Erosion Resistance and Rainfall-Induced Slope Stability of a Cement-Free Surface Stabilizer Based on Industrial By-Products

Shallow failure and surface erosion account for the majority of rainfall-induced slope damage on Korean expressways, yet conventional countermeasures such as anchors and soil nailing are designed against deep-seated circular failure. This study evaluates the erosion resistance and slope-scale performance of a cement-free surface stabilizer composed of ground granulated blast furnace slag and hemihydrate gypsum as binders, with high-calcium fly ash, magnesium sulfate and a nutrient additive. Two selected mixes for water contents above and below 30%, C-91-6.41-30 and C-94-4.276-25, were examined; at the total agent content specified in the applicable construction specification, their binder contents were 28.8% and 28.7% lower, respectively. Consolidated drained direct shear tests at 3 days of curing gave cohesion increases of 160% and 65.6% over the parent soil, with the friction angle increasing by 71% in the former and by only 0.3% in the latter, indicating a cohesion-dominated improvement at the lower water content. Under simulated rainfall of 30 mm/h for 60 min, the parent soil lost 70.92–77.39% of its mass and failed by sliding, whereas the improved slopes lost only 0.23–1.27%, a reduction of 98.2–99.7%, with only localized surface erosion and no progression to global sliding. Coupled transient seepage and limit-equilibrium analyses were conducted for the geometry and stratigraphy of an actual failed slope under a conservative upper-bound infiltration scenario. The maximum hourly rainfall intensity recorded on each day during the antecedent rainfall period was applied as a sustained 24 h surface flux to impose a severe hydraulic loading condition. Under this assumption, 0.5 m improved layers gave factors of safety of 1.416 and 1.475, compared with 0.759 for the unreinforced section. A closed-form check at the field stress level shows that 85–93% of the mobilized shear resistance of the improved layer is contributed by cohesion, indicating that the calculated shear-strength improvement at this stress level is predominantly attributable to the cohesion component rather than the high apparent friction angle. The numerical analysis was subject to limitations associated with differences between the laboratory parent soil and the field soil, the limited hydraulic characterization of the improved layer, the simplified treatment of surface runoff, and the artificial upper-bound rainfall boundary condition. Therefore, the numerical results should be interpreted as a comparative parametric evaluation of the mechanical contribution of surface stabilization under severe infiltration conditions rather than as a reproduction of an observed rainfall event or a prediction of the actual failure timing.