Dispersive soils are common problematic geomaterials in road and hydraulic engineering. In seasonally frozen regions, they exhibit structural loosening and strength degradation under repeated freeze–thaw (F–T) cycles. Fiber-microbially induced carbonate precipitation (MICP) synergistic reinforcement is a promising green soil improvement technique. However, the mechanical response and microstructural evolution of polyvinyl alcohol (PVA) fiber – MICP-treated dispersive soils under F–T cycling remain poorly understood. In this study, laboratory-simulated F–T cycling tests, direct shear tests, unconfined compressive strength (UCS) tests, scanning electron microscopy (SEM), and nuclear magnetic resonance (NMR) analyses were conducted to investigate the shear stress–horizontal displacement relationship, normalized peak stress ratio ( η p / η 0 ), normalized cohesion ( c / c 0 ), normalized internal friction angle ( φ / φ 0 ), UCS loss rate, and microstructural mechanisms of PVA fiber – MICP-treated dispersive soils. The results indicated that the UCS decreased exponentially with increasing F–T cycles. After 20 F–T cycles, the UCS of the MICP-0.3% specimen increased by 924.5 kPa compared with the untreated soil, and its UCS loss rate decreased by 34.01%. At 100 kPa normal stress, the peak shear stress of the MICP-0.3% specimen was 66.49% higher than that of the untreated soil and remained 34.31% higher after 20 F–T cycles. η p / η 0 remained above 1, suggesting good shear strength retention. c / c 0 declined significantly during the initial F–T cycles and stabilized later, whereas φ / φ 0 varied little throughout. Additionally, empirical models were established and validated to predict UCS and shear strength considering fiber content and F–T cycling effects. The microstructural mechanisms were attributed to: (1) fiber–particle interlocking, suppressing structural loosening; (2) fiber–CaCO 3 crystal–particle composite structure, improving microstructural compactness; (3) fiber networks dispersing frost-induced stress; and (4) physical barriers limiting ice crystal expansion. These findings provide essential experimental support and theoretical basis for designing F–T-resistant reinforcement strategies for dispersive soils in seasonally frozen regions.
Zhang et al. (Sun,) studied this question.