Deep soil cementation is widely used for strengthening weak soils; however, the effectiveness of injection mortars largely depends on their penetration ability and interaction with in-situ soil conditions. This study aims to evaluate the performance of an injection mortar through large-scale model testing under conditions as realistic as possible. The experiments were conducted at a scale of 1:25 using loamy soils representative of Astana’s engineering-geological conditions. The model soil was prepared by reproducing natural density and moisture parameters, including an average natural density of 1.85 g/cm³ and moisture content of 13.3%, while overburden pressure was simulated using a rigidly fixed cover. Laboratory tests determined consistency limits (plastic limit 14.42%, liquid limit 19.75%), indicating a semi-solid to solid soil state. A strong correlation (R = 0.99) between added water and achieved moisture content was established, with an optimal water volume of 18,900 cm³ required to achieve target conditions. Compaction tests showed that the required density was achieved with 20-50 roller passes depending on layer depth, resulting in final densities of 1.85-1.88 g/cm³. The results confirm that the proposed modeling approach reliably reproduces in-situ conditions and can be effectively used to assess injection mortar performance in weak soil stabilization.

Deep soil cementation is widely used for strengthening weak soils; however, the effectiveness of injection mortars largely depends on their penetration ability and interaction with in situ soil conditions. This study aims to develop and validate a methodology for preparing equivalent loamy soil for large-scale tray model tests intended for subsequent investigations of deep soil cementation and injection technologies. The study focuses on reproducing natural soil density, water content and stress conditions representative of engineering–geological conditions of Astana. The model soil was prepared by reproducing natural density and moisture parameters including a target natural bulk density of 1.85 g/cm³ and an average measured natural density of 1.88 g/cm³, with a water content of 13.3%, while overburden pressure was simulated using a rigidly fixed cover. Laboratory tests determined consistency limits (plastic limit 14.42%, liquid limit 19.75%), indicating a semi-solid to solid soil state. A strong correlation (R = 0.99) between added water and achieved water content was established, with an optimal water volume of 18,900 cm³ required to achieve target conditions. Compaction tests showed that the required density was achieved with 20–50 roller passes depending on layer depth, resulting in final densities of 1.85–1.88 g/cm³. The results confirm that the proposed modeling approach reliably reproduces in situ conditions and can be effectively used to assess injection mortar performance in weak soil stabilization.