Introduction

In general, piles serve the purpose of transmitting the superstructure’s loads into the deep soil layers, especially when the surface-level soil is weak and unsuitable for bearing heavier loads. Moreover, there are many types of piles used by engineers and designers, viz., reinforced cement concrete (RCC) piles, steel helical piles, under-reamed piles, timber piles, sheet piles, etc. However, steel helical piles were considered the most effective pile system in foundation engineering during the mid-nineteenth century [1,2,3]. The steel helical piles are extensively used for enhancing the uplift and lateral load resistance for many structures.

Steel helical piles consist of a steel shaft attached to one or more helical (spiral) steel plates at designed spacing. These plates typically resemble spirals and play a crucial role in the interaction of the pile with the soil. The installation process involves the use of hydraulic machines to screw the steel helical piles into the soil. This method offers benefits such as ease of installation, making it a convenient and efficient approach. Steel helical piles are associated with enhanced load-carrying capacity as compared the piles without helices. This means they can support heavier loads compared to traditional piles or other kinds of foundations [4,5,6,7,8,9,10,11]. The helical plates contribute to this enhanced load-carrying capacity. The helical plates on the steel shaft serve dual functions. First, they act as end bearings at the location of each plate, providing support at different levels. Second, they create an anchoring effect with the surrounding soil, thereby enhancing the stability and load-carrying capacity of the pile. The enhanced performance of steel helical piles is attributed to the interlocking characteristics of the helices with the surrounding soil [12,13,14,15,16,17,18,19,20,21]. As the helical plates are screwed into the soil, they create a secure and interlocked connection, improving the pile–soil interaction.

However, steel helical piles are susceptible to corrosion, especially in harsh soil conditions. The corrosion can result in a reduction in the load-carrying capacity of the piles [22,23,24]. This is a significant concern as it may impact the long-term performance and stability of structures supported by these piles. The cost of steel is higher than that of other materials commonly used in foundation systems. While steel helical piles may incur higher material costs, the overall construction cost may still be reduced due to the advantages of rapid installation. The ease and speed of installation can be advantageous, potentially offsetting the higher material costs. Unlike driven piles, steel helical piles cannot be reversed once installed in position. This lack of reversibility might limit the flexibility in adjusting or relocating the piles, which could be a consideration in certain construction scenarios.

RCC piles are more cost-effective than steel helical piles. Cost-effectiveness is a crucial factor in construction projects, and the affordability of RCC piles makes them an attractive choice, especially for large-scale infrastructure projects. RCC piles are deemed reliable for huge infrastructure projects. Their strength, durability, and load-bearing capacity make them suitable for supporting substantial structures and handling the demands of large-scale construction. RCC piles are less susceptible to corrosion compared to steel helical piles. Corrosion resistance is a critical property, especially in environments with aggressive soil conditions or exposure to moisture, where corrosion can compromise the integrity of structural elements. The information implies that RCC piles experience less degradation over time compared to steel helical piles [25]. This characteristic contributes to the long-term stability and reliability of RCC piles, making them a durable foundation solution.

The installation process for RCC piles is time-consuming compared to steel helical piles. The time factor is a crucial consideration in construction projects, and longer installation times can impact project schedules and the associated timelines. RCC piles pose challenges in transportation due to their heavier weight compared to steel helical piles. The weight of the piles can influence logistics and transportation requirements, potentially adding complexity to the construction process. Quality control of RCC piles is more challenging compared to steel helical piles. The difficulty is attributed to factors such as mix proportions and the placement of concrete in position. Maintaining consistent quality in the production and installation of RCC piles is essential for their structural performance.

In this study, the helical grooves are introduced into RCC piles to mitigate the drawbacks associated with both steel helical piles and traditional RCC piles. This suggests an innovative solution to combine the advantages of both pile types. The helical grooves are introduced to incorporate anchoring and enhanced interlocking with the surrounding soil to improve the soil–pile interaction. This implies that the helical grooves are expected to provide similar advantages, such as enhanced load-carrying capacity and potentially faster installation [26] while retaining the beneficial properties of RCC piles. The amalgamation of RCC piles with helical grooves is expected to improve the efficiency of the RCC helical piles. The degradation of RCC is much slower compared to steel in harsh soil conditions, thus ensuring the longevity of pile foundations. This consideration implies that over time, RCC piles with helical grooves are expected to provide better load-carrying capacity compared to degraded steel helical piles. Additionally, these piles find application in challenging soil conditions, showcasing their versatility in varying environmental contexts.

To assess the performance of RCC piles with helical grooves, it is essential to subject the piles to various loading conditions. Loading of piles involves applying external forces, including vertical and lateral loads. Vertical loading is applied axially, which comprises the dead load of the structure and any live loads. Axial loads are transmitted through the pile via end bearing and skin friction. In certain cases, vertical loads may be applied as tension loads, testing the pull-out strength of the pile. Lateral or horizontal loads play a crucial role in tall structures where the lateral deflection of the structure is critical. These loads result from various factors, including wind loads, seismic forces, and earth pressure acting on the piles. In many practical scenarios, piles are exposed to combined axial and lateral loads due to a combination of dead loads, wind loads, wave pressure, and seismic forces. Assessing the performance under these diverse loading conditions is crucial for understanding the robustness and efficiency of RCC piles with helical grooves in real-world applications. Most of the researchers have studied the performance of piles subjected to axial loading and lateral loading independently [12, 20, 21, 27,28,29,30,31,32]. The combined axial and lateral loading studies are very important, as pile behavior under combined loading provides more realistic results for offshore and earthquake-resistant structures. However, the combined axial and lateral loading studies are limited [33,34,35,36,37].

Piles are categorized into two groups: short and long. The classification is determined by the length-to-diameter (L/d) ratio. The L/d ratio, representing the ratio of the length of the pile to its diameter, is a common parameter used to evaluate the behavior of piles under different loading conditions. Typically, when L/d is greater than 30, the pile is categorized as a long pile, while when L/d is less than 20, the pile is considered a short pile [38, 39]. Short piles behave as rigid bodies under lateral load [40]. This means that their response to lateral forces is more analogous to that of a solid, unyielding structure. In contrast, long piles lose their significance beyond a particular depth under lateral load. This implies that, beyond a certain depth or L/d ratio, the behavior of long piles differs from that of short piles, and they may not exhibit the same rigid body response. The behavior of piles, whether short or long, is also governed by the characteristics of the surrounding soil [12, 37]. Soil properties play a crucial role in how piles interact with the soil and respond to lateral loads. This is particularly important in geotechnical engineering, where the soil-structure interaction is a critical aspect of foundation design.

In the present study, the performance of RCC helical piles subjected to combined axial and lateral loading is studied. Also, the L/d ratio of the pile and soil elastic modulus is varied to study the behavior of the pile.

Integrating all these variations of pile and soil properties requires complex experimental techniques or numerical methods. However, the finite element method (FEM) significantly simplifies the simulation environment and provides a versatile platform to study pile-soil interaction problems. In the early 1980’s, a two-dimensional FEM approach was used [41] for simulation of flexible piles under lateral loading. Furthermore, advancement in FEM three-dimensional analysis is carried out [42] for lateral load analysis in layered soil. To perform a three-dimensional pile-soil interaction analysis in this study, Abaqus software is used.

Numerical analysis

Abaqus is a finite element analysis (FEA) software widely used for simulating and analyzing complex structural and geotechnical behavior. This study is employed to conduct numerical simulations and analyze the behavior of RCC plain piles and RCC helical piles. The focus of the study is on the combined effect of axial and lateral loading on both types of piles. Combined loading scenarios are often encountered in practical applications, and understanding the behavior under these conditions is crucial for designing reliable foundation systems. The behavior of piles under such combined loading scenarios can be complex and may require advanced numerical methods for accurate representation. The use of Abaqus for numerical simulation provides satisfactory results for the behavior of RCC helical piles under combined loading.

The numerical simulation of helical piles involves five major steps, outlined as follows:

  • Step 1: Creation of solid models (pile and soil).

  • Step 2: Assigning material properties to the solid models.

  • Step 3: Assigning boundary and loading conditions to the pile and soil.

  • Step 4: Meshing of the solid models and analysis of the model

  • Step 5: Obtaining the results in post-processing mode.

In step one, the plain pile of a 0.5 m diameter circle is drawn and extruded to a length of 5 m as shown in Fig. 1a and the reinforcement cage is modeled by using line element as shown in Fig. 1b. The geometrical properties of different pile configurations, considered for the present study are shown in Table 1. In total, six different pile configurations are selected to investigate the combined axial and lateral loading effects. Out of these, there are three plain piles, and the remaining three are helical piles with a helical pitch of 250 mm. Figure 1c represents the plain pile with a reinforcement cage embedded, and Fig. 1d represents the helical pile along with the reinforcement cage.

Figure 1
figure 1

a Plain pile, b Reinforcement cage, c Plain pile with reinforcement cage, and d Helical pile with pitch 250 mm and reinforcement cage.

Table 1 Geometrical properties of the pile

The pile is modeled by using the C3D10 tetrahedral element as a deformable body with linear elastic properties [35, 36]. The reinforcement cage of the pile is modeled by using T3D2, a 2-node linear 3-D truss element [43, 44]. The pile is strengthened with main reinforcement and lateral ties. The main reinforcement provided for the pile is 1% of the gross sectional area [45, 46] of the pile. The main reinforcement is provided in the form of 8 bars of 25 mm diameter, tied with lateral reinforcement at 150 mm center-to-center spacing, along the depth. In step two, the physical properties of the pile and steel are assigned as presented in Table 2.

Table 2 Physical properties considered for steel and pile

Step three involves the study of the interaction between concrete and steel, specifically referring to the reinforcement cage, as it is a critical aspect of the RCC pile model. The interaction is explicitly modeled by establishing an embedded connection. An embedded connection is a modeling technique that ensures a strong bond or attachment between different components in a structural system. In this case, it is applied to simulate the connection between the concrete of the pile and the steel reinforcement cage. The embedded connection is configured in a way that prevents any slip between the concrete and the reinforcement cage [43, 44]. This condition is essential for accurately representing the behavior of the reinforced concrete pile, where the steel reinforcement is intended to enhance the structural integrity by resisting tensile forces.

The soil is modeled using a three-dimensional continuum model. This approach considers the soil as a continuous medium, allowing for the simulation of its behavior in three dimensions. The effective size of the soil considered for the study is specified as 10 m (length) × 10 m (width) × 15 m (height). This defines the dimensions of the soil domain in the numerical model. The size of the soil is 20 times the diameter of the pile. This is a significant consideration as it helps in mitigating boundary effects. When the size of the soil domain is sufficiently large compared to the size of the pile, the influence of stress reflections from the boundaries is reduced [11].

The soil is modeled in Abaqus software by using the C3D8R element as a Mohr–Coulomb plasticity model [16, 34,35,36, 42, 47,48,49]. In the classical Mohr–Coulomb model, the friction angle and dilatancy angle are treated as constant material parameters. Nevertheless, it is crucial to recognize that these parameters may exhibit variability depending on the initial relative density and stress level of the sand being simulated. The specific values assigned to these angles are responsive to the conditions of the sand material, and changes in density and stress level prompt corresponding adjustments in the friction and dilatancy angles within the model [50,51,52,53]. The soil elastic modulus was varied at three different levels low, medium, and high range to account for the behavior of piles in different cohesionless soil grades.

The properties of the soil are assigned as per Table 3, and elastic modulus is varied for all the models. All the elements are assembled in the assembly step of the software, where the coordinates of soil, pile, and reinforcement cage are matched in such a way that they are in proper connection with all the points of contact as shown in Fig. 2.

Table 3 Soil properties
Fig. 2
figure 2

Assembled model

In the Abaqus boundary condition manager, the soil is fixed at the bottom and all side faces are restrained for horizontal movement as shown in Fig. 3 [35, 36, 54,55,56]. The interaction between the pile and soil surface is assigned to transfer the loads from the pile to the soil, the interaction is simulated by employing the penalty method with tangential behavior [35,36,37, 54] and a tangential frictional coefficient of 0.43, which is two-thirds of the friction angle of the cohesionless soil. The interaction is surface-to-surface type by considering the master surface as the pile exterior surface and the slave surface as the soil’s internal surface as shown in Fig. 4. The axial and lateral loads are applied at a reference point, where the whole pile is connected to the reference point to enable the transfer of loads. The loads are applied as gradually increasing linear loads to simulate the actual loading scenario of the pile loading. The axial load considered for all the models is 1 × 103 kN and a lateral load of 0.25 × 103 kN as per [57] as shown in Fig. 5.

Fig. 3
figure 3

Boundary conditions

Fig. 4
figure 4

Surface interaction

Fig. 5
figure 5

Combined axial and lateral load

Mesh dependency test

The mesh dependency test in FEM analysis plays a crucial role as the results are contingent on the chosen mesh size. A coarser mesh, characterized by larger element sizes, yields fewer elements in the model, resulting in less accurate results and reduced computation time. Conversely, a finer mesh, with smaller element sizes, increases the number of elements, leading to more accurate results but requiring more computational resources and time.

In this study, the mesh dependency test was initiated with a 1000-mm coarser mesh and progressed to a 250-mm finer mesh. The 1000 mm mesh generated 3116 soil elements, while the 250 mm mesh produced 96,704 soil elements. As depicted in Fig. 6, a mesh size of 500 mm resulted in 16,166 soil elements, and the Von Mises stress in the soil closely approached that of the 250 mm mesh size. Therefore, for this study, a mesh size of 500 mm is deemed appropriate, striking a balance between accuracy and computational efficiency from a similar study [58].

Fig. 6
figure 6

Number of elements v/s Von Mises stress in soil

As per the mesh dependency test, in step four the meshing of all the parts is done by providing a mesh size of 500 mm, after meshing the individual parts are assembled as shown in Fig. 7.

Fig. 7
figure 7

Meshed model

In this study, two types of piles, namely, RCC plain piles and RCC helical piles with a helix pitch of 250 mm, have been investigated as per Table 4. The evaluation of their performance encompasses a range of scenarios, including varying the L/d ratio of the piles and examining the influence of soil elastic modulus. The loading conditions considered are pure axial loading, pure lateral loading, and the combined effect of axial and lateral loading. These comprehensive analyses aim to provide a thorough understanding of how these different factors impact the behavior and performance of the piles, offering valuable insights for engineering and design applications.

Table 4 Models considered for simulation

Validation of the numerical model

The validation of the Abaqus software model is crucial for conducting a parametric study. In this study, to validate the numerical model, a scaled-down version is created using Abaqus software, and its results are compared to those obtained from a laboratory-scale experimental model.

The experimental setup, as illustrated in Fig. 8, involves a masonry tank measuring 1.25 m in length, 1.25 m in width, and 1.50 m in height. The tank is filled with cohesionless sand in three layers, each 0.4 m in thickness. The sand is vibrated to achieve a relative density of 85%, following the method outlined by [59]. The poorly graded sand has a fineness modulus of 2.75, falls under zone II, and an angle of internal friction of 35°.

Fig. 8
figure 8

Experimental setup for axial load testing

Axial loading is applied using a loading frame fabricated from channel sections sized 100 × 50 × 5 mm. The loading is executed through the lever arm method to minimize the requirement for dead load, as recommended by [60]. The experimental pile considered has a diameter of 80 mm and a length of 800 mm, resulting in an L/d ratio of 10. The numerical model is created to match the experimental conditions and dimensions as shown in Fig. 9. The methodology used to model is a similar way as outlined in Sect. “Numerical analysis” of this paper.

Fig. 9
figure 9

Numerical model and mesh configuration used for simulating the laboratory model

The comparison of the numerical and experimental results specifically focuses on axial loading and settlement characteristics. The findings, as presented in Fig. 10, reveal that the numerical results for the pile under axial loading closely resemble the outcomes observed in the experimental setting. However, the numerical model results are slightly 11% greater than the experimental results. This alignment between numerical and experimental results suggests a successful validation of the numerical model.

Fig. 10
figure 10

Validation of numerical model using Abaqus with experimental results

Analysis of results

Pure axial loading effect on settlement of piles

The piles are subjected to pure axial loading. Pure axial loading implies that the load is applied along the axis of the pile, without any lateral load component. This loading condition is fundamental for evaluating the response of the piles under axial forces. The behavior of the piles is assessed based on several parameters. These parameters include:

Displacement: The movement or deformation of the pile, likely measured at specific points or along its length.

Stresses in the Soil: The distribution and magnitude of stresses within the surrounding soil. This provides insights into how the soil reacts to the applied axial load.

Contact Pressure Between Pile and Soil: The pressure at the interface between the pile and the soil. This is crucial for understanding the interaction and load transfer mechanism between the pile and the surrounding soil. The assessment is conducted.

In Fig. 11a, the vertical displacement of the plain pile under pure axial loading is depicted, representing the settlement of the pile. The maximum observed settlement for the plain pile is 9.01 mm. In Fig. 11b, the settlement of the helical pile under axial loading is illustrated, with the maximum settlement recorded at 6.37 mm. Notably, this settlement is nearly 30% less than that observed in the plain pile subjected to similar loading conditions.

Figure 11
figure 11

a Plain pile subjected to axial load showing displacement of soil particles b Helical pile subjected to axial load showing displacement of soil particles c Plain pile subjected to axial load showing stresses in the soil d Helical pile subjected to axial load showing stresses in the soil e Plain pile subjected to axial load showing contact pressure in the soil and pile f Helical pile subjected to axial load showing contact pressure in the soil and pile.

In Fig. 11c, the axial stresses in the soil under pure axial loading for the plain pile are illustrated. The maximum stress recorded is 4.608 × 10–01 MPa. Conversely, in Fig. 11d, the stress recorded with the helical pile is 2.977 × 10–01 MPa, indicating a lower stress level compared to the plain pile. This reduction in stresses is attributed to the helical pile design, where each helical groove interlocks with the soil, resulting in the distribution of loads and stresses.

Additionally, it is observed that in the case of the plain pile, the stresses in the soil are concentrated at the pile bottom. However, in the helical pile configuration, the stresses are distributed at different levels due to the interlocking effect of the helical grooves on the soil similar study by [59, 61].

In Fig. 11e, the contact pressure between the pile and soil for the plain pile configuration is depicted, with a recorded value of 1.90 MPa. Conversely, in Fig. 11f, the contact pressure for the helical pile configuration is illustrated, and the recorded value is 3.85 MPa. The increase in contact pressure in the helical pile configuration can be attributed to the presence of helical grooves, which results in a greater surface area for contact between the pile and the soil.

The piles are simulated under pure axial loading, with an L/d ratio of 10, soil elastic modulus of 10 MPa, and helix pitch of 250 mm. Figure 12 illustrates the response of the plain and the helical pile subjected to pure axial loading. The helical pile offers a 26.5% lower settlement as compared to a plain pile. This is because the helix interlocks with the surrounding soil, leading to enhanced skin friction and end-bearing similar trend of results are obtained in [59, 61, 62].

Fig. 12
figure 12

Pure axial load v/s settlement

The influence of the L/d ratio on the response of both plain and helical piles is depicted in Fig. 13. As the L/d ratio of the pile increases, the settlement of the pile decreases, primarily due to the additional skin friction provided by the extended length of the pile.

Fig. 13
figure 13

Pure axial load v/s settlement for L/d ratio variation

For the plain pile, having L/d ratios of 15 and 20, there is a reduction in settlement of 40 and 87%, respectively, compared to the plain pile with an L/d ratio of 10. In the case of the helical pile with an L/d ratio of 10, there is a 26.5% decrease in settlement compared to the plain pile with a similar L/d ratio. On the other hand, helical piles with L/d ratios of 15 and 20 exhibit greater reductions in settlement, with values of 28 and 32%, respectively, compared to the settlement observed in plain piles.

The effect of soil elastic modulus on the response of plain and helical piles is shown in Fig. 14. The L/d ratio of the pile is considered as 10. The settlement of the pile reduces, with the increase in the elastic modulus of soil (E). For soil elastic modulus of 30 and 80 MPa, the settlement of plain pile decreases by 3 and 7.5%, respectively, as compared to that of 10 MPa.

Fig. 14
figure 14

Pure axial load v/s settlement for soil elastic modulus variation

A similar trend is observed in helical piles, but the percentage reduction in the settlement is higher than in the plain piles. For soil elastic modulus of 10 MPa, it is observed that the settlement is 26% less than that of the plain pile. Similarly, for soil elastic modulus of 30 and 80 MPa, the settlement decreases by 33 and 42%, respectively.

Pure lateral loading effect on the lateral displacement of piles

When piles are subjected to lateral loading without any axial loading component, the behavior of piles is observed based on the lateral displacement of the pile and the stresses in the soil. The lateral behavior of piles is of critical importance in many structures. It provides insights into how piles respond to horizontal forces, which is crucial for assessing the stability and performance of various engineering projects.

In Fig. 15a, the lateral behavior of the plain pile under lateral loading is illustrated, revealing that the maximum lateral displacement of the pile is 6.06 mm. However, in Fig. 15b, the helical pile subjected to similar loading conditions exhibits a maximum displacement of 6.68 mm. The increase in lateral displacement of the helical pile is attributed to the presence of helical grooves, where soil particles pass horizontally through the groove area.

Figure 15
figure 15

a Plain pile subjected to lateral load showing displacement b Helical pile subjected to lateral load showing displacement c Plain pile subjected to lateral load showing stresses in the soil d Helical pile subjected to lateral load showing stresses in the soil.

In contrast to the lateral displacement of the helical piles, the stresses in the soil decrease, as observed in Fig. 15d. This reduction in stresses is attributed to the increased surface area of the helical pile, allowing for more interaction with the surrounding soil. Consequently, the helical pile exhibits lower stresses in the soil compared to the plain pile, as illustrated in Fig. 15c.

Helical piles of pitch 250 mm and plain piles are subjected to pure lateral loading, with soil elastic modulus as 10 MPa to assess the response of the piles. Figure 16 illustrates that, as the L/d ratio of the pile increases, the lateral displacement of the pile decreases.

Fig. 16
figure 16

Pure lateral load v/s lateral displacement for L/d ratio variation

The helical piles with L/d ratios 15 and 20 exhibit reductions in lateral displacement of 28 and 22%, respectively, compared to plain piles. However, a helical pile of L/d ratio 10 experiences a lateral displacement of 8% greater than that of a plain pile.

The effect of soil elastic modulus on the response of plain and helical piles is shown in Fig. 17. The L/d ratio of the pile is considered as 10. Figure 17 depicts that, as the elastic modulus of the soil increases the lateral displacement of the pile decreases. However, it is observed that the helical pile exhibits higher lateral displacement as compared to a plain pile.

Fig. 17
figure 17

Pure lateral load v/s lateral displacement for soil elastic modulus variation

Combined loading effect on the settlement of piles

Piles are subjected to combined axial and lateral loading, with a soil elastic modulus of 10 MPa and a helical pitch of 250 mm. The effect of the L/d ratio on the settlement of the pile is shown in Fig. 18. The settlement diminishes as the L/d ratio of the pile increases under the combined loading effect. It is observed that the settlement of the plain pile and helical pile reduces by 3–5%, respectively, as compared to that in the case of pure axial loading. This reduction is due to the lateral load which tries to lift the pile in a vertical direction.

Fig. 18
figure 18

Combined loading effect on axial response of pile v/s settlement for L/d ratio variation

The effect of soil elastic modulus on the response of plain and helical piles is shown in Fig. 19. The L/d ratio of the pile is considered as 10. Figure 19 illustrates that the settlement decreases as the soil elastic modulus increases as in the case of pure axial loading. For soil elastic modulus 10 MPa, the settlement of the helical pile is 18% less than that of the plain pile. Similarly, for soil elastic modulus of 30 and 80 MPa, the settlement decreases by 24 and 34%, respectively.

Fig. 19
figure 19

Combined loading effect on axial response of pile v/s settlement for soil elastic modulus variation

Combined loading effect on lateral displacement of piles

The lateral behavior of piles with soil elastic modulus of 10 MPa is illustrated under combined axial and lateral loading. The helical pitch is considered as 250 mm. Figure 20 illustrates that the lateral displacement decreases as the L/d ratio increases as in the case of pure lateral loading.

Fig. 20
figure 20

Combined loading effect on lateral response of pile v/s lateral displacement for L/d ratio variation

However, for combined loading the lateral displacements are lower than that observed under pure lateral loading. The helical piles of L/d ratios 15 and 20 exhibit 29 and 21% lower lateral displacement, respectively, as compared to that of plain piles. Moreover, helical piles with an L/d ratio of 10 exhibits 9% higher lateral displacement as compared to plain piles.

The effect of soil elastic modulus on the response of plain and helical piles is shown in Fig. 21. The L/d ratio of the pile is considered as 10. As the soil elastic modulus increases lateral displacement decreases. However, it is observed that helical piles exhibit higher lateral displacement as compared to plain piles.

Fig. 21
figure 21

Combined loading effect on lateral response of pile v/s lateral displacement for L/d ratio variation

Thus, this study clearly shows that helical piles exhibit superior performance compared to plain piles. With an increase in the L/d ratio of the pile and the elastic modulus of the soil, there is a reduction in settlement and lateral displacement under various loading conditions. The combined axial and lateral loading provide realistic behavior of the pile.

Conclusions

In this paper, a three-dimensional pile-soil interaction model has been developed utilizing Abaqus software. A comprehensive set of numerical simulations has been conducted on RCC plain and helical piles under combined axial and lateral loading conditions, by varying L/d ratio and soil elastic modulus. The following conclusions are drawn from the results of these numerical studies:

  • The piles with higher L/d ratios tend to experience less settlement under the given loading conditions. The helical pile with an L/d ratio of 10, 15, and 20 exhibits 26.5, 28, and 32% reduction in settlement, respectively, as compared to that of plain piles subjected to pure axial loading.

  • When helical piles subjected to pure axial loading, the stresses in the surrounding soil decreases. The decrease in stresses is attributed to the presence of helical grooves in the helical piles. The helical grooves likely influence the way the piles interact with the soil, resulting in a reduction in stresses compared to plain piles.

  • As the elastic modulus of soil increases, it was observed that there is reduced settlement under pure axial loading conditions. Despite the general trend of reduced settlement with increasing soil stiffness, helical piles exhibit a higher percentage reduction in settlement compared to plain piles.

  • Similar to the axial loading condition, it was observed that as the L/d ratio of the pile increases, the lateral displacement of the pile decreases under pure lateral loading.

  • As the elastic modulus of soil increases, the lateral displacement decreases. Despite the general trend of reduced lateral displacement with increasing soil stiffness, the helical piles exhibit higher lateral displacement compared to plain piles.

  • Under combined loading conditions, both settlement and lateral displacement decreases with an increase in the L/d ratio of the pile. Additionally, it was observed that the settlement of the pile decreased by 3–5% as compared to pure axial loading.

  • As the soil elastic modulus increases there is a higher decrease in settlement for combined loading compared to pure axial loading. The presence of lateral loading influences settlement behavior, and the impact is more pronounced with increasing soil stiffness.

Future scope of the study

  • The study can be extended to explore the group effect by examining various pile group combinations.

  • The behavior of both plain and helical piles can be assessed under dynamic loads, providing insights into their performance in comparison to static loading conditions.