The FA utilized in this study was sourced from a coal-fired power plant located in Ninghai, Zhejiang Province. The median particle size (D50) of GGBS and FA were 8.32 µm and 5.88 µm, respectively. X-ray fluorescence (XRF) analysis was performed on both the FA and GGBS, with the results presented in Table 1. The sodium hydroxide employed had a purity exceeding 98.5%, and the initial modulus of sodium silicate (defined as the molar ratio of SiO₂ to Na₂O) was 3.3. Detailed specifications for the sodium silicate are provided in Table 2. Sodium hydroxide and sodium silicate were used in combination as alkaline activators, both of which were procured from professional chemical reagent manufacturers.

The soil utilized in this experiment was collected from the K185 + 000 section of the Yongjinqu expressway in Quzhou, Zhejiang Province. A series of tests were performed on the soil, including compaction, UCS, CBR, and SEM-EDS microscopic analysis. The basic properties of the soil are summarized in Table 3, which classifies the selected material as high liquid limit soil according to the “Specifications for Soil Test in Highway Engineering” (JTG3430-2020). As stipulated in the “Specifications for Subgrade Soil in China” (JTG D30-2015), high liquid limit soils must be treated with inorganic binders prior to their application in subgrade fill.

Under standard curing conditions (20°C ± 2°C, 95% relative humidity), the study investigated the influence of various mixing ratios of precursor materials (FA and GGBS) and the modulus of the alkaline activator on the resulting strength (Liu et al., 2025). Table 4 details the different mixing ratios of the geopolymer curing agent used in this investigation. The experimental design comprised 22 groups, each containing one test sample and two control samples. For each group, the average strength of the three samples was calculated to ensure reliability and consistency of the results (as illustrated in Figures 1, 2). The mixing ratio that produced the highest strength was identified as the Optimum mixing ratio. Which was then used in subsequent experiments.

As shown in Figures 2, 3, the UCS of FA geopolymers did not exceed 20 MPa. However, substituting 10% of the FA with GGBS significantly enhanced the strength. Several factors contributed to this phenomenon: a. GGBS primarily consisted of calcium silicate and calcium aluminate, among other components. These constituents exhibited high reactivity and could chemically interact with the active silica and alumina in FA to form additional geopolymer gels, thereby augmenting the material’s strength. b. Particle Characteristics of GGBS: The particles of GGBS were typically very fine, resulting in a large specific surface area. This characteristic increased the contact area between GGBS and reactive geopolymer components, promoting higher reaction rates and enhanced strength development. c. Improved Microstructure: The incorporation of GGBS improved the microstructure of the geopolymer. The formation of gel structures, such as calcium aluminosilicate gel, through the reaction of GGBS within the geopolymer matrix increased the material’s density and hardness, thus contributing to higher strength.

The UCS of FA - GGBS based polymer was shown in Figure 3. In the low modulus interval (1.0–1.2), as the modulus increased from 1.0 to 1.2, the strength enhancement at the same dosage was minimal. For example, the 7-day strength increased from 25 MPa to 27 MPa at a 30% dosage, decreased slightly from 31 MPa to 28 MPa at a 35% dosage, and remained at 28 MPa at a 40% dosage. In the medium - high modulus range (1.4–1.6), as the modulus increased from 1.4 to 1.6, the strength increased significantly. For example, at a 40% dosage, the 7-day strength decreased slightly from 38 MPa to 35 MPa but remained higher than that of the low modulus group; at a 35% dosage, it increased from 29 MPa to 33 MPa. There was a peak phenomenon. At a modulus of 1.4 and a 40% dosage, the strength reached the highest value of 38 MPa, suggesting that the combination of intermediate modulus (1.4) and high dosage was more favorable for strength development. This was because the higher alkali content (Na₂O) at low modulus (1.0–1.2) provided a strong alkaline environment initially, promoting the dissolution of Al₂O₃ and SiO₂ in FA and GGBS. However, the low polymerization degree of SiO₂ resulted in a high concentration of silicate ions, which readily formed a low polymerization gel phase with Al³⁺, but the structure was loose and the strength enhancement was limited. In the medium-high modulus range (1.4–1.6), the polymerization degree of SiO₂ increased, the alkali content decreased relatively, and the initial dissolution rate slowed down. However, silicate ions readily formed long - chain polymerization structures and three-dimensional network gel phases with Al³⁺ (such as higher polymerization degree N-A-S-H gels), improving structural densification and significantly increasing strength. At a modulus of 1.4, the alkali content and SiO₂ polymerization degree reached a balance, promoting raw material dissolution and stable network structure formation, thus peaking strength at high dosage.

The incorporation of FA-GGBS geopolymers showed an inverse relationship with the OMC of high liquid limit soil. This phenomenon could be attributed to the formation of geopolymeric gels when FA-GGBS geopolymers were combined with high liquid limit soil. These gels filled the voids between soil particles, enhancing their bonding strength and compactness, which in turn increased the MDD. Moreover, since geopolymer materials had inherent water absorption capabilities, they effectively reduced the moisture content within the soil by absorbing some of this water. This reduction not only lowered the OMC but also decreased the amount of free water between soil particles, facilitating more intimate contact among them. Consequently, these interactions contributed to a higher MDD. The results of the Proctor tests, as shown in Figure 4, demonstrated these effects. Subsequent experiments were carried out at the determined OMC and maximum dry densities corresponding to various geopolymer contents.

Figure 5 illustrated the UCS of FA-GGBS geopolymer stabilized soil under various curing temperatures. Under standard curing conditions, the UCS of the stabilized soil at 7 days was higher than that of the geopolymer - stabilized soil cured at 0°C but lower than that cured at 40°C. At high curing temperatures (40°C), the reaction rate of the geopolymer significantly increased, accelerating the formation process and leading to a greater quantity of geopolymer gel being produced in a shorter period. This contributed to achieving higher strength more rapidly. However, excessively high temperatures might have compromised the stability of certain reaction products, potentially resulting in long term structural instability of the gel and affecting its ultimate long term strength performance. Conversely, at low curing temperatures (0°C), the chemical reaction rate of the geopolymer markedly decreased, hindering the hydration and gelation processes. This inhibition led to incomplete reactions between the geopolymer and the soil’s raw materials, thereby producing an insufficient amount of gel and negatively impacting strength development. Consequently, the UCS under low temperature curing was lower compared to that under standard curing conditions. These findings emphasized the critical role of curing temperature in influencing the strength development of geopolymer stabilized soils. Understanding these effects provided valuable insights for optimizing curing conditions to achieve desired mechanical properties in practical applications.

At 28 days, the UCS of the stabilized soil under standard curing conditions was the highest, whereas the geopolymer stabilized soil cured at 40°C exhibited greater strength than that cured at 0°C. This observation could be explained as follows: Under standard curing conditions, the geopolymer stabilized soil had achieved full reaction completion and stable hydration products by 28 days, resulting in maximal strength development. In contrast, although the geopolymer stabilized soil cured at high temperature (40°C) exhibited higher strength at 7 days, due to the accelerated reaction rate and increased gel production, its strength dropped below the standard curing conditions at 28 days. This reduction was attributed to potential structural instability caused by high curing temperatures, which might have compromised long term performance. Meanwhile, the geopolymer stabilized soil under low temperature (0°C) curing conditions demonstrated some improvement in strength at 28 days but still recorded the lowest UCS values among the tested conditions.

In this study, unsoaked CBR tests and soaked CBR tests (after 4 and 8 days of soaking) were carried out on stabilized soil samples with four different geopolymer contents at both 7 and 28 days of curing. The effects of geopolymer content and soaking duration on the CBR values were analyzed. The results were presented in Figures 6, 7. The findings indicated that as the curing time extended, the CBR values of all stabilized soil samples increased. This enhancement was attributed to the continuous progression of geopolymerization reactions, which strengthened the internal structure of the soil. Conversely, the CBR values of geopolymer stabilized soil significantly decreased with extended soaking periods. This reduction was due to the gradual increase in soil moisture content as soaking time increased. When the soil became fully saturated, pore water pressure rose, leading to a decrease in effective stress within the soil matrix, thus diminishing its bearing capacity. During the soaking process, water penetrated the soil pores, softening the soil and consequently affecting its CBR value. Furthermore, prolonged soaking could induce structural changes in the soil, such as weakening the bond between soil particles and altering their relative positions. These alterations resulted in reduced bearing capacity, further contributing to the decline in CBR values.

Figure 8 shows the resilient modulus trends for geopolymer-stabilized soils under varying stresses. Resilient modulus decreased with rising deviatoric stress at constant confining pressure. For instance, unstabilized soils dropped by 6.34%, 11.15%, and 17.24% as deviatoric stress rose from 30 kPa to 55 kPa, 75 kPa, and 105 kPa. Similarly, geopolymer-treated soils (4%, 8%, 12%) saw declines of 10.11%, 13.21%, and 10.01% under the same stress increments. This occurred because higher deviatoric stress increased vertical deformation, reducing modulus when strain outpaced stress growth. Confining pressure positively influenced modulus. At 30 kPa deviatoric stress, unstabilized soil modulus rose 12.94%, 21.90%, and 29.36% as confining pressure increased from 15 kPa to 30 kPa, 45 kPa, and 60 kPa. Geopolymer-treated soils (4%, 8%, 12%) showed boosts of 9.69%, 18.68%, and 31.77% under identical pressure changes. Greater lateral confinement from higher confining pressure hardened the soil, enhancing modulus.

Geopolymer dosage significantly impacted modulus. At 60 kPa confining pressure and 30 kPa deviatoric stress, increasing geopolymer from 0% to 4%, 8%, and 12% raised modulus by 17.84%, 56.22%, and 75.13%. The relationship followed a “weak-strong-weak” pattern: modest gains at low dosages (4%), maximum improvement at 8%, and diminishing returns at 12%. This trend is explained by geopolymer gel behavior. Low dosages (4%) produce limited gel, insufficient to fully fill soil voids. As dosage rises to 8%, gel fully occupies voids, strengthening particle bonds and embedding. Beyond 8%, excess gel adds minimal structural benefit, leading to weaker modulus gains. Overall, optimal geopolymer content balances void filling and particle reinforcement to maximize resilient modulus.

Figure 9 illustrated the microstructure of unstabilized soil and FA-GGBS geopolymer stabilized soil. The images revealed that the unstabilized soil primarily consisted of solid particles, pores, and internal cracks. After the addition of the FA-GGBS geopolymer, the generated geopolymer gel filled these pores and cracks, effectively binding the solid particles together. It was observed that the stabilized soil with a 4% geopolymer content still showed a considerable number of pores and cracks. However, as the geopolymer content increased, the porosity of the stabilized soil decreased significantly. This reduction in porosity accounted for the higher CBR and UCS values observed in soils stabilized with 8% and 12% FA-GGBS geopolymer compared to those stabilized with 4% geopolymer. This proves that the mechanical properties of stabilized soils are related to microscopic porosity, and as soil porosity decreases, the mechanical properties get improved.

As shown in Figure 10, EDS analysis indicated that the dominant elements within the geopolymer gel were oxygen (O), silicon (Si), aluminum (Al), calcium (Ca), and sodium (Na). These findings suggested that the primary types of gel formed in FA-GGBS geopolymer stabilized soil were likely to be calcium aluminosilicate hydrates (C-A-S-H) and calcium silicate hydrates (C-S-H). Additionally, sodium aluminosilicate hydrate (N-A-S-H) gel was also present in the geopolymer stabilized soil, which contributed to its enhanced mechanical properties.