Effects of Specimen Size and Loading Boundary on the Strength Evaluation of Cemented Phosphate Tailings Backfill and Corresponding Correction Methods

Specimen size and loading boundary conditions affect the uniaxial compressive strength (UCS) measured for cemented backfill and hinder direct comparison among test configurations. Cubic cemented phosphate tailings backfill specimens (40–200 mm) were tested at 3, 7, 14, and 28 d; 200 mm specimens were also subjected to partial loading at contact-area ratios η of 0.063–1.000. The mixture had a binder-to-tailings ratio of 1:8 and a slurry mass concentration of 68%. Using the 70.7 mm specimen and full-section loading as references, size- and loading-boundary correction coefficients, Ks and Kc, were defined. At 3 d, Ks for the 200 mm specimen was 0.394; at 7 and 14 d, Ks ranged from 0.709 to 1.114 and 0.889 to 1.000, respectively. At 28 d, the 100–200 mm specimens exceeded the reference-group point estimate, while the reference group had a coefficient of variation of 41.1%, indicating an apparent response affected by curing boundary and variability. As η decreased, peak load and full-section equivalent strength decreased, whereas area-averaged nominal contact stress increased; at 28 d and η = 0.063, Kc = 0.105 and the nominal contact stress was 9.080 MPa. Qualitative COMSOL comparisons and exploratory data-driven analysis supported physical interpretation and variable-sensitivity assessment. These coefficients improve comparability under matched conditions.