Mechanical behaviour and SEM-based microstructural interpretation of Grewia optiva fiber-reinforced expansive soil

Abstract

Introduction:

Expansive black cotton (BC) soils exhibit poor strength and significant volumetric instability, creating challenges for infrastructure development. This study evaluated Bhimal (Grewia optiva) natural fiber as a sustainable discrete reinforcement for improving the geotechnical response of BC soil.

Methods:

Fiber contents of 0.25, 0.50, 0.75, 1.00 and 1.25% and fiber lengths of 15, 20 and 25 mm were investigated using Modified Proctor, direct shear and unconfined compression tests. Two-way analysis of variance (ANOVA) and regression analysis were used to evaluate the effects of fiber content and length, while scanning electron microscopy (SEM) provided qualitative support for soil–fiber interaction mechanisms.

Results:

The selected optimum configuration, M13 (0.50% fiber, 25 mm), increased maximum dry density from 1.53 to 1.69 g/cm³ (10.46%) and optimum moisture content from 11.00 to 15.50% (40.91%). UCS increased from 55.6 to 68.0 kPa (22.30%) and cohesion from 51.0 to 60.5 kPa (18.63%). ANOVA showed significant effects of fiber parameters, with fiber length exerting the stronger influence. SEM observations provided localized qualitative evidence of soil–fiber interlocking, particle attachment, crack bridging and pull-out mechanisms.

Discussion:

The classical UCS ≈ 2c relationship showed weak agreement with the reinforced-soil data (R² = 0.088), indicating that this simplified cohesive-soil correlation is not directly applicable to Bhimal fiber-reinforced expansive soil. The results support Bhimal fiber as a promising sustainable reinforcement while highlighting the need to account for altered stress-transfer and failure mechanisms in fiber-reinforced soils.

1 Introduction

Black cotton (BC) soil is common in many areas of southern and western India; it poses a challenge to the entire civil engineering infrastructure because of its large volume change behavior [1]. They have high swelling and shrinkage capacities owing to changes in moisture; hence, they have low bearing capacity and shear strength and progressively disrupt shallow foundations, pavements and embankments [2]. Traditional soil stabilization methods, such as lime, fly ash, GGBS, cement and chemical additives, have been widely used to improve the engineering performance of problematic soils [3–5]. Although these methods are quite successful in improving the strength and reducing the swelling, their use is frequently constrained by environmental issues, high material costs, and the growth of carbon emissions related to the use of binders [6]. In addition, the stabilized soil exhibited brittle behavior and low ductility under cyclic loading [7]. This has prompted researchers to explore alternative, sustainable, cost-effective, and environmentally friendly soil improvement methods [8]. Here comes the natural fiber reinforcement as an environmentally acceptable method of soil stabilization, which has received considerable attention.

Several natural fibers including jute, coir, sisal, banana and bamboo have been used to improve the behaviour of weak and expansive soils [9,10]. In jute reinforced expansive soil, randomly distributed fibers can improve the tensile strength, crack bridging ability and ductility [11,12]. Similar improvements have been reported for coir and sisal reinforced black cotton soils also, where inclusion of fiber improved the load resistance, reduced brittle failure and improved the composite behaviour of soil matrix [13]; [14,15]. Studies have indicated that natural fibers can serve as discrete tensile reinforcements in expansive soils [16]. However, the efficiency of the natural fibers is greatly influenced by fiber type, aspect ratio, surface texture, dosage, length, interaction with moisture and soil-fiber bond [17].

The Bhimal fiber (Grewia optiva) is a lignocellulosic fiber found in the Himalayas and traditionally used in rural applications due to its tensile resistance and flexibility [18,19]. Grewia optiva soil has received comparatively limited attention as a soil-reinforcement material [20,21] and previous studies have been conducted on polymer composites and hybrid bio-composites. Hence, the usage of Bhimal fiber as a reinforcement material for expansive black cotton soil is not much studied. This gap is important because the mechanical response of natural fiber reinforced soil cannot be generalized from a fiber type to other types of fibers it depends on fiber morphology, water absorption, surface roughness and interaction with clay particles. This study addresses this gap by investigating the effect of fiber content and length of Bhimal fiber on compaction, shear strength, UCS response and soil-fiber microstructural interaction of expansive black cotton soil.

The experimental program systematically varied the fiber content (0.25%–1.25% on dry weight of soil) and fiber length (15 mm–25 mm). The behaviour of reinforced soil was studied using heavy Proctor compaction, direct shear and unconfined compressive strength tests. Microstructural analysis was carried out to study the interfacial mechanism between the fibers and soil through scanning electron microscopy (SEM). Two-way analysis of variance (ANOVA) and regression analyses were performed.

The present work is on expansive black cotton soil reinforced with the Bhimal (Grewia optiva) fiber as a soil fiber composite system instead of conventional stabilized soil. The primary physical aim is to examine the effect of randomly distributed natural fibers on the deformation and failure response of expansive soil through interfacial friction, mechanical anchorage, crack-bridging, fiber pull-out and stress redistribution. The experimental program is to study the effect of fiber content and fiber length on the compaction behaviour, shear strength, unconfined compressive strength and failure characteristics of fiber reinforced soil. The macroscopic response observed is related to interaction mechanisms at the soil-fiber particle scale as revealed by SEM observations. Therefore, the present study aims to provide a physical insight into the brittle-to-ductile transition and strength enhancement of a sustainable natural fiber reinforced geomaterial.

2 Materials and methodology

2.1 Materials

2.1.1 Soil collection and characterization

Black cotton (BC) soil was used for the research, which is considered highly plastic [22]. Soil was obtained from Rajkot, Gujarat, India at a depth of 1.0 m on a site to reduce the chances of organic matter. For laboratory testing, the Soil was air-dried and passed through a 90 μm (0.090 mm) to obtain a uniform sample as shown in Figure 1, and the properties of black cotton soil are listed in Table 1.

FIGURE 1

2.1.2 Bhimal fiber

The natural reinforcement used in this study was Bhimal fiber (Grewia optiva), which was obtained from bhimal plant shown in Figure 2 [18,20]. The extracted fibers were manually cleansed, separated, and cut into three selected lengths: 15 mm (short), 20 mm (medium), and 25 mm (long). The physical properties of the Bhimal fiber were tested in the laboratory and are listed in Table 2.

FIGURE 2

The fiber length and aspect ratio in this table are shown as the extracted conditions of Bhimal fiber. Then the fibers were cut into lengths of 15, 20 and 25 mm respectively as per the experimental mix design for soil reinforcement. Bhimal fiber has high aspect ratio, tensile strength and surface irregularity that is responsible for its role as a mechanical reinforcement. High water absorption explains the increased moisture demand in reinforced soil mixtures.

The measured fiber properties demonstrate that Bhimal fiber is sufficiently strong and stiff to be used as discrete reinforcement in the expansive soil matrix. The large variation in tensile strength is expected for natural lignocellulosic fibers due to variation in diameter, maturity, extraction condition and local defect. In the reinforced soil, the contribution of these fibers is mainly by mechanical anchorage, frictional resistance and crack-bridging rather than by chemical bonding.

2.2 Sample preparation

To achieve the required moisture content for testing, distilled water was added to the sieved, air-dried soil. For the fiber reinforced specimens, predetermined weights of Bhimal fiber as a percentage of the mass of dry soil were slowly added to the wet soil. The mixture was manually mixed for about 10 min to achieve an even fiber distribution and to reduce fiber agglomeration. The mixtures were then transferred to sealed sampling bags, wrapped in two layers of polythene, stored for 24 h to allow moisture equilibration and then tested.

2.3 Experimental program and design

A systematic experimental program was planned to study the effects of two independent variables, fiber content (0.25%, 0.5%, 0.75%, 1.0%, and 1.25% by dry weight of soil) and fiber length (15, 20, and 25 mm) on UCS. Sixteen different mixtures including an unreinforced control specimen were prepared and tested. The experimental matrix is presented in Table 3.

Specimen preparation and testing were performed with repeatability in mind. For each mix, three cylindrical specimens were prepared and tested for UCS test and the mean value was taken for reporting and interpretation. The direct shear test was conducted on each mix at three normal stress levels (50, 100 and 150 kPa) to determine the Mohr–Coulomb failure envelope. The tests are for different loading conditions and not repeated tests at the same normal stress. The moisture content–dry density relationship was established for each mix in the Modified Proctor test and the test was repeated where inconsistent or scattered compaction points were observed. The UCS results were supported by triplicate testing, and the direct shear and Proctor results were used to support trends within the experimental matrix.

2.4 Testing procedures

All laboratory tests were performed in accordance with the relevant Indian Standard (IS) codes of practice for soil characteristics and ASTM for fiber tensile characteristics.

2.4.1 Modified proctor test

The compaction characteristics, namely, the maximum dry density (MDD) and optimum moisture content (OMC), of all 16 mixtures were determined according with IS Code (IS 2720 P-8 1984). The Modified Proctor test was conducted using a 4.89 kg rammer with a drop height of 450 mm. The soil was compacted in five equal layers within a 1000 cm³ mould, with each layer receiving 25 blows to achieve the desired compactive effort. Compaction curves were plotted to determine the MDD and corresponding OMC for each mixture. The resulting variations of MDD and OMC with fiber content and fiber length are presented in Figure 3.

FIGURE 3

2.4.2 Direct shear test

The shear strength parameters, that is, cohesion (c) and angle of internal friction (ϕ), were obtained for all mixtures using a standard direct shear apparatus that complied with the requirements of IS Code (IS 2720 P-13 1986). The sample was compacted in a shear box (60 mm × 60 mm × 25 mm) at the respective maximum dry density (MDD), and optimum moisture content (OMC).

The tests were conducted at three normal stress levels of 50, 100, and 150 kPa, at a constant strain rate of 0.625 mm/min. The Mohr-Coulomb failure envelope derived using the values of the maximum shear stress was used to obtain the shear strength parameters.

2.4.3 Unconfined compressive strength (UCS) test

The unconfined compressive strength (UCS) test was performed by following the guidelines of (IS 2720 P-10 1991) on cylindrical specimens with a diameter of 38 mm and a height of 76 mm. The specimens were prepared at the maximum dry density (MDD) and optimum moisture content (OMC) using static compaction. The specimens were stored in a desiccator before testing to balance the moisture. For each mix, three replicate UCS specimens were tested, and the average UCS value was used in the results and statistical analysis.

The tests were performed in a strain-controlled loading frame at a constant strain rate of 1.25 mm/min. The peak loading at breaking was noted, and the UCS was obtained based on Equation 1:

Where P is the compressive load at failure (Newton) and A is the initial cross-sectional area of the specimen (mm²).

2.5 Microstructural examination via scanning electron microscopy (SEM)

Representative fragments of the optimum reinforced mixture M13 (25 mm fiber length and 0.50% fiber content) were examined by scanning electron microscopy (SEM). SEM examination gave qualitative mechanistic support to interpret soil–fiber contact, particle attachment, crack-bridging behavior, interfacial debonding and fiber pull-out features in the optimum reinforced soil matrix. Comparative SEM observations of the unreinforced control soil and other reinforced mixtures were not performed in the present study and are identified as a limitation and suggested for future work. Small fragments were carefully collected from the failed UCS specimens to preserve the failure surface and the soil–fiber interface zone.

The collected fragments were oven-dried at 60 °C to remove residual moisture while minimizing disturbance to the soil–fiber contact features. The dried samples were mounted on aluminium stubs using conductive adhesive tape and coated with a thin layer of gold to improve surface conductivity during imaging. SEM observations were carried out at an accelerating voltage of 10–15 kV and at magnifications ranging from ×50 to 5000×. Particular attention was given to fiber surface texture, soil particle attachment, interfacial contact, crack propagation, fiber pull-out, and the evidence of mechanical interlocking between the soil matrix and Bhimal fibers.

2.6 Statistical analysis

Statistical analysis was performed to assess, in measurable terms, whether the Bhimal fiber reinforcement and geotechnical properties of black cotton soil were statistically significant. Several statistical tests were used in the analysis to evaluate the individual and interactive effects of the experimental variables [23].

The statistical analysis was interpreted considering the repeatability of the experiment. The UCS data was supported by three replicate tests for each mix. Direct shear tests were conducted at three normal stress levels in order to obtain the Mohr–Coulomb failure envelope and Modified Proctor tests were repeated when inconsistent compaction points were observed. Thus, ANOVA, effect size and power analysis were used to reinforce trends observed within the present experimental matrix, rather than to propose a universal predictive relationship.

2.7 Two-way analysis of variance (ANOVA) and post-hoc analysis (Tukey’s HSD test)

When ANOVA indicated a statistically significant difference among treatment groups, Tukey’s honestly significant difference (HSD) test was applied for pairwise comparison of group means [24]. This post-hoc test was used to identify which specific fiber-content or fiber-length groups differed significantly while controlling the family-wise error rate associated with multiple comparisons. The Tukey HSD results were used to support the interpretation of optimum fiber combinations and to verify whether the observed differences among selected mixtures were statistically meaningful.

2.8 Regression modeling

Multilinear regression analysis was performed to build predictive correlations between fiber parameters and strength properties [25]. The overall form of the model is as per Equation 2:

Y is the response variable of interest, C is the fiber content expressed as a percentage, L is the fiber length measured in mm, and βᵢ is the regression coefficient.

The coefficient of determination (R2), adjusted R2, and significance of the regression coefficients were used to determine model adequacy. Pearson correlation coefficients were calculated to investigate the correlation between compaction and strength parameters.

3 Results and discussion

The experimental results of all 16 mixtures (M1-M16) are reported and discussed to evaluate the effect of the Bhimal fiber content and length on the compaction and strength properties of black cotton (BC) soil.

The improvement of strength and ductility observed is consistent with previous studies of jute-, coir- and sisal-reinforced expansive soils in which natural fibers acted as discrete tensile inclusions and improved the response of the soil by crack bridging, interfacial friction and resistance to brittle failure. However, the present work differs from most previous studies by focusing on Bhimal (Grewia optiva) fiber, for which prior applications have mainly been reported in polymer composites rather than expansive soil reinforcement.

3.1 Compaction characteristics

The highest MDD was obtained for M13 with 0.50% fiber 25 mm in length. The MDD increased from 1.53 g/cm³ for the control soil to 1.69 g/cm³ for M13, which corresponds to an increase of 10.46%. The 15 mm fiber mixtures showed M4 to have the highest MDD at 1.680 g/cm3 with 0.75% fiber content. In the 20 mm fiber series, M10 displayed the highest MDD of 1.66 g/cm³ with 1.00% fiber content. This indicates that the optimum fiber content for maximum densification was a function of fiber length.

The initial increase in dry density may be due to the better mechanical interlocking between soil particles and fibers which resulted in better packing arrangement of soil-fiber matrix. However, MDD decreased above the optimum fiber content for a given fiber length. This decrease is probably related to the lower density of natural fibers relative to soil solids and the higher probability of fiber clustering at higher fiber contents. This clustering can increase void spaces and reduce the compaction efficiency.

OMC was also increased after fiber inclusion. The control soil had an OMC of 11.00% while M13 had an OMC of 15.50% which is 40.91% increase. In 20 mm fiber series, some mixes showed higher moisture demand such as M7 with OMC 21.50% and M10 with OMC 18.50%. This increase is mainly due to the hydrophilic nature and water absorption capacity of Bhimal fibers which demands additional water during mixing and compaction, this interpretation is also consistent with the measured 24 h water absorption value of fiber reported in Table 2 and with earlier studies on the natural fiber behaviour [17,18]

The compaction results reveal that the fiber length and fiber content should be optimized jointly. The 20 mm fiber series revealed higher OMC values for some combinations, but M13 was found to be the most balanced in terms of compaction response by achieving the highest MDD with a moderate OMC compared with several other reinforced mixtures.

3.2 Shear strength parameters

The basic parameters (cohesion (c) and angle of shearing resistance (ϕ)) of shear strength obtained from direct shear tests are also used to evaluate the stability of slope and the performance of foundation. The addition of Bhimal fiber had a significant effect on the shear strength behavior of the BC soil, as shown in Figure 4. Cohesion varied with fiber length and content; several reinforced mixtures showed improvement, with the maximum value of 60.5 kPa obtained for M13.

FIGURE 4

This improvement in the level of cohesion can be explained by the better soil–fiber interaction mechanisms. The Bhimal fibers have a rough surface at the interface, which promotes interparticle and frictional resistance between the soil and fibers. The random distribution of the fibers under shear loading provides tensile inclusion between the possible planes of failure, mobilizes further resistance, and increases the apparent soil cohesion [11,12,16,17].

A significant improvement was also observed in the angle of shearing resistance (ϕ). The angle of internal friction of the virgin soil was 3.43°, which is small, representing only 26.16% of the maximum angle of the mix at the highest content (20 mm fiber length, 6.47°). However, in the case of mixed M13, ϕ was 4.19°, indicating that cohesion was maximized at this combination; however, the friction angle was not.

A higher friction angle of M10 indicates that the effectiveness of the fibers in increasing frictional resistance is related to the fiber length-content interaction and the aspect ratio. An optimal aspect ratio facilitates enhanced interlocking and mobilization of frictional resistance within the soil matrix. When the dosage of an ingredient is low, an extreme length is likely to increase the bonding (cohesion) component more than an intermediate length would increase the frictional component.

In general, it can be concluded that Bhimal fiber reinforcement improves both the cohesive and frictional elements of the shear strength, and fiber combinations have different impacts on these factors. M13 exhibited maximum cohesion, and the friction angle was the highest for M10.

3.3 Unconfined compressive strength (UCS)

The unconfined compressive strength (UCS) is a basic parameter used to determine the capacity of cohesive soils to carry weight while keeping the soil unconfined. This is of particular importance for short term stability evaluation and subgrade performance. The strength of black cotton (BC) soil is increased by the reinforcement of Bhimal fiber as shown in Figure 5. The UCS of the control specimen (M1) was 55.6 kPa. UCS varied with fiber length and content, as summarized in Table 4, and the highest value of 68.0 kPa was obtained for M13 (25 mm fiber length with 0.50% fiber content). In particular, the maximum strength was obtained for an intermediate fiber content and not at the highest loading, indicating an optimum relationship between fiber length and content. This shows the need for a balance between the fiber dispersion in the soil matrix and the reinforcement’s effectiveness. Using too much fiber is likely to result in clustering and stress concentration, which reduces the effectiveness of the entire structure.

FIGURE 5

3.3.1 Failure behavior and ductility

In addition to increasing the peak strength, fiber reinforcement altered the failure behavior of the soil. The unreinforced control specimen (M1) exhibited lower post-peak resistance and energy absorption capacity as compared to the reinforced specimens. In the control curve, the drop was not sharp but the stress–strain response indicated the limited capacity to sustain the deformation after the peak stress. On the other hand, the post-peak response of the reinforced specimens especially M5, M10 and M13 was improved and energy absorption capacity was increased. The randomly dispersed fibers were responsible for arresting and bridging the microcracks, carrying tensile stresses across the developing microcracks and therefore retarding the crack propagation. The major improvement obtained from fiber addition was not only in peak strength but in a better deformation tolerance and a higher energy absorption during loading. This type of behavior is particularly useful for pavement subgrades and low-loaded foundations, where tolerance to deformation is important.

Overall, the findings indicate that Bhimal fiber reinforcement improves not only the compressive strength but also the deformation response and failure behaviour of expansive BC soil.

3.4 Synthesis of results and optimal mix identification

Table 5 presents the combined analysis of the compaction and strength properties, which allows for the determination of the best fiber combination. The addition of fiber led to a slight change in the maximum dry density (MDD) of 1.53 g/cm³ (control) to 1.69 g/cm³ with mix M13 (25 mm, 0.50%), amounting to an increase of 10.46%. Conversely, the maximum moisture content level also increased significantly between 11.00% and 15.50% (40.91%), which shows the hydrophilic nature of Bhimal fibers and their capacity to absorb water. Improvements in compaction were not as significant as those in strength enhancement. The unconfined compressive strength of untreated soil was 55.6 kPa, whereas that of M13 was 68.0 kPa, which was 22.30% higher. Similarly, the cohesion value increased to 60.5 kPa (18.63%), which is consistent with the added cohesion in the presence of fiber reinforcement using the crack-bridging and soil–fiber interaction processes.

M13 recorded an increase of 22.16% in its angle of internal friction (friction angle increased to 4.19°). However, the most striking point is that the highest friction angle (6.47°) was recorded in mix M10 (20 mm, 1.00%), thus showing that the aspect ratio and content of a fiber shape the frictional resistance more than cohesion itself.

Although M10 exhibited the largest friction angle, M13 demonstrated the most balanced performance as a whole in terms of the highest UCS, highest level of cohesion, better dry unit weight, and medium moisture demand compared to other reinforced mixes. Thus, based on both increased strength and compaction efficiency, it can be concluded that the mixed 25 mm fiber length at a 0.50% content (M13) yielded the best results in enhancing the strength of black cotton soil.

3.5 Statistical analysis

3.5.1 Two-way ANOVA results for black cotton soil properties

Two-way analysis of variance (ANOVA) was used to test the individual and interactive effects of fiber content and fiber length on the engineering behavior of black cotton soil.

3.5.1.1 Unconfined compressive strength (UCS)

According to the results (Table 6; Figure 6) of ANOVA, UCS, fiber content (p = 0.002) and fiber length (p = 0.001) significantly affected the compressive strength at a 95-percentage level. The strength-enhancing effect of fiber length showed a stronger influence (F = 13.37) than that of fiber content (F = 8.94), indicating that length is the most important variable in enhancing strength. The Fiber length-content interaction was non-significant (p = 0.122), indicating relative independence in the effects of the two factors on UCS. These results verify the experimentally determined optimum blend of 25 mm long fibers with 0.50% content.

FIGURE 6

3.5.1.2 Cohesion (c)

In terms of cohesion the strongest effect was observed for fiber length (F = 14.08, p < 0.001) (Table 7; Figure 7). The fiber content was borderline significant (p = 0.048); therefore, the effect was moderate but significant. The non-significant interaction term (p = 0.086) indicates that the increase in cohesion was mainly associated with the individual effects of fiber length and fiber content rather than their combined interaction. The highest level of cohesion improvement of 18.63% was observed for the optimum combination (M13).

FIGURE 7

3.5.1.3 Interaction analysis across properties

The interaction p-values for all investigated properties (Table 8; Figure 8) were greater than 0.05.

FIGURE 8

3.5.1.4 Effect size and statistical power

The effect size (η²) and statistical power (1 − β) were evaluated (Table 9), and the F-value comparison across properties is shown in Figure 9.

FIGURE 9

Because η² > 0.14 indicates large effects, all tested properties exhibited large practical significance. The statistical power exceeded 0.80 in all cases, indicating adequate sensitivity for detecting true treatment effects.

3.5.2 Regression analysis and model limitations

The relationship between the UCS and the fiber content was quantified using linear regression analysis as shown in Figure 10. Regression Equation 3 for individual fiber lengths is written below Equations 4–6:

FIGURE 10

where UCS is the unconfined compressive strength in kPa and C is the fiber content in percent.

The positive slopes indicate that increasing the fiber content enhances the compressive strength; however, the magnitude of improvement varies with the fiber length.

A multiple linear regression model has been developed using the 15 reinforced soil mixtures to describe a combined effect of fiber content and fiber length on UCS. This regression did not include the control mix because it did not contain fiber reinforcement. The combined multiple linear regression model is expressed in Equation 7:

where UCS is unconfined compressive strength in kPa, C is fiber content (%), and L is fiber length in mm. The combined model provided R2 = 0.588 and adjusted R2 = 0.519. The model was significant in the tested data set with a model p-value of 0.0049. The coefficients of fiber content and fiber length were significant as well with p-values of 0.009 and 0.019 respectively. The residual error was moderate, RMSE = 3.96 kPa and MAE = 3.07 kPa. These values show that the model captures the general effect of fiber content and fiber length on UCS but do not explain all the variability in the measured strength.

There was no strong systematic trend in the residuals but several mixtures had conspicuous deviations from the fitted line. This is expected since the UCS response of natural fiber reinforced soil is a function of fiber content, fiber length, fiber orientation, local dispersion, soil-fiber contact, specimen heterogeneity and variability of compaction. Therefore, the regression equation should be interpreted as a trend-supporting relationship within the tested fiber content range of 0.25%–1.25% and fiber length range of 15–25 mm. It should not be extrapolated beyond these limits or used as a general design equation without additional experimental data. This model suggests that both fiber content and length contribute to the UCS response within the tested experimental range.

3.5.3 Optimal mix vs. control comparison

Table 10 and Figure 11 provide a direct statistical comparison between the control soil (M1) and the optimum mix (M13). The parameters showed statistically significant changes (p < 0.05). The effect size (Cohen’s d) values were medium and large (UCS, (d = 1.85) and OMC (d = 2.45)) and had the strongest treatment effects. These findings provide a statistical support of the choice of M13 (25 mm, 0.50%) as the optimal mix.

FIGURE 11

3.6 Microstructural analysis (SEM)

The SEM analysis was carried out on optimum reinforced mixture M13 to provide qualitative evidence for the soil-fiber interaction mechanisms leading to the observed mechanical improvement. The SEM results are not provided as a comparative microstructural validation of all mixtures, because the SEM images of the control soil and other reinforced mixtures were not available. Rather, the SEM observations serve to back the interpretation of particle attachment, interfacial contact, mechanical interlocking, crack bridging, fiber pull-out and interfacial debonding in the optimum reinforced matrix.

The reinforcement process of black cotton soil with Bhimal fiber M13 was examined using (SEM) analysis. Before loading the machine, Figure 12 shows the strong interfacial bond between the fiber strands and soil particles recorded before mechanical loading. When viewed under high magnification (x4500), the soil particles are firmly attached to the top of the rough fibers, which shows that there is strong mechanical interlocking as well as friction resistance.

FIGURE 12

The SEM image at low magnification (×140) shows a localized area of a Bhimal fiber embedded in the surrounding soil matrix. This observation indicates local contact between soil and fiber and particle attachment around the fiber, but should not be interpreted as evidence of homogeneous fiber distribution throughout the whole specimen. Thus, the SEM observations are considered as qualitative support for the proposed soil-fiber interaction mechanism only.

The post-failure micrographs Figure 13 indicate that the behavior changed to fiber-bridged ductile failure of the soil. The fracture surface (×75) indicates fiber pull-out and crack bridging, suggesting that the fibers successfully restricted crack propagation and increased energy dissipation. With increased magnification (×220) of the interfaces, there is evidence of interfacial debonding and damaged ends of the fibers, which demonstrates that failure was mainly due to frictional pull-out of the fibers, but not by sudden rupture of the fibers. These microscopic observations largely indicate the macroscopic enhancements in UCS (22.30%) and cohesion (18.63%), in addition to supporting the statistically significant action of enhancement by 25 mm Bhimal fibers at 0.5% content.

FIGURE 13

The microstructural interpretation of the present study is based on the SEM observations only. Thus, the term microstructural behaviour is used to denote the SEM-based evidence of soil–fiber contact, particle attachment, interfacial debonding, crack bridging and fiber pull-out. Additional microstructural techniques such as XRD, FTIR, EDS mapping, MIP or pore-size analysis were not performed in the present study and are recommended for future work to provide a more comprehensive understanding of mineralogical and pore-structure changes.

3.7 Qualitative link between SEM observations and mechanical behaviour

The SEM observations qualitatively support the mechanical behaviour of the best reinforced mixture M13. Increased UCS and cohesion of M13 can be correlated to the visible soil–fiber contact, particle attachment, mechanical interlocking, crack bridging and frictional pull-out features observed in the SEM micrographs. These mechanisms are helpful to explain the improvement in stress transfer, delay in crack propagation and change in failure response of expansive black cotton soil from brittle to more ductile behavior due to randomly distributed Bhimal fibers.

SEM micrographs of M13 before mechanical loading showed the soil particle attachment on the rough Bhimal fiber surface indicating the interfacial contact and mechanical anchorage developed in the reinforced matrix. These features are in line with the observed increase in UCS and cohesion for the optimum mix. SEM micrographs after UCS testing showed fiber pull-out, crack bridging, interfacial debonding and damaged fiber ends. Such post-failure features demonstrate that the fibers could restrain the crack growth and dissipate energy during the deformation.

However, the SEM observations cannot be considered as an independent statistical confirmation of the ANOVA, regression or effect-size results because SEM was performed only on the optimum reinforced mixture M13. The statistical analysis establishes the effect of fiber content and fiber length in the experimental matrix. SEM qualitatively explains the mechanism of reinforcement at the soil-fiber interface. The SEM results are used solely for mechanistic support of the observed mechanical improvement and not for comparative microstructural validation of all mixtures.

In future work, comparative SEM analysis of the unreinforced control soil, suboptimal reinforced mixtures and optimum reinforced mixtures should be conducted to better establish the microstructural evolution across the different fiber contents and lengths. Other complementary techniques such as EDS mapping, FTIR, XRD, mercury intrusion porosimetry or pore-size analysis may also be used to provide a more complete understanding of soil–fiber interaction and microstructural change.

4 Limitation of the classical UCS–Cohesion relationship

The relationship between unconfined compressive strength (UCS) and cohesion (c) was examined only as a reference comparison with the classical cohesive-soil assumption. In saturated undrained cohesive soils, the simplified relationship UCS ≈ 2c is commonly used when the friction angle is assumed to be approximately zero. However, this assumption is not fully applicable to fiber-reinforced expansive soil because fiber inclusion changes the stress-transfer mechanism, failure-plane development, and mobilization of shear resistance. Therefore, the present comparison was not treated as a validation of the classical relationship, but as an assessment of its limitation for Bhimal fiber-reinforced black cotton soil. The classical relationship is expressed as Equation 8:

Table 11 and Figure 14 show the comparison between the experimental UCS and the predicted UCS values obtained from 2c. For all mixes, the predicted UCS values were much higher than the measured UCS values, which indicates a systematic overestimation by the classical relationship. Poor correlation between experimental UCS and predicted UCS (R² = 0.088, p = 0.326) showed that UCS ≈ 2c assumption is not sufficient to describe the behavior of the reinforced soil mixtures. This poor agreement is expected as the strength response of fiber-reinforced soil is governed not only by apparent cohesion but also by fiber pull-out, interfacial friction, crack bridging, fiber orientation and non-uniform failure development.

FIGURE 14

4.1 Theoretical basis of the deviation

The classical UCS ≈ 2c relationship is based on the assumption of purely cohesive soil condition with ϕ ≈ 0, ideal undrained behaviour and relatively uniform stress distribution during failure. In the present study, the black cotton soil and reinforced mixes had non-zero friction angles, and the randomly distributed Bhimal fibers modified the failure mechanism. The UCS and direct shear tests also mobilize strength along different stress paths. In direct shear tests, the failure plane is prescribed along the horizontal shear plane, which can mobilize more apparent cohesion as fiber bridging and interfacial resistance are activated along the imposed shear plane. In contrast, UCS test allows the specimen to fail along its weakest natural plane, where local defects, fiber orientation, specimen heterogeneity and stress concentration may reduce the measured compressive strength.

This difference explains why cohesion obtained from direct shear testing cannot be directly converted into UCS using the simplified 2c relationship. The deviation is therefore not an experimental error, but a result of the different loading paths and failure mechanisms involved in the two tests.

4.2 Influence of fiber reinforcement

Therefore, the comparison of the cohesion values based on the DST presents a higher relative improvement over the UCS values.

These observations are also supported by the SEM results (Section 3.6), in which the influence of fiber pull-out and interfacial friction were found to be the leading factors. In shear loading, these mechanisms are mobilized more efficiently than in pure axial compression.

4.3 Effect of moisture and compaction

The UCS is highly sensitive to the Moisture content, Dry density and Degree of saturation. The OMC increased by 40.91 percent for the optimum combination, as indicated in Table 10. High moisture levels can lead to decreased effective stress and compressive resistance, which is at least one factor explaining why the calculated UCS is no longer proportional to cohesion. However, direct shear cohesion may be significantly less sensitive to these preparation variables.

4.4 Significance of observed difference

Similar deviations have been reported in fiber-reinforced cohesive soils, where the mobilization of cohesion and compressive strength may differ because of stress redistribution, fiber orientation, interfacial friction, crack bridging, and pull-out resistance [26]. Therefore, the present UCS–cohesion comparison should be interpreted as a limitation of the classical UCS ≈ 2c relationship for Bhimal fiber-reinforced expansive soil, rather than as a validation of the relationship. The simplified correlation may be used only as a reference check, and not as a direct predictive equation for reinforced expansive soils without considering the altered failure mechanisms.

4.5 Limitations and future research needs

The present study shows the short-term mechanical improvement of expansive black cotton soil by using Bhimal fiber under controlled laboratory conditions but still has two important limitations. Firstly, Bhimal fiber is lignocellulosic in nature, and it may be affected by long-term exposure to water, wetting–drying cycles, microbial activity and acidic or alkaline soil environments, which may reduce the long-term soil–fiber bond and tensile resistance. Second, SEM observations were only conducted for the optimum reinforced mixture M13. Therefore, SEM results only provide qualitative mechanistic support and should not be interpreted as comparative microstructural validation across all mixtures. Future work should consist of long-term durability testing, field exposure studies and comparative microstructural analysis of control, suboptimal and optimum mixes using SEM and complementary techniques such as EDS, FTIR, XRD or pore-structure analysis.

5 Conclusion

References

• 1.

  AbhishekAGuharayAHataT. Enhanced microstructural analysis of black cotton soil stabilized through rice Husk ash integration. Soils and Foundations (2025) 65(6):101698. 10.1016/j.sandf.2025.101698

  • CrossRef
  • Google Scholar
  • View reference in article

• 2.

  Al-KaliliAAliASAl-TaieAJ. A review on expansive soils stabilized with different pozzolanic materials. J Eng (2022) 28(1):1–18. 10.31026/j.eng.2022.01.01

  • CrossRef
  • Google Scholar
  • View reference in article

• 3.

  AlmuaythirSZainiMSIHasanM. Shear strength, compressibility, and consolidation behaviour of expansive clay soil stabilized with lime and Silica Fume. Scientific Rep (2025) 15(1):26185. 10.1038/s41598-025-10642-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar
  • View reference in article

• 4.

  AdnanMKumarSGargNGuptaKKKumar DasS. Soil stabilization using waste ‘bagasse ash and lime’: a review. In: Materials Today: Proceedings, Ahead of Print (2023). 10.1016/j.matpr.2023.05.018

  • CrossRef
  • Google Scholar
  • View reference in article

• 5.

  ParhizkarANazarpourAKhayatN. Investigation of geotechnical and microstructure characteristics of gypsum soil using ground granulated blast-furnace slag (GGBS), fly ash, and lime. Construction Building Mater (2024) 418(February):135358. 10.1016/j.conbuildmat.2024.135358

  • CrossRef
  • Google Scholar
  • View reference in article

• 6.

  AnburuvelA. The engineering behind soil stabilization with additives: a state-of-the-art review. Geotechnical Geol Eng (2023) 42(1):1–42. 10.1007/s10706-023-02554-x

  • CrossRef
  • Google Scholar
  • View reference in article

• 7.

  MartinsLACorreiaAASOliveiraPJVLemosLJL. Effect of the number of loading cycles on the behavior of a stabilized fiber-reinforced soft soil: energy evaluation. J Mater Civil Eng (2023) 35(11):04023392. 10.1061/jmcee7.mteng-15231

  • CrossRef
  • Google Scholar
  • View reference in article

• 8.

  PuttegowdaM. Eco-friendly composites: exploring the potential of natural fiber reinforcement. Discover Appl Sci (2025) 7(5):401. 10.1007/s42452-025-06981-8

  • CrossRef
  • Google Scholar
  • View reference in article

• 9.

  BizualemLGideboFA. Performance assessment of bamboo and sisal fiber reinforcement in red Ash stabilized expansive black cotton soil. Results Eng (2026) 29(March):108743. 10.1016/j.rineng.2025.108743

  • CrossRef
  • Google Scholar
  • View reference in article

• 10.

  ZafarTAnsariMAHusainA. Soil stabilization by reinforcing natural and synthetic fibers – a state of the art review. In: Materials Today: Proceedings, Ahead of Print, April 1, 2023 (2023). 10.1016/j.matpr.2023.03.503

  • CrossRef
  • Google Scholar
  • View reference in article

• 11.

  GhavamiKToledo FilhoRDBarbosaNP. Behaviour of composite soil reinforced with natural fibres. Cement and Concrete Composites (1999) 21(1):39–48. 10.1016/s0958-9465(98)00033-x

  • CrossRef
  • Google Scholar
  • View reference in article

• 12.

  WangY-XGuoP-PRenW-XYuanB-XYuanH-PYan-LinZet alLaboratory investigation on strength characteristics of expansive soil treated with jute fiber reinforcement. Int J Geomechanics (2017) 17(11):04017101. 10.1061/(asce)gm.1943-5622.0000998

  • CrossRef
  • Google Scholar
  • View reference in article

• 13.

  JairajJPrathap KumarMT. Long-term performance studies on strength characteristics of black cotton soil reinforced with untreated and treated coir Fibre. Int J Geosynthetics Ground Eng (2018) 4(3):25. 10.1007/s40891-018-0143-9

  • CrossRef
  • Google Scholar

• 14.

  MagrayARSinghP. Improvement in the properties of soil by utilization of Jute fiber with black carbon as a soil reinforcement. Mater Today Proc (2023). 10.1016/j.matpr.2023.04.024

  • CrossRef
  • Google Scholar
  • View reference in article

• 15.

  MozumderRSChandra BhumikMSheikhP. Influence of bamboo fiber reinforcement on the mechanical behavior of soft soil. Discover Civil Eng (2025) 2(1):181. 10.1007/s44290-025-00339-0

  • CrossRef
  • Google Scholar
  • View reference in article

• 16.

  SharmaVVinayakHKMarwahaBM. Enhancing compressive strength of soil using natural fibers. Construction Building Mater (2015) 93(May):943–9. 10.1016/j.conbuildmat.2015.05.065

  • CrossRef
  • Google Scholar
  • View reference in article

• 17.

  Medina-MartinezCJSandoval-HerazoLCZamora-CastroSAVivar-OcampoRReyes-GonzalezD. Natural fibers: an alternative for the reinforcement of expansive soils. Sustainability (2022) 14(15):9275. 10.3390/su14159275

  • CrossRef
  • Google Scholar
  • View reference in article

• 18.

  MurasingKKKumarMKumarA. Effect of altitude on the mechanical strength of Grewia optiva fiber in garhwal himalaya, India. J Nat Fibers (2021) 19(13):6638–47. 10.1080/15440478.2021.1929652

  • CrossRef
  • Google Scholar
  • View reference in article

• 19.

  KumarHTejyanSKannanM. Himalayan plants natural fibres-based Woven mat reinforced polymer composites: a thermo-mechanical and abrasive wear study. J Indian Acad Wood Sci (2026) 23(March):192–210. 10.1007/s13196-026-00408-6

  • CrossRef
  • Google Scholar
  • View reference in article

• 20.

  VermaDGuptaDRimdusitSJawaidMChaudharyAKOkhawilaiM. Opportunities and challenges of Grewia optiva (bhimal) fiber as a sustainable resource for polymer bio-composites. J Reinforced Plastics Composites (2025):07316844251397001. 10.1177/07316844251397001

  • CrossRef
  • Google Scholar
  • View reference in article

• 21.

  RajaTDevarajanYVickramS. Evaluation of Grewia Optiva fiber as a sustainable and high-performance reinforcement material for composite applications. Results Eng (2025) 25(March):104096. 10.1016/j.rineng.2025.104096

  • CrossRef
  • Google Scholar
  • View reference in article

• 22.

  SinghMJWeiqiangFDong-ShengXBoranaL. Experimental Study of compression behavior of Indian Black cotton soil in oedometer condition. Int J Geosynthetics Ground Eng (2020) 6(2):30. 10.1007/s40891-020-00207-0

  • CrossRef
  • Google Scholar
  • View reference in article

• 23.

  PratamaBPutraMNurhalizaN. Experimental research Methods‘Anava and factorial experiment psychology research design. Psikologiya J (2024) 1(1):7–15. 10.62872/rvbnnb40

  • CrossRef
  • Google Scholar
  • View reference in article

• 24.

  NandaAMohapatraDBBMahapatraAPKMahapatraAPKMahapatraAPK. Multiple comparison test by tukey’s honestly Significant difference (HSD): do the confident level control type I error. Int J Stat Appl Mathematics (2021) 6(1):59–65. 10.22271/maths.2021.v6.i1a.636

  • CrossRef
  • Google Scholar
  • View reference in article

• 25.

  ZulkifliMNFMYahyaMF. Estimating mechanical tensile strength of single fiber composites by adopting multiple linear regression. Int J Res Innovation Social Sci (2024) 8(11):3008–14. 10.47772/ijriss.2024.8110232

  • CrossRef
  • Google Scholar
  • View reference in article

• 26.

  LiuZWusimanNJiliQ. Mechanical properties and micromechanisms of natural fiber-modified saline soil. Scientific Rep (2026) 16(1):2736. 10.1038/s41598-025-32424-w

  • CrossRef
  • Google Scholar
  • View reference in article