1 Introduction

Ground improvement is a crucial aspect of geotechnical engineering, particularly when the soil at a selected site lacks the desired mechanical properties such as cohesion, internal angle of friction, and bearing capacity [1]. Poorly compacted soil can lead to catastrophic consequences, including slope failures [2], foundation sinkage [3], and even the complete collapse of structures like tunnels, mine dumps, and buildings [3]. Several methods are used to improve ground conditions for safe and reliable construction [4]. Based on the treatment method, the engineering techniques of ground improvement can be broadly grouped into mechanical [4] and chemical stabilization [3]. Among these, mechanical stabilization is the most common and oldest technique of ground improvement [5], where the soil's density is increased by applying mechanical force and compacting the surface layers by static and dynamic loading [6]. Mechanical methods are more complex and expensive [5, 6] than chemical processes [7].

In chemical treatment for soil improvement, ground improvement is achieved by mixing various chemicals, for example, cement and lime, with soil to develop desirable characteristics [6]. These methods have shown a long-term change in ground properties. Still, some environmental concerns are usually associated with them [1, 5]. Cement serves as a crucial binding material in the construction industry [7]. However, it is characterized by its high cost, substantial energy consumption, depletion of natural resources, and significant emission of greenhouse gases into the atmosphere [8]. The production of one ton of cement results in approximately one ton of CO2 gas emissions, leading to severe environmental contamination and various pollution-related issues [9]. Many studies have been conducted to search for cementitious [10], low-priced [11], readily available cement replacement materials from the cement industry to minimize greenhouse gases for environmental protection [12]. Predominantly, one of the cement replacement materials gaining research attention is rice husk ash made from rice husks, a product of rice milling [13]. As per FAOSTAT data from 2020 [14], the global paddy production reached 782 million tons in 2018. Extrapolating from this information, approximately 172 million tons of rice husks were generated that year, potentially producing 34 million tons of Rice husks (RHA) if all rice husks were incinerated [15]. However, due to underutilization, low market value, and logistical challenges in transportation, this substantial quantity of RHA is often indiscriminately disposed of in landfills, agricultural fields, riverbanks, [14] and open areas near populated regions. This unregulated disposal leads to air, soil, and water pollution, [16] releasing toxic substances and posing risks to public health.

RHA comprises the highest amount of amorphous silica among all agricultural wastes [14]. The content of SiO2 in RHA is more than 70% [15], and it can go up to more than 90%, especially under controlled combustion temperature conditions [17, 18]. Moreover, the morphology of SiO2 in RHA subjected to low-temperature combustion consists of nano-sized amorphous particles [16], which exhibit very high pozzolanic activity similar to silica fume [19]. This means that RHA has a potential gelling activity to promote the generation of cementitious hydration products as C–S–H/C–A–S–H under certain conditions [20]. Many researchers have discovered that adequately utilizing locally obtainable RHA as a pozzolanic material in ground improvement can decrease environmental degradation and construction price and increase soil strength properties [21]. The amorphous silica (SiO2) in RHA chemically reacts with free calcium hydroxide Ca(OH)2 in the secondary hydration reaction of cement [22] and forms the calcium silicate hydrate (C–S–H) gel, which improves the strength and durability of soil [23, 24]. Researchers have found that the utilization of RHA in construction projects can diminish CO2 emissions by reducing the production of cement [25], waste disposal problems, environmental pollution, construction costs, and consumption of natural resources [26], and can increase the strength and durability of concrete and soil [27].

Additionally, RHA can effectively improve soil shear strength [28], indicating the potential for economic and environmental benefits [29]. Moreover, Incorporating RHA into geotechnical engineering practices allows us to contribute to a sustainable future by mitigating waste [30], reducing carbon emissions [31], and fostering the principles of sustainable development [32]. Furthermore, our study uniquely emphasizes minimal cement dosage, a distinguishing feature that differentiates it from the prevailing literature. Most existing research has primarily focused on the impact of combining rice husk ash with cement to enhance the cohesion of cohesionless soils, leaving exploring its effects on weak clay soils and subgrade utilization relatively underexplored [33,34,35]. This distinctive approach underscores our commitment to providing innovative solutions for sustainable ground improvement while efficiently managing construction materials, making it an essential contribution to the field. On the other hand, investigating subgrade layers and their influence on the overall shear strength of RHA-cement treated soils has remained a relatively unexplored area in the existing literature.

An experimental study was conducted to evaluate the shear strength of this new construction material. Sixteen specimen combinations of weak expansive with well-defined subgrade layers of soil-RHA-cement were investigated to understand the effects of RHA and cement on the shear strength parameters. The admixtures for each subgrade were prepared by mixing soil with 2%, 4%, and 6% RHA and 2%, 4%, and 6% cement. The upper, bottom, and subgrade double layers were used. A comprehensive analysis of the shear stress–shear displacement relationship, cohesion, and internal friction angle was done.

2 Materials and methods

2.1 Materials

2.1.1 Soil

The soil utilized in this study was collected from the Handa area, Tsu City, Mie Prefecture, Japan, located approximately 22 km away from Mie University. The exact location of soil sampling is shown in Fig. 1. The selection of this location was motivated by its characterization as an area facing challenges associated with weak, expansive, and loose soils. Following standard geotechnical practices [36], systematic soil sampling was conducted to obtain representative samples. The soil used in this research was air-dried for three weeks [37] and then sieved through the 2 mm sieve [38] before the triaxial test. Air drying the soil for three weeks ensured the moisture content was consistent across all samples. This was crucial to obtaining more accurate, reproducible[39], and comparable test results, as inconsistent moisture content could lead to variations in soil behavior during testing. Sieving through the 2 mm sieve removed coarse particles and aggregates, ensuring that the soil used in the triaxial test was relatively homogeneous. This enhanced the accuracy of the test results[40] and ensured that the soil conforms to the specified particle size requirements and can be easily weighed, mixed, and compacted into the desired specimen shape for the triaxial test. Subsequently, various analyses were carried out, including sieve, hydrometer, liquid, and plastic limit analyses to classify the soil. The liquid and plastic limits of the soil were measured to be 48.5% and 28.594%, respectively. The soil composition comprised 0.02% sand, 57.28% silt, and 42.7% clay. According to the American Association of State Highway and Transportation Officials (AASHTO) classification, the soil was identified as A-7–6 (5) clayey soil. These soil characteristics indicate that it belongs to the category of clay soil. A detailed grain size analysis for soil, rice husk ash (RHA), and cement is presented in Fig. 2. Testing procedures followed Japan Industrial Standards JIS A 1204 [41] for the sieve and hydrometer test and JIS A 1205 [41] for the liquid and plastic limit tests. Additional geotechnical properties are detailed in Table 1.

Fig. 1
figure 1

The exact location of soil sampling used in this study

Full size image
Fig. 2
figure 2

Particle size distribution for Soil, Cement, and RHA

Full size image
Table 1 Properties of Soil and RHA
Full size table

2.1.2 Rice husk ash

A controlled burned RHA at 650–700℃ with a high silica content of 91.10% was provided by Make Integrated Technology Co., Ltd, Osaka, Japan. The high silica content was obtained by burning the rice husks for 27 h in a computer-controlled self-combustion Ethical Star ash production machine. The RHA particle sizes ranged from 0.07 to 0.3 mm. Detailed particle size distribution of soil and RHA is shown in Fig. 2. A detailed physical and chemical property chart for rice husk ash is illustrated in Table 1.

2.1.3 Portland cement

Ordinary Portland Cement (OPC) was also used as a binder element. The physical and chemical properties of OPC cement are illustrated in Table 2. [41]

Table 2 Properties of OPC
Full size table

2.2 Specimen preparation

A series of specimens were prepared using an experimental sample to investigate the effects of soil-RHA-cement mixtures on shear strength (Fig. 3a). The entire sample had a total weight of 500 g and a height of 12.5 cm. The weight of each specimen was determined by dividing the total weight of the sample by the height, resulting in a ratio of 40 g per 1 cm.

Fig. 3
figure 3

a Specimen preparation for laboratory test (a) Lower layers, (b) Upper layers, (c) Double layers. b Specimen preparation for different layers

Full size image

Approximately 160 g of soil was used for each specimen to form each layer. In the layer designated for experimentation, a mixture of 40 g of soil, RHA, and cement was added to 120 g of soil alone, creating a layer with a height of 1 cm composed explicitly of the soil-RHA-cement mixture to be tested (Fig. 3b). The specimens were manually compacted inside a mold with a height of 12.5 cm and a diameter of 5.0 cm. The compaction process involved compacting the specimens in three layers using a 4.9 cm diameter hand rammer with a mass of 1 kg. Each layer was compacted by applying 20 blows with a falling height of 30 cm. This ensured the uniform and consistent compaction of the soil-RHA-cement mixture within the specimens. The components were thoroughly mixed in a large tray to achieve a homogeneous mixture of soil-RHA-cement. Water was added gently to the dry mix, ensuring proper hydration without excessive saturation. The amount of water added to each specimen was determined based on the optimum moisture content (OMC). All testing procedures followed the Japan Industrial Standards (JIS) A 1210 standards [42] calculated from the compaction test. This meticulous specimen preparation and compaction method ensured the integrity and reliability of the subsequent shear strength testing using specific weight ratios and standardized compaction techniques.

In this study, the dosage of RHA was selected as 2%, 4%, and 6% based on the maximum utilization of amorphous silica at low percentages, as indicated by previous research [17, 41]. The maximum utilization of amorphous silica was 5%; therefore, we chose 2%, 4%, and 6% to investigate the effects of increasing the dosage by 1% and decreasing it by 1%. In addition, it is essential to highlight that this study primarily focuses on treating the soil to improve its properties while incorporating a small amount of cement 2%, 4%, and 6% [17, 28]. Including a minimal cement dosage complements the soil treatment process and ensures optimal proportions of Ca+ ions during pozzolanic reactions. This approach not only aligns with the objectives of environmental conservation and carbon emission reduction discussed in the introduction section but also emphasizes the soil treatment aspect of the study rather than focusing solely on concrete.

2.3 Testing method

2.3.1 Compaction test

For each material blend shown in Table 3, a standard proctor compaction test was done to determine the optimum moisture content (wopt) and the corresponding maximum dry density (ρdmax). Each blend was compacted into a cylindrical mold measuring 10 cm in diameter and 12.73 cm in height using a 2.5 kg rammer at a falling height of 30 cm while maintaining a constant compaction effort for each of the three layers. This process was repeated for various known moisture contents, and the dry densities were calculated for each. The plots of the dry density to the moisture content also called the compaction curves, were then used to obtain the peak points representing the maximum dry density and the optimum moisture content. All testing procedures followed the Japan Industrial Standards (JIS) A 1210 standards [42]. The results of optimum moisture content and dry density for soil and the soil blends are shown in Fig. 4.

Table 3 Detailed blend combinations
Full size table
Fig. 4
figure 4

Effects of RHA and cement on compaction characteristics

Full size image

2.3.2 Triaxial compression test

The experimental procedure followed the Japan Geotechnical Society 0520 ~ 0524 standards [17]. Consolidated drained (CD) triaxial tests were conducted on cylindrical specimens with a diameter of 5 cm and a height of 12.5 cm [43]. To prepare the specimens, each blend was compacted in three layers using a 1 kg rammer dropped from a height of 30 cm into a mold of similar dimensions. The specimens were then placed in the triaxial test pressure chamber and sealed with a rubber sleeve, as depicted in Fig. 4. The experimental setup for the consolidated drained triaxial tests included several key components: (1) a load cell for measuring the axial load, (2) a loading piston to apply axial loading on top of the specimen, (3) a dial gauge for measuring axial displacement, (5) a pressure cylinder to hold the chamber fluid, (6) chamber fluid to exert hydrostatic chamber pressure (σ3), (9) a route for measuring drainage/pore water pressure, (10) a path for supplying the confining fluid, and (13) an inlet for air pressure. Figure 5 provides a detailed illustration with labeled components.

Fig. 5
figure 5

Triaxial testing apparatus

Full size image

This study used three confining pressures: 50 kPa, 100 kPa, and 150 kPa. The drainage/pore water pressure measurement route was kept open to ensure a drained condition within the specimen, preventing the development of pore water pressure within the composite structure. Shearing the specimens occurred slowly at a rate of 0.5 mm/min. Deviatoric stresses and axial strains were recorded throughout the experiment using a data logger. The experiment was terminated when the peak deviatoric stress value was reached or the axial strain got two-thirds or 10% of its initial value, whichever occurred first.

2.3.3 Microstructural study on shear strength development

A comprehensive investigation of shear strength development in the reinforced and Treated soil composite necessitated using microstructural analysis techniques, including scanning electron microscopy (SEM) imaging and Energy-Dispersive X-ray Spectroscopy (EDS). SEM imaging was a critical tool to examine the soil composite at a micrometer scale. It utilized low-energy secondary and backscattered electrons to visualize the structural alterations occurring within the new composite material [44]. The three-dimensional images produced by the SEM test effectively portrayed the influence of additives in modifying the surface morphology of the stabilized soil at a micro-level [45, 46]. EDS, on the other hand, played a pivotal role in identifying crystalline patterns, offering a comprehensive analysis of the extent to which chemical stabilizing agents were incorporated into the treated soil composite. More specifically, EDS facilitated a deeper comprehension of the various chemical transformations in the treated soil due to pozzolanic reactions [47,48,49].

3 Results and discussion

3.1 The relationship between axial strain, deviatoric stress, and specimen combination

Figures 6, 7, and 8 depict the correlation between axial strain and specimen combinations for the lower, upper, and double layers. These combinations consisted of varying percentages (2%, 4%, 6%) of RHA and (2%, 4%, 6%) cement. When 2% cement was added to the lower and double layers, the axial strain ranged from 6 to 7%. By increasing the cement content to 4%, the axial strain decreased to 5% for the double layer and 6% for the lower layer. Notably, with 6% cement, the axial strain was slightly increased across all sub-grade layers. On the other hand, focusing on specimens containing RHA alone, the axial strains slightly increased to 7% due to the compressibility of the porous RHA microstructure. However, 2% to 6% of cement content reduced axial strains by 4% for all subgrade layers, attributed to the formation of cementitious compounds within the specimen structure.

Fig. 6
figure 6

The relationship between axial strain and specimen combination for the lower layer

Full size image
Fig. 7
figure 7

The relationship between axial strain and specimen combination for the Upper layer

Full size image
Fig. 8
figure 8

The relationship between axial strain and specimen combination for the Double-layer

Full size image

Figures 9, 10, and 11 illustrate the correlation between deviatoric stress and specimen combinations for the lower, upper, and double layers. The results demonstrated a clear trend of increasing deviatoric stresses with higher confining pressures ranging from 50 to 150 kPa. In this study, the double-layered specimens exhibited the highest deviatoric stresses compared to the lower and upper layers. For instance, when examining the specimen with a combination of 2% RHA and 6% Cement (2%RHA6%C), the double layer achieved maximum deviatoric stresses of 383 kN/m2. In contrast, the lower and upper layers attained 368 kN/m2 and 338 kN/m2, respectively. The outstanding increase in deviatoric stresses can be attributed to the improved resistance to deformation resulting from the pozzolanic reactions between RHA and cement, as discussed in the introduction, a phenomenon corroborated by findings similar to those reported by Rachmawati [28]. Introducing 2% RHA across all the specimen combinations led to improved deviatoric stresses for all subgrade layers compared to the control specimens and those with cement only. However, a slight reduction in deviatoric stresses was observed as the RHA content increased from 2 to 6%, specifically in the double layer.

Fig. 9
figure 9

The relationship between Deviatoric stress and specimen combination for the Lower layer

Full size image
Fig. 10
figure 10

The relationship between Deviatoric stress and specimen combination for the Upper layer

Full size image
Fig. 11
figure 11

The relationship between Deviatoric stress and specimen combination for the Double-layer

Full size image

In conclusion, the highest deviatoric stresses were achieved by the double layer in the specimen containing 2% RHA and 6% cement, as evident from Fig. 11. Furthermore, in our quest to comprehend how unreinforced and RHA-Cement-reinforced specimens respond mechanically to shear, we harnessed photographic data that unveils the failure patterns during testing. These revealing visuals are illustrated in Fig. 12.

Fig. 12
figure 12

Specimens failure planes at 150 kPa confining pressure a Soil control, b Upper layers, c Double layers, d Lower layers

Full size image

Notably, it proved intriguing that the composition of RHA and Cement within this subgrade soil significantly influenced the initiation points of failure cracks. Consequently, the position from which these cracks emanate hinges primarily on the arrangement of RHA-Cement mixtures and the number of layers. For instance, when one layer of 2RHA + 6C or 4RHA + 6C is embedded in the upper layer, the fractures are observed to initiate solely within the upper region and propagate towards the middle, as depicted in Fig. 12b. Additionally, incorporating 6RHA + 6C in the upper layer similarly results in fractures originating in the upper section and extending into the lower layer, as presented in Fig. 12b. However when double layers of 2RHA + 6C and 4RHA + 6C are embedded, this zone doesn't exhibit the same level of reinforcement, thus leading to fractures starting in the weaker upper portion and extending into the middle soil layers, as indicated in Fig. 12c. Conversely, embedding 6RHA + 6C in the double layers produces a robustly stiffened zone.

On the contrary, a notable reduction in visible vertical planes of weakness (cracks) within the lower layers is observed when embedding in the lower layer, as evidenced in Fig. 12d. This decrease emphasizes the existence of augmented elasticity within the specimens, consistent with similar outcomes reported in other studies [19, 31]. They affirmed the RHA-cement composite's effectiveness in curbing cracks within RHA-cement composite soil composites.

3.2 Effects of RHA and cement subgrade layers on shear strength parameters

The present study aimed to quantify the shear parameters of the stabilized expansive soil using the Mohr–Coulomb failure theory, which describes the relationship between shear stress (τ) and effective normal stress (σ). The shear strength parameters regarding the angle of internal friction (φ) are summarized in Figs. 13, 14, and 15. Incorporating RHA and cement resulted in an enhanced angle of internal friction for the soil-RHA-cement composite in the lower Fig. 13 and double subgrade layers Fig. 15. This improvement can be attributed to the irregular shapes formed during the pozzolanic reactions between RHA and cement particles, which contributed to increased frictional resistance within these layers of the specimens. However, slight variations in the angle of internal friction were observed, specifically in the lower layers. These differences can be attributed to the location of the subgrade layer within the specimen, as the shear plane was positioned well above the subgrade layer.

Fig. 13
figure 13

The relationship between the angle of internal friction (φ) and specimen combination for the Lower layer

Full size image
Fig. 14
figure 14

The relationship between the angle of internal friction (φ) and specimen combination for the Upper layer

Full size image
Fig. 15
figure 15

The relationship between the angle of internal friction (φ) and specimen combination for the Double layer

Full size image

The shear strength parameters were also presented, as shown in Figs. 16, 17, and 18. Concerning the cohesive strength observed in all the examined specimens, it is noteworthy that the inclusion of cement and RHA exerted a substantial impact compared to the control, aligning with findings akin to those reported by other studies[41, 50]. The highest cohesion values were achieved with 6% cement, particularly demonstrating eminent advancements in the lower Fig. 16 and double subgrade layers Fig. 18. For instance, specimens 6%RHA6%C, 4%RHA6%C, and 2%RHA6%C exhibited cohesion values of 143 kN/m2, 138 kN/m2, and 140 kN/m2, respectively, as shown in Fig. 16. Similarly, specimen 6%RHA6%C attained the highest cohesive strength when considering the double subgrade layer with a value of 134kN/m2, as depicted in Fig. 18.

Fig. 16
figure 16

The relationship between the cohesion (c) and specimen combination for the Lower layer

Full size image
Fig. 17
figure 17

The relationship between the cohesion (c) and specimen combination for the Upper layer

Full size image
Fig. 18
figure 18

The relationship between the cohesion (c) and specimen combination for the Double layer

Full size image

Figure 17 shows a slight reduction in the cohesion value across all the specimen combinations. This observation suggests that the upper subgrade layer possessed the lowest cohesion values, indicating weaker bonding between its particles and the composition of the subgrade. It is worth noting that this layer was located along the shear plane, thereby experiencing direct initial contact with the axial load.

3.3 SEM and EDS analysis

In this study, incorporating a mixture comprising RHA, cement, and soil impacted the failure behavior under shear forces. Therefore, it is essential to comprehend the reinforcing mechanism to substantiate their potential in preventing the formation of weak planes within the composite materials. To begin with, we conducted a microstructural analysis to illustrate the structural alterations in specimens containing only soil. Subsequently, we examined the overall composite structure after introducing RHA and cement. It became apparent that micro-cracks formation was a common occurrence in the unreinforced specimens. Porous morphology (micro-pores) and micro-cracks indicated an increased susceptibility to shear failure during loading due to the proliferation of weak planes, as depicted in Fig. 19 for the control sample (S Control). However, for the samples containing 2RHA + 6C and 4RHA + 6C, as shown in Figs. 20 and 21. there was a notable reduction in micro-crack formation and the development of a more cohesive cemented composite structure. Furthermore, adding modest amounts of cement and RHA (6RHA + 6C) enhanced the degree of cementation within the composite structure, decreasing the development of micro-pores and micro-cracks, as illustrated in Fig. 22 Other relevant studies have reported similar findings [51].

Fig. 19
figure 19

SEM image for Soil-only specimen at × 500

Full size image
Fig. 20
figure 20

SEM image for 2RHA + 6C specimen at × 500

Full size image
Fig. 21
figure 21

SEM image for 4RHA + 6C specimen at × 500

Full size image
Fig. 22
figure 22

SEM image for 6RHA + 6C specimen at × 500

Full size image

The specimens underwent EDS analysis to determine their elemental chemical composition and relative abundance. Figures 23, 24, 25, 26 present the EDS spectra and elemental compositions of the soil, 2RHA6C, 4RHA6C, and 6RHA6C samples. The influence of introducing RHA and cement as stabilizing agents is apparent in these figures. As the RHA content increased, there was a corresponding increase in the SiO2 percentage because RHA is rich in SiO2. This is evident from the higher SiO2 levels in the EDS graphs when RHA increased from 2 to 6%. When the RHA content is increased with cement, a substantial amount of C–S–H (calcium–silicate–hydrate) gel formation is observed. This C–S–H gel formation can be attributed to the pozzolanic activity between RHA, cement, and soil. Interestingly, it was observed that Aluminum ions were incorporated into the C–S–H gel. This Aluminum ion uptake occurred both on the surface and in the interlayer of the gel, leading to the formation of stronger C–A–S–H gel bonds referred to as Calcium Alumino Hexasilicate Tetrahydrate [(CaO)(Al2O3)(SiO2)6 4(H2O)], as illustrated in Figs. 25 and 26. Other pertinent research has documented comparable outcomes [41, 52, 53].

Fig. 23
figure 23

EDS spectrum of soil and its elemental composition

Full size image
Fig. 24
figure 24

EDS spectrum of 2RHA6C and its elemental composition

Full size image
Fig. 25
figure 25

EDS spectrum of 4RHA6C and its elemental composition

Full size image
Fig. 26
figure 26

EDS spectrum of 6RHA6C and its elemental composition

Full size image

4 Conclusion

This paper comprehensively investigated the effects of incorporating RHA and cement into clay soils as subgrade layers, shedding light on their mechanical behavior and microstructural characteristics. Sixteen samples were meticulously examined through triaxial tests, and the results were thoughtfully analyzed in terms of deviatoric stress, axial strain, angle of internal friction (φ), and cohesion (c). Meanwhile, the double layer achieved maximum deviatoric stresses of 383 kN/m2 in the specimen (2%RHA6%C). Remarkably, the angle of internal friction exhibited a noticeable increase with higher RHA-Cement content in the subgrade layer. This observed phenomenon can be attributed to the pozzolanic reactions within the subgrade, contributing to increased friction and enhanced interlocking between soil particles. It is essential to note that conventional triaxial test results supported these outcomes and were consistent with the SEM analysis, which provided valuable insights into the microstructural changes occurring within the subgrade layers. The SEM analysis revealed the development of distinct structures and bonds due to the RHA-Cement treatment, corroborating the observed increase in the angle of internal friction.

Furthermore, the increase in cohesion, a critical parameter for soil improvement, was particularly pronounced in the lower subgrade layer. Specimen 6%RHA6%C demonstrated the highest cohesive strength with cohesion values of 143 kN/m2, compared to the soil control 95 kN/m2, corroborated by our EDS analysis, which provided elemental composition data supporting the enhanced cohesion. These findings underline the effectiveness of our innovative subgrade construction material comprising 6%RHA6%C, 4%RHA6%C, and 2%RHA6%C for sustainable ground improvement projects, a conclusion reinforced by both mechanical testing and the complementary SEM and EDS analyses. This comprehensive approach enhances our understanding of the material and demonstrates its potential to revolutionize the field of geotechnical engineering by combining mechanical performance with microstructural insights.