Basalt Fibers versus Plant Fibers: The Effect of Fiber-Reinforced Red Clay on Shear Strength and Thermophysical Properties under Freeze–Thaw Conditions

Abstract

Freeze–thaw cycling has a significant impact on the energy utilization and stability of roadbed fill. Given the good performance of basalt fiber (BF) and plant fiber (PF), a series of indoor tests are conducted on fiber-reinforced red clay (RC) specimens to analyze the shear strength, thermophysical, and microstructural changes and damage mechanisms of the RC under the freeze–thaw cycle–BF coupling, meanwhile, comparing the improvement effect of PF. The results indicate that the RC cohesion (c) first increases and then decreases with the increasing fiber content under BF improvement, reaching the maximum value at the content of 2%, and the change in the internal friction angle (φ) is relatively small. As the number of freeze–thaw cycles increases, cohesion (c) first decreases and then gradually stabilizes. The thermal conductivity increases with increasing moisture content, and the thermal effusivity increases and then decreases with increasing moisture content and fiber content. The heat storage capacity reaches the optimum level at a moisture content of 22.5% and a fiber content of 1%. Microanalysis reveals that at 2% fiber content, a fiber network structure is initially formed, and the gripping effect is optimal. The shear strength of PF-improved soil is higher than that of BF at a fiber content of 4–6%, and the thermal conductivity is better than that of BF. At the same fiber content, the heat storage and insulation capacity of BF-improved soil is significantly higher than that of PF.

1. Introduction

Under the global “dual-carbon” goal and the green concept of sustainable development, clean materials and new energy industries are central to achieving green and sustainable development. Using environmentally acceptable green renewable fiber to reinforce and improve special soils is a key method for realizing this concept. This approach aims to improve the utilization of construction materials, increase the lifespan of buildings and road foundations, enhance their stability, reduce pollution, and achieve green, sustainable development. Meanwhile, the development and utilization of geothermal energy resources, a new energy industry, is a popular research topic both domestically and internationally. Greenhouse gas emissions from geothermal energy typically have a relatively low environmental impact [1,2,3], leading to increased geothermal energy extraction and utilization in recent years. Geothermal energy systems contribute to environmentally friendly and green sustainable energy use in industrial processing, enhanced power generation, building heating/cooling, and energy tunneling techniques [4,5,6,7]. Geothermal energy is exploited in closed-loop geothermal systems using Ground Source Heat Pump (GSHP) units [8,9], where backfill material is required to fill the pores formed between the pipes and the surrounding soil or rock, making the extraction and delivery of geothermal energy more efficient [10]. The main factors affecting the performance of GSHP are the hydrogeological conditions and the thermophysical properties of the backfill material, which are important for the selection of admixtures and backfill for the efficient use of geothermal energy [11,12]. Therefore, it is necessary to study the shear strength and thermophysical properties of fiber-reinforced RC to inform the selection of backfill materials for geothermal energy systems and contribute to green and sustainable development.

RC is a highly plastic soil overlying the bedrock, as one of the special clays, characterized by high natural moisture content, a large pore ratio, ease of agglomeration, weak acidity, softening by water absorption, cracking by water loss, and strong water sensitivity. In alpine zones, RC is often used as a backfill for construction and roadbeds, as well as a fill material for geothermal energy transport. It may negatively impact construction roadbeds and geothermal-insulating soil, making it crucial to improve its mechanical properties and thermal insulation capacity.

BF is a natural fiber drawn from basalt at a high temperature of 1450–1500 °C. It is an inorganic silicate material with excellent mechanical properties [13], good compatibility, high-temperature resistance [14], and acid and alkali resistance. BF is extremely chemically stable, does not degrade easily under normal conditions [15,16], does not harm the soil, and is an environmentally acceptable and sustainable green material for polymer composites. They are also inexpensive and readily available [17]. Currently, more studies focus on the application of BFs as an external admixture for improving concrete [18,19,20], while fewer research results are available on their application to special RC soils. In alpine environments, construction and roadbed backfill will face freeze–thaw cycle conditions, and there are fewer research results on the strength changes and micro-mechanisms of BFs when incorporated into RC soils. Meanwhile, lignin fiber (LF), an organic fiber, is an extremely stable, green, non-polluting, and sustainable, green, and renewable material with great reserves in nature, making it a purely natural and good improvement material [21]. Comparing BF-improved soils with PF improves the utilization of PFs, protecting the environment and maximizing resource use for sustainable development.

In recent years, fiber-reinforced soil has become a popular direction of research in the field of geotechnical engineering, which is a technique to improve the properties of soil by mixing dispersed fiber filaments uniformly into the soil with a view to improving the mechanical properties of the soil [22,23]. Due to the structural characteristics of the fibers themselves, which make them discrete in the soil, the mechanical properties are basically close to isotropic after mixing with the soil, which can effectively make up for the shortcomings of the traditional reinforced soil [24,25]. The shear strength of fiber-reinforced soils depends on many factors, such as type of fiber, content, thickness, length, etc., where the effect of content is most prominent [26,27]. Wang et al. [28] found that the shear strength and dry density of polypropylene fiber-improved soil increased with fiber content, with a relatively small enhancement of the internal friction angle (φ). Niu et al. [29] observed that the strength of reinforced soil increased and then decreased with the length of BFs and glass fibers, with an optimum length of 9 mm. Wei et al. [30,31] found that the compressive strength and destructive strain of fiber- and lime-reinforced cured soils decreased in stages with increased freeze–thaw cycles, with fiber reinforcement significantly improving soil deformation resistance. Zhou et al. [32] found that the reinforcing effect of basalt stone powder and BFs with a content of 0.4% is optimal for all working conditions when added to pulverized clay. Nitin et al. [33] found that the addition of polypropylene fibers and silica fume to expansive soils effectively increased their unconfined compressive strength. Almajed et al. [34] conducted tests considering polypropylene fiber length, content, and soil conditioning time, finding that each test corresponded to different optimal lengths, content, and conditioning times. Shaker et al. [35] used fiber filaments and strips to improve expansive soils, finding a positive effect on soil hydraulic conductivity. Li et al. [36] found that the compressive strength of lime-improved soil increased and then decreased with the increase of coir fiber, which produced larger friction with the soil body, limiting soil deformation and inhibiting damage expansion. Wei et al. [37] found that the compressive and tensile strengths of polypropylene fiber, BF, and palm fiber-reinforced cured soils decreased in stages with increased freeze–thaw cycles, with higher destructive strains than lime-cured soils. Jiang et al. [38] found that incorporating palm fibers in pulverized clay significantly increased shear strength and deformation resistance, peaking at a fiber length of 20 mm, which is 172.9% higher than that of plain soil. Wu et al. [39] found that the highest shear strength is achieved with 0.2% polypropylene fiber at a length of 12 mm. Tang et al. [40] found that the shear strength of reinforced soils increases with an increase in dry density. Xiong et al. [41] scanned glass fiber-reinforced soil and found that glass fibers are poorly embedded with loess particles. Niu et al. [42] found a significant critical point for breaking strength and shear strength of improved soils with the increased loading rate, with higher BF content showing a more significant strain rate effect on breaking strength. Fan et al. [43] found that six wet and dry cycles of combined BF-lime-volcanic ash-cured silt increased unconfined compressive strength by 15.8% to 59.7% with 0.8% BFs. Chai et al. [44] found that randomly distributed and interwoven polypropylene fibers in soil provided spatial confinement, improving freeze–thaw resistance. Zhang et al. [45] found that the unconfined compressive strength of improved sandy soils decreased with increased wet and dry cycles, with minimal strength loss at 0.2% fiber content. Liu and Qu et al. [46,47] selected BFs, hemp fibers, and other composites to add to chalky clay soils in different regions, finding that fiber admixture significantly affected soil strength. Mohamad et al. [48] found that using natural hemp fiber-reinforced clay, the foaming force increased by 60%, and the cohesion (c) and friction angle (φ) increased by 7–10 kPa and 3–7°, respectively. Salimi et al. [49] used glass fibers to form interlocking zones between soil particles to enhance the unconfined compressive strength and indirect tensile strength of lime-nano clay solidified lime soil. Wang et al. [50] established a prediction model based on the concept of equivalent perimeter pressure, which can better describe the bias stress-strain and pore water pressure-strain relationships of fiber-reinforced clay, and can effectively predict the shear strength of fiber-reinforced clay under different perimeter pressures. Tao et al. [51] found that both single and composite components of nano-SiO₂ and sisal fiber improved the impermeability of modified clay, while the optimum content of nano-SiO₂ is 3%. Mohamad et al. [52] found that clay strength and stiffness increased by more than 50% with 7% cement, and cumulative mass loss was reduced by about 50% with 0.4% fiber. Few studies have been conducted on content and fiber length, and most focus on conventional soils with high moisture content, with fewer studies on special soils and reinforced soils subjected to freeze–thaw cycles.

Currently, many scholars have studied soil thermal conductivity. Wang et al. [53] found that thermal conductivity increased linearly with dry density and decreased exponentially with age in lime-improved RC. Xu et al. [54] found that the thermal conductivity of quartz sand, quartz powder, and kaolin increased rapidly with the increase of moisture content before the critical moisture content, after which the increase slowed down and gradually stabilized. Salomone et al. [55] found that the moisture content has a large effect on the thermal conductivity of pulverized clay below the critical moisture content at different dry densities. Nikiforova et al. [56] studied the thermophysical properties of soils, analyzed different soil types, and found correlations between thermal conductivity and humidity. Ye et al. [57] and Dong et al. [58] found that the thermal conductivity of loess at room temperature is positively correlated with moisture content and dry density, and the effect of moisture content is greater. Tang et al. [59] found that increasing microencapsulated phase change materials from 5% to 10% improved silt sandy soils’ thermal conductivity. Feng et al. [60] found that copper tailing sand improved RC thermal conductivity, with linear growth with sand mixing and moisture content, and reduced growth rate beyond a dry density of 1.65 g/cm³. Dong et al. [61] found that the thermal conductivity, thermal diffusivity, and specific heat of LF improved loess increased with the increase of dry density, and at a content of 5%, the sensitivity of the improved loess to changes in heat was low and the strength of the specimen reached its maximum value. Wang et al. [62] found that when the soil temperature is negative, the lower the temperature, the greater the thermal conductivity of the soil. If the initial moisture content is lower than 10%, the thermal conductivity of the soil is not affected by temperature; if the initial moisture content is greater than 15%, the thermal conductivity of the soil is more affected by temperature. Kuila et al. [63] investigated CaCO₃-modified epoxy/BF composites and found that BFs effectively improved the thermal properties of the composites and increased their stability. Aziz et al. [64] found that the optimum improvement in thermophysical and mechanical properties of Bricks made from clay is 41.5% for doum fiber content of 10%. Yassine et al. [65] found that when the hemp volume fraction increased, the thermal conductivity, thermal diffusivity, and thermal effusivity capacity all decreased by approximately 52%, 27%, and 35%, respectively. Sayouba et al. [66] found that the thermal properties, such as thermal conductivity and thermal diffusivity of plain clay–PF mixture samples decreased with increasing fiber content.

In summary, scholars have mainly focused on studying shear strength and thermal properties of conventional viscous and silt soils after reinforcement and analyzing thermal conductivity under different moisture content, porosities, and soil textures. In cold regions, the impact of freeze–thaw cycles is a concern, and few studies have explored BF and PF-modified RC as construction and subgrade backfill soil and geothermal energy system backfill material. Additionally, improving special soils’ thermal conductivity and studying RC’s characteristics and BFs’ insulating properties require further research.

This study aims to explore changes in shear strength and thermodynamic performance of modified soil under different moisture content and fiber content under freeze–thaw cycle conditions, as well as changes in microstructure and damage mechanisms. Additionally, it compares PF with BF to provide references for improving RC in cold regions, such as subgrade and construction backfill soil, energy tunnel construction materials, and backfill materials for geothermal energy extraction.

2. Materials and Experimental Methods

2.1. Materials

The RC used in the experiments is sourced from the vicinity of Heling New District in Hohhot, Inner Mongolia. The natural state of the collected RC is shown in Figure 1a. The clay is air-dried outdoors, crushed using a pulverizer, sifted through a 2 mm sieve, and then dried in an oven set to 105 °C for 12 h. After cooling, it is stored in sealed zip-lock bags for later use. The basic physical properties of the processed RC are determined according to the “Geotechnical Test Method Standard” (GB/T 50123-2019) [67], with the basic physical indicators shown in

Table 1. Basic physical parameters of RC.

Soil Sample	ωP (%)	ωL (%)	ωopt (%)	ρdmax (g/cm3)	Gs
RC	25.7	44.7	18.1	1.72	2.62

ω_P: plastic limit; ω_L: plastic limit; ρ_dmax: the maximum dry density; ω_opt: optimum moisture content; Gs: particle relative density.

.

The BFs used in the experiments, shown in Figure 1c, are purchased from Changsha, Hunan, with their specific parameters listed in

Table 2. Basic physical indexes of BF.

Materials	ρ (g/cm3)	E (GPa)	Rm (N/tex)	ε (%)	L (mm)	D (μm)
BF	2.94	96.82	0.41	2.83	6	17

ρ: density; E: elastic modulus; Rm: tensile strength; ε: elongation; L: lengths; D: caliber.

. These fibers exhibit very high acid and alkali resistance, strong low-temperature performance, and high-temperature endurance, along with a relatively low water absorption capacity. The LF used in the test is shown in Figure 1b, which is purchased from Shijiazhuang, Hebei, and is white in color, odorless, contains some crumbly particles, is extremely stable, has a small specific gravity, a large comparative area, and has strong toughness and dispersion. Its specific parameters are shown in

Table 3. Basic physical indexes of LF.

Materials	L (mm)	D (μm)	PH	ω (%)	T (°C)	Ash (%)
LF	1 ± 0.5	40	7.0 ± 0.5	<5	220	18 ± 2

L: lengths; D: Fiber diameter; T: heat resistance.

.

2.2. Shear Strength Test

2.2.1. Experimental Program

The shear strength indices of BF-modified RC are obtained through direct shear tests. To investigate the impact of freeze–thaw cycles and BF content on the shear strength of RC and comparison with PF, the prepared samples are controlled at a dry density of 1.71 g/cm³ with a moisture content of 22.5%. BFs are added to the RC in dry soil mass percentages of 0%, 0.5%, 1%, 2%, and 3%. Five groups of samples are prepared, including plain RC and RC with different amounts of fiber reinforcement. Each group consisted of three parallel samples, with four incremental vertical pressures of 100 kPa applied, totaling 15 samples. Four sets of samples are prepared for each fiber content to undergo 1, 2, 3, and 4 freeze–thaw cycles, respectively. Each set also consisted of three parallel samples, which are tested in the same manner as described above, totaling 75 samples. Under the same conditions of dry density and moisture content as the PF, LFs are added to the RC at 0%, 1%, 2%, 4%, 6%, and 8% by mass of dry soil, for a total of six groups of samples to be tested, and three parallel tests are also set up for each group of samples.

2.2.2. Specimen Preparation

In the preparation of direct shear test samples, the method of incorporation significantly affects the results. In previous experiments, such as those conducted by Li et al. [68], air-dried soil samples are first prepared to the target moisture content, and then wet soil samples are uniformly mixed and pressed with fibers. Adding water first can lead to a significant loss of moisture in the soil, causing considerable errors in the experiment. This paper adopts a post-water addition mixing method, manually dispersing the BFs and PFs before mixing them with the soil. Although this process takes longer, it ensures that the moisture content of the samples is consistent with the targeted moisture content. Each sample’s required amount of soil and BFs and PFs is divided into three parts, mixed separately, and then combined. Water is then added to adjust to the target moisture content, ensuring uniform distribution of BFs and PFs within the soil and reducing moisture loss, thus guaranteeing the actual moisture content reaches the target value. After the soil sample preparation, the samples are placed in zip-lock bags, sealed, and cured for 24 h to allow for an even distribution of moisture throughout the samples. After curing, the samples are placed in a ring cutter with an inner diameter of 61.8 mm and a height of 20 mm, compacted to the target dry density, and immediately sealed with cling film to preserve them, as shown in part of Figure 1d. The samples are then stored at room temperature in a sample storage box until testing. The same method is applied to prepare samples for four freeze–thaw cycles, with some samples shown in Figure 1e. The prepared samples are then placed in a freeze–thaw test chamber and subjected to varying numbers of freeze–thaw cycles.

2.2.3. Experimental Process

According to the direct shear test standard, using the SDJ-1 three-speed equal strain direct shear apparatus produced by Nanjing Speico Test Instruments Co., Ltd. (Nanjing, China), the preparation of the instrument is completed by zeroing the dial gauge of the force ring. Then, vertical loading is carried out according to the standard (GB/T 50123-2019) [67], with vertical pressures set at 100, 200, 300, and 400 kPa, respectively. The shearing rate is set to 1.2 mm/min. During the shearing process, the occurrence of a significant rebound in the load cell reading serves as the indicator of shear failure. Typically, the test is halted when the shear deformation reaches 4 mm or extends to 6 mm if no shear failure is observed. The entire shearing process lasts about 5 min. After shearing, the shear stress and vertical pressure are removed, the upper and lower shear boxes are separated, and the percentage gauge readings at different vertical pressures are recorded for data analysis.

2.3. Thermophysical Property Test

2.3.1. Experimental Instruments and Principles

Thermal conductivity is a physical quantity that indicates the thermal conductivity of a substance. It refers to the amount of heat transferred through an area of 1 m² within 1 s under the conditions of unit thickness and unit temperature gradient. It is expressed by the symbol K, with the unit of W/(m·K). In this paper, the thermophysical parameters of the modified reinforced soil are determined by utilizing the transient planar heat source method of the thermal conductivity tester, as illustrated in Figure 1f,g. The test range is 0.01~300 W/(m·K) with a resolution of 0.0005 W/(m·K), accuracy ≤ 3%, and repeatability ≤ 3%. The solution formula is:

$$\lambda ={\displaystyle \frac{q}{4\pi }}/{\displaystyle \frac{d\theta }{d\ln t}}$$

(1)

In the formula, $\lambda$ is the sample thermal conductivity, $q$ is the heating power, $\theta$ is the probe temperature rise, and $t$ is the heating time. The transient plane heat source method of the thermal conductivity tester employs the screening data method to obtain data on the linear section of the linear fit, therefore enabling the calculation of thermal conductivity. This method can also be used to determine the specific heat, thermal effusivity, and other thermophysical parameters.

2.3.2. Experimental Program

To investigate the effect of moisture content and fiber incorporation on the thermophysical properties of RC, a moisture content gradient is established based on the liquid and plastic limits of RC. The BF content is 0.5%, 1%, 2%, and 3%, and the PF content is 1%, 2%, 4%, and 6%. At the same time, the dry density of the improved soil is controlled to 1.71 g/cm³, which is compared with the PF-improved RC at optimum moisture content, as shown in

Table 4. Test Scheme.

Serial Number	BF Content (%)	Mass Moisture Content (%)
HN-1	0	18.5, 22.5, 26.5, 30.5
HN-2	0.5	18.5, 22.5, 26.5, 30.5
HN-3	1.0	18.5, 22.5, 26.5, 30.5
HN-4	2.0	18.5, 22.5, 26.5, 30.5
HN-5	3.0	18.5, 22.5, 26.5, 30.5

for the experimental protocol. Three identical specimens per group are prepared for parallel tests to analyze the relationship between the thermophysical properties of RC and moisture content and fiber admixture, and the mineralogical composition of RC is analyzed using X-Ray Diffraction (XRD).

2.3.3. Experimental Steps

First, BF, PF, and RC are weighed according to the moisture content and dry density set in the test program, and the fiber and soil are mixed uniformly. Then distilled water is added for mixing again, and the mixing is carried out in a sealed plastic bag to prevent the loss of water during the mixing process and the mixing is left for 24 h after mixing uniformly, and then the amount of soil required for the test specimen is weighed to the precision of 0.01 g, and placed in a specimen mold with specifications of Φ 61.8 mm × 20 mm and a volume of 60.0 cm³ to make a standard specimen, control the dry density to keep the same, make a standard specimen, and immediately seal and preserve it with cling film, and then put the specimen into a sample preservation box to preserve the specimen at room temperature, and wait for the test to measure the samples. The hot disk probe of the thermal conductivity tester is then placed in the middle of the specimen, the specimen is fixed with the probe using a jig, and the above operation is carried out using gloves with heat-insulating functions to prevent the test results from being affected by the temperature in the hand. Finally, during the test (20–30 min), a constant temperature (around 20 °C) is controlled, and the windows and doors are closed to prevent air convection from affecting the temperature at the cold end of the sensor. Wait 45 min, repeat the test, and take the average value after 3 tests, i.e., the measured thermal property parameters.

2.4. XRD and SEM Test

XRD analysis and SEM scanning are conducted on BF-modified RC specimens to analyze the mineral composition of RC from the Hohhot City area in Inner Mongolia. Samples of RC with varying BF content and numbers of freeze–thaw cycles are prepared. Representative soil samples are then selected from the shear damage surface, dried, sliced in two dimensions, and kept flat to ensure that the electron beam can hit the surface of the samples more uniformly, resulting in a clearer final image. The processed specimens are then connected to a glass sheet through conductive adhesive, vacuumed, and gold-plated. Once the specimen has been vacuumed and gold-plated, it is scanned and analyzed using a SEM model S-3400N, as illustrated in Figure 1h. This microscope can produce backscattered electron images of different solid specimens, surface morphology images using secondary electrons, mixed images of two images, and image processing. The microstructural characteristics of RC and modified RC after the freeze–thaw cycle are elucidated from the microscopic level. The contact interface between BFs and soil is observed through SEM, revealing the mechanical transfer mode and mechanism of action between fibers and soil medium. Furthermore, the influence law of BFs on the micro-surface structure and micro-pore structure of RC is visually analyzed.

3. Results and Discussions

3.1. Effect of BF Content and Number of Freeze–Thaw Cycles on Shear Strength

Following the addition of BFs to the RC, a series of shear tests are conducted at room temperature and under freeze–thaw conditions with varying content. The objective is to analyze the two key parameters of soil shear strength, cohesion (c), and angle of internal friction (φ). The shear strength indexes of the BF-reinforced RC specimens are obtained at the control of the dry density (1.71 g/cm³) and moisture content (22.5%) as a function of the number of freeze–thaw cycles and the content of the fibers. These changes are illustrated in Figure 2. As illustrated in Figure 2a, the straight shear test is conducted at room temperature without a freeze–thaw cycle. The results demonstrated that the incorporation of BF could effectively enhance the cohesion (c) of RC. Cohesion (c) exhibited a tendency of increasing and then decreasing in conjunction with the content of BF. The results are compared with the study of Niu et al. [29], and the pattern of change is close to that of the simultaneous addition of BFs and glass fibers, but the optimum fiber content is different. They determine the optimum fiber length to be 9 mm and find two peaks with increasing fiber content. The addition of the fiber results in an increase in the toughness of the soil body, with the shear stress appearing to increase slowly and continuously during the shear process.

The maximal shear stress Is observed to Increase continuously, with the greatest increase in cohesion (c) occurring at a content of 0.5~1%. The increase ratio is found to be 60.91%. At a content of 2%, the cohesion (c) reached its maximum value, with the amplification effect reaching the best value. This is accompanied by a significant increase in the cohesion (c) value, which reached 58.19 kPa. This value is higher than that of plain RC, with an average increase ratio of 259.86%. Comparing the results of Mohamad et al. [48] research, BFs enhance the shear strength better than hemp fibers. Jiang et al. [38] find that the incorporation of palm fibers in pulverized clay can significantly improve the shear strength and deformation resistance of the soil, with the peak value reached at 20 mm fiber length, which is 172.9% higher than that of the plain soil. The improvement of BFs over palm fibers is more pronounced and the increase in shear strength is relatively more significant compared to that of plain clay. The image of the whole roughly fits into the quadratic parabola. At the content of 2%, the BF and the soil particles are tightly bonded. The shear strength is remarkable. With the increase in fiber content, the cementing effect of soil particles decreases, and the reinforcing effect begins to decline significantly. The internal friction angle (φ) in fiber content is one of decreasing and then increasing. The internal friction angle (φ) is the smallest in fiber content at 2%, and the overall change in the internal friction angle (φ) is relatively gentle. Comparing the results of Wang et al. [28] research, a close pattern appeared with the addition of polypropylene fibers and BFs, and the change in the angle of internal friction of the soil improved by both types of fibers is smaller. In fiber content at 3%, there is a certain increase, but this is less than the internal friction angle (φ) of plain RC. The error bars in the graphs indicate that the data from the three sets of parallel tests is generally within reasonable limits.

Figure 2b,c illustrate that the cohesion (c) of the RC exhibits a tendency to increase and then decrease with the addition of BFs following several freeze–thaw cycles. Concurrently, the curves observed under different freeze–thaw cycles exhibit a quadratic parabola-like shape. However, from the initial freeze–thaw cycle, the curves exhibit a flattening trend compared to the unfrozen conditions, accompanied by a decrease in the growth rate of cohesion (c). The extent to which the average growth rate of fiber content decreased from 0%~2% after the first freeze–thaw cycle is 97.3%, which is a decrease of 162.56% compared to the unfrozen conditions. Concurrently, the rate of decrease is also slowed down in comparison to the unfrozen thaw at 2~3% content, and the freeze–thaw cycle exerts a less pronounced effect on the change in cohesion (c) at 3% fiber content. The internal friction angle (φ) of plain RC and different fiber content decreased significantly with the first freeze–thaw cycle, except for the fiber content of 2%. In contrast, the overall decrease in the decrease ratio is in a decreasing trend with the increase in fiber content, with an average decrease ratio of 38.445%. The change in the internal friction angle (φ) is very small and gradually tends to be stabilized in the subsequent freeze–thaw cycle.

Figure 2. (a) Variation of shear strength index with BF contents (b) Variation of internal friction angle (φ) with BF contents and number of freeze–thaw cycles (c) Cohesion (c) with BF contents (d) Cohesion (c) with number of freeze–thaw cycles.

Figure 2. (a) Variation of shear strength index with BF contents (b) Variation of internal friction angle (φ) with BF contents and number of freeze–thaw cycles (c) Cohesion (c) with BF contents (d) Cohesion (c) with number of freeze–thaw cycles.

In Figure 2d, it can be observed that as the number of freeze–thaw cycles increases, the cohesion (c) of both the plain RC and the RC with different fiber content displays a downward trend, followed by a gradual stabilization. After the initial freeze–thaw cycle, the cohesion (c) of both the plain RC and the RC with different fiber content exhibited a significant decrease, with an average reduction of 35.71%. Furthermore, the decrease ratio demonstrated a trend of increasing and then decreasing with the content of BF, with the maximum reduction observed at 51.76% when 1% of BF is added to the soil. The similarity with the results of Wei et al. [30,31] is that after freeze–thaw cycles, the strength of the improved soils all shows a decrease, but the shear strength of BF-improved special soils gradually stabilizes with the increase in the number of cycles. Following the second freeze–thaw cycle, the cohesion (c) of RC under each fiber content exhibits a slight increase. With the number of freeze–thaw cycles increasing, cohesion (c) gradually becomes more stable, maintaining a BF content of 2%. The maximum value of cohesion (c) is observed at this content, and with the increase of fiber content, the slope of the fluctuation of the curve appears to be the trend of the first increase and then decrease.

The freeze–thaw cycle test revealed that following the initial freeze–thaw cycle, a considerable quantity of pore water accumulated within the specimen, resulting in the specimen’s surface becoming inundated with water. The morphological transformation of the pore water is identified as the primary factor contributing to the observed change in shear strength. The strength index of the RC is found to be significantly reduced as a result of the water turning into ice. This process involved the hydrogen bonding in the water becoming smaller and the spacing of the water molecules becoming bigger, which resulted in the volume of the ice becoming bigger. This generated a freezing expansion force that destroyed the original structure of the soil body. In the specimen subjected to the straight shear test, the shear stress in the upper and lower boxes reached 2 mm, with the value of the measuring force ring reaching its maximum. There is no similar unfrozen specimen, with the shear stress in the process of slowly increasing. However, the value began to decline rapidly, resulting in a significant reduction in the specimen’s toughness. The specimen, following the straight shear test, exhibited a deterioration in its internal structure due to the precipitation of pore water on the soil surface, leading to the destruction of the internal structure and the generation of through-pore space. This increased the brittleness of the specimen, as illustrated in Figure 3. Comparing the results of Shaker et al. [35], it is surprising to find that the addition of fibers had a favorable effect on the pore water flow in the soil, with the reinforcement of fibers reducing the pore water flow phenomenon due to freeze–thaw cycles. The incorporation of BFs and the elevation of their content resulted in a considerable enhancement in the shear strength of the soil despite the damage to its internal cemented structure caused by the freeze–thaw cycle. Moreover, the reinforcing effect of the fibers continued to yield positive outcomes even after the freeze–thaw cycle.

3.2. XRD Whole Rock Mineral Analyses

To investigate the mineral composition of RC, modified RC samples underwent XRD spectroscopy analysis. The XRD spectra of the RC samples are analyzed to obtain the composition and content of the RC. The main components are identified as quartz (45.4%), potassium feldspar (2.7%), plagioclase (8.4%), calcite (2.6%), and clay minerals (40.8%). Details of the XRD images are shown in Figure 4d.

In Hohhot RC, quartz is the predominant component, characterized by its stability and poor hydrophilicity. At room temperature, its thermal conductivity is 1.3 W/(m·K), and it maintains good thermal conductivity even in high-temperature environments. This thermal performance is due to the Si-O bonds in the quartz structure, which effectively transmit thermal energy. The next major components are clay minerals, including kaolinite and montmorillonite, known for their strong hydrophilicity. Lv et al. [69] use XRD, the K-value method, and the Bogue RH method to determine the main minerals of RC from Guilin and Wuming, Guangxi, as kaolinite, Goethite, and Gibbsite. The mineral content in Wuming RC is kaolinite (74.0%), gibbsite (12.53%), goethite (>2 mm, 8.37%) and goethite (<2 mm, 2.22%); the mineral content in Guilin RC includes kaolinite (56.59%), goethite (15.61%), gibbsite (11.44%) and quartz (12.45%). Compared to the RC in Guangxi, the Hohhot area in the north has relatively fewer clay minerals and more quartz, making up nearly half of its content, indicating that Hohhot RC has relatively weaker water absorption properties and higher strength.

Figure 4. (a) Thermal conductivity versus moisture content (b) Thermal effusivity versus moisture content (c) Thermal diffusivity versus moisture content (d) XRD.

Figure 4. (a) Thermal conductivity versus moisture content (b) Thermal effusivity versus moisture content (c) Thermal diffusivity versus moisture content (d) XRD.

3.3. Effect of Moisture Content and BF Blending on Thermal Conductivity Properties

The moisture content of soils significantly impacts various physical properties, including thermal conductivity, etc. The fitted curve of thermal conductivity versus moisture content for BF-modified RC at the same dry density is shown in Figure 4a. From the figure, it can be observed that the thermal conductivity of veg-RC increases gradually with the increase of moisture content at the same dry density (1.71 g/cm³). In the range of moisture content from 18.5%~30.5%, the thermal conductivity exhibited a linear increase with the increase of moisture content, with the following values: 0.0412, 0.0834, and 0.1719 W/(m·K), with an average increase of 10.78%. Comparing the results of Xu et al. [54] and Salomone et al. [55] research. The RC in Hohhot region contains a large amount of quartz material, which produces a near-consistent pattern. For special RC, the thermal conductivity maintains a linear increase with moisture content up to the bounding moisture content. After exceeding the bounding moisture content, it gradually becomes flat. Furthermore, the improved soil with BF addition demonstrates a sustained increase in thermal conductivity with the increase of moisture content at varying fiber content, with the curves exhibiting a roughly exponential distribution.

As moisture content increases, the pores within the soil are filled with water, decreasing the volume of air, which has a lower thermal conductivity compared to water. Therefore, within a certain range of moisture content, the thermal conductivity of BF-modified soil increases as the moisture content increases. This is due to the thickening of water films between soil particles, reduction in porosity, and increased area for heat transfer, resulting in an increased thermal conductivity of the soil. Also, with reference to the findings of Xu et al. [54], the thermal conductivity increases with increasing moisture content. Notably, the increase in thermal conductivity is most pronounced at moisture content between 26.5% and 30.5%.

Soil is a porous medium, and its thermal conductivity is primarily related to its porosity and the materials filling the pores. Upon adding BFs, the thermal effusivity (S), which reflects the material’s capacity to store heat, changes, as shown in Figure 4b. Thermal effusivity initially increases and then decreases as moisture content rises. At 22.5% moisture content, both the pure RC and fiber-modified soils reach their optimal heat storage capacity and thermal stability. Beyond 22.5% moisture content, the heat storage capacity begins to decline, gradually stabilizing between 26.5%~30.5%. Ye et al. [57] and Dong et al. [61] only analyze the thermal conductivity of loess, and there is no in-depth study on the thermal storage capacity. According to the above analysis, RC shows good thermal storage capacity. In addition, the heat storage capacity of the plain RC varied less with increasing moisture content, and the rate of increase was very significant in the modified RC with the addition of BFs.

The thermal diffusivity is influenced by the thermal conductivity, the density of the material, and the specific heat. This coefficient reflects the ability of the material to reach the same temperature in each part when it is heated or cooled. The variation of this value with moisture content is shown in Figure 4c. It can be observed that the thermal diffusivity of plain RC and modified clay exhibit a tendency to decrease and then increase as the moisture content increases, reaching a minimum of 22.5% and then beginning to increase continuously. At 30.5% moisture content, the value of thermal diffusivity is significantly higher than that of the initial moisture content, and the thermal insulation property is gradually weakened while the heat dissipation capacity is gradually increased.

The BF exhibits the same outstanding performance in terms of thermodynamics. It has good thermal insulation properties at high temperatures and possesses superior water stability. The effect of different BF content on the thermal conductivity is shown in Figure 5a, where the same dry density (1.71 g/cm³) is used. The analysis indicates that with an increase in the content of BF, there is a decrease in thermal conductivity, followed by an increase with a slight change in the curve. After the content rate of 2%, a slow increase in the trend begins to appear. The addition of microencapsulated phase change materials (mPCM) increases the thermal conductivity of the sandy clay in comparison with the results of Tang et al. [59], who change the temperature to analyze the change in thermal conductivity change, but do not analyze the change in thermal storage capacity, BFs do not improve the thermal conductivity of RC, but greatly strengthen the thermal storage and insulation capacity of RC.

Figure 5b illustrates that with the increase in BF content, the thermal effusivity initially increases and then decreases in a linear fashion. The growth rate is particularly evident in the fiber content range of 0.5%~1%, with an average increase ratio of 28.15%. At a fiber content level of 1%, the thermal storage capacity reaches its maximum. In conjunction with the previous analysis of the effects of different moisture content, it can be seen that the thermal effusivity is at its maximum, and the ability to store heat is at a relatively optimum level, with a moisture content of 22.5% and a fiber content of 1%.

The relationship between different fiber content and thermal diffusivity is illustrated in Figure 5c. It shows that with an increase in BF content, the thermal diffusivity appears to decrease and then increase, and it tends to be stable when the content is 2~3%. Combined with the above analysis of the thermal diffusivity of different moisture content, it can be seen that the thermal diffusivity of red clay is significantly reduced after adding fiber. At the same time, as the fiber content increases, the thermal diffusivity gradually becomes stable. At a moisture content of 22.5% and a fiber content of 1%, the thermal diffusivity reaches a low level, and the heat preservation effect is relatively high, in contrast to the study of Yassine et al. [65]. Hemp fibers have a large effect on thermal conductivity, thermal diffusivity, and thermal effusivity. They both decrease with increasing hemp fiber.

Figure 5. (a) BF contents versus thermal conductivity (b) BF contents versus thermal effusivity (c) BF contents versus thermal diffusivity.

Figure 5. (a) BF contents versus thermal conductivity (b) BF contents versus thermal effusivity (c) BF contents versus thermal diffusivity.

3.4. Electron Microscope Scanning and Analysis

The microstructure of the soil body is the fundamental factor affecting the mechanical properties and basic physical properties of the soil body. To analyze the microstructural changes in RC under the action of freeze–thaw cycles and different content of BFs, some of the specimens are selected after shearing, and SEM is carried out on the shear surfaces. Representative SEM images are then analyzed.

It displays representative SEM images taken after the experiment in Figure 6. For the 0% fiber content RC, Figure 6a,b show that the soil particles have a rough surface, and the sample mainly consists of variably sized particles bonded together without a regular arrangement. Smaller particles adhere to larger ones, primarily through face-to-face contact and fewer point-to-face contacts, which is characteristic of the fine-grained structure of clay particles. The images also reveal a fragmented shear fracture surface with large and numerous pore openings, indicating poor compaction. Gao et al. [13,16] carry out some research on BF-improved southern RC and analyze the microcosmic situation. Comparing the microcosmic situation of RC in different regions without fiber mixing, the difference in the contact mode between soil particles is relatively small, and the RC in the Hohhot region is relatively looser, with more internal pores and relatively low moisture content. The samples underwent shear failure, fundamentally caused by the disruption of the bond between soil particles and the water films, reducing cohesion (c) and dislocating the particles, leading to a more dispersed arrangement and internal microstructural changes. Adding fibers effectively reduces the contact damage between soil particles.

Figure 6. (a,b) Vegetative RC; (c,d) Content of 1%; (e,f) Content of 2%; (g,h) Content of 3%; (i) Content of 0% after 4 freeze–thaw cycles; (j) Content of 2% after 4 freeze–thaw cycles.

Figure 6. (a,b) Vegetative RC; (c,d) Content of 1%; (e,f) Content of 2%; (g,h) Content of 3%; (i) Content of 0% after 4 freeze–thaw cycles; (j) Content of 2% after 4 freeze–thaw cycles.

It shows the SEM images of BFs doped with 1% content in Figure 6c,d. As can be seen from the images, the low content of blended fibers results in a more uniform distribution of fibers in the soil, with a higher degree of dispersion, mainly individual fibers wrapped in the soil or simple crossings between individual fibers and bare fibers containing a small number of soil particles, too. The main manifestation of this content is the gripping effect of fiber reinforcement, i.e., the friction and adhesion generated by the gripping effect between the fibers and the soil particles, which improves the shear strength of fiber-reinforced soils. Tang et al. [70] suggest that two main mechanical interactions occur at the fiber/soil interface: adhesion and friction, both resulting from the gripping effect. Figure 6c illustrates that due to shear forces, the soil particles on the surface are compressed, reducing porosity and making the soil more compact and smoother.

It illustrates the results of 2% fiber content in Figure 6e,f. The BF and soil particles are observed to be closely cemented, resulting in a significant increase in shear strength, and the maximum shear strength is reached at this point. It can be seen from the figure that the fibers are no longer distributed individually. After the number of fibers increases, a plane intersecting shape appears. The fiber distribution begins to become concentrated, the discreteness is reduced to a certain extent, and the density is moderate. The contact between the fibers is no longer in the form of simple individual intersections. The intersection of the fibers became more complex with the emergence of a number of different angles of the cross, the staggered overlap phenomenon, and the formation of a simple fiber network structure. This structure can still effectively fill the soil void and adsorb soil particles on the surface of the fiber, enhancing the soil and fiber gripping effect between them. The structure can effectively increase the force area of the soil body so that the force of the soil body is more uniform and improves the integrity of the soil body. When the fiber network is subjected to shear force, the local effect is transferred to the whole fiber network to halt the trend of this force. The superposition of the transfer in the fiber network can better resist the role of shear force. At this time, the mechanical properties of the soil body are enhanced to the greatest extent. Comparing the results of Jiang et al. [38], it can be seen that the grooves on the surface of palm fiber can increase the interfacial friction and occlusion between fiber and soil particles. The interfacial friction and occlusion between fibers and soil particles reached the optimum at 2% BF content, which is close to the microscopic situation of palm fibers, but in this paper, the microscopic mechanism is further studied and analyzed in depth.

In Figure 6g,h, the fiber content is 3%. It can be seen in the figure that a complex fiber network structure has been formed. This process has also produced a certain degree of embracing phenomenon, with a low degree of discrete fibers and high local densities. Some fibers are not even cemented with soil particles. Excessive fiber agglomeration cannot fill the soil pores, and the gripping effect begins to decrease. The role of the fibers in gripping the soil begins to decline, and the reinforcing effect begins to wane. However, the role of the fiber network structure remained prominent. Furthermore, at this stage, the soil shear strength is not 2% fiber-modified RC high, but in comparison to the fiber content of 0% of the RC, there is still a certain degree of enhancement. Xiong et al. [41] carried out SEM of glass fiber-reinforced soil, and the poor cementation of glass fibers with non-cohesive soils, similar to the overmixing of BFs, resulted in uneven distribution of fibers, which makes it difficult to play a role in reinforcing the soil and is unable to form a uniformly distributed mesh structure in the soil. However, BFs have superior dispersibility to glass fibers.

Freeze–thaw cycles can damage the internal structure of the soil body, resulting in a reduction of its shear strength. Figure 6i illustrates the RC after four times freeze–thaw cycles. In comparison to Figure 6a,b, which shows unfrozen and thawed RC, it is evident that the pore space is significantly larger, and the number of pores is also larger. This results in the formation of local cracks and holes, and the soil particles become more fragmented and looser, along with the change in contact mode from face-to-face to point-to-face, leading to the formation of localized clusters, which in turn resulted in a decrease in the cohesion (c) of the RC. During the freeze–thaw cycle, multiple changes in pore water morphology result in the generation of freezing and expansion forces within the soil body, which reduce the adhesion between soil particles and damage the soil skeleton. As the temperature increased, pore water flowed more easily in the dilated pore space, eventually aggregating on the surface of the specimen, which led to a significant enhancement of the deterioration degree of the soil body.

In Figure 6j, a comparison of the images with an unfrozen fiber content of 2% shows that it can be observed that the freeze–thaw cycle causes BFs to rearrange within the specimen, resulting in a transition from an original staggered overlap to a parallel arrangement of fibers. At the same time, there has been an increase in fiber density, which has produced some clumping and some fibers directly and closely in contact with each other. This results in the formation of a localized fiber network, which is separated from the soil particles by a significant fissure. The width and size of the cracks and pores are evidently strengthened, while the gripping effect between the fibers and the soil is significantly reduced. This is comparable to Figure 6g. Following freezing and thawing, the rearrangement of fibers and movement of pore water resulted in the formation of larger and longer cracks than those observed in Figure 6f. However, the presence of the fiber network structure effectively constrained the flow of pore water and the movement of soil particles due to the freezing and thawing cycles. Even though the resistance to shear damage remains significantly higher than that of 0% RC following numerous freeze–thaw cycles, the shear strength reaches the optimum at 2% with an increase in fiber content. Wei et al. [30,31] find that after freeze–thaw cycles, the spatial constraint effect of fibers on soil and the friction effect of reinforced soil limits the increase of pore diameter and porosity and improves the shear resistance and freeze–thaw resistance of the soil. Through the microscopic study of BF-improved soil, BFs effectively inhibit the movement phenomenon of soil particles after freeze–thaw cycles, which microscopically verifies the analysis of Wei et al. Fiber reinforcement enhances the freeze–thaw resistance and toughness of the soil.

As the content of BF increases, the RC’s heat storage capacity is also effectively enhanced. At 1% fiber content, the dispersion of fibers is high, resulting in a more uniform distribution within the soil and forming a simple crossing structure that was strongly gripping effect with the soil, effectively utilizing the thermal insulation properties of BF. As the fiber content increases to 2–3%, the preliminary and maturing fiber network structure forms and some fiber clumping occurs, leading to a decrease in the uniformity of the soil’s heat storage ability, though it remains higher than that of pure RC.

The mechanism of fiber improvement in RC is illustrated in Figure 7. The freeze–thaw cycle significantly reduces the shear strength of specialized RC and causes volumetric changes. By adding different content of BF to RC, the effects of enhancing RC performance are explored. At a 2% content, the shear strength of RC reaches its maximum, effectively enhancing its freeze–thaw resistance. The primary reason for this improvement is the addition of BF, which reinforces the internal structure of RC, resulting in a more compact internal structure and the formation of a simplified fiber network. The bonding effect of the fiber effectively resists the freezing and expansion forces generated during the freeze–thaw cycle. Simultaneously, the mechanical properties of RC are greatly enhanced, reducing internal pore space and minimizing crack formation, therefore enhancing integrity and completeness.

Meanwhile, comparing the PF in terms of the micro-mechanism of the reinforcing action, it is basically the same as the BF, and it is sufficient to refer to Figure 7 without further elaboration.

3.5. Comparison with LF Improved RC

The shear strength and thermophysical properties of the BF-modified RC are analyzed macroscopically and microscopically in the previous section, and based on the above tests, the shear strength and thermophysical properties of the RC modified by the addition of PF: LFs at a moisture content of 22.5% and a dry density of 1.71 g/cm³ are shown in Figure 8.

In Figure 8a, it can be seen that compared with BF, after the addition of LF, the cohesion (c) of RC increases and then decreases with the increase of fiber content. In the fiber content of 2%, the cohesion (c) is 38.18 kPa, and under the same content, the reinforcing effect is worse than that of BF. In the content of 4%, the cohesion (c) appears to increase significantly, and then the pores of the soil body are completely filled with PF, and the reinforcing effect increases significantly. The reinforcing effect of the LFs is significantly better than that of the BF at a content of 6%, while the cohesion (c) reaches its maximum value at a content of 6%, with an increased ratio of 324.18%. The internal friction angle (φ) appeared to decrease and then increase with increasing fiber content, and the overall change is small. Although the effect of BF improvement is superior to that of LF at the same content of 2%, the optimum effect of LF improvement is much higher than that of BF, and the effect of LF on the shear strength of RC is better than that of BF in a comprehensive analysis of improved effect and waste utilization perspectives. At the same time, depending on the content and strength improvement effect of the two fibers, they have different advantages in different application scenarios.

By comparing the thermophysical properties of BF-modified RC at 22.5% moisture content and 1.71 g/cm³ dry density with the addition of LF, the thermophysical properties of the modified RC are shown in Figure 7 and Figure 8b,d. In Figure 8b, it can be seen that with the increase of LF content, the thermal conductivity seems to increase and then decrease, and the thermal conductivity reaches a maximum value of 1.1772 W/m·K at a fiber content of 2%, with an increase ratio of 18.44%, and the incorporation of LF effectively improves the thermal conductivity of RC. Feng et al. [60] add copper tailing sands to RC, which effectively enhances the thermal conductivity of RC, and find the bounding dry density of 1.65 g/cm³ for strengthening the thermal conductivity of RC, from the economic, environment, and sustainable development point of view, PFs are more advantageous. Dong et al. [61] also observe a significant increase in thermal conductivity when LF is added to loess. However, the thermal conductivity of RC reaches its maximum value when LF is added at 2%, which is significantly better than the amount of loess, and the effect is better.

In Figure 8c, the thermal effusivity appears to increase and then decrease with increasing fiber content, and the heat storage capacity reaches the maximum value of 16.478 W/(m²·K) at 2% content. Compared with BF, the heat storage capacity of the two types of fiber is almost the same at the same content, but BF is significantly better than LF and achieves the best effect at 1% content.

In Figure 8d, the thermal diffusivity does not change much with the fiber content, and in the content of 6%, the thermal diffusivity has a large improvement to 0.8902 mm²/s, and the specific heat decreases significantly, with the specific heat of 0.634 kJ/kg·k. The comprehensive analysis shows that the BF effectively strengthens the heat preservation and storage capacity of the soil body, and the decrease in the thermal conductivity capacity is small. The LF effectively strengthens the thermal conductivity of the RC soil, and the thermal storage capacity has been improved, which has different reference values for energy development and utilization in alpine areas, energy tunnel construction, and so on.

Figure 8. (a) LF contents versus cohesive forces (b) LF contents versus thermal conductivity (c) LF contents versus thermal effusivity (d) LF contents versus thermal diffusivity.

Figure 8. (a) LF contents versus cohesive forces (b) LF contents versus thermal conductivity (c) LF contents versus thermal effusivity (d) LF contents versus thermal diffusivity.

4. Conclusions

This investigation aims to explore the changes in shear strength and thermodynamic performance of modified soil under different moisture content and fiber content, as well as changes in microstructure and damage mechanisms. The following conclusions could be obtained:

(1) At room temperature, the cohesion (c) of the RC increases and then decreases with the increase of BF content, reaching the maximum value at a content of 2%. The internal friction angle (φ) is less affected by the BF content. With the increase in the number of freeze–thaw cycles, the cohesion (c) of RC decreases first and then gradually stabilizes. Meanwhile, with the increase of BF content, after different numbers of freeze–thaw cycles, cohesion (c) maintains the law of first increasing and then decreasing, and all of them reach the maximum value when the content is 2%. The angle of internal friction decreases significantly after the first freeze–thaw cycle and is less affected by the number of freeze–thaw cycles after that.

(2) The XRD and SEM analyses indicate that the main components of RC in the Hohhot area are quartz and clay minerals. At a content level of 1%, the BFs are more discrete. At 2% content, the BFs had an optimal gripping effect with the soil particles, forming a simple fiber network structure. After the freeze–thaw cycle, a large number of cracks appear inside the RC, the pores increase and become more numerous, and the contact mode of the soil particles is dominated by point-surface contact. At 3% content, a clustering phenomenon will be produced for BFs, from stagger overlap into a parallel arrangement, and the gripping effect is reduced.

(3) The thermal conductivity of BF-improved RC increases gradually with the increase of moisture content, with an average increase of 10.78%, and decreases and then increases with the increase of BF content, with a flatter curve. With the increase of moisture and BF content, the thermal effusivity increases and then decreases, and the heat storage capacity reaches a relatively optimal level with the moisture and BF content at 22.5% and 1%, respectively, the thermal diffusivity shows a pattern of decreasing and then increasing. BF-improved RC significantly improves the heat storage and insulation capacity of RC to some extent and reduces the degree of thermal diffusion of the soil.

(4) PFs are added to RC to compare with BFs, revealing that LFs have a weaker cohesive effect than BFs in improving RC at the same content of 2%. However, there is a significant increase in cohesion (c) when the LF content is 4–6%, where the best effect is reached when the content is 6%. The improvement effect of PF is significantly higher than that of BF. The effect of both fibers on the angle of internal friction of the RC is relatively small. Thermal conductivity and thermal effusivity have a tendency to increase and then decrease with the increase of LF content, at a content of 2%, reaching its maximum value. At the same content, the thermal conductivity of LF is significantly higher than that of BF. When the BF content is 1%, the heat storage capacity is significantly higher than LF. From economic, effect, and other perspectives, it can be concluded that PF improves the RC shear strength and thermal conductivity capacity effect better than BF, but BF improves RC thermal insulation and heat storage capacity more than LF.

The research results can be used as a reference for engineering in cold regions subjected to freeze–thaw cycles, improving the stability of backfill soils for buildings and roadbeds, construction of energy tunnels, selection of backfill materials for geothermal energy systems, as well as improving the thermal insulation capacity of soils, thus contributing to the concept of green and sustainable development around the globe.