Abstract

The swelling properties of expansive soils can be reduced by the addition of modifiers. Nevertheless, the performance deterioration after modification occurs when weathering for a long term. Therefore, in this study, the effect of drying-wetting cycles on swelling behaviour and compressibility of modified expansive soils with the iron tailing sand and calcium carbide slag has been investigated. The swelling potential initially increases and subsequently decreases with the increasing number of cycles, reaches the peak at the seventh cycle, and tends to equilibrium after the tenth cycle. These results show that drying-wetting cycles will destroy the soil structure. The compressibility of modified expansive soils increases with the drying-wetting cycles, where an empirical formula between compressibility and the cycle number was established. Microstructural analysis is performed using mercury intrusion porosimetry (MIP) and scanning electron microscopy (SEM). The results of microstructural analysis show a tendency of degradation process.

1. Introduction

Expansive soils are highly plastic soils and are mainly composed of hydrophilic clay minerals (i.e., montmorillonite group minerals). The behaviour of expansive soils changes with the water mass, i.e., expanding when absorbing water and shrinking when discharging water [1, 2]. The long-term evolution in climate and hydrological conditions will cause progressive alteration in deformation and strength, leading to the degradation of geo-based structures, including the highway, railway, slope, and foundation, and also causing a significant financial loss annually in the world [35].

Many improvement technologies have been proposed and applied in engineering practices, such as replacement, compaction, moisture control, chemical modification, and thermodynamics methods. Chemical additives such as cement, lime, and fly ash have been confirmed to effectively change the plasticity and swelling characteristics of the expansive soils [69]. Despite that cement and lime have been extensively used as they are cost effective, many industrial wastes, such as recycled fiber, glass, iron tailing sand, and carbide slag, have been regarded as the substitution of traditional modifiers for environmental protection and economic benefits [1014]. Iron tailing sand and calcium carbide slag are the byproducts of iron mines and calcium carbide industries, respectively [15, 16], which bring serious environmental problems when disposing these wastes. Calcium carbide slag with high Ca(OH)2 content may be a sustainable agent to substitute lime and cement when modifying expansive soil. Iron tailing sand composes of the fine and stable fraction, which could be used to optimize the particle size distribution [17, 18], the swelling pressure [19], and the shear strength [20] of expansive soil. Therefore, it is applicable to recycle these wastes as expansive soil modifiers.

Previous studies highlight that the stabilization of modified expansive soils depend on the type and content of modifier, curing period, temperature, and moisture control. Note that stabilization quality in working condition should be focused, especially in the alternate region between rainy and dry seasons. This alternation will weaken the stabilization quality and then lead to the engineering problems (deformation and cracks of geo-based structures). The beneficial effects of expansive soils modified by lime are partially lost after several drying-wetting cycles [2125]. These results showed that irreversible damage of the improved geo-based structures could be triggered after experiencing the drying-wetting cycles [2628]. In addition, for the geo-based structures (i.e., foundation and roadbed), the compressibility is another concentration for the long-term stability and safety [2932].

In this study, the effects of drying-wetting cycles on the free swelling ratio, the swelling pressure, and the compressibility of expansive soils improved by the iron tailing sand and calcium carbide slag were explored. Furthermore, microstructure was investigated by the mercury intrusion porosimeter (MIP) and scanning electron microscope (SEM) tests as well to clarify the internal mechanism.

2. Materials and Methods

2.1. Materials

The expansive soil used in this study was sampled from Hefei, the central part of Anhui province, China. The basic physical properties of the soil are summarized in Table 1. Swelling potential was classified as “weak” according to the previous research studies [33]. Its maximum dry density of 1.84 g/cm3 and the optimum water content of 16.6% are identified by the compaction test [34]. The particle size distribution was determined by the hydrometer analysis obeying ASTM D421-85 [35], and the clay fraction of the sample after drying-wetting cycles was also measured. The iron tailing sand (hereinafter abbreviated as ITS) and calcium carbide slag (hereinafter abbreviated as CCS) were both sampled in Anhui, China. The composition of the iron tailing sand was mainly primary minerals with stable structure, including quartz and feldspar. The main chemical composition of calcium carbide slag is shown in Table 2. The main chemical composition and the particle size distribution of iron tailing sand are shown in Tables 3 and 4, respectively.

During sample preparation, the expansive soil and the industrial wastes were oven-dried at 105°C for 36 h and then ground into powders and sieved through a 0.5 mm sieve. According to Ye et al. [13], the iron tailing sand content of 30% can effectively inhibit swelling and achieve the requirements for engineering construction. Therefore, the soil mixed with iron tailing sand at a mass proportion (aITS) of 30% and with calcium carbide slag at a mass proportion (aCCS) of 6%, 8%, 10%, 12%, and 14% was designed respectively. Hereinafter, the water content of mixtures was adjusted to the optimal water content depending on aITS and aCCS (Table 4). After 24 h of curing period in an airtight container, the mixtures were statically compacted in a mould with 20 mm height and 61.8 mm diameter to reach the maximum dry density (Table 5). Finally, the prepared samples were cured under standard curing conditions (temperature of 20°C and relative humidity of 95%) for 28 days.

2.2. Drying-Wetting Cycle Tests

Cycled drying-wetting experiments were performed according to ASTM D4843-88 [36]. During the drying process, each sample was placed in an oven with temperature at 60 ± 1°C for 23 h. In the subsequent wetting process, the sample was placed on a porous stone in an immersion chamber with temperature controlled at 20 ± 1°C for 1 h. Distilled water was used to saturate the porous stone and maintained for 23 h. After 1, 3, 5, 7, 10, and 15 cycles of the drying-wetting process, the free swelling ratio, swelling pressure, and the compressibility tests were performed in accordance with ASTM D4546 [37] and ASTM D 2435 [38].

2.3. Swelling Potential Tests

Swelling potential reflects the volume change or the pressure required to prevent swelling [39]. The swelling ratio was adopted to describe the swelling potential, which is defined as the ratio of the swelling amount to the original thickness in percentage. The swelling ratio tests were carried out using a one-dimensional oedometer apparatus. The sample in the ring was placed between two porous stones, and filter papers were added between the sample and the porous stones. A static pressure of 0.7 kPa was applied. The sample was then soaked in water and allowed to swell under initial static pressure. The dial gauge displacement was recorded until no further swelling, where the swelling potential was considered as full completion. Based on the measured results, the free swelling ratio was calculated using the following equation:where δN (%) and HN (mm) are the free swelling ratio and height of the sample after swelling, respectively, and H0 (mm) is the initial height.

2.4. Swelling Pressure Tests

The swelling pressure tests were measured using the constant volume method. The sample installed in the 1D oedometer apparatus was the same as the swelling ratio test. The water-soaked sample was subjected to a specific load to keep constant volume. The load was tracked when no further swelling was observed. Swelling pressure was calculated using the following equation:where Pe (kPa) is the swelling pressure, W (N) is the total load, and A (cm2) is the area of sample.

2.5. Compressibility Tests

The compressibility tests were carried out by using an automatic pneumatic 1D oedometer and computer data acquisition and processing system. The step-step loads (25, 50, 100, 200, 400, and 800 kPa) were applied until 24 hrs. The sample installation in the 1-D oedometer apparatus was the same as the swelling ratio test. The compressibility can be evaluated using the coefficient of compressibility from 100 kPa to 200 kPa of the effective vertical stress [40]. The coefficient of compressibility was calculated using the following equation:where (kPa−1) is the coefficient of compressibility and e1 and e2 are the void ratio under P1 (100 kPa) and P2 (200 kPa), respectively.

2.6. Scanning Electron Microscopy (SEM) and Mercury Intrusion Porosimetry (MIP)

Scanning electron microscopy (SEM) and mercury intrusion porosimetry (MIP) were used to investigate the microstructure and the pore characteristics of the modified soils. In the SEM procedure, the samples were cut into pieces with approximately 1 cm3, gold coated, and then scanned by using a high-resolution scanning electron microscope. MIP tests were carried out on an automatic mercury intrusion instrument. According to MIP requirement, the samples were cut into small pieces and the lyophilization technique should be first adopted.

3. Results and Discussions

3.1. Swelling Potential and Pressure

The evolutions of the free swelling ratio and swelling pressure with the number of drying-wetting cycles in the present work are shown in Figures 1 and 2, respectively. From these figures, it can be seen that both the free swelling ratio and swelling pressure initially increase and subsequently decrease with the increasing cycles. As shown in Figures 1 and 2, the tendency in swelling potential is the same in clay fraction (Figure 3) when encountering the drying-wetting cycles. The tendency indicates that the clay fraction increases the swelling ratio. After the process of the drying-wetting cycles, this clay existence is probably released from the modified soil aggregates for the partial breakdown and disintegration, in agreement with the observations of previous studies [23, 24]. After the seventh cycle, the maximum values of the swelling ratio and swelling pressure occur. Meanwhile, the maximum value of the free swelling ratio is below 30% (limit line in Figure 1), which satisfies the requirement of backfilling materials in Jianghuai area of Anhui province [40]. The value of swelling pressure with optimal CCS of 10% is below 45 kPa (limit line in Figure 2), which depends on the overburden vertical stress of applied subgrade layer. Note that the free swelling ratio, the swelling pressure, and the clay fraction all decrease after seventh cycles, which is different with the observation of the free swelling ratio and the swelling pressure in the previous studies. This trend may be related to loss of clay fraction or transformation of clay mineral by CCS. The excessive drying-wetting cycles may lead to the serious damage of soil structure and the porosity, and the clay fraction releasing from the aggregates is lost from the pores with the infiltration and flow during the wetting-drying cycles [24].

3.2. Compressibility

The effect of drying-wetting cycles on compressibility is presented in Figure 4. The compressibility can be divided into two stages. In the first stage, the compressibility increases rapidly. In the second stage, the compressibility increases slowly and reaches to a relatively stable state after the tenth drying-wetting cycle. This increase trend is contributed to the irresistible deformation under volumetric stress which caused by the destruction of cementitious products and loose matrix structure. This stable state reveals that the effects of drying-wetting cycles are resisted, which indicates some stronger cementitious compounds after drying-wetting cycles are regenerated. The compressibility evolution is consistent with the results of Stoltz et al. [25]. It is worth noting that the coefficient of compressibility with aCCS of 10% is lower than that of 6%, 8%, 12%, and 14%, suggesting that the structure with aCCS of 10% is more compact than that of 6%, 8%, 12%, and 14%.

With these results, evolution of the compressibility with the drying-wetting cycles can be described using equation (4), with the specific parameters listed in Table 5:where is the compressibility (MPa−1) and N is the number of drying-wetting cycles.

As shown in Table 6, parameters A, B, and C change very little with aCCS, but dependent on aCCS..

It is clearly shown that the compressibility is strongly related to the cycle numbers, as the parallel curves shown in Figure 5. Therefore, an empirical relationship can be used to predict the compressibility of the expansive soil treated with iron tailing sand and calcium carbide slag experiencing drying-wetting cycles.

3.3. Microstructure by SEM and MIP

The pore size distribution of mercury intrusion porosimetry tests is presented in Figure 6. From Figure 6(a), the PSD distribution shows the typical bimodal characteristics, indicating the presence of two groups of pores: macropores and micropores. Note that the averages of total volume in macropores and micropores slightly increase with the number of drying-wetting cycles in Figure 6(b). The final intrusion volume keeps increasing during the wetting-drying process, which is also consistent with the tendency of compressibility encountering the drying-wetting cycles.

To differentiate the porosity within the modified soils, the pore group would be classified. Kong et al. [41] divided the pores into micropores (<0.007 µm) within the particles, small pores (0.007 µm∼0.9 µm) between the particles, middle pores (0.9 µm∼35 µm) within the aggregates, large pores (35 µm∼300 µm) between the aggregates, and the macropores (>300 µm). Figure 6 shows the pore sizes subjected to drying-wetting cycles which mainly lie in 0.007∼0.9 µm and 0.9∼35 µm, suggesting that small and medium pores are majority (more than 85%). Note that the volume of small and middle pores increases from the first cycle to the tenth cycle.

The evolution of stabilized soil during the drying-wetting process is presented from the SEM results in Figure 7. After the first cycle, the microstructure of stabilized soils is compact and contains needle-like morphology. The aggregates are big, and the pores in improved soil are mainly composed of the medium pores. After the fifth cycle, the binding of cementitious compounds and protection of cementation were mostly retained. The small and middle pores changed trivially. After the seventh cycle, the aggregates are severely damaged and become smaller in dimension. The pore volume between aggregates increases significantly. After the tenth cycle, the aggregates become weak and the pores within aggregates increase seriously.

For the compacted soils, the microstructure should involve two structural levels, i.e., among and within aggregates [4244]. This classification (double porosity) plays an important role in the hydraulic strength and deformation behaviours [4547]. This model of double porosity [48, 49] was attempted to clarify the structure characteristic of original and stabilized soils in the present work shown in Figure 8. According to analysis by SEM images, the structure among aggregates became looser and that within aggregates is destroyed and dispersed with the increase number of drying-wetting cycles. The disintegration of soluble substances and the effects of swelling and shrinkage lead to the destruction of partial gelatinous substance and the generation of cracks and pores. In view of the skeleton structure, macroaggregates contribute to the limitation of the swelling potential and the swelling pressure more than microaggregates, which suggests that the change in macroaggregates is consistent with that in macroscopic swelling characteristics.

4. Conclusions

The impacts of successive drying-wetting cycles on the swelling properties, the compressibility, and the microstructure of expansive soil improved by iron tailing sand and calcium carbide slag have been investigated in this study. Effects of changes in iron tailing sand and calcium carbide content and curing time were discussed. The main conclusions have been drawn as follows:(1)The variation of swelling potential and the swelling pressure is closely related to the clay fraction. The swelling behaviour of modified expansive soils subjected to the drying-wetting cycles is a gradual destructive process and can be predicted by the function of the cycle number.(2)Despite the negative effect of drying-wetting cycles, the iron tailing sand and calcium carbide are effective in stabilizing the expansive soil.(3)Drying-wetting cycles enrich the porosity within and among the aggregates, suggesting the progressive damage process of the soil structure.(4)Aggregates play an important role in improving expansive soil. The wetting-drying cycle has an influence on the modified expansive soils in the transformation of the aggregates.

Data Availability

The data used to support the findings of this study are included within the article.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

This study was supported by the National Natural Science Foundation of China (Grant nos. 51878159 and 41577280).