Short-term high-humidity response of 3D-printed earthen repair materials: strength retention and multiscale pore characteristics

Rammed-earth materials are susceptible to moisture-related deterioration, which can weaken interparticle interactions and compromise mechanical performance. Although the moisture sensitivity of conventional earthen materials has been extensively investigated, the short-term high-humidity response of 3D-printed earthen repair materials and its relationship with multiscale pore characteristics remain insufficiently understood. This study investigated the compressive-strength retention and structural characteristics of a 3D-printed earthen repair material in comparison with an original rammed-earth reference material. Following 28 days of curing, specimens were exposed to 95% relative humidity at 20 ± 2 °C for 0, 3, and 6 h. Compressive-strength testing and post-compression failure observations were combined with scanning electron microscopy (SEM), X-ray diffraction (XRD), thermogravimetric analysis (TG/DTG), mercury intrusion porosimetry (MIP), and nitrogen adsorption. The original rammed-earth material exhibited a progressive decrease in mean compressive strength, with retention rates of 96.8% and 87.9% after 3 and 6 h, respectively. The printed material retained 102.6% of its initial mean strength after 3 h, indicating no observed reduction in mean strength, followed by a decrease to 77.6% after 6 h, corresponding to an approximately 22.4% reduction relative to the unexposed condition. Two-way ANOVA identified a significant overall effect of exposure duration (p = 0.0186), whereas the material type and material × exposure-duration interaction were not statistically significant. No individual pairwise comparison reached statistical significance under the six-group Tukey HSD procedure. Post-compression observations revealed cracking and particle detachment in the rammed-earth material, whereas the printed material exhibited localized crushing and predominantly trans-layer cracking. SEM revealed fine-particle filling, lamellar overlap, and localized matrix continuity in the printed material, while XRD and TG/DTG identified compositional differences between the two systems. MIP indicated greater mercury-accessible porosity in the printed material (38.01% versus 33.53%), despite its lower nitrogen-accessible pore volume at finer pore scales. These findings demonstrate that the investigated materials exhibit distinct descriptive strength-retention trends and scale-dependent pore characteristics under short-term high-humidity exposure. The coexistence of localized matrix continuity and greater mercury-accessible porosity highlights the importance of evaluating mechanical performance alongside complementary pore-structure measurements. The results provide material-level evidence to inform the formulation and further development of 3D-printed earthen repair systems for heritage conservation.