Keywords

1 Introduction

Mineral quantification plays a vital role when dealing with expansive soils, as they grossly account for the consistency and swelling behavior. However, the quantification of clay minerals is a challenging issue due to their complex chemical compositions and effect of particle orientations [1, 2]. Compared to all associated minerals in expansive soils, montmorillonite is having a high surface and high cation exchange capacity. Considering it, several studies also report that mineral content can be identified indirectly from cation exchange capacity [3] and chemical balance method [4], where the latter method was successful up to the mark. On the other hand, Chittoori [5] developed a new regression model for the quantification of clay minerals based on chemical soil properties such as cation exchange capacity (CEC), total potassium, and specific surface area (SSA). Analytically, X-ray diffraction (XRD), transmission electron microscopy (TEM), infrared analysis (IR), and differential thermal analysis (DTA) are in use for the identification of clay minerals in soils [6], which are expensive and available at a limited number of laboratories [4]. Nevertheless, XRD is the only technique that in practice for the quantification of minerals from early 1960s.

Needless to state, the enormity in consistency limits of expansive soils largely depends on different parameters relevant to physical, chemical, and mineralogy [7]. But, undoubtedly, not much attention was paid to comprehend the exact influence of MMC on geotechnical properties, in particular on consistency limits. This paper is themed to address important points related to the quantification of montmorillonite content for understanding its impact on swelling behavior and consistency limits. For the study purpose, expansive soil samples were collected from nine different regions across India. MMC, and consistency limits of these soils were determined, and appropriate interrelations were developed. The compilation of data was prompted to develop separate sets of correlations for bentonite soils (BS), having extreme consistency values, and natural expansive soils (NES), having low to very high consistency values, and the same was accomplished in the study.

2 Materials and Testing Methodology

Experimental investigations were conducted on expansive soil samples collected from different regions across India. The clay content of the soil samples was determined by hydrometer analysis (<75 μm) (ASTM D7928-17) [8], and consistency limits (liquid limit, wL, plastic limit, wP, plasticity index, wPI, shrinkage limit, wSL and shrinkage index, wSI) were determined as per the ASTM D4318-00 [9] and ASTM D427–98 [10] standards. The swelling potential of the samples was determined as per the guidelines of ASTM D2435–04 [11]. The physical properties of soil samples used in the current study are listed in Table 1.

Table 1 Physical properties of soil samples used in the current study
Full size table

2.1 XRD Analysis

The mineralogy of soil samples used in the current study is analyzed by employing to D8 Advanced X-ray powder diffraction device (make, BRUKER, USA). For testing purpose, the clay content of the soil samples was separated by following the procedure of Rao et al. [12]. The obtained samples were placed in the equipment and scanned with a voltage of 40 kV and a current of 40 mA. Further, the 2θ in the range of 5–80°, step size of 0.025°, and time interval of 0.5 s are maintained, and a copper X-ray tube (i.e., Cu-Kα radiation) was used. The presence of different minerals, especially targeting montmorillonite content (MMC) in the soil samples, was analyzed with the help of software DIFFRAC. EVA, which is already equipped with the equipment. Further, the quantity of MMC is identified by using the software TOPAS 4.2, which performs a whole pattern followed by Rietveld analysis. The MMC of soil samples determined from the analysis is listed in Table 1. The detailed procedure for the quantification of minerals can be found in Rao et al. [12].

3 Results and Discussion

Consistency limits are the mostly used parameters for identifying the presence of clay minerals in fine-grained soils. Among them, liquid limit, wL and plasticity index, wPI are the best for approximation of mineralogy of fine-grained soils [7, 13]. With this in mind, an attempt was made to deduce the relationship between consistency limits and montmorillonite content, MMC as depicted in Fig. 1a, b. The maximum values of wL and wPI of the soils used in the study are measured as 93 and 57%. In order to validate the present results, the related data have been collected from the literature and superimposed on Fig. 1.

Fig. 1
The image depicts two graphs: liquid limit Versus M M C and plasticity index Versus M M C. Both show straight lines with less than 45-degree angles. It uses some geometrical shapes for different places.

Influence of MMC on liquid limit and plasticity index of soils used in the study

Full size image

As depicted in Fig. 1a, b, both wL and wPI increased linearly with an increase in MC. As MMC increases so does wL and wPI. Prominently, the data of Fig. 1a, b reflect both NES and BS. It is observed that the values of wL and wPI of BS are considerably higher vis-à-vis with those of NES. It is seen from Fig. 1a, b that the value of wL as high as 993% and wPI up to 950% have been reported, especially, for BS. It is, in general, seen from Fig. 1a, b that MMC in BS is reported significantly high (>50%) vis-à-vis with those of NES, for which MMC is measured below 50%. Consequently, it is prudent to make an inference that soils rich in MMC of above 50% can exhibit an average wL of above 100% and wPI of above 60%, respectively. These observations well corroborate with the results of Croft [14], who has reported that soils with wL of greater than 60% and wPI of greater than 25% evidently consist of expansive clay minerals such as montmorillonite.

As such, a linear fit model was employed to mathematically correlate wL and wPI with MMC. The empirical equations derived based on fitting functions are also printed in the same graph, alongside the values of constants. Incidentally, the values of regression coefficients, R2, pertinent to wL are 0.78 and 0.82 for wPI. It is obvious from the trends shown in Fig. 1a, b that when MMC is extrapolated to zero, soils yet exhibited wL of 32% and wPI of 19%. Based on these observations, it can be theorized that soils without comprising of expansive minerals like montmorillonite, they still exhibit consistency behavior. This may be attributed to the fact that soils in addition to possessing minerals, like montmorillonite, might also constitute with mixed layer minerals such as illite–smectite, which would cause soils to exhibit consistency behavior. The findings of studies by Prakash and Sridharan [13] and Reddy et al. [2], who reported that the mixed layer minerals could induce swelling to the expansive soils, excellently validate this statement.

Shrinkage index, wSI of soil can be defined as the difference between plastic limit, wPI and shrinkage limit, wSL. A close observation of Fig. 2a, b shows that wSL and wSI decreases with an increase in MMC. These observations are contrary to the results of Fig. 1. The maximum and minimum values of wSL and wSI of soils are measured as 8.5–15.32% and 46.81–110.15%, respectively. Based on the measured wSL values, the volume change behavior of soils used in the study can be categorized as: non-critical to critical [15]. Furthermore, as per IS 1498 that classifies the soil based on wSI, soils used in the study exhibited high to very high swell potential [16]. Many studies report that wSL and wSI may not be better parameters for predicting the swelling behavior of expansive or fine-grained soils. It is because these properties are not considered as plasticity characteristics, and the mechanism behind these properties is different from swelling. The swelling is due to the expansion of clay mineral lattice, whereas, in the case of wSL, it is chiefly managed by the relative grain size distribution of expansive or fine-grained soils [17].

Fig. 2
The illustration depicts two graph. Graph 1 illustrates shrinkage limit values from 0 to 20, with an interval of 4, Versus M M C. Graph 2 illustrates the shrinkage index from 0 to 120, with the interval of 20, Versus M M C from 0 to 60 minutes, with an interval of 10. They have some geometrical shapes that are used for different places.

Influence of MMC on shrinkage limit and shrinkage index of soils used in the study

Full size image
Fig. 3
A bar chart of Liquidity index values from minus 0.25 to 0.5 Versus M M C values from 0 to 50. All the charts show negative data where each bar is used for a place.

Variation of liquidity index with MMC of soils used in the study

Full size image

Liquidity index (wLI) merely indicates the moisture condition of a soil. It becomes negative when the natural moisture content of soils lies below wPI, which generally happens for overconsolidated clays or semi-plastic solids. Figure 3 depicts the relationship for wLI versus MMC of respective soil samples. Observance of both +ve and –ve values is possible from Fig. 3. A close examination of Fig. 3, in general, reveals that wLI could predominantly be –ve for expansive type soils. It is clearly seen that most of the data fall below zero (i.e., –ve), but a few data points are even +ve (i.e., above zero). This highlights the variability in consistency of expansive soils used in the present study. These statements well corroborate with the observation of Chen [18], who reported that expansive soils exhibit wLI of –ve. It is seen from Fig. 3 that wLI decreased with an increase in MMC. As such, the results presented in Fig. 3 are found to matching excellently with those reported by Shi et al. [19].

Many studies confirmed that MMC and its associated minerals, such as illite–smectite, kaolinite-smectite, which are categorized as mixed layer minerals, inherently govern the swelling behavior of expansive soils [2]. With this in mind, an attempt is made to correlate MMC with swelling potential, Sa. From Fig. 4, it can be observed that Sa increased linearly with an increase in MMC in the expansive soils. It is obvious that as MMC increases so does Sa. In order to validate the present results, the relevant data have been collected from the literature, and the same is used to validate the data produced by the present study. Additionally, a linear fit is employed, and empirical equation along with values of constants is printed in the figure. The regression coefficient, R2, value of 0.58 was obtained for MMC versus Sa. The following observations can further be made from the results presented in Fig. 4: (a) the maximum MMC measured in NES is 42%; (b) the maximum Sa measured of these soils is 15%; (c) Sa consistently increased with MMC, for its whole measured range from 0 to 42%, and (d) MMC has a significant effect on Sa of expansive soils. The similarity in trends of Figs. 1a, b and 4 reveals that there exists a relationship between these parameters.

Fig. 4
A graph of swelling potential from 0 to 30 Versus M M C from 0 to 80 shows a straight line with less than a 45-degree angle. Some geometrical shapes are used for different places.

Variation of swelling potential with MMC of soils used in the study

Full size image

4 Conclusions

In this study, different expansive soil samples were collected from different regions across India and were characterized for montmorillonite content. From the extensive experimental data, several correlations separately for natural expansive soils and mineral-rich bentonite soils were developed, and the same were validated with the literature data. Based on the correlations and further interpretations of the results, the following conclusions were derived:

  1. 1.

    Demonstrably, it is the montmorillonite mineral and its content is a prime controlling factor prompting consistency and swelling behavior in expansive soils.

  2. 2.

    Interpretation of the results evidently manifests that variability in the data of montmorillonite content, consistency limits, and swelling potential is intrinsic, in particular, pertinent to expansive soils.

  3. 3.

    The maximum values of MMC, wL, wPI, wSL, wSI, wLI , and Sa of expansive soils used in the study are measured as: 44%, 93%, 57.68%, 15.32%, 110.15%, −0.05, and 18%, respectively.