Surface Settlement Induced by Excavation of Double-Line Shield Tunnel

Abstract

In the construction of the Harbin Metro Line 3 twin-bore tunnel project, the ABAQUS numerical simulation method was employed to investigate the impact of various grouting thicknesses, burial depth ratios, tunnel spacing, and excavation pressures on the geological conditions of the project. Corresponding construction control measures were proposed. The study revealed: (1) Grouting thickness is inversely proportional to ground settlement, with an optimal thickness of 0.3 m; (2) Lateral ground settlement increases with an increasing depth ratio, showing an initial increase followed by a decrease, closely related to soil arching effects; (3) As the distance (L) between two parallel tunnels increases, the peak ground settlement of the left and right tunnels decreases. Smaller tunnel spacing leads to stronger mutual interactions, resulting in a “V”-shaped settlement deformation within the overlapping disturbance zone. When the tunnel spacing is sufficiently large, the mutual interference between the two tunnels diminishes, and the ground settlement deformation is considered the superimposition of two independent tunnel settlements, forming a “W”-shaped settlement trench; (4) Excavation pressure exceeds the static soil pressure, causing soil disturbance, leading to an uplift of the soil in front of the tunnel face and an expansion of the plastic zone in the surrounding soil, ultimately resulting in increased subsequent surface settlement.

1 Introduction

The underground railway engineering is a highly anticipated project characterized by its massive construction scale, lengthy duration, substantial investment, and significant societal benefits. Therefore, it is essential to conduct specialized research on safety and geological issues that may arise during subway construction. In the construction of urban rail transit in major cities, it has gained increasing popularity due to its fast construction pace, high level of automation, safety, and environmental features. However, irrespective of the measures taken, they can result in disturbances to the surrounding soil and have an impact on the ecological environment. For instance, this can lead to issues such as tilting of nearby buildings, damage to underground pipelines, and ground subsidence, among others [1].

With the development of tunnel boring technology, the issue of ground subsidence caused by tunnel excavation has increasingly captured people's attention. Scholars from both domestic and international backgrounds have conducted extensive research on the surface settlement induced by shield tunnel construction [2,3,4,5]. Xin et al. [6], based on two tunnels within the Hangzhou Rail Transit Line 3 project, the Industrial University Station to Liuhe Station section, utilized a 3D numerical simulation method to analyze the deformation and the effects of tunnel excavation on adjacent bridge pier foundations. Ngoc et al. [7] employed numerical simulation methods to analyze the impact of subway construction at different distances on subway construction and the surrounding ground subsidence in operational subway projects. Alessandra et al. [8], using two deep-buried tunnels located in fine-grained sandy soil layers as an example, improved the model’s predictive accuracy by adjusting geological parameters with observation data, and conducted numerical simulations and calculations. In the Beijing Metro Line 12 project, Wang et al. [9] adopted FLAC three-dimensional numerical simulation method to obtain the deformation characteristics of the strata and the surface when the shield traverses the existing railway. Zhang et al. [10] used the finite element method to analyze the influence of various factors on the surrounding environment in the shield construction process, and analyzed the influence of various factors on the surrounding environment, providing a theoretical basis for engineering construction. Based on the sandy soil foundation of a section of Nanchang subway, Zhao et al. [11] made a theoretical analysis and research on the ground subsidence before and after tunnel excavation by combining the method of mathematical regression analysis and numerical simulation, and obtained the changes of ground subsidence caused by tunnel excavation. Zhifu [12] analyzed the cause of surface subsidence during tunnel excavation and proposed the treatment method based on the measured data of the first phase of the first phase of Zhengzhou Metro Line 5.Based on the measured data of a rail transit tunnel construction site in Hangzhou, Gao [13] took surface settlement, tunnel settlement and tunnel settlement as the main research content, and used mathematical statistics to study the common rules of silt layer settlement and deformation under the condition of high water head. Jiahui et al. [14] conducted construction technology and engineering survey of EPB shield in Liancheng Town of Huaian East Railway Station shield working well in the Yellow River flood area, and obtained deformation characteristics of soil after the high-speed rail underpass through the wall. Dezhi et al. [15] took the existing buildings in urban areas through the subway shield as an example, and studied the effect of subway shield construction on existing buildings on the ground by combining measured data and numerical simulation. On this basis, the settlement control standards for existing buildings were proposed. Aiming at the ground subsidence caused by subway shield construction, Liu et al. [16] studied the change and interaction mechanism of ground subsidence under its influence by constructing a three-dimensional finite element simulation model combined with engineering observation data. Jun et al. [17] studied the problem of adjacent ground subsidence caused by shield construction. By constructing a finite element simulation model and combining with engineering monitoring data, Xiu et al. [18] revealed its action mechanism in the formation.

In summary, previous studies have predominantly focused on investigating the patterns of surface settlement caused by shield tunnel construction, with an emphasis on varying construction methods for different geological strata. However, there has been limited research on the surface settlement patterns resulting from shield tunneling through fine-grained sandy layers. In this study, the authors conducted three-dimensional numerical simulations using ABAQUS software during the construction of the double-track tunnels of the Harbin Rail Transit Line 3. They investigated the impact of grouting layer thickness (r), excavation pressure (p), tunnel depth ratio (H/D), and the distance between the double tunnels (L) on surface settlement. The study provides relevant recommendations for construction control in this context.

2 Engineering Background and Model Construction

This article is based on the civil engineering project of the second phase of Harbin Rail Transit Line 3. The project spans from the Dingxiang Park Station to the Sports Park Station, forming an overall inverted V-shaped slope longitudinally. Both the left and right line segments have a length of 819.789 m, resulting in a total length of 1639.578 m. The tunnel passes through fine sand layers and medium-coarse sand layers. Construction is carried out using a shield machine, with a designed tunnel outer diameter (including lining thickness) of 6 m, an inner diameter of 5.4 m, and a lining segment thickness of 0.3 m. Each segment is 1.2 m wide and assembled with staggered joints. The tunnel center is approximately 13 m below the ground. The geological cross-section and geological structure of this section are depicted in Fig. 1.

Fig. 1

Geological cross-section and geological structure map of the section

3 Calculation Model

In this study, the mechanical simulation model for shield tunnel construction is implemented using the ABAQUS finite element program. ABAQUS is known for its high accuracy in numerical simulations of shield tunnel excavation, making it one of the preferred programs for this purpose.

3.1 Basic Model Assumptions

In order to maintain consistency with actual construction, several computational assumptions need to be made before the study:

• (1) Neglect small variations in the terrain and all soil layers along the tunnel alignment, simplifying them as uniform horizontal layers;

• (2) Assume the rock mass as an ideal elastic–plastic medium;

• (3) Disregard time effects of rock deformation and the influence of groundwater seepage;

• (4) Ignore the in-situ stress within the soil layers and stress concentration near the geological boundaries;

• (5) Consider the rock mass as an isotropic and homogeneous material;

• (6) Set the initial stress field as the initial stress field of the surrounding rock;

• (7) Model the excavation-induced disturbance by reducing the modulus of the disturbed rock mass in the disturbance zone.

3.2 Parameter Selection

When conducting numerical simulation studies, it is necessary to select representative sections for modeling. Based on previous research, the lateral range boundary that affects soil deformation is roughly about 4 times the tunnel diameter away from the tunnel's side, while the vertical range boundary is approximately 3 times the tunnel diameter away from the tunnel bottom. Beyond these boundaries, the impact is nearly zero. Combining these findings with design parameters such as tunnel depth, the final dimensions of the ABAQUS model were determined to be 100 m in the lateral (x) direction, 100 m in the advancing (y) direction for the rock tunnel, and 70 m in the vertical (z) direction. Six cross-sections of the model required the setup of essential boundary conditions, with the top surface set as a free surface, and the other surfaces receiving corresponding displacement constraints. The ABAQUS model constructed for the research project is depicted in Fig. 2.

Fig. 2

ABAQUS model

The soil layers in ABAQUS were simulated using the Mohr–Coulomb constitutive model. The lining segments and the shield machine’s outer shell were simulated using shell structure elements and set as elastic materials. The overall finite element analysis used linear hexahedral elements, with a total model divided into 196,802 nodes and 210,764 elements of type C3D8R. The soil properties in the layers at the tunnel depth were determined based on geological reports and material properties to meet the requirements for accurately reflecting the construction conditions. The mechanical parameters for each soil layer are presented in Table 1.

Table 1 Physical and mechanical parameters of each soil layer

3.3 Simulating the Shield Tunnel Construction Process

Following the working principles of the shield tunneling machine, the simulated construction steps are as follows:

• (1) Establish material models for the soil layers, the shield tunneling machine, the waiting layer, and lining segments. Categorize these materials and assign the corresponding material parameters. Assemble these material models on the assembly board. For the initial stress zone, consider only self-weight effects and apply deformation corrections.

• (2) Based on the “live” and “dead” element method, create the analysis steps. During excavation of each layer of soil, activate the waiting layer, lining segments, and grouting layer. Simulate the support provided by the cutterhead to the rear soil using the positive-pressure method during excavation, thereby achieving support for the rear soil. Determine the numerical value of the static pressure that the construction process bears.

• (3) After each week of the tunnel boring machine's operation, remove the support from the previous week, generate new support, and adjust the stiffness and other parameters of the shield tunnel support material. Utilize synchronous grouting to implement grouting around the tunnel's surrounding rock using the grouting method. Apply grouting pressure to the surrounding rock and set corresponding operational procedures. Meanwhile, while maintaining a constant grouting pressure, layer the grouting material and conduct numerical simulations.

• (4) According to the numerical model, set the overall excavation into 80 excavation steps. During the 80 excavation steps, the shield shell supports the surrounding rock. After the construction is completed, the tunnel is fully penetrated.

3.4 Numerical Simulation Validation

Currently, there are several methods for predicting settlement in subway engineering, including statistical methods using measurement data (such as the Peck formula) and numerical methods using finite elements and boundary elements. This article employs a comparative analysis between measured data and numerical simulation data to validate the results.

This article uses the actual measured ground settlement data during the construction of the double tunnels of Harbin Metro Line 3 as reference, and conducts a comparative analysis with the numerical solution results, as shown in Fig. 3, the surface settlement along the tunnel excavation direction exhibits a good fit between measured and simulated data. The maximum fitting error between numerical simulation and actual settlement on the left tunnel side is 12.61%, while on the right tunnel side, it is 10.56%. These results validate the accuracy of the numerical simulation in this article, indicating that using ABAQUS finite element software for simulating shield tunnel construction can provide valuable guidance for practical engineering projects.

Fig. 3

Comparison between actual settlement curve and numerical simulation settlement curve

As depicted in Fig. 4, it shows settlement cross-sections of the tunnel at different excavation steps. During the shield tunneling process, the excavation of soil disturbs the surrounding soil, resulting in settlement of the surface soil. As the advancement continues, the unexcavated soil experiences compression, leading to a certain degree of uplift in the unexcavated area. In the region where excavation support has been completed, the peak surface settlement occurs directly above the tunnel. Through support and reinforcement measures, the excavated area forms a stable settlement region.

Fig. 4

Multi-analysis step slice displacement nephogram

4 Analysis of the Impact of Shield Tunneling Construction Parameters on Surface Settlement

4.1 Scheme Design

Three factors were selected for the study: grouting thickness (r), depth-to-diameter ratio (H/D, where H represents the vertical distance from the tunnel top to the ground surface, and D represents the tunnel excavation diameter), thrust pressure (P), and the distance between twin tunnels (L). The study focused on observing lateral surface settlement and, in turn, delving into the impact of each parameter on surface settlement during shield tunnel construction. The specific scheme settings are detailed in Table 2.

Table 2 Tunnel boring machine construction parameter configuration plan

4.2 Influence of Different Grouting Thickness on Surface Settlement During Shield Tunnel Excavation

In order to study the influence of different grouting thicknesses on surface settlement during tunnel construction, we set various tunnel grouting thicknesses (r), including 0, 0.1, 0.2 and 0.3 m. The distance between the two tunnels (L) was set at 8 m, and the depth from the tunnel crown to the ground surface (H/D) was 10 m, which corresponds to a depth-to-diameter ratio of 1.67. We observed that as the grouting thickness increased, surface settlement decreased, with the maximum displacement settlement occurring directly above the tunnel crown. After the tunnel was fully penetrated, we generated lateral surface settlement values for each grouting thickness and obtained corresponding lateral surface settlement curves for different grouting thicknesses, as shown in Fig. 5.

Fig. 5

Lateral surface settlement curves corresponding to tunnel boring machine excavation with different grouting thicknesses

Based on the research results from Fig. 5, we observed that the peak of surface settlement typically occurs directly above the tunnel. With other construction parameters held constant, increasing the grouting layer thickness effectively reduces the deformation of the surface soil, and the peak surface settlement decreases as the grouting layer thickness increases. Specifically, when there is no grouting, the numerical simulation indicates a peak surface settlement of 13.32 mm, directly above the tunnel. When the grouting layer thickness is 0.1 m, the lateral surface settlement peak is 9.47 mm; at 0.2 m grouting layer thickness, the lateral surface settlement peak is 7.59 mm; at 0.3 m grouting layer thickness, the lateral surface settlement peak is 6.04 mm.

The results indicate that within a certain range, increasing the grouting layer thickness effectively fills the gap between the shield tail and reduces ground settlement during shield tunnel excavation. However, based on practical engineering experience, an excessively thick grouting layer may have an impact on the surrounding soil mechanics, leading to soil disturbance and being detrimental to controlling surface settlement. Therefore, it is important to carefully select the grouting layer thickness in real-world engineering. The process from no grouting to injecting a 0.3 m thick grouting layer results in a significant reduction in peak surface settlement. Taking various factors into consideration, a grouting layer thickness of 0.31 m is found to meet the practical requirements in the project.

4.3 The Impact of Twin Tunnel Spacing on Surface Settlement

This subsection aims to investigate the influence of different twin tunnel spacing on surface settlement. Various tunnel distances (L) were selected, including 4, 8, 12, and 16 m, for numerical simulation and analysis. In the numerical simulation parameters for shield tunnel construction, a grouting layer thickness of 0.31 m, a tunnel excavation step length of 1.2 m, and a depth-to-diameter ratio (H/D) of 1.67 (corresponding to a crown depth of 10 m) were chosen.In the numerical models, after the completion of tunneling, lateral surface settlement curves for different spacing were generated, as illustrated in Fig. 6.

Fig. 6

Lateral surface settlement curves corresponding to tunnel boring machine excavation with different spacings

Based on the observations from Fig. 6, it is evident that as the distance (L) between the two parallel tunnels increases, the surface settlement for both the left and right tunnels gradually decreases. With the increasing distance (L) between the two parallel tunnels, the surface settlement curves expand outward, and the settlement area also becomes larger. The original “V”-shaped settlement trough gradually evolves into a broader “W” shape, becoming shallower and wider. When the distance (L) between the twin tunnels increases from 4 to 8 m, the change in the settlement distribution curve transforms from the initial “V” shape to a “W” shape.

This change can be primarily attributed to the fact that when the distance between the tunnels is small, the mutual interference between them gradually intensifies. This leads to an increased deformation of the surface settlement in the common disturbance area, thus forming a “V"-shaped settlement trough. Conversely, when the distance between the tunnels is larger, the mutual interference caused by construction weakens. In this case, the surface settlement is considered as the superposition of the independent settlement and deformation effects of the two separate tunnels, resulting in the “W”-shaped settlement trough.

4.4 Impact of Different Depth-To-Diameter Ratios on Surface Settlement

This section aims to investigate the influence of different depth-to-diameter ratios (H/D) on surface settlement during tunnel construction. We considered various H/D ratios, specifically 2, 3, 4, and 5, which correspond to crown depths of 12, 18, 24, and 30 m, respectively. Numerical simulations were conducted with a 0.31m grouting layer thickness and a tunnel excavation step of 1.2 m. Once the tunneling was complete, we generated lateral surface settlement curves for each H/D ratio, as depicted in Fig. 7.

Fig. 7

Lateral surface settlement curves corresponding to tunnel boring machine excavation with different depth-to-diameter ratios

Based on the observations from Fig. 7, we can conclude that the maximum surface settlement of shield tunnels exhibits an increasing trend initially and then decreases as the depth-to-diameter ratio (H/D) increases. When H/D increases from 1.67 to 2, the maximum surface settlement rises from 6.03 to 9.47 mm. When H/D increases from 2 to 3, the maximum surface settlement decreases from 9.47 to 5.63 mm. As H/D increases from 3 to 4, the maximum surface settlement further decreases to 5.17 mm. Finally, as H/D increases from 4 to 5, the maximum surface settlement decreases to 4.22 mm.

The numerical simulation results for deeply buried tunnels show that the maximum surface settlement exhibits a trend of increasing and then decreasing. This phenomenon occurs because tunnels with relatively small depth-to-diameter ratios (H/D = 1.67) have less disturbance to the surrounding soil during excavation, and during the excavation process, the soil naturally forms an arch, which reduces surface settlement. As the depth-to-diameter ratio increases (from H/D = 1.67 to 2), the influence zone of the soil arch above the tunnel is limited. With the increase in depth, the soil exerts greater pressure on the tunnel. However, at this point, the soil arch above the tunnel cannot provide sufficient support, leading to increased surface settlement. When the depth-to-diameter ratio continues to increase (from 2 to 5), the soil arch gradually forms, and the self-supporting capacity of the surrounding rock is greater than the self-weight stress of the soil. Consequently, the deformation of the soil layer decreases.

4.5 Effects of Different Excavation Pressures on Surface Settlement

In this study, the grouting layer thickness of the tunnel is 0.31m, the burial depth is 10m (the burial depth ratio H/D = 1.67), and the excavation step of the tunnel is 1.2m. We considered four different excavation pressures: 100, 142, 200, and 300kPa. After the tunnel is completely excavated, we generated lateral surface settlement curves for various excavation pressures, and these curves are shown in Fig. 8, revealing the lateral surface settlement patterns associated with different excavation pressures.

Fig. 8

Shows the lateral surface settlement curves corresponding to different excavation pressures during shield tunneling

According to the data from Fig. 8, it is evident that, with all other construction parameters held constant, increasing the excavation pressure leads to an increase in the peak of the lateral surface settlement curve. Specifically, when the excavation pressures are 100, 142, 200, and 300 kPa, the corresponding numerical simulation results indicate that the peak of the lateral surface settlement is 4.27, 6.03, 8.46, and 12.64 mm.

When the excavation pressure increases from 100 to 300kPa, the excavation pressure exceeds the static earth pressure of the face, resulting in increased soil disturbance and, in turn, causing an uplift of the soil ahead of the face. This leads to a larger plastic zone in the surrounding soil, resulting in later-stage settlement deformation and an increase in surface settlement.

According to theoretical analysis and engineering experience, the excavation pressure should be slightly less than the stress relief value of the face soil. The differences in disturbances caused by different excavation pressures are relatively small. Therefore, considering both lateral deformation and economic benefits, an excavation pressure of 142 kPa was chosen for the project, which meets the practical requirements.

5 Optimization Analysis of Surface Settlement Deformation

The accuracy of the ABAQUS numerical simulation model was verified based on actual surface settlement data. By varying the selected parameters, we examined the impact of different grouting thicknesses, burial depth ratios, and the distance between the two tunnels on surface settlement. This allowed us to perform an optimization analysis of surface settlement deformation.

• (1) Within a certain range, a thicker grouting layer results in better filling of the annular gap at the shield tail, leading to reduced ground settlement during shield excavation. However, based on practical engineering experience, the grouting layer should not be excessively thick, as it can impact the stress on the surrounding soil and hinder the control of surface settlement. Numerical simulations show that the peak settlement values for grouting radii of 0.3 and 0.31 m are very close, measuring 6.03 and 6.04 mm, respectively. Taking all factors into account, an optimal grouting layer thickness of 0.3 m is recommended. Nevertheless, an actual construction choice of 0.31 m meets construction requirements but results in some material loss.

• (2) When the distance between tunnels is small, the mutual interference between them gradually intensifies, leading to increased settlement and deformation in their common disturbance area. In contrast, when the distance between tunnels is larger, the mutual interference effects caused by construction diminish. In this case, ground settlement is considered as the superposition of the independent settlement and deformation effects of the two tunnels. In actual engineering, with a distance of 8m between the two tunnels, the ground settlement is minimal and meets practical requirements.

• (3) The maximum settlement due to shield tunnel excavation occurs at the crown of the tunnel. Construction activities disturb the geological layers, and the displacement and deformation of the surface soil decrease with increasing depth. In the actual construction with a depth-to-diameter ratio (H/D) of 1.67, the impact on surface settlement is relatively small. If the depth is increased, it is recommended to have a depth-to-diameter ratio of 5 or higher. With greater depth, it is advantageous for the formation of a soil arch, which leads to minimal impact on surface settlement.

• (4) Excavation pressure plays a significant role in shield tunnel excavation. The excavation pressure of the shield machine should be lower than the static earth pressure of the face soil. As the excavation pressure increases, it causes less disturbance to the soil, resulting in smaller soil settlement. When the excavation pressure exceeds the static earth pressure of the face soil, it leads to increased soil disturbance and, in turn, causes the uplift of the soil in front of the face, enlarges the plastic zone of the surrounding soil, and results in later-stage settlement and deformation, increasing surface settlement. In this study, an excavation pressure of 142 kPa meets the requirements of the working conditions.

6 Conclusion

This study utilized the ABAQUS finite element software to establish a finite element calculation model for the construction of double-line tunnels for Harbin Metro Line 3. Through the analysis, it was found that the longitudinal surface settlement curves in the numerical simulation data were similar to the actual observed results, with maximum errors in longitudinal surface settlement of 12.61 and 10.56% for the left and right tunnels, respectively. It has been verified that the numerical simulation in this study is reliable. In the context of practical engineering, this study explored the influence of grouting layer thickness, depth-to-diameter ratio, spacing between the two tunnels, and excavation pressure on surface settlement, and the following conclusions were drawn:

• (1) During shield tunnel excavation, ground subsidence is inevitable. Through the analysis of grouting layer thickness, excavation pressure, depth-to-diameter ratio, and the spacing between the two tunnels, the calculated surface settlement curves closely matched the actual settlement curves observed in the field. The maximum vertical settlement was consistently located directly above the tunnel crown. To avoid excessive settlement, during the actual construction process, underground reinforcement techniques such as grouting or soil freezing can be employed to enhance the mechanical properties of the soil, thereby reducing the extent of settlement.

• (2) Increasing the grouting layer thickness appropriately can reduce the impact on the ground during shield tunnel construction, resulting in a greater reduction in surface settlement. However, it should be noted that a thicker grouting layer is not necessarily better, as an excessively thick grouting layer can affect the strength of the soil and may lead to resource wastage.

• (3) The morphology of ground settlement distribution curves during shield construction of parallel twin tunnels is significantly influenced by the proximity and distance, exhibiting a transition from a “V” shape to a “W” shape. Therefore, in shield tunneling practice, it is recommended to pay special attention to the impact of the relative spacing between twin tunnels on subsurface deformations in order to reduce the impact of shield construction on nearby buildings and the surrounding environment. In the actual construction process, it is recommended to moderately increase the spacing between the two tunnels to reduce the impact on nearby tunnels and surface settlement.

• (4) The maximum settlement caused by shield tunnel excavation occurs at the crown position of the tunnel. Construction with a shield machine extends soil disturbance to the surface, and the displacement and deformation of the soil above the tunnel decrease with decreasing soil depth. With an increase in depth, when the soil above the tunnel is thick enough to form a stable soil arch, the peak surface settlement reduces.

• (5) During shield tunnel excavation, the excavation pressure significantly exceeds the static earth pressure of the soil, causing soil disturbance and resulting in the uplift of the soil in front of the tunnel face and the expansion of the plastic zone in the surrounding soil. This ultimately leads to increased subsequent surface settlement. Therefore, during the construction process, it is necessary to ensure that the tunneling pressure matches the static soil pressure at the face to minimize surface settlement.

This study solely focuses on factors such as grouting layer thickness, excavation pressure, burial depth ratio, and tunnel spacing, without considering various factors like different soil layers, permeability, excavation speed, among others. Furthermore, due to the complexity of subsurface structures and the non-uniformity of material properties, there is a certain degree of deviation between theoretical results and actual results. Future research will incorporate a comprehensive analysis of various potential influencing factors and conduct in-depth investigations into the impact of these parameters on both TBM and shaft sinking methods.