-to-binder ratio and 28 d compressive strength. The pooled coefficient, paper-equally-weighted coefficient, and median within-study coefficient were −0.493, −0.422, and −0.516, respectively, supporting the use of the water-to-binder ratio as a priority screening variable rather than as a universally transferable strength predictor. Observed interquartile ranges in the cement-containing subset described the concentration of reported proportions. Scatterplots and sensitivity analyses for 12 variables showed that zero handling and sample boundaries affected the direction of some associations between mass concentrations. Five studies provided 28 comparable within-study results across 27 condition groups, each containing only one independent study. Four studies linked successful printing to mechanical specimens from the same mixture and process, yielding seven pairs; one included a four-mixture comparative series. Prediction data for nine printability outcomes and four strength ages failed all three eligibility scenarios. Thus, current evidence supports initial variable screening and study-specific material hypotheses, but independent replication and more complete process–specimen linkage are needed for transferable prediction and resource-efficient mixture research.