Study on the compaction and dynamic properties of loess enhanced by waste tyre rubber particles

2.1. Test program

The experiment considers four types of rubber particle sizes ($S$), namely 10 mesh, 20 mesh, 40 mesh, and 100 mesh, and takes into account four rubber particle content ($C$) namely 5 %, 10 %, 15 %, and 20 %, with a volume ratio used for the content. Rubber particles are mixed into loess by external mixing method, the freeze-thaw specimen are subjected to dynamic triaxial tests at unsaturated moisture content to determine the dynamic shear modulus ($G$) and damping ratio ($\lambda$).

The test is divided into three stages, the first stage is determination of mix proportion moisture content of rubber particle-loess. According to Article 13 of the “Standard for Soil Test Methods” (GBT50123-2019), a light compaction test is conducted to determine the optimum moisture content and saturated moisture content. Take one-third of the difference between the saturated moisture content and the optimum moisture content as a gradient. The four moisture content values obtained by subtracting the gradient from the saturated moisture content for four consecutive times are used as the mix proportion moisture content (as shown in Table 1). And the density of rubber particle-loess is determined under the conditions of light compaction test. The second stage is the freeze-thaw cycle test. The temperature range of freeze-thaw is –20 ℃-20 ℃, and the number of freeze-thaw cycles ($FT$) is divided into five types: 0, 1, 3, 6, and 9. Each cycle lasts for 12 hours, including 6 hours of freezing and 6 hours of melting. The third stage is dynamic triaxial loading stage. At this stage, the confining pressure (${\sigma }_z$) is applied at 50 kPa, 100 kPa, and 200 kPa, and the axial loading is controlled by deviator stress. The deviator stress loading is divided into six levels: 3 kPa, 5 kPa, 10 kPa, 20 kPa, 40 kPa, and 240 kPa, with 10 vibrations per loading, a loading frequency of 1Hz, and sine wave control. In the test, the remolded loess is designed as a reference, and its mix proportion moisture content is 11.14 %, 12.20 %, 13.26 %, and 14.32 %, respectively.

It should be noted that the dynamic triaxial and freeze-thaw cycle test conditions in this study are carefully designed to simulate the environmental conditions that infrastructure in Inner Mongolia’s loess areas may encounter. The confining pressures applied in the dynamic triaxial tests (50 kPa, 100 kPa, 200 kPa) cover the stress range that shallow to deep foundations may bear, effectively simulating the stress state of foundation soil at different depths. The number of freeze-thaw cycles (0, 1, 3, 6, 9) is set based on the severe and large temperature difference climate in Inner Mongolia's winter. Each cycle lasting 12 hours (6 hours freezing, 6 hours thawing) aims to simulate the freeze-thaw effect caused by temperature changes in a day, thus studying the performance changes of rubber particle-loess under different freeze-thaw cycles.

Moreover, the selected rubber particle content range (5 %, 10 %, 15 %, 20 %) and size range (10 mesh, 20 mesh, 40 mesh, 100 mesh) are determined based on relevant literature. Comprehensive research findings indicate that the selection of rubber particle size and content in loess improvement should balance engineering requirements and economic benefits. A content range of 5 % to 15 % generally enhances the compactness and strength of loess [19], with 40-mesh rubber particles proving superior in improving soil structure, shear strength, and compactness [20]. Lower contents (5 %-10 %) can significantly reduce the usage of other agents and enhance soil properties, offering economic viability [14]. However, contents exceeding 20 % can lead to increased porosity, reduced compactness, and strength due to the formation of a separate rubber particle framework. Therefore, controlling the rubber particle content within 5 % to 20 % optimizes performance, avoids degradation, and ensures environmental sustainability and cost-effectiveness [21]. The selected rubber particle sizes of 10 mesh, 20 mesh, 40 mesh, and 100 mesh aim to comprehensively evaluate the effects on loess properties. Notably, 40 mesh has demonstrated superior performance in previous studies, while the others span a spectrum from coarse to fine particles. Therefore, the selected rubber particle content and size range in this study have a practical significance, aiming to provide scientific and reasonable parameter guidance for the improvement of loess in Inner Mongolia using waste tire rubber particles.

2.2. Test material

The rubber particles are taken from the finished rubber particles of waste tires produced by Jiubaoshi Powder Factory in Hebei Province, as shown in Fig. 1. The test loess is taken from the Helingeer area of Hohhot city, China, and the soil extraction site is shown in Fig. 2. The soil particle density of loess is 2.62, the plastic limit is 15.5%, the liquid limit at a cone depth of 10 mm is 24.7 %, the liquid index is 9.2, and the natural moisture content is 7.25 %. It is determined that the soil is silt. The grading curves of loess and rubber particles are shown in Fig. 3.

Fig. 1Rubber particles

a) 10mesh

b) 20 mesh

c) 40 mesh

d) 100 mesh

Fig. 2Sampling point layout of loess

Fig. 3Grading curve

2.3. Test equipment

The Dynamic Triaxial Testing System (DYNTTS) is developed and manufactured by GDS in the UK, as shown in Fig. 4(a). Its maximum axial dynamic loading is 6 kN, accuracy is 1 kPa, axial displacement stroke is ±50 mm, axial displacement measurement resolution is 0.08 μm, axial displacement measurement accuracy is 0.07 %, axial force measurement accuracy is greater than 0.1 % FRO, and the confining pressure control subsystem is completed by the DYNTTS digital pressure/volume (200 cc/2 MPa) controller, with a resolution of up to 3 kPa. The maximum frequency of dynamic loading is 5 Hz, with sine wave dynamic control, and unconsolidated undrained shear test is adopted.

The freeze-thaw cycle test equipment utilizes the LRHS-225D high-low temperature alternating and humidity test chamber, as shown in Fig. 4(b), with a working temperature range of –70-150 ℃ and a humidity range of 0-10 0%.

2.4. Specimen preparation

The air-dried loess is sieved through a 2 mm mesh and mixed evenly with rubber particles at the designed content, as shown in Fig. 5(a), and is mixed with water using mix proportion moisture content and allowed to cure for 24 hours. By using quality control method, the soil mixture is loaded into the triaxial compaction mold in three times and compacted. When adding in stages, chiseling treatment is required. After compaction, the specimen has a diameter of 50 mm and a height of 100 mm, as shown in Fig. 5(b). The molded specimen is wrapped in a fresh-keeping bag and subjected to freeze-thaw cycles according to the set number of freeze-thaw cycles. Post freeze-thaw treatment, the specimen (as shown in Fig. 5(c)) can be used for dynamic triaxial tests.

Fig. 4Test equipment

a) DYNTTS

b) Freeze-thaw meter

Fig. 5Triaxial specimen preparation

a) Loess mixed with rubber particles

b) Triaxial specimen

c) Post freeze-thaw specimen

3.1. Compactness analysis

When rubber particles of the same density are mixed into loess of the same dry density, the compactness of the specimen varies under the same compaction work but different moisture content conditions. The density reflects its compactness to a certain extent, and the higher the density, the better the compactness. This experiment used a light compaction test to determine the densities of loess specimens with different content and size of added rubber particles. Fig. 6(a-d) show relationship curve between density and moisture content of rubber particle-loess under different content conditions. According to the figure, the density of rubber particle-loess shows a tendency of increasing first and then decreasing with the increase of moisture content. Under different rubber particle content conditions, the maximum density of rubber particle-loess corresponds to different rubber particle sizes. The density of rubber particle-loess is maximum at 5 % rubber particle content and 40 mesh particle size. Therefore, from the perspective of compactness, the combination with a rubber particle content of 5 % and a particle size of 40 mesh has the best effect. In this state, rubber particle-loess is the most dense, with high strength and resistance to deformation.

The reason for this is that the addition of rubber particles first changes the grading curve of loess. This change is because rubber particles, as a new type of particle, are of different sizes from those of the loess particles (As shown in Fig. 1), thus forming a new combination of particles after mixing. When rubber particles are evenly dispersed in loess, they can effectively fill the gaps between soil particles. Especially the 40 mesh rubber particles, which are of moderate size, can not only fill the gaps between larger soil particles, but also form good interlocking with smaller soil particles. This interlocking effect creates a more tightly packed structure between particles of different sizes, resulting in a decrease in soil porosity and an increase in density.

Fig. 6The relationship curve between density and moisture content of rubber particle-loess under different rubber content conditions. (Note: Moisture content (Low, Optimum, Medium and High) correspond to mix proportion moisture content in Table 1)

a)$C=$5 %

b)$C=$10 %

c)$C=$15 %

d)$C=$20 %

3.2. Dynamic shear modulus analysis

Dynamic shear modulus is an indication of the ability of soil to resist dynamic shear deformation, and is also an important indicator for evaluating the dynamic shear deformation of soil. Fig. 7 shows the relationship curve between dynamic shear modulus and dynamic strain, according to the figure, the dynamic shear modulus decays with the increase of dynamic strain, similar to the variation of dynamic shear modulus of loess [22]. The dynamic stress and strain exhibit a strain softening behavior, which is distinct from the static properties of loess [23]. At small strains (less than 0.8 %-1 %), the dynamic shear modulus decays more rapidly than at large strains (exceeding 1 %), where its decay rate gradually slows down and eventually becomes flat. It indicates that the rubber particle-loess is initially relatively hard, but as the deformation increases, the soil structure gradually adjusts, and its stiffness subsequently decreases.

Fig. 7(a) shows the relationship between dynamic shear modulus and dynamic strain at different rubber particle content conditions. According to the figure, although the curves for 10 % and 15 % rubber particle content intersect to some extent, the overall position of the curve shifts downward as the rubber particle content increases. Under the same dynamic strain conditions, the specimen with 5 % rubber particle content exhibits the highest dynamic shear modulus. Fig. 7(b) shows the relationship between dynamic shear modulus and dynamic strain under different rubber particle sizes conditions. It indicates that under the same dynamic strain conditions, the dynamic shear modulus is highest for 40 mesh rubber particles, followed by 10 mesh, with an alternating situation between 10 mesh and 100 mesh. Fig. 7(c) shows the relationship between dynamic shear modulus and dynamic strain under different freeze-thaw cycles. It indicates that under the same dynamic strain conditions, as the number of freeze-thaw cycles increases, the dynamic shear modulus decreases. Fig. 7(d) shows the relationship between dynamic shear modulus and dynamic strain under different confining pressure conditions. It indicates that, under the same dynamic strain conditions, the greater the confining pressure, the greater the dynamic shear modulus, which conforms to the variation law of loess dynamic properties [24]. Adding rubber particles with a content of 5 % and a particle size of 40 mesh to loess results in a relatively high dynamic shear modulus. Therefore, a content of 5 % and a particle size of 40 mesh are the ideal content and particle size.

Fig. 7The relationship curve between dynamic shear modulus and dynamic strain

a) Different particle content (${\sigma }_z=$ 100 kPa, $S=$40 mesh, $FT$= 0)

b) Different particle sizes (${\sigma }_z=$100 kPa, $C=$5 %, $FT=$0)

c) Different number of freeze-thaw cycles (${\sigma }_z=$100 kPa, $C=$5 %, $S=$ 40 mesh)

d) Different confining pressure ($C=$5 %, $S=$ 40 mesh, $FT=$0)

3.3. Damping ratio analysis

The damping ratio, which characterizes the energy dissipation of soil under cyclic loading, significantly influences the vibration response of soil [25] and the ground vibration parameters of the site [26]. According to the dynamic triaxial test, the dynamic stress-dynamic strain hysteresis loop is drawn, as shown in Fig. 8. The axial damping ratio can be calculated according to Eq. (1), and then the relationship between the damping ratio of rubber particle-loess and moisture content, freeze-thaw cycle times, confining pressure, and original soil damping ratio can be determined:

where: $\lambda$ – damping ratio; $A_s$ – dynamic stress-dynamic strain hysteresis loop area (cm²), area of the shaded part in Fig. 8; $A_t$– area (cm²) of right triangle abc in Fig. 8.

Fig. 8Dynamic stress-dynamic strain hysteresis loop

3.3.2. Relationship between damping ratio and number of freeze-thaw cycles

Fig. 9(b) shows the relationship between damping ratio and the number of freeze-thaw cycles under different moisture content conditions. According to the figure, under the same moisture content conditions, the damping ratio increases with the number of freeze-thaw cycles. This indicates that freeze-thaw action enhances the dynamic load absorption capacity of rubber particle-loess, thereby improving its seismic performance. This is due to the freeze-thaw cycles that lead to the destruction of soil structure [27] and changes in pore water pressure. Freezing and thawing can cause water to freeze and thaw, alter the characteristics of pore water in soil, increase water fluidity, and promote changes in pore water. At the same time, freeze-thaw cycles can cause the rearrangement and loosening of soil particles, reducing the soil's internal friction angle and increasing damping. As the number of freeze-thaw cycles increases, the soil may experience more microscopic damage, which will also improve its damping performance. The damping ratio curves for the 3rd and 4th mix proportion moisture contents are significantly higher than those for the other two mix proportion, and thus are more affected by freeze-thaw cycles at higher moisture contents, where water plays a significant role in the changes.

Fig. 9The relationship curves between damping ratio and various factors. (Note: Moisture content (Low, Optimum, Medium and High) corresponds to mix proportion moisture content in Table 1)

a) Different moisture content (${\sigma }_z=$100 kPa)

b) Different number of freeze-thaw cycles (${\sigma }_z=$100 kPa)

c) Different confining pressure (Optimum moisture content)

d) Rubber particle-loess vs. loess ($FT=$0)

4.1. Analysis of microstructure changes under freeze-thaw action

The SEM images reveal the microstructural changes of rubber particle-loess with increasing freeze-thaw cycles (As shown in Fig. 10). As freeze-thaw cycles increases, the contact between rubber and loess particles becomes looser. At FT = 0, rubber and loess particles are tightly bonded in full contact, forming a dense structure with few pores. This tight contact enhances the dynamic shear modulus of rubber particle-loess, giving it high stiffness and strength. As freeze - thaw cycles increase (FT = 1, 6, and 9), the bond weakens, and contact types change from surface contact to partial contact and point contact. These changes lead to a looser structure and larger pores. At FT = 9, the contact is very loose, with many pores and mainly point contact. The looser structure reduces the contact area and friction between particles, lowering the dynamic shear modulus. Meanwhile, larger pores and looser particle arrangement increase energy dissipation, raising the damping ratio.

4.2. Analysis of microstructure changes under the influence of moisture content

The SEM images illustrate how the microstructure of rubber particle-loess changes with varying moisture content (As shown in Fig. 11). Under optimum moisture content conditions, rubber particles are tightly bonded with loess particles, forming a dense structure with minimal pores. Rubber particles are evenly distributed and effectively fill the gaps between loess particles (As shown in Fig. 11(b)). The dense microstructure results in a high dynamic shear modulus, while the tight structure limits particle displacement, reducing energy dissipation and leading to a low damping ratio. Under low moisture content conditions, the bond between rubber and loess particles weakens slightly, creating small pores and loosening. Rubber particles still provide some filling but less effectively than at optimum moisture content (As shown in Fig. 11(a)). The slight loosening reduces the dynamic shear modulus, while the appearance of small pores increases energy dissipation, causing the damping ratio to rise. At higher moisture content, the bond between rubber and loess particles significantly weakens, resulting in larger pores and more pronounced loosening. The filling effect of rubber particles diminishes, and separation may occur locally (As shown in Fig. 11(c) and Fig. 11(d)). The substantial loosening leads to a significant decrease in dynamic shear modulus, while the increased pore size enhances energy dissipation, resulting in a high damping ratio.

Fig. 10SEM images of rubber particle-loess under different freeze-thaw cycles: a) full contact, b) surface contact, c) partial contact, d) point contact

a) FT = 0

b) FT = 1

c) FT = 6

d) FT = 9

These microstructural changes provide insights into the mechanisms underlying the observed macroscopic trends in dynamic shear modulus and damping ratio with freeze-thaw action and varying moisture content.