Permeability Characteristics and Mechanism of Silicone-Hydrophobic-Powder-Modified Compacted Loess

Abstract

This paper proposes the addition of an environmentally friendly silicone hydrophobic powder (SHP) to compacted loess to enhance its impermeability and reduce the settlement caused by wetting deformation. A flexible wall permeability test was conducted to examine the permeability coefficient of compacted loess with different dry densities mixed with different proportions of the SHP. Water drop infiltration testing, low-temperature drying testing, low-field nuclear magnetic resonance testing, and scanning electron microscopy were performed to clarify the occurrence state and permeability reduction mechanism of the SHP between loess particles from the microscopic viewpoint of the loess structure. The permeability coefficient of the compacted loess decreases significantly after adding 1–3% SHP. Specifically, the permeability coefficient of the loess with low compaction degree (1.35 g/cm³) decreases by 87% after the addition of 3% SHP. Moreover, the permeability coefficient of the loess with a compaction degree of 1.50 g/cm³ decreases from 1.58 × 10⁻⁶ to 7.9 × 10⁻⁸ cm/s. When the SHP is added to compacted loess, filling and adsorption effects occur: The filling effect decreases the proportion of large pores in the loess. The adsorption effect allows the SHP to encapsulate the loess particles to form a hydrophobic film layer, which leads to the formation of a non-closed hydrophobic structure. This phenomenon decreases the water flow capacity while ensuring the water evaporation capacity inside the loess. The findings demonstrate the potential of adding SHP to enhance the loess impermeability and facilitate and optimize the anti-seepage treatment of various loess filling projects.

1. Introduction

In loess filling projects, post-construction settlement and potential wetting deformation risk are typically alleviated through the remodelling and compaction of the filler. At present, large-thickness loess filling projects, such as the first phase of Yan’an New Area and Lanzhou New Area in China, are entering the operation stage. Relevant monitoring and experimental research has shown that the loess filling body with large filling thickness, complex original terrain, and local hydrogeological conditions is expected to affect the filling site in the long-term operation stage [1,2,3]. Owing to surface water infiltration and groundwater field rebalancing, uneven settlement of the fill site, local cracking, and collapse of excavation and fill joint frequently occur [4,5,6]. Considering the creep characteristics of filling loess [7] and the humidification deformation mechanism [8], the improvement of loess strength and permeability has emerged as a promising method for controlling settlement development and alleviating the risks in filling engineering. Many researchers have used microbial reinforcement technologies [9] and introduced lime, fly ash, and other additives to improve the compressive strength [10], shear strength, impermeability of loess [11], and the loess surface electrochemical properties [12]. Notably, as a porous medium, the permeability of loess is closely related to its pore structure and distribution [13]. Therefore, the optimization of the internal structure of loess by introducing additives to reduce the moisture migration in the loess and, thus, the humidification deformation of loess has attracted significant research attention. Hu et al. [14] obtained the loess water characteristic curve of lime-modified loess and predicted the unsaturated seepage coefficient. The unsaturated seepage coefficient of the loess with lime was smaller than that of compacted loess without lime. Zheng [15] performed an infiltration test to determine the optimal content of lime and fly ash for improving the impermeability of loess. Gao et al. [16] and Zhang et al. [17] reported that the impermeability of fly ash- and lime-improved loess is attributable to the hydration, filling, and adsorption of free water of fly ash and lime.

In large-thickness loess filling projects, the exhaust conditions in the consolidation of unsaturated compacted loess directly affect its settlement development [18]. The improvement of remolded loess, such as by the addition of lime, cement, and fly ash, decreases the water circulation ability by blocking the connected pores of loess. This phenomenon changes the basic physical and mechanical properties of loess, with the effect being similar to that of curing, and the corresponding environmental impacts are difficult to estimate. Furthermore, the loess permeability is lowered and the air and water flow velocity in the large-thickness fill body is reduced, which increases the complexity of the post-construction deformation problem of the large-thickness loess fill body in the unsaturated state. To address these problems, some scholars have adopted new materials, such as nano-clay [19,20,21], silica [22,23], and hydrophobic curing agents [24], to reduce the permeability of loess while maintaining the permeability of its pore structure.

As a green hydrophobic material that can enhance the impermeability of materials, silicone hydrophobic agents have been successfully applied to improve cement mortar, concrete, and pavement materials. These agents can effectively reduce the capillary water absorption capacity of concrete and improve the waterproof and anti-permeability performance [25,26]. When silicone hydrophobic materials are coated on the surface of concrete specimens as a protective agent, the nature of the specimen surface changes from hydrophilic to hydrophobic [27] and the penetration depth is significantly reduced [28]. Notably, the existing research on the application of silicone hydrophobic materials for loess improvement is limited. Xu [29] applied silicone hydrophobic agents to the surface loess of the central green belt of an expressway and reported the satisfactory ground seepage control effect. Zhao et al. [30] applied silicone materials for loess slope protection. Choi et al. [31] and Haquie et al. [32] used hydrophobic materials, such as organic silane, to improve the permeability of kaolin. The authors reported that organic silane changed the particle size distribution, enhanced the adhesion between particles, and reduced the water absorption.

In general, silicone materials have been widely applied for concrete protection. However, only a few researchers have focused on the permeability characteristics of compacted loess modified using silicone hydrophobic agents. To examine the feasibility of using silicone hydrophobic materials for improving loess properties, the corresponding effect and mechanism must be understood.

Therefore, in this study, the permeability coefficient of loess with different dry densities and organic silicone contents was measured through flexible wall permeability tests. These tests quantitatively analyze the improvement effect of hydrophobic materials on loess impermeability, determine the optimal ratio of hydrophobic materials, and reveal the improvement mechanism from the perspective of macroscopic surface contact characteristics and microstructure changes. The findings can provide guidance for the anti-seepage design and risk control of various loess filling projects. These aspects can help alleviate the engineering risks associated with water erosion or internal water field changes in the loess fill site.

2. Materials and Methods

2.1. Materials

The test loess material was extracted from the filled remolded loess at a depth of 3 m in the filling body in Yan’an New Area. The loess sample was crushed and screened and the required water content was configured. The loess sample was then placed in a sealed bag that was in turn placed in a plastic storage box and sealed. The box was allowed to stand for 24 h to ensure uniform water distribution in the loess. The water content in the loess sample was then determined. The initial water content was measured. Moreover, the basic physical and mechanical indexes of the remolded loess were measured through an indoor geotechnical test. The basic parameters and compaction curves of the samples are presented in

Table 1. Basic physical properties of test loess.

Moisture Content of Specimens (%)	Specific Gravity	Optimum Moisture Content (%)	Max. Dry Density (g/cm3)	Liquid Limit (%)	Plasticity Index
16.5	2.7	16.9	1.71	34	14.5

and Figure 1, respectively.

The silicone hydrophobic material used in the test was SHP60+ (Silicone Hydrophobic Powder 60+) produced by Dow Corning, Midland, MI, USA, hereinafter referred to as SHP. SHP is an odourless, non-toxic, non-polluting white powder with a free-flowing particle size of 0.001–0.1 mm (Figure 2). The main component is siloxane resin. Compared with other modified materials, SHP can be easily mixed loess particle materials. It has a high dispersibility and high storage stability owing to the presence of active ingredients. Samples with dry densities of 1.35, 1.50, and 1.65 g/cm³ were designed and the organic silicone contents were 0%, 1%, 2%, and 3%.

2.2. Methods

2.2.1. Flexible Wall Seepage Test

When the permeability coefficient of a porous material is less than 10⁻⁵ cm/s, side wall leakage of the rigid wall permeameter is expected to directly affect the accuracy of permeability coefficient measurement. Therefore, we adopted two methods of flexible wall infiltration and rigid wall infiltration to test the permeability coefficient of organosilicon-modified loess to increase the authenticity of the experimental data. As mentioned previously, remolded loess samples with dry densities of 1.35, 1.5, and 1.65, organic silicone content of 0%, 1%, 2%, and 3%, and initial water content of 16.5% were prepared. Six samples were set in each group for synchronous testing. Three of the samples were used to determine the permeability coefficient under confining pressure through flexible wall permeability testing (Figure 3). The samples were saturated by back pressure after vacuum saturation for 48 h and a confining pressure of 200 kPa was applied for consolidation. The process of back pressure saturation and consolidation was continued for 10 d to ensure that the samples were completely saturated and reached a stable state under the consolidation confining pressure [33]. Infiltration was then allowed to occur over 100 h under the confining pressure of 200 kPa and data were collected every 1 h. The permeability coefficient was calculated every 10 h according to the Geotechnical Test Procedure (SL237-1999) [34]. The other three samples were subjected to parallel variable head permeability testing as a reference.

2.2.2. Water Drop Infiltration Test

Compacted loess samples with a diameter and height of 7.24 cm and 2 cm, respectively, were prepared. From a height of 5 mm from the surface of the loess samples, distilled water (0.05 mL) was dropped onto the samples by a standard burette. The water droplets exhibited a reflective phenomenon on the surface of the loess sample. The time at which this reflective phenomenon disappeared was considered to be the time at which the water droplets completely infiltrated the loess sample. Ten groups of parallel tests were conducted and the average water droplet infiltration time was used as the result. The water repellence was classified using the boundary value classification method [35], as indicated in

Table 2. Water repellence grade of loess.

Time (s)	Water Repellence Grade
<5	Hydrophilic
5–60	Slight water repellence
60–600	Strong water repellence
600–3600	Severe water repellence
>3600	Super water repellence

.

2.2.3. Water Evaporation Capacity Test of Loess

An intelligent artificial climate simulation box was used to simulate the evaporation capacity of the pore water of modified loess improved by organic silicone in Yan’an area. Firstly, the loess samples with dry density of 1.65 g/cm³ and different silicone contents were filled in a PVC pipe with a diameter of 6 cm and height of 10 cm. After compaction and saturation, the bottom and side walls were wrapped with thermal insulation materials, such that only the top surface could exchange water with the external environment. The samples were placed on an electronic scale and this framework was placed in the intelligent artificial climate simulation box. According to the actual meteorological conditions in Yan’an, the water evaporation test was conducted for 15 h under the local annual average temperature, hottest monthly average temperature, and extreme maximum temperature (10, 23, and 40 °C, respectively). Moreover, considering the test results of Zhang et al. [11] and Zheng [15], the optimal dosages of lime and fly ash for improving the permeability of loess were used to prepare samples, while water evaporation tests were conducted in the same conditions to compare the performances of the loess modified by silicone and traditional materials. Electronic scale data were recorded every hour and the water content was expressed in terms of the change in the loess sample quality. Each sample was subjected to three groups of parallel tests and the average value was used as the water content corresponding to the loess sample at a certain time.

2.2.4. Nuclear Magnetic Resonance Test

The NMRC12-010V nuclear magnetic resonance pore analysis system was used to conduct low-field nuclear magnetic resonance tests after the vacuum saturation of modified loess with different dry densities and different silicone contents subjected to saturated permeability testing. The relaxation time distribution curve was obtained by inversion of the test results; the pore structure and size distribution characteristics of the silicone-modified loess were then studied.

2.2.5. Scanning Electron Microscopy

A Phenom desktop scanning electron microscope was used to obtain images of the modified loess subjected to saturated infiltration. The samples were cut into 10 mm $\times$ 10 mm $\times$ 10 mm bits, air-dried, and broken to generate a fresh surface. The loose loess debris on the surface of the loess sample was blown off. A conductive adhesive was applied to the surface and sprayed with gold. Subsequently, scanning electron microscopy was performed to evaluate the microstructure and pore morphology of the modified loess with different dry densities and different silicone contents, as well as to explore the meso-mechanism of silicone hydrophobic materials in remolded compacted loess.

3. Results

3.1. Analysis of Permeability Characteristics of Improved Loess with Different Dry Densities

Figure 4 shows the variation in the permeability coefficient of modified loess with different silicone content. The dry density considerably affects the permeability coefficient of modified loess. Specifically, the permeability coefficient of the modified loess with the same organic silicone content decreases with the increase in the dry density. When the dry density is 1.50 g/cm³ (compaction degree 87.72%), the permeability coefficients are 49.23%, 20.69%, 10.83%, and 19.71% for those values at 1.35 g/cm³ (compaction degree 78.94%). When the dry density is 1.65 g/cm³ (compaction degree 96.49%), the permeability coefficients are 12.29%, 7.89%, 2.27%, and 4.82% for those values at 1.35 g/cm³. In other words, for a given organic silicone content, when the compaction degree of the modified loess is 87.72%, the permeability coefficient of the modified loess decreases by more than 50% compared with that at the compaction degree of 78.94%. When the compaction degree reaches 96.49%, the reduction in the permeability coefficient is more than 88%. These findings demonstrate that increasing the compaction degree of the improved loess can decrease its permeability coefficient at a given silicone content.

3.2. Analysis of Permeability Characteristics of Improved Loess with Different Organic Silicone Contents

Figure 5 shows the permeability coefficients of compacted loess with different proportions of silicone and obtained through flexible wall permeability testing and rigid wall permeability testing. The addition of silicone hydrophobic material significantly reduces the saturated permeability of the compacted loess. Compared with the test results of the flexible wall permeability test, the permeability coefficient of the organosilicon modified loess measured by the rigid wall variable head permeability test is larger. In fact, the rigid wall sidewall is rigid and its sidewall is prone to leakage. The side wall of the flexible wall is flexible, which can fit the sample well and can effectively avoid the problem of side wall leakage. Therefore, compared with the rigid wall permeability test, the flexible wall permeability test can ensure the accuracy and validity of the sample permeability coefficient. Focusing on the results of the flexible wall permeability test, the permeability coefficients of remolded compacted loess without SHP and with SHP were compared, as shown in

Table 3. Permeability coefficients of improved loess with different SHP contents.

Dry Density (g/cm3)	Mixing Amount (%)	Flexible Wall Penetration (%)	Rigid Wall Penetration (%)
1.35	0	/	/
	1	45	43
	2	60	56
	3	87	85
1.50	0	/	/
	1	77	69
	2	91	88
	3	95	92
1.65	0	/	/
	1	65	59
	2	93	89
	3	95	93

.

As the amount of SHP increases, the permeability coefficient decreases exponentially. In the low compaction condition (1.35 g/cm³), the permeability coefficient decreases by 60% and 87% after adding 2% and 3% silicone, respectively. When the dry density is 1.50 g/cm³, the permeability coefficient decreases from 1.58 × 10⁻⁶ to 7.9 × 10⁻⁸ cm/s when the SHP content is 2%. In other words, the permeability coefficient does not significantly decrease with further increase in the SHP content. After adding 3% SHP, the permeability coefficient decreases to 5% of that of the non-modified loess. In comparison to the conventional variable head test results, the decrease in the permeability coefficient of the silicone-modified loess in the flexible wall permeability test is 2 to 8% higher. This phenomenon is attributable to the influence of the in situ stress simulated by the confining pressure in the flexible wall test. Compared with those in the conventional variable head test, the loess pores are compressed to a certain extent under the confining pressure. The permeability test results demonstrate that the optimal addition ratio of SHP depends on the dry densities. When the dry density is 1.50 g/cm³ and 1.65 g/cm³, the dosage of 2% yields the best results.

3.3. Analysis of Water Inflow Capacity

The infiltration time of water droplets can reflect the water-intake capacity of the surface layer of the loess sample, while the water repellence of the loess sample can be characterized by the time at which the water droplets completely infiltrate the loess sample. More time required by the droplets to completely penetrate the loess sample corresponds to a stronger water repellence. Figure 6a,b shows the infiltration state of the silicone-modified loess with a dry density of 1.65 g/cm³ and organic silicone contents of 0% and 3% after 60 s. With the incorporation of 3% silicone, the water droplets exhibit obvious reflection phenomenon when they contact the surface of the loess sample after 60 s. In contrast, the water droplets on the surface of the pure loess sample are completely infiltrated.

Figure 6c shows the water droplet infiltration time curve of loess with different organic silicone contents (dry density 1.65 g/cm³ and water content 16.5%). This figure, along with the data presented in Table 2, shows the water droplet infiltration time of plain loess is less than 5 s, corresponding to hydrophilic behavior. When organic silicone is added to the loess, the infiltration time of the water droplets increases to more than 60 s, corresponding to strong water repellence. The water infiltration time of modified loess increases with the increase in the organic silicone content. The complete infiltration time of modified loess with 1% organic silicone content increases by 105 s compared with that of plain loess. The corresponding improvements for 2% and 3% organic silicone contents are 138 s and 153 s, respectively, indicating the improved ability of compacted loess for resisting water infiltration.

3.4. Analysis of Water Vapor Evaporation Capacity

The results of water vapor evaporation test and slope of evaporation curve of modified loess with different dosages at different temperatures are shown in Figure 7 and

Table 4. Slopes of water content curve of different types of improved loess at different temperatures.

Atmospheric Temperature (°C)	Mixing Amount (%)				Fly Ash	Lime
	0	1	2	3
10	−0.14	−0.17	−0.19	−0.23	−0.09	−0.08
23	−1.11	−1.15	−1.19	−1.19	−0.32	−0.17
40	−1.61	−1.62	−1.62	−1.60	−1.17	−0.79

. The results of traditional loess permeability improvement materials, such as lime and fly ash, are provided as references.

The atmospheric temperature considerably affects the evaporation rate of water vapor in loess. A higher atmospheric temperature is associated with a faster decrease in the loess moisture content, that is, a higher evaporation rate of loess water vapor. At a given temperature, the slope of the water content curve of the modified loess is consistent with that of the non-added loess, while the water content curve of the fly ash- and lime-modified loess exhibits a more gradual curve. This finding shows that although organic silicone improves the loess permeability, it exerts less influence on the water vapor evaporation capacity of loess compared with lime, fly ash, and other improvement methods; the water vapor discharge capacity inside the loess is the same as that in the case without any modification.

3.5. Analysis of Nuclear Magnetic Resonance Results of Modified Loess

According to the relaxation mechanism of low-field nuclear magnetic resonance, the relationship between the relaxation time of loess pore fluid and the pore structure [36] can be expressed as follows:

$$\frac{1}{T_2}={\rho }_2{\left(\frac{S}{V}\right)}_{proe}$$

(1)

$$r=9.0T_2$$

(2)

where T₂ is the relaxation time, ${\rho }_2$ is the relaxation strength (transverse relaxation rate), S/V is the ratio of the pore surface area to volume in the loess sample, and $r$ is the pore diameter. The low-field nuclear magnetic resonance system was used to test the silicone-compacted loess samples. Figure 8 shows the results of loess samples with different dry densities and silicone contents.

According to the principle of low-field nuclear magnetic resonance, the change in the loess pore volume can be reflected by the change in the T₂ spectrum. Considering the structural characteristics of the sample and shape characteristics of the pore distribution, the relaxation strength in Equation (1) was set as 3.0 μm/ms, ${\left(S/V\right)}_{pore}=r/3$ [36]. The correlation between the relaxation time and porosity of the sample, Equation (2), was obtained by simplifying Equation (1). According to the existing classification criteria and methods of loess pore size analysis [36], the pore size of the sample was divided into four categories: small pores [0, 1), medium pores [1, 20], large pores (20, 1000], and ultra-large pores (1000, 6000] (unit: μm). The relaxation time distribution of the T₂ spectrum was transformed into the pore size distribution of the sample [37]. Combined with the spectral area of different relaxation time values, the percentage content of each pore size distribution was calculated. The corresponding histogram and pore distribution are shown in Figure 8 and

Table 5. Pore distribution of silicone-modified loess.

Dry Density (g/cm3)	Mixing Amount (%)	Pore Proportion (%)
		Small Pore [0, 1)	Medium Pore [1, 20]	Large Pore (20, 1000]	Extra-Large Pore (1000, 6000]
1.35	0	0.01	52.15	47.13	0.59
	1	0.07	55.54	43.80	0.48
	2	0.10	60.36	39.07	0.38
	3	0.38	65.47	33.89	0.25
1.50	0	0.49	61.38	37.39	0.28
	1	0.52	63.13	36.22	0.12
	2	0.61	64.90	34.46	0.03
	3	1.12	65.99	32.79	0.005
1.65	0	1.13	76.52	22.19	0.12
	1	1.38	77.51	21.07	0.02
	2	1.39	79.25	19.26	0.001
	3	1.42	80.18	18.38	0.0005

, respectively.

The T₂ spectrum of the silicone-modified loess exhibits one main peak and two secondary peaks. The main peak area occupies a large part of the T₂ spectrum area. The amplitude of the main peak signal increases with the increase in the dry density, whereas the amplitude of the secondary peaks Ⅰ and Ⅱ decreases. At a given dry density, with the increase in the silicone content, the main peak signal amplitude gradually increases and the amplitude of the secondary peaks Ⅰ and Ⅱ gradually decrease. Most of the pores in the loess are medium and large. At a given dry density, with the increase in the silicone content, the proportion of small and medium pores increases and that of large and super large pores decreases. For example, consider samples with dry densities of 1.35 g/cm³ and 1.65 g/cm³: When the content of organic silicone is increased to 3%, the proportion of large pores decreases by 5.18% and 0.88%, respectively, whereas the proportion of medium pores increases by 5.11% and 0.93%, respectively. After the incorporation of silicone hydrophobic materials, some of the large pores in the loess are converted to medium pores. Compared with the addition of lime, silicone hydrophobic materials achieve similar or superior anti-seepage effects. Moreover, the reduction in the large pores is significantly smaller with the SHP, which is also the main reason why the water vapor evaporation capacity of the loess is not significantly affected, as highlighted in Section 3.4.

3.6. Analysis of Scanning Electron Microscopy Results of Modified Loess

Based on the analysis of the pore characteristics by nuclear magnetic resonance, four analysis points were defined for the samples subjected to the flexible wall permeability test, while the action state of the hydrophobic materials on the loess particles and pores was revealed by scanning electron microscopy. Owing to space limitations, the 300× and 5000× magnified images for only the samples with dry densities of 1.35 g/cm³ and 1.65 g/cm³ are presented in Figure 9 and Figure 10.

According to the 5000× magnified images, the SHP exerts simultaneous filling and adsorption effects.

Table 6. Pore analysis of improved loess with different SHP contents.

Dry Density (g/cm3)	Mixing Amount (%)	Pore Proportion (%)	Average Pore Length (μm)	Average Pore Width (μm)
1.35	0	38.16	22.14	12.57
	1	37.54	21.54	11.63
	2	37.18	20.68	10.85
	3	36.82	20.21	10.12
1.50	0	35.39	21.15	10.88
	1	34.67	20.67	9.97
	2	33.25	19.48	9.16
	3	33.87	18.46	8.87
1.65	0	31.27	20.45	10.71
	1	30.75	18.94	9.52
	2	30.49	16.62	8.35
	3	30.18	16.31	8.11

summarizes the results of the pore analysis based on the 300× images of the loess with different organic silicone contents. At a given dry density, when the organic silicone content increases, the average length and width of the loess pores decrease. A large amount of organic silicone is adsorbed on the surface of the loess particles and the pores between the loess particles are not cemented and filled. Specifically, at a constant dry density (compaction degree), the addition of silicone hydrophobic material to loess changes only the pore size of pores to a certain extent; it does not considerably change the original pore structure of loess, which is consistent with the results of the nuclear magnetic analysis.

4. Discussion

The results of the flexible wall permeability test, water droplet infiltration test, water evaporation test, low-field nuclear magnetic resonance test, and scanning electron microscopy test of silicone-modified loess were analyzed to clarify the mechanism through which the silicone hydrophobic material improves the permeability of compacted loess, considering the improvement in the internal microstructure of loess. After the incorporation of silicone hydrophobic materials into loess, adsorption and interstitial filling occur, slightly reducing the proportion of large and medium pores and resulting in the formation of a non-closed structure. The silicone active group combines with the Si-O(OH) group on the loess surface through hydrogen or chemical bonding to form a water repellent film. This phenomenon increases the contact angle between the water and loess particles and significantly decreases the permeability coefficient. Unlike the traditional methods of improving the permeability of loess by cementation and filling pores, the air and water vapor channels inside the loess are not significantly affected and, thus, permeability inside the loess is maintained. The micro-effect diagram is shown in Figure 11.

Most of the loess fills lie in the unsaturated zone above the groundwater level, while their engineering settlement development and site stability are significantly affected by the change in the internal water field. Ground collapse and uneven settlement problems are typically encountered in weak positions, such as the excavation and filling area of water erosion, new and old interfaces, and interfaces with different compactness [38]. After the large thickness loess fill changes the original gully runoff, falling spring, and other water discharge channels, water evaporation becomes an important path for balancing the water field inside the fill. Results of field monitoring show that post-construction settlement development is directly affected by the exhaust conditions and capacity [18]. In this context, it is necessary to control the water infiltration at the interface of ground, excavation, and filling through anti-seepage technologies to avoid the groundwater field uplift. Furthermore, it is necessary to maintain the water evaporation capacity to ensure the balance of the water field inside the large thickness filling body. The incorporation of optimal amounts of silicone hydrophobic materials into excavation-filling joints and other positions susceptible to water erosion can effectively improve the impermeability of loess while ensuring the evaporation capacity of water vapor inside the fill. These aspects can help alleviate the engineering risks associated with water erosion or internal water field changes in the loess fill site.

5. Conclusions

Silicone hydrophobic materials were added to compacted loess to improve its impermeability and the surface contact characteristics and internal microscopic improvement mechanism of the loess were comprehensively investigated. The following conclusions were derived:

(1)

The addition of 1–3% silicone hydrophobic material to remolded loess results in a significant anti-permeability effect. The permeability coefficient of modified loess decreases exponentially with the increase in the silicone content. When the remolded loess is compacted to a low dry density (1.35 g/cm³), the permeability coefficient decreases by approximately 87% after the addition of 3% SHP. When the dry density increases to 1.50 and 1.65 g/cm³, the permeability coefficient of compacted loess with 3% content decreases to approximately 5% of that without hydrophobic material.

(2)

With the addition of SHP, the large pores in part of the loess are converted to small and medium pores without cementation filling. The silicone particles are adsorbed on the surface of the loess skeleton to form a hydrophobic film. This phenomenon enhances the water repellence of the loess particles and generates a non-closed hydrophobic structure, thereby effectively reducing the saturated permeability without significantly affecting the water evaporation capacity.

(3)

After the silicone hydrophobic material is incorporated into the loess, several silicone particles adhere to the surface of the loess skeleton, thereby forming a hydrophobic film and improving the water repellence of the loess particles. Consequently, the surface of the silicone-modified loess is transformed from hydrophilic to strongly water repellent.

(4)

By optimizing the additional contents of silicone hydrophobic materials into remolded loess for compaction and filling, the permeability coefficient of compacted loess can be effectively reduced without increasing the degree of compaction, while the ability of water vapor evaporation inside the loess can be guaranteed. The findings can help promote post-construction water field control and anti-seepage treatment of weak parts in loess filling projects.