1 Introduction

Stabilization of expansive clays is an effective method to modify the hydro-mechanical behavior of these problematic soils in order to improve their performance. High swell-shrinkage potential of expansive soils makes them susceptible to moisture changes due to seasonal climatic conditions. Change in water table level or broken underground drains may result in considerable damage to structures. Foundation settlements, subsidence and bulging of roads, cracking in pavements and buildings are some of the structural failures that may be associated with expansive soils. Therefore, it is very important to study the expansive soil characteristics and mitigate its properties in order to achieve the desired engineering requirements for any civil engineering project [27, 39, 44, 59, 74, 76].

One of the methods of soil stabilization is the use of natural, chemical or by-product materials as additives in the soil [31, 40, 44, 47, 50, 56]. Utilization of waste materials as stabilizing agents grew more interest among civil engineers in the recent years to manage waste. Whether the source of the waste materials is industrial, such as micro-silica, fly ash, or agricultural, such as rise husk ash, oyster shells, olive seed, or from construction wastes (calcined clay), it is rapidly becoming a priority for geotechnical engineers to utilize these materials in order to enhance the soil properties, while achieving an effective and sufficient method of waste management [1,2,3,4, 8, 10, 21, 32, 34, 37, 42, 44, 46, 53, 55, 57, 68,69,70].

In the previous studies, the effect of traditional soil stabilizers such as hydrated lime, limestone dust and cement has been studied [1, 2, 28, 38, 58, 60, 75]. Schanz and Elsawy [64] studied the effect of limestone and hydrated lime on the behavior of expansive soil. Their results indicated that while both calcium-based additives positively affected the swelling potential of soils, due to more abundant calcium ions in hydrated lime (Ca(OH)2), the effect of hydrated lime on reduction of swelling pressure was much more significant than limestone (CaCO3). Also, the result of their study showed that limestone negatively affected strength, while hydrated lime increased the strength of expansive soil. These results affirm that the chemical composition of different types of lime plays an important role in effectiveness of reaction of lime with soil. Therefore, hydrated lime is clearly a better choice for expansive soil stabilization. Another factor that closely influences the result of soil treatment with lime is the amount of added lime. Previous researches have documented expansive soil treatment with different variations of lime content (from 1 up to 10%). According to Bell [23], the optimum amount of lime required to achieve maximum modification of soil is between 1 and 3% of lime added by weight. Up to this point which is known as lime fixation point, the addition of lime only satisfies the affinity of the diffuse double layer of clay particles for lime’s divalent Ca2+ ions. Therefore, the positive ions that are adsorbed by clay particles are not available to participate in other reactions, and they only affect the plasticity of the soil due to cation exchange between surface of negatively charged clay platelets and calcium ions. It is only after this point that the further addition of lime can participate in pozzolanic reaction with soil. However, Schanz and Elsawy [64] found that increasing lime content of up to 5% and more has positive effects on swell pressure and swell time. Generally, it is observed that increasing the lime content and curing period positively affects soil strength, swelling and Young’s modulus [23, 25, 29, 58, 64].

However, there are limitations associated with calcium-based additives for stabilization of expansive soils. Considerable costs are necessary for the repair and renovation of damages regarding sulfate-induced heaves that occurs after stabilization. There is limited information available in the literature on how to eliminate the problems associated with expansive soil treatment with such materials. The findings from literature review indicate that in recent years the addition of waste materials, such as micro-silica, to clay soils and their effect on swelling potential, hydraulic conductivity, volume change, development of desiccation cracks and unconfined compressive strength have been investigated. The results show that the addition of pozzolanic waste materials such as micro-silica improves these properties of expansive soils.

Micro-silica (MS), also known as silica fume or condensed silica, is a by-product of calcium silicon, ferrosilicon alloys and silicon metal production, which results from the reduction process of high purity quartz, and therefore does not need any further processing. Hence, from environmental perspective, using micro-silica as a stabilizer saves energy, because its production does not consume any extra energy [4, 9, 11, 33, 41, 44, 46]. MS is composed of mostly amorphous silicon dioxide (SiO2), and due to its extremely small particle size, it has low unit weight. These characteristics combined with large surface area of the particles places MS in the category of highly reactive pozzolanic materials [11, 61, 73]. MS has been successfully used to enhance the durability, strength and electrical resistivity of concrete [9, 21, 30]. Others investigated the performance of micro-silica on expansive soil [8, 22, 24, 43,44,45]. Kalkan [44] studied the effect of micro-silica on swell behavior of expansive soils subjected to wetting–drying cycles and showed that both swell pressure and swell potential decreased with the increase in the cycles. It was also observed that micro-silica contents of up to 20% reduced the swell potential, thus swelling pressure, whereas no significant effect occurred with higher amounts of micro-silica. Kalkan [43] has observed that micro-silica contents of 0–25% reduced the development of desiccation cracks, whereas higher micro-silica contents did not have any considerable effect on the extent of development of desiccation cracks.

Goodarzi et al. [41] studied the effect of lime–micro-silica on highly expansive smectite clay. They indicate that although addition of micro-silica has positive effect on soil properties, the addition of micro-silica and lime combination to highly expansive soil improves soil strength, swelling and permeability more than the addition of lime or micro-silica alone. They also found that combination of micro-silica and lime enhances the soil strength with lower amount of lime and shorter curing time in comparison with samples that are treated with lime alone. Moayyeri et al. [53] examined the effect of lime–micro-silica mixture on geotechnical properties of low plasticity gypsiferous clay soil. Their results show that addition of lime and micro-silica increases the stability against soaking and compressive strength of gypsiferous soil. Alrubaye et al. [6, 7] investigated the effect of lime–micro-silica mixture on soft kaolin clay. Their results show that addition of lime and micro-silica reduces the degree of permeability and coefficient of consolidation and increases the shear strength of kaolin soil.

According to ASTM C 618 standard, micro-silica is considered as a pozzolanic material. The reaction between reactive silica and alumina oxides of the pozzolan and the calcium oxide present in lime develops pozzolanic reaction in the presence of water. Calcium silicate hydrates (CSH) and calcium aluminate hydrates (CAH) are the cementitious compounds resulting from pozzolanic reaction. The mechanism of pozzolanic reaction can be explained as in Eqs. (1) and (2):

$${\text{Ca}}^{2 + } + 2({\text{OH}})^{ - } + {\text{SiO}}_{2} \to {\text{C}}{-}{\text{S}}{-}{\text{H}}$$
(1)
$${\text{Ca}}^{2 + } + 2({\text{OH}})^{ - } + {\text{Al}}_{2} {\text{O}}_{3} \to {\text{C}}{-}{\text{A}}{-}{\text{H}}$$
(2)

Hydration of lime with water liberates (OH)− ions which will increase the pH of the soil solution. This will increase the solubility potential of silica and alumina and leads to pozzolanic reaction.

The pozzolanic reaction can be separated into two phases: in the first phase with the addition of lime an immediate cation exchange will occur in the soil in which the high valance cations of the lime like Ca2+ instantly isolate the clay anions from other monovalent ions in order to attach to the negatively charged surface of the clay. This reaction reduces the thickness of the diffuse double layer which will cause flocculation and reduction in the plasticity of the soil [52]. In the second phase, the cementing compound (CSH and CAH) gradually bonds the clay particles together. Therefore, this phase takes place over long time scale. These pozzolanic compounds improve the mechanical behavior of the soil by binding the soil particles together. Also, since the pozzolanic reaction uses some of the pore water, it will make the soil stiffer and reduce the swell and shrinkage potential [12, 13, 54, 59, 62, 65, 75].

The present research aims to investigate the reutilization of a waste material in the field of geotechnics. To achieve this, the effect of micro-silica as an industrial waste on modifying the engineering characteristics of expansive soils was investigated. As inferred from previous works, the use of MS–lime combination in expansive soils is quite scarce and usually does not cover study of shrinkage behavior, as well as durability of the suggested cementing mixture over prolonged time. Therefore, shrinkage behavior, which is important in the study of swelling clays in semiarid climates, is included in this study. Durability of the cemented soil is also an important issue which was investigated up to 28 days in previous research. 90 days of curing period for unconfined compressive strength specimens has been incorporated in this study, in order to assess the sustainability of the method over a longer time period. The experimental study included tests on swell-shrinkage behavior, compressibility and unconfined compressive strength. The twofold purpose of using MS, to stabilize a local expansive soil, as well as recycling it, is assessed together with lime addition required for pozzolanic reaction, yet a reduced amount of lime to eliminate the undesirable effects when used alone.

2 Materials and methods

2.1 Materials

The expansive soil was collected from Famagusta, Cyprus. According to the Unified Soil Classification System [16], it is classified as clay with high plasticity (CH) consisting of mostly clay particles with an appreciable amount of silt. Physical properties of the soil are presented in Table 1. Mineralogical composition of soil was determined by X-ray diffraction (XRD), and the results are illustrated in Fig. 1 depicting that the expansive soil includes calcite, illite, montmorillonite, quartz, kaolinite and albite which are clay minerals as well as non-clay minerals of calcite.

Table 1 Physical properties of soil
Full size table
Fig. 1
figure 1

X-ray diffraction pattern of expansive soil

Full size image

The micro-silica (MS) used in this study is a white ultrafine powder which was obtained from Semra Ltd. Company in Northern Cyprus. It contains more than 95% purity of silicon dioxide. Based on hydrometer test, MS has 52% clay size particles and 20.8% silt size particles. Table 2 presents the chemical composition of the soil and micro-silica obtained from wet chemistry analysis. The total percentage of SiO2, Al2O3 and Fe2O3 in micro-silica is more than 70%, which is an indicator of pozzolanicity according to ASTM [13].

Table 2 Chemical properties of materials
Full size table

2.2 Methods

The expansive soil was oven-dried at 50 °C, pulverized and was separately mixed with different percentages of micro-silica, lime and combinations of both under dry conditions. The amounts of micro-silica were selected to be 10% and 20% of the dry mass of the expansive soil and was mixed with 3% and 5% lime as a secondary additive. The dry mixtures were then blended with the required amount of water for each combination to reach the optimum moisture content (Fig. 3). Soil mixtures were left to mellow for 24 h to ensure a homogeneous distribution of moisture and were then compacted according to ASTM [19]. Based on different requirements for each test, the samples were wrapped and placed in desiccators for different curing periods.

Reactivity of the pozzolan (micro-silica) was studied in accordance with the method proposed by Luxán et al. [49].

Specific gravity [20] tests were conducted for expansive soil, micro-silica and lime.

The effect of different additives on Atterberg limits was investigated as a basis for the assessment of mechanical properties for each soil-additive mixture [17].

One-dimensional swell test was performed according to ASTM [18] followed by consolidation test [15].

Shrinkage test was conducted on samples confined in the consolidation rings of 75 mm diameter and 15 mm height. After full swell was achieved in one-dimensional swell, the samples were placed in a temperature-controlled room to dry. Mass, diameter and height of the samples were continuously measured at different time intervals until measurements showed no change in volume.

Unconfined compressive strength (UCS) tests were performed according to ASTM [14], on non-cured and cured (1, 28 and 90 days) specimens of 38 mm diameter and 76 mm height with compression rate of 0.76 mm/min.

The X-ray diffraction (XRD) analysis was carried out using Bruker D8 XRD instrument with a copper sealed tube X-ray source producing Cu. Samples were first milled into fine powder and then scanned with a 2θ value ranging from 5.0° to 90.02° with the step time of 1 s at a temperature of 25 °C.

Scanning electron microscopy (SEM) analysis was performed on treated and untreated samples for further assessment of the interactions between different additives and clay particles. The SEM analysis was carried out using Philips XL30 scanning electron microscope, with high vacuum pressure. Air-dried samples collected from mixed batches prior to compaction test were used for XRD and SEM analyses.

3 Experimental results and discussions

3.1 Reactivity of pozzolan

The pozzolanicity of clay, micro-silica and their combinations was evaluated by the method presented by Luxan et al. [49]. According to this method variation in conductivity within the first 2 min is taken as a measure of pozzolanic activity. Materials possess good pozzolanicity when this variation is greater than 1.2 mS/cm, variable pozzolanicity between 0.4 and 1.2 mS/cm and non-pozzolanic below 0.4. The results presented in Fig. 2a, b show that micro-silica and its combinations with clay have higher pozzolanicity than clay, which indicates that it has more dissolved silica and consequently governs a stronger pozzolanic reaction.

Fig. 2
figure 2

a Reactivity test results for clay, MS and clay + MS mixtures, b degree of pozzolanicity according to Luxan’s classification

Full size image

3.2 Compaction characteristics

The compaction curves obtained by standard Proctor compaction test are illustrated in Fig. 3. It can be observed that the addition of 10% and 20% MS has an insignificant effect on optimum water content and maximum dry density, whereas addition of 3% and 5% lime changed the optimum water content from 25 to 34% and 35.5% and maximum dry density from 1.52 to 1.31 g/cm3 and 1.35 g/cm3, respectively. A significant change was observed in maximum dry density, from 1.52 to 1.25 kg/m3 when minimum amount of additives, 10% MS + 3% lime, was added to the soil, with optimum water content changing from 25 to 27%. Addition of the maximum quantity of additives, 20% MS + 5% lime, changed the optimum water content from 25 to 38% and maximum dry density from 1.52 to 1.23 kg/m3.

Fig. 3
figure 3

Compaction curves of treated and untreated samples

Full size image

3.3 Atterberg limits

The effect of additives on plasticity index of natural and stabilized clay samples is presented in Fig. 4. It is observed that the addition of micro-silica has a very small effect on plasticity index (PI). However, when 3% and 5% lime were introduced to the soil-pozzolan mixtures, PI decreased by 65% and 72%, respectively, compared to untreated sample (PI = 34.5%). However, it was observed that while 5% lime has the lower plasticity index, almost the same results can be seen in samples with lower percentage of lime when combined with micro-silica. For example, the plasticity index of samples containing 10% + 3% lime micro-silica decreased by 70% compared to the untreated soil. This reduction is due to the occurrence of pozzolanic reaction which takes place in higher pH environment in the soil solution that is formed by addition of (OH)− ions to the soil. Therefore, by the addition of lime–micro-silica (LMS) mixtures to the soil the concentration of soil solution increases also, because of instantaneous cation exchange rate of clay minerals in the soil, cations with higher valance like additional Ca2+ ions substitute the monovalent ions and attach to the negatively charged surface of the clay almost immediately. This will ultimately lead to reduction of diffuse double layer thickness. Therefore, as the water retention capacity of the soil declines, the plasticity of the soil also decreases.

Fig. 4
figure 4

Plasticity index of treated and untreated soil

Full size image

3.4 Volume change

3.4.1 One-dimensional swell

Figure 5 depicts the effect of different quantities of lime and micro-silica on the swelling potential of treated specimens. As presented in the figure, expansive soil possesses a high swell potential of 4.86%. The results show that the addition of 3% and 5% lime reduced the swell potential by 87% and 89% to 0.61% and 0.56%, respectively. This shows that the increase in additive in lime treatment significantly reduces the swell capacity of the expansive soil. These results are in good agreement with the findings of Atterberg limits tests which showed that lime treatment of expansive soil leads to reduction of water uptake potential due to the increase in the ion concentration and instantaneous cation exchange which eventually results in reduction of the swelling potential.

Fig. 5
figure 5

Percent swell versus time curves

Full size image

Figure 5 shows that addition of 10% and 20% MS reduces the swell potential. However, based on Snethen [66] classification, the soil still possesses high swell potential. The results indicate that specimens containing combination of lime and MS show a substantially lower swelling capacity. It is also observed that the addition of MS to the mixture not only could reduce the consumption of lime, but also yields lower swelling potential. For example, swell potential of specimens treated with 5% lime is 0.56%, whereas when the mixtures of 10% MS + 3% lime and 10% MS + 5% lime are added to the natural soil, the swell potential reduces to 0.41% and 0.45%, respectively. Therefore, utilization of LMS mixture has a better effect on overcoming the swelling tendency of the expansive soil compared with lime alone. This behavior can be explained by the pozzolanic reaction between lime and silica taking place after the separation of silica from clay minerals, which takes a longer time for pozzolanic compounds to develop and bind the particles together. On the other hand, when a higher concentration of silica is introduced to the soil through the addition of micro-silica, the reaction between Ca2+ of lime and SiO2 of micro-silica takes place at a higher rate. Consequently, CSH compound developed immediately begins to coat and bind the soil particles together. Hence, a higher reduction in swell potential is obtained from LMS-treated samples compared to lime-treated ones.

It is also illustrated in Fig. 5 that secondary swell rate is dramatically reduced in the specimens treated with LMS mixture. Overall, based on the results of swelling test, it can be concluded that stabilization of expansive clay with LMS significantly improves the swelling characteristics. These findings are in good agreement with the observations from SEM micrographs to be described in later sections, which show formation of flocculated structure due to LMS addition and development of cementitious compound. As mentioned before, MS is a waste product that is produced all around the world; therefore, utilization of it is greatly beneficial for the environment. It is also found that by addition of MS, lesser amount of lime is required for the pozzolanic reaction. Based on the results of this study, 3% lime addition would be sufficient to achieve low swell potential. Therefore, the risks associated with lime stabilization, such as formation of ettringite and sulfate-induced heaves, are also reduced [26, 36, 48, 63].

3.5 Compressibility

In order to better understand the compressibility behavior of treated samples, a normalization method was applied to the one-dimensional consolidation test results. Normalization method helps to create the same initial condition for specimens treated with different additives. Figure 6 demonstrates the normalized one-dimensional consolidation curves which account for the variation in initial void ratio prior to loading. As observed from the results, addition of MS alone does not have a considerable effect on compressive behavior of the soil; however, compressibility was reduced upon the addition of LMS mixtures. The reduction in compressibility is 45% and 70% after the addition of 10% MS + 3% lime and 10% MS + 5% lime, respectively.

Fig. 6
figure 6

Normalized consolidation curves

Full size image

The rebound index (Cr) and preconsolidation pressure (\(\sigma_{\text{p}}^{\prime }\)) are also calculated from void ratio versus logarithm of effective stress graphs and presented in Table 3, which depicts that rebound index of the soil decreased by 43% after the addition 10% MS + 3% lime. This improvement is also attributed to the formation of pozzolanic compounds which binds the clay particles together. Also, preconsolidation pressure increased by threefold after the addition of 10% MS + 3% lime, which confirms that the addition of LMS develops cementitious bonds, rapidly occupying the clay voids and enhancing the strength of the soil.

Table 3 Compressibility characteristics of treated and untreated specimens
Full size table

Consolidation test results were also used to determine coefficient of consolidation (cv) and coefficient of volume compressibility (mv). These parameters were used in Eq. (3) to determine the saturated hydraulic conductivity (ksat) under 7 different ranges of consolidation pressures.

$$k_{\text{sat}} = c_{\text{v}} m_{\text{v}} \gamma_{\text{w}}$$
(3)

where \(\gamma_{\text{w}}\) is the unit weight of water.

Saturated hydraulic conductivity values under different ranges of consolidation pressure are illustrated in Fig. 7. The results show that after the addition of lime–pozzolan mixtures to the expansive soil, saturated hydraulic conductivity increases. 20% MS + 3% lime mixture shows the highest ks value at all pressure ranges. This can be attributed to the formation of flocculated structure of clay particles at short curing time which increases the size of voids between clay particles [41]. The selected optimum combination required to produce a notable effect on the swell-shrink potential, 10% MS + 3% lime, however, yielded approximately a threefold increment in hydraulic conductivity within stress range of 7–220 kPa, whereas higher increments are observed at higher stresses with respect to the hydraulic conductivity value of clay alone.

Fig. 7
figure 7

ksat values under different consolidation pressures

Full size image

3.6 Shrinkage

Figure 8 presents the results of shrinkage tests in terms of diametral (∆D/D0), axial (∆H/H0) and volumetric (∆V/V0), shrinkage strains of different mixtures. The addition of 10% MS does not have a significant effect on volumetric shrinkage strain though 20% MS reduced the volumetric shrinkage strain by 38.9% from 0.18 to 0.11. Also, 39% and 61% reduction in volumetric shrinkage strain was achieved when 10% MS + 3% lime and 20% MS + 5% lime was added to the mixture, respectively. This can be attributed to the formation of pozzolanic compounds that binds the particles together and reduces the volume change in dry seasons.

Fig. 8
figure 8

Volumetric, axial and diametral shrinkage strains

Full size image

In addition, the hyperbolic model by Fredlund et al. [35] given in Eq. (4) was used to model volumetric shrinkage test results.

$$e(w) = a_{\text{sh}} \left[ {\frac{{w^{{c_{\text{sh}} }} }}{{b_{\text{sh}}^{{c_{\text{sh}} }} }} + 1} \right]^{{\left( {\frac{1}{{c_{\text{sh}} }}} \right)}}$$
(4)

here w is gravimetric water content, e(w) is the void ratio at a given water content, ash is minimum void ratio, bsh is shrinkage limit, and csh is curvature of the hyperbola.

Shrinkage curves are shown in Fig. 9, and fitting parameters are presented in Table 4. The results show that the addition of MS increased the shrinkage limit (bsh parameter); however, when lime was introduced to the mixture, a higher increase in shrinkage limit was observed. For example, the addition of 20% MS increased the shrinkage limit by 42.8%, after the addition of 20% MS + 5% lime the shrinkage limit increased by 157.1%. The same trend was observed with the final void ratio (ash parameter), the addition of MS and LMS increased the void ratio by 38.5% and 130.7%, respectively, affirming the considerable reduction in volumetric shrinkage.

Fig. 9
figure 9

Shrinkage curves: a untreated clay, b 20% MS, c 20% MS + 5% lime

Full size image
Table 4 Hyperbolic fitting parameters of the shrinkage curves
Full size table

3.7 Unconfined compressive strength

Figure 10 demonstrates the unconfined compressive strength (UCS) test results for untreated and treated samples cured for 28 and 90 days. The results indicate that the addition of MS alone does not have a considerable effect on enhancing the strength regardless of the curing time. Addition of 3% and 5% lime increases the unconfined compressive strength by 127% and 129%, respectively, after 90 days of curing. Samples containing 20% MS + 5% lime show the highest strength value after 28 days with UCS of 1245.8 kPa, followed by only 14.9% strength increment to 1431.5 kPa after 90 days. On the other hand, 10% MS + 5% lime mixtures demonstrate 81.5% increase in strength from 940.7 kPa after 28-day curing to 1707.8 kPa after 90 days of curing exhibiting the highest strength after long-term curing which is due to long-term pozzolanic reaction and development of cementitious pozzolanic compounds [65, 73]. This behavior can be attributed to the microstructural changes observed in micrographs obtained from SEM tests as explained in the next section. The flocculated structure is observed to occur after the addition of LMS which binds the soil particles together and results in higher strength in the soil. These results show that higher percentage of MS increases the amount of soluble silica which accelerates the pozzolanic reaction rate [59]. However, if the amount of MS is more than the required amount for pozzolanic reaction, after the reaction has established, the excess amount of silica will no longer affect the microstructural changes of the soil.

Fig. 10
figure 10

Effect of curing time and different additives on UCS values of MS, lime and LMS treated samples: a 28-day curing and b 90-day curing

Full size image

On the other hand, the addition of 10% MS + 5% lime improves the soil strength by threefold after 28 days compared to untreated sample while the required curing time for 5% lime treated specimens to reach to the similar strength is 90 days. This again shows that by addition of reactive silica to the soil in the form of MS, Ca2+ ion in the lime and silica immediately engage in pozzolanic reaction. Therefore, the rate of pozzolanic compounds formation increases, whereas when there are no additional reactive silica and/or alumina, development of the pozzolanic reaction only starts as silica or alumina dissociate from clay particles.

As inferred from the results, 10% MS + 5% lime mixture demonstrates 311% increase in strength from 415.6 to 1707.8 kPa after 90 days, addition of 10% MS + 3% lime also increased the strength by 144% from 415.6 to 1013.5 kPa exhibiting a satisfactory result [67]. Therefore, it can be concluded that even though the mixture of 10% SF + 5% lime gives the highest result, 10% MS + 3% lime is selected as the optimal amount of additive for the intended expansive clay. Figure 11 demonstrates the relationship between the additive content and unconfined compressive strength.

Fig. 11
figure 11

Unconfined compressive strength of different additives

Full size image

The stiffness of treated and untreated samples was determined by calculating the secant modulus (E50). The variation of secant modulus for different mixtures after two curing periods (28 days and 90 days) is presented in Table 5. It can be noted that the addition of 10% MS + 5% lime has the highest effect on the stiffness in both curing times. With this mixture, secant modulus increased by threefolds after 90 days of curing. However, a satisfactory result was also achieved with lesser amount of lime with the addition of 10% MS + 3% lime which led to a 257% increase in secant modulus after 90 days of curing.

Table 5 Secant modulus for different mixtures in 28 days and 90 days of curing
Full size table

3.8 Scanning electron microscopy analysis

In order to further assess the interactions between different additives and clay particles, scanning electron microscopy (SEM) analysis was performed on treated and untreated samples.

The SEM micrographs of untreated expansive clay, 5% lime and combination of 20% MS + 5% lime are presented in Fig. 12a–c, respectively. As observed from the figures, clay SEM micrograph demonstrates thin sheets and dispersed structure. Although the particles appear more integrated after the addition of lime, cementitious compounds formed from pozzolanic reaction are more significantly observed in LMS-stabilized sample. This structural transformation of clay particles results in higher compressive strength and improvement of swell potential of treated soil. These findings are in good agreement with Goodarzi et al. [41], Kalkan [44] and Kalkan and Akbulut [45].

Fig. 12
figure 12

SEM micrographs of a untreated clay, b 5% lime, and c 20% MS + 5% lime

Full size image

3.9 The X-ray diffraction (XRD) analysis

XRD results of untreated expansive clay and combination of 20% MS + 5% lime are presented in Fig. 13. The results indicate that the cementitious pozzolanic are formed after the addition of 20% MS and 5% lime. A careful comparison of the XRD patterns shows the presence of reflections in the range 2θ 20.0°–31.0° that could be attributed to calcium aluminate hydrate (CAH) phase. Also, XRD pattern reveals a reflection at 2θ 11.0°–12.0° assigned to a calcium silicate hydrate (CSH) phase [5, 51, 71, 72].

Fig. 13
figure 13

X-ray diffraction of untreated and treated soil

Full size image

4 Conclusions

On the basis of the results obtained from this research it can be concluded that:

  1. 1.

    Stabilization of expansive clay with a combination of 10% MS + 3% lime reduced the swell potential by 92% from 4.86 to 0.41%, Therefore, selecting the optimum mixture not only would result in lower swelling potential but also would reduce the consumption of lime.

  2. 2.

    Hydraulic conductivity values of samples treated with 10% MS + 3% lime yielded approximately threefold increment in hydraulic conductivity values compared to untreated clay, under stresses of 7–220 kPa.

  3. 3.

    Maximum volumetric shrinkage results showed that although the addition of MS increased the shrinkage limit, due to the formation of pozzolanic compounds which bind the particles together, when lime was introduced to the mixture, a higher increase in shrinkage limit was observed.

  4. 4-

    10% MS + 3% lime mixture demonstrated 144% increase in strength after 90 days and was selected as the optimal amount of additive for the intended expansive clay. Consequently, addition of 10% MS + 3% lime increased the stiffness, demonstrated by secant modulus (E50), by 257% after 90 days of curing.

It is finally concluded that while the addition of micro-silica (MS) alone has less effect on the engineering parameters of expansive clay, the addition of lime–micro-silica (LMS) provided promising results in stabilization of expansive soil. According to different tests, reaction between Ca2+ of lime and SiO2 of micro-silica and the formation of a cementitious pozzolanic compound bind the clay particles together. Therefore, by stabilizing the soil with LMS not only the pozzolanic reaction takes place at a faster pace, but a lower percentage of lime is required which lowers the risks associated with lime stabilization, such as formation of ettringite and sulfate-induced heaves. Therefore, utilization of MS, which is an industrial waste, enhances the engineering properties of soils as well as providing an application for recycling against degradation of the environment.