High-liquid limit red clay has poor engineering characteristics, namely poor water stability, low strength, and large expansion and contraction deformation. The clay may be reinforced with an F1 ionic soil stabilizer. The engineering characteristics of this reinforced clay were studied, specifically concerning its basic physical parameters, shear strength parameters, and micropore structure. The F1 ionic soil stabilizer significantly improved the water sensitivity, compaction characteristics, and shear strength of red clay. We determined that the optimal F1 ionic soil stabilizer mix was 0.5 L/m³, resulting in a reinforced clay with plastic limit increased by 45.74%, optimal moisture content increased by 12.12%, maximum dry density increased by 5.8%, liquid limit reduced by 8.4%, plasticity index reduced by 43.8%, infiltration coefficient reduced by 41.8%, cohesion increased 1.64-fold, and internal friction angle increased 1.30-fold. Freeze-thaw cycles reduced the shear strength parameters of the reinforced red clay, although even after 15 cycles, it still had 18.4% higher cohesion and 57.1% higher internal friction angle than undisturbed red clay. The F1 ionic soil stabilizer significantly reduced the pore size and area of red clay, the complex connected pore structure is adjusted to a more regular structure. The reinforced clay had 56.64% lower pore area ratio, 32.27% lower average Feret diameter, and 2.43% lower fractal dimension.

Bored piles comprise an advanced pile foundation technology that has the advantages of high bearing capacity, fast construction speed, stable construction technology, and no noise or mud pollution. To study the applicability of bored piles to collapsible loess sites, the compaction effect and load-bearing characteristics of bored piles before and after immersion were studied via a full-scale field test combined with the theory of hole expansion. The results indicate that when the pile spacing is 1.0, 1.25, and 1.5 m, the average dry density of the soil between piles increases by 23.8%, 18.5%, and 3.1%, respectively, compared with that of untreated foundation soil. When bored piles are used to treat deep collapsible loess foundations, the reasonable pile spacing to eliminate the collapsibility of the loess foundation is 2.5 times the pile diameter. It is feasible to estimate the effective compaction range using the pore expansion theory, and the effective compaction coefficients of similar sites are given. The positive friction of bored piles in the collapsible loess area is more than 95.5 kPa, which increases by more than 48.5% compared with that of non-extruded piles. Therefore, the bearing capacity of a single pile is significantly improved, and it is an effective treatment method for collapsible loess areas. Under immersion, the pile side negative friction did not change significantly with a pile diameter of approximately 27 kPa, and the increase was approximately 14% compared with that of non-extruded piles. Consequently, to avoid the adverse effects of negative friction resistance on the bearing capacity of pile foundations and to fully utilize the technical advantages of bored piles, it is necessary to eliminate or partially eliminate site collapsibility before applying bored piles. The results can provide experimental support and theoretical guidance for the popularization and application of screw–squeeze piles in deep, collapsible loess areas.

To address the engineering problems of high cement content, high brittleness, and weak frost resistance of cement-improved loess in the seasonal frozen soil area of Northwest China, F1 ion curing agent (F1) and cement composite improved loess (FCIL) were used in this paper. Through unconfined compressive (UC) strength tests, consolidated undrained (CU) triaxial shear tests, and microscopic pore characteristics analysis, the mechanical properties, freeze–thaw cycle deterioration law, and microscopic pore structure of FCIL were studied. The effects of cement content (C_c), F1 dosage (C_F), number of freeze–thaw cycles (N_F-T), and confining pressure (σ₃) on the strength, deformation behavior, and pore characteristics of FCIL were analyzed. The synergistic improvement mechanism of FCIL, as well as the freeze–thaw damage mechanism, was elucidated. The results show that C_c is the primary factor controlling the strength of improved loess. The incorporation of F1 can further increase UCS and markedly enhance the failure strain (ε_f), thereby achieving simultaneous improvements in strength and ductility. An appropriate mix proportion was identified as C_F = 0.2 L/m³ and C_c = 6%. After 7 d curing, FCIL exhibited a UCS of 1.35 MPa, a cohesion (c) of 205 kPa, an internal friction angle (φ) of 36.2°, and ε_f 1.8 times that of loess improved with C_c = 6% cement alone. CU triaxial shear tests indicate that, under all tested conditions, the stress–strain responses of FCIL exhibit σ3-sensitive strain-softening behavior. As C_c and σ₃ increase, triaxial peak strength (q_max) and secant modulus (E₅₀) increase significantly. Compared with natural loess (NL), FCIL shows a markedly lower porosity (n), a substantial increase in the proportion of micropores, and reductions in medium and small pores. After multiple freeze–thaw cycles, the evolution of the pore structure is effectively restrained. This indicates that the combined use of F1 and cement promotes the formation of a dense layered stacking structure, significantly improves the microscopic pore-size distribution, and enhances the mechanical performance of loess under freeze–thaw environments.