Performance Study of Casing Piles in Expansive Soil Foundations: Model Testing and Analysis

Abstract

This study investigates the critical behavioral characteristics of pile foundations in expansive soil foundations through a series of model tests, including settlement, axial force, and side frictional resistance. The experiment initially utilized sand, bentonite, and gypsum as the fundamental materials for the preparation and composition research of expansive soil simulant materials. Subsequently, the performance of different types of model piles under various loads and water immersion conditions was analyzed. The results indicate that non-cased piles exhibit typical friction pile behavior, while PVC-cased and steel-cased piles effectively reduce side frictional resistance, resulting in a more uniform distribution of axial force along the pile. After immersion, the model test materials experience expansion, with a faster initial expansion rate and a more gradual later expansion rate. Different types of model piles exhibit different displacement characteristics, and the presence of an outer casing can reduce the uplift of the inner pile. Furthermore, PVC casing demonstrates better performance in handling negative frictional effects. This study provides valuable insights for the design and construction of pile foundations in expansive soil foundations.

1. Introduction

Expansive soil is a special type of clayey soil, whose main clay composition is hydrophilic minerals, so it has the characteristics of water-absorbing expansion and water-losing contraction [1,2]. The expansion and contraction cycle of this expanded soil foundation often leads to uneven settlement of buildings or structures, which, in turn, leads to cracking, tilting, and other destructive phenomena. Expansive soil, which is widely distributed across the globe, has been identified as one of the important challenges facing infrastructure and buildings [3]. Pile foundation is widely used in expansive soil foundation, but the expansion behavior of expansive soil will reduce the bearing capacity of pile foundation, resulting in uneven and uniform settlement of foundation [4,5]. Therefore, studying the influence of expansive soil properties on pile settlement and bearing capacity in expansive soil regions is of significant importance.

In recent years, extensive research has been conducted by scholars into the mechanical behavior of pile foundations in expansive soil regions, encompassing diverse aspects such as theoretical studies, model experiments, and numerical simulations [6,7,8]. Ter-Martirosyan and Ter-Martirosyan (2018) [9] introduced an analytical solution for determining the stress–strain characteristics of soil–pile interaction under conditions of non-uniform moisture content. Jiang et al. (2020) [4] presented a non-linear analysis method for the bearing capacity of single piles in expanded soil after immersion, employing the shear displacement approach. Yang et al. (2023) [10] investigated soil pressure and displacement variations on pile surfaces in expansive soil through repeated immersion, utilizing a large-scale soil centrifuge. Soundara and Robinson (2017) [11] identified through experiments that the interface shear stress between piles in expansive soil does not remain constant with depth during the swelling process. Mohamedzein et al. (1999) [12] established a two-dimensional axisymmetric finite element analysis model for pile–soil systems in expansive soil foundations and compared numerical results with field test outcomes. Wu and Vanapalli (2022) [13] utilized ABAQUS software to develop a novel water–soil coupled model for unsaturated expansive soil and a constitutive model for pile–soil interfaces, conducting three-dimensional finite element analysis of single piles during the infiltration process. The outcomes suggested that a reduction in matrix suction contributes to a decrease in pile frictional resistance, allowing for more efficient load transmission to the pile and subsequently resulting in increased pile head settlement and bottom resistance. However, there are limited reports on model tests concerning the load-bearing characteristics of piles in expansive soils. Therefore, it is necessary to analyze the influence of lateral friction on piles under submerged conditions through model testing [7].

To address the challenges posed by expansive soil, scholars have proposed various innovative foundation techniques, including the implementation of sand cushion layers, non-expansive adhesive layers, physical modifications, and chemical stabilization [14]. Phanikumar et al. (2008) [15] investigated the behavior of granular pile anchors in expansive soil through laboratory model tests and field studies, revealing that granular pile anchors effectively control uplift in expansive soil. Albusoda and Abbase (2017) [16] assessed the performance of single and multiple helical piles in expansive soil through model experiments and conducted a systematic examination of the effects of helix length, helix diameter, and quantity on pile performance. Despite numerous prior research efforts, the complexity of the mechanical properties of expansive soils has resulted in intricate load transfer characteristics within pile foundations. Therefore, it is necessary to conduct a comprehensive analysis of the variations in parameters such as axial load on the pile, lateral resistance, and negative skin friction before and after immersion. Furthermore, further research is needed to enhance the stability of pile foundations in expansive soil by exploring innovative pile designs.

Therefore, the objectives of this study are as follows: (1) Conducting a series of non-immersion and immersion tests on pile in expansive soil under controlled conditions; (2) investigating the axial force and side frictional resistance of cased steel piles under pile head loading and non-immersion conditions; (3) analyzing the expansion characteristics of expansive soil layers under immersion conditions and studying the bearing characteristics and displacement variations of pile with or without additional surface loading. This study aims to provide theoretical support for the application of cased piles in expansive soil foundations.

2. Experimental Setup

2.1. Preparation of Expansive Soil

In this study, our objective was to develop a similar model material for expansive soil. Initially, we conducted a comprehensive analysis of the physical and mechanical properties of sand, bentonite, and gypsum. Considering the distinctive characteristics of expansive soil, such as swelling–shrinkage behavior, fissuring, and superconsolidation, we opted for sand as the primary aggregate, bentonite as the expansive component, and gypsum as the binding material. The selected bentonite here is nanoscale montmorillonite, with a montmorillonite content of 93%. Sand in the experiment was sieved to a particle size of 0.5 mm for size control. The gypsum used in the experiment primarily consists of calcium sulfate, and upon exposure to water, it exhibits a cementitious effect and a certain degree of expansiveness.

Five different mixtures with various mass ratios were configured, with the mass ratios of sand, bentonite, and gypsum being 7:6:1, 6:6:2, 5:6:3, 5:7:2, and 4:7:3, respectively. Through a series of experiments and comparisons, the optimal mixture ratio was determined. A PVC pipe with a height of 30 cm and a diameter of 20 cm served as the experimental model bucket. Maintaining consistent moisture content but differing initial dry densities, the filling and compaction were conducted in three stages to reach the design height. After leveling the top surface, a thin glass plate was placed, and an instrument for measurement was installed, with a tray at the bottom. Once the experiment commenced, distilled water was gradually added to the bottom tray. Hourly recordings of the measuring instrument readings and the amount of water added were taken until the difference between consecutive hourly readings did not exceed 0.01 mm.

When the initial moisture content and dry density are 11% and 1.2 g/cm³, respectively, the relationship between the expansion strain rate of artificially prepared expansive soil and time is shown in Figure 1. After the mixture reaches a stable expansion, the mixture with a ratio of 4:7:3 exhibits the highest expansion strain rate, reaching 12.3%. It is worth noting that, in this proportion, the mass proportion of sand was only 28.57%. Considering the influence of the sand skeleton on the mixture’s strength, it was necessary to ensure sufficient strength in the mixture. Therefore, the mixture with an expansion strain rate of 10.5% and a ratio of 5:6:3 was selected as the expansive soil material for the model test. Through compaction experiments, the maximum dry density of the sand–bentonite–gypsum model test material was found to be 1.81 g/cm³, which is similar to the maximum dry density of moderately expansive soils in Hanyin County, Shaanxi Province. Comparing the physical mechanical properties of the obtained natural expansive soil with the artificially prepared expansive soil indicates that the specifications for the artificially prepared expansive soil are similar to those of natural expansive soil. Furthermore, the free swell test determined that the model test material has a free swell percentage ranging from 65% to 90%, classifying it as a moderately expansive soil. This result further confirms the rationality of material selection and proportion design in the preparation of artificial expansive soil.

2.2. Model Piles

As shown in Figure 2, the experiment comprised eight model piles, all constructed using steel pipes with a diameter of 108 mm. Piles numbered 1 to 7 had a length of 2160 mm, while pile number 8 had a length of 1440 mm, with specific details provided in

Table 1. Detailed information of model piles.

Number of Piles		1	2	3	4	5	6	7	8
Type		steel	steel	steel	steel	steel	steel	steel	steel
Diameter (mm)		108	108	108	108	108	108	108	108
Length (mm)		2160	2160	2160	2160	2160	2160	2160	1440
Casing	Type	/	PVC	steel	steel	/	PVC	steel	steel
	Length (mm)	/	2160	2160	2160	/	2160	2160	1440
	Diameter (mm)	/	158	158	158	/	158	158	158

. The casing materials included PVC pipes and steel pipes. With the exception of piles 1 and 5, all other piles were fitted with casings. The casings for piles 2, 3, 4, 6, and 7 had a length of 2160 mm, while pile 8 had a casing length of 1440 mm, with a casing diameter of 158 mm.

Strain gauges were affixed longitudinally along the model piles to measure axial forces. The first strain gauge was positioned on the side of the pile at the top, with strain gauges 1 and 2 spaced 6 cm apart, followed by additional strain gauges placed at 30 cm intervals. Each side of the model piles was equipped with a total of 9 strain gauges. Load sensors were evenly distributed at both the base and top of the piles to measure the loads borne by the piles. Additionally, four layers of settlement instruments, totaling 15 instruments, were deployed to monitor real-time soil settlement and establish the relationship between pile top loads and immersion time. The locations where these settlement instruments were placed can be seen in Figure 3.

2.3. Experimental Design and Steps

The experiment was conducted in the Geotechnical Laboratory at Lanzhou Jiaotong University. The dimensions of the experimental model tank were 3 m (length) × 2.6 m (width) × 2.6 m (height). To ensure the impermeability of the model tank and facilitate the immersion test, plastic films were placed on the bottom and sidewalls of the model tank as a waterproofing boundary treatment, as shown in Figure 4.

The main experimental steps are as follows:

(1)

In the model tank, a 44 mm-thick layer of gravel was first placed at the bottom as a load-bearing layer. An electric compaction machine was used to lightly compact the sand layer at intervals of 110 mm to achieve a density of 1.7 g/cm³. Prefabricated model piles were suspended in their designated positions by ropes. Limiting frames were installed at the bottom of the test tank to provide stable support for the piles. The verticality of each pile was ensured using the drop hammer suspension method. After the installation of the model piles was completed, a 2160 mm-thick layer of artificially prepared expansive soil was placed. The compaction degree of the expansive soil was controlled to 0.95 using the ring cutting method. To ensure uniform percolation of water without surface ponding, a 50 mm layer of crushed stone was placed on the surface of the expansive soil.

(2)

Slow loading maintenance was applied to piles 1–4. Pile 1 was directly loaded, while piles 2, 3, and 4 were loaded through their inner piles. A loading system consisting of a reaction beam and hydraulic jacks was used to gradually apply the experimental load. Pile 1 was loaded at 2 kN per stage, while piles 2, 3, and 4 were loaded at 1 kN per stage. For pile 8, a mechanical jack was set up to measure the upward expansion force exerted on the pile body by the expansive soil model material upon contact with water.

(3)

The immersion test was initiated when the pile top settlements reached an average of 4 mm, indicating that the piles were under normal load conditions. Settlement measurements at the pile top were taken at the 5th, 15th, 30th, 45th, and 60th minutes. Subsequently, settlement measurements were taken every 30 min. When the settlement rate at the pile top reached a relatively stable state (settling at less than 0.1 mm/h for two consecutive measurements), the next level of load is applied.

3. Results and Discussion

3.1. Pile Bearing Characteristics before Immersion

The axial force curves of model piles 1, 2, 3, and 4 were calculated based on the strain measurements, as shown in Figure 5. Under the action of the pile head load, the axial forces within the piles gradually decrease along the length of the piles. For without casing piles (pile 1), at the beginning of loading, the axial force distribution along the pile was relatively uniform. As the load increased, most of the axial force concentrated in the upper part of the pile, and the axial force transmitted to the pile base was relatively small, indicating the characteristics of a friction pile. In the case of cased piles (piles 2, 3, and 4), at lower loads (loads < 3 kN), the axial forces along the depth of the pile remained nearly constant, with minimal axial force decay. With the increase in load (loads > 3 kN), the soil exerted squeezing forces on the casing, further transferring these forces to the inner pile, resulting in a reduction in axial force at the support locations of the inner and outer casing. The ultimate bearing capacity of the uncased pile (pile 1) is 16 kN, while the ultimate bearing capacity of the cased piles (piles 2, 3, and 4) is 9 kN. The test results indicate that the uncased pile (pile 1) has the highest bearing capacity, exceeding that of the cased piles (piles 2, 3, and 4) by 7 kN. It is noteworthy that cased piles exhibited significantly lower reductions in axial force along the depth under different loadings compared to uncased piles. The bearing behavior of the inner piles resembled that of end-bearing piles, indicating that both PVC and steel casings effectively eliminated the side friction resistance from the surrounding soil.

3.2. Deformation of the Pile–Soil System after Immersion

3.2.1. Soil Expansion Characteristics

The model test material exhibits significant expansion when immersed in water, and during the soil expansion process, friction occurs at the interface between the soil and the pile, causing vertical displacement of the pile and subsequently altering the stress within the pile. Vertical displacements at the pile top and around the pile were measured using dial gauges, providing insights into the material’s expansion effects. Throughout the experiment, the soil was soaked for 30 days until the expansion in each layer stabilized.

Figure 6 illustrates the relationship between soil expansion and time at depths of 2.16 m, 1.62 m, 1.08 m, and 0.54 m below the surface. As time progressed, the soil’s expansion increased gradually. Closer to the surface, the soil exhibited greater expansion, with a maximum expansion of approximately 7.1 mm. During the initial stages of immersion, the soil expansion rate was relatively high. Within the first 5 days, the expansion rates at depths of 2.16 m, 1.62 m, 1.08 m, and 0.54 m were 0.318 mm/day, 0.34 mm/day, 0.752 mm/day, and 0.84 mm/day, respectively. However, from days 5 to 30, the expansion rates gradually slowed and stabilized. The decay rates of expansion at depths of 2.16 m, 1.62 m, 1.08 m, and 0.54 m were 0.071 mm/day, 0.081 mm/day, 0.094 mm/day, and 0.114 mm/day, respectively. Therefore, the rate of expansion in different soil layers within the first 5 days is more than four times the rate of expansion between 5 and 30 days. The expansion within the initial 5 days accounts for more than 45% of the total expansion occurring within 20 days.

3.2.2. Pile Head Displacement

The displacement of piles in expansive foundation soils is closely related to the expansion of the soil layers. In expansive soils, piles are subjected to an uplifting force caused by the expansion of the soil, resulting in significant uplift displacement. The pile head displacement curves for piles 1, 2, and 3 under the applied loads are shown in Figure 7. From Figure 7a, it can be observed that different types of model piles exhibit varying pile head displacement changes under the same load. Pile 1 (without casing) experienced the smallest variation in pile head displacement, while the displacement changes for pile 2 (steel casing) and pile 3 (PVC casing) were similar. After 30 days of soil immersion, the maximum change in pile head displacement was 8.3 mm, while the minimum was 1.9 mm. The pile head displacement change for cased piles was approximately four times that of uncased piles. This is primarily due to the absence of casing, which allows the soil to expand due to water absorption, generating upward side frictional resistance that reduces the force applied to the pile head, resulting in smaller pile head displacement changes. In contrast, when using casing piles, the casing can eliminate side frictional resistance from the surrounding soil, causing a greater force to be applied to the pile head and, therefore, larger pile head displacement.

The pile head displacement of cased piles 2 and 3 is illustrated in Figure 7b. Casings experience upward frictional resistance because they are in direct contact with the expansive soil layer. During the first 5 days of immersion, the soil layer expands rapidly, leading to a quick rise of the casing. On the fifth day, the steel casing of pile 3 exhibited an ascent of approximately 0.7 mm, while the PVC casing of pile 2 rose by about 2.5 mm. Over the subsequent 5 to 30 days, pile head displacement gradually stabilized. The final rise of the steel casing of pile 3 and the PVC casing of pile 2 reached 2.66 mm and 4.55 mm, respectively. The initial rise of the pile head is attributed to the rapid expansion of the surface soil layer during immersion. The mid-section sinking of the pile may be due to water entering the lower middle sand layer through the drainage sand well, causing the gap to be refilled with particles and leading to subsidence. In the later stages, as water gradually penetrates into the interior of the soil through the sand well channel, the lower soil expands, and the pile gradually rises to a stable position.

Figure 8 displays the displacement variation curves at the pile heads of piles 6, 7, and 8 when no load was applied. As immersion continued, the outer casings of all uncoupled piles experienced a certain degree of uplift. At 30 days, the PVC casing of pile 6 rose about 4.67 mm, and the steel casing of pile 7 rose about 2.9 mm. The rise of the PVC casing was closer to the average expansion of the soil layer, indicating that PVC casings can better handle the ground uplift caused by expansive soil. Additionally, the experiment compared the trend of inner and outer pile head displacements of pile 8. After the soil expansion in the submerged soil stabilized, the inner pile had an upward displacement of approximately 0.82 mm, while the outer casing rose by 3.10 mm. This indicates that the outer casing counteracted the displacement caused by soil absorption and expansion, thereby reducing the upward displacement of the inner pile.

3.3. Pile Bearing Characteristics after Immersion

3.3.1. Without Additional Surface Loading

Under immersion conditions, the axial force curves of piles 5, 6, and 7 without additional load are shown in Figure 9. When the water immersion began, the soil experienced expansion from top to bottom. Between days 1 and 20, there was a significant variation in the axial force of each pile, while between days 20 and 30, the axial force of each pile remained relatively stable. Taking uncased pile number 5 as an example, its peak axial force increased from 0.78 kN on the first day to 1.35 kN on the twentieth day, with a peak axial force increase rate of approximately 0.1 kN/day. However, from the twentieth day to the thirtieth day, the peak axial force of pile number 5 only increased by approximately 0.1 kN, with a peak axial force increase rate of 0.01 kN/day. The peak axial force increase rate decreased by about 10 times. This indicates that after 20 days of immersion, the soil expansion deformation is essentially completed, and the axial forces on the piles are also stabilizing. Furthermore, there was a certain pattern in the changes of axial force curves for each pile. Among them, the variations in axial force of pile number 5 and the outer casing piles of piles 6 and 7 exhibited a similar trend, showing an initial increase followed by a decrease along the length of the piles. The peak axial force moved downward along the length of the piles with an increase in immersion time. After 30 days of submersion, the peak axial forces for pile number 5, PVC casing pile (pile 6), and steel casing pile (pile 7) were 3.5 kN, 4.6 kN, and 3.5 kN, respectively.

The inner pile axial force curve for pile 8 is shown in Figure 10. Due to the casing’s isolating effect on frictional resistance, the distribution of axial force within Pile 8 is relatively uniform. On the first day of immersion, the axial force within the pile is essentially 0 kN, possibly because the lower soil layers beneath the pile have not yet expanded or have expanded only minimally. As the immersion time increases, the lower soil layers beneath pile 8 gradually begin to expand and deform, leading to an increase in the expansion force exerted on the pile base, resulting in a gradual increase in the axial force within the pile. After the soil expansion stabilizes, the measured axial force at the pile top is 2.13 kN. Because pile 8 is entirely located within the expansive soil layer, this force represents the expansion force exerted by the 720 mm thick expanded soil at the pile base. The stress magnitude of this force at the pile base is 108 kPa. This implies that under these model conditions, the model pile can only counteract the effect of the upward expansion force when the upper load exceeds 1.08 kPa.

After immersion, the soil expanded, and under the upward expansion of the soil, frictional resistance was generated on the piles. Figure 11 displays the side frictional resistance curves for piles 5, 6, and 7. For piles without additional loads, there exists both positive and negative side frictional resistance. This leads to the appearance of a neutral point, and the location of the neutral point is essentially the same for all model piles. When immersed for 30 days, the point where the side frictional resistance becomes 0 is found at 160 cm (0.74 pile length). Therefore, the upper portion of the pile experiences an uplift force at this point, while the lower portion of the pile serves as an anchor. Furthermore, the peak values of both positive and negative side frictional resistance increase with an increase in immersion time. The peak negative side frictional resistance for pile number 5 and PVC-casing pile (pile 6) is approximately −32.5 kPa, while the peak negative side frictional resistance for the steel-casing pile (pile 7) is only −25 kPa. This indicates that both types of casings are effective in counteracting potential negative frictional effects, but the PVC casing’s effect is more significant.

3.3.2. With Additional Surface Loading

Pile 1 was loaded to 16 kN and maintained a constant pile top load while undergoing a water immersion test. During the immersion process, measurements of the axial force of pile number 1 were taken. Figure 12 presents the axial force and side frictional resistance curves of pile number 1 during the immersion process. Under the application of a 16 kN top load, pile number 1 exhibited characteristics entirely distinct from those of unloaded upper sections. Under this top load, the side frictional resistance of pile number 1 was entirely positive frictional. In the submerged condition, the axial force within pile number 1 decreased rapidly with depth due to the effect of soil expansion. The side frictional resistance of the pile increased as the immersion time increased, reaching a maximum of 23 kPa. The maximum side frictional resistance occurred at a depth of 50 cm within the surface expanded soil layer.

4. Conclusions

In this study, a series of model experiments were conducted using expansive soil model materials to analyze the expansion deformation of soil layers and the displacement, axial force, and lateral friction resistance of piles in expansive soil foundations. The main conclusions are as follows:

(1)

Before immersion, uncased piles under loading exhibited typical friction pile characteristics, with axial forces primarily concentrated in the upper portion. In contrast, PVC-casing and steel-casing piles effectively reduced side frictional resistance, resulting in a more uniform distribution of axial force along the pile shaft.

(2)

After immersion, the model test materials underwent expansion. Soil expansion deformation was divided into two phases: an initial rapid rise phase with a significant expansion rate and a later gradual rise phase where the expansion rate stabilized. The uncased pile (pile 1) has the highest ultimate bearing capacity, exceeding the cased piles (piles 2, 3, and 4) by 7 kN.

(3)

Under the immersion condition with without additional pile top loads, the pile displacement caused by soil absorption expansion increases, and the presence of an outer casing reduces the upward displacement of the inner pile. The axial forces in the casing piles exhibit an initial increase followed by a decrease, with the peak axial force being highest in the PVC casing. The expansive soil, 720 mm thick, exerts an expansion stress of 108 kPa on the model piles. The positive and negative frictional resistance of the casing piles alternate along the length of the pile, with the neutral point of the model pile located at 160 mm from the top.

(4)

Under the immersion condition with additional pile top load, with additional pile top loads, uncased pile displacement changes the least, while cased piles (including steel casing and PVC casing piles) exhibit displacement changes approximately four times that of uncased piles. The axial forces along the depth of pile 1 decrease rapidly, and the side frictional resistance of the pile is positive friction and increases with the increase in immersion time.

(5)

From the analysis of pile–soil deformation, it is evident that casing piles effectively mitigate the impact of expansive soil on pile displacement, reducing the risk of foundation damage in expansive soil areas. Among the casing types, PVC casings provided the most effective treatment compared to bare and steel-casing piles.