The Influence of Construction Methods on the Stability of Tunnels and Ground Structures in the Construction of Urban Intersection Tunnels

Abstract

The construction of intersection tunnels in urban induces multiple stress redistribution in the surrounding rock, leading to engineering disasters such as instability in rock strata during excavation, disturbance of supporting structures in existing tunnels, and subsidence of ground adjacent buildings. Employing an appropriate construction method is crucial in circumventing excessive stress concentrations and large-scale rock strata subsidence, making it a key aspect of urban intersection tunnel engineering. In this paper, a numerical model for an urban intersection tunnel is developed based on an underground circular road project in a central business district. We conduct numerical simulations of the excavation processes using the full-section method, step method, and center cross diagram (CRD) method, respectively. The findings indicate that while different construction methods do not change the variation trends of surrounding rock stress and displacement, adjacent ground building deformation, and existing tunnel convergence, they affect the variation degrees. The maximum compressive and tensile stresses in the surrounding rock caused by the CRD method are the smallest, which are 3.56 MPa and 0.76 MPa, respectively. The maximum arch subsidence affected the amount, and horizontal convergence affected the amount of branch tunnel #1 caused by the CRD method are the smallest too, which respectively are 1.428 mm and 0.931 mm. The foundation subsidence and overall inclination of the ground building resulting from the three methods are identical. Then, we discuss the construction safety of the three methods and obtain the influence order on construction stability, which is as follows: full-section method > step method > CRD method. It is concluded that the CRD method is the most suitable for urban intersection tunnel engineering in terms of safety. This study could offer valuable insights for selecting construction methods in urban intersection tunnel engineering and provide a foundation for evaluating the safety and stability of tunnel construction.

1. Introduction

Amid rapid economic growth and continuous urbanization, cities are witnessing an influx of people, which leads to urban syndromes such as an over-saturated population, congested building spaces, and dwindling green areas [1,2]. Underground spaces provide a safe and comfortable supplementary environment for human life; their effective utilization can alleviate urban syndromes while generating substantial economic benefits. Since the 21st century, China has experienced accelerated underground space construction, particularly in urban tunnels, with 406 urban tunnels built across 82 cities by 2021 [3], as shown in Figure 1. However, due to the complexity of city environments and limited underground space availability, numerous urban tunnels must coexist within a single region, resulting in the emergence of a new tunnel structure form—the urban intersection tunnel. With its ability to meet actual engineering construction requirements, reduce engineering volume, and accommodate diverse terrain conditions [4,5], intersection tunnel construction is progressively increasing in cities, including Beijing, Guangdong, and Chongqing, where they have been incorporated into underground engineering projects.

Constructing urban intersection tunnels is more challenging compared to ordinary highway tunnels. On the one hand, the unique structure of urban intersection tunnels leads to repeated disturbances to the surrounding rock during construction, impacting the stability of existing and under-construction tunnels. On the other hand, these projects have distinctive geological environments, with rock strata deformation induced by construction activities posing threats to the safety of ground buildings. Consequently, ensuring the stability and safety of existing tunnels, under-construction tunnels, and ground buildings is a critical challenge in urban intersection tunnel construction.

Numerous scholars have investigated the construction stability of urban intersection tunnels, focusing on mechanical behavior, spatial position relationships, intersection angles, and excavation methods through theoretical analysis, on-site measurements, and numerical simulation methods [6,7,8]. Song and Wu [9] carried out dynamic monitoring of the cross-section of a soft rock tunnel with a small vertical distance and analyzed the stress and strain characteristics of the surrounding rock and supporting structure. Based on the Zou Magang highway tunnel project, Chen et al. [10] conducted dynamic monitoring measurements on excavating tunnel and existing tunnel, respectively, during cross-section tunnel construction, then obtained mechanical behavior characteristics of surrounding rock and supporting structure of the tunnels. By means of numerical simulation, Li et al. [11] studied variations of stress, displacement, and plastic zone of surrounding rock before and after deep-buried intersection tunnel excavation in long-span highway tunnel engineering. It was considered that the angle between axial and horizontal stress and the cross angle were the two most important factors affecting the stability of the surrounding rock in the cross-section. Nonomura et al. [12] analyzed the stress state of the surrounding rock and supporting structure of the cross-section tunnel and obtained a variation range of stress, displacement, and plastic zone of surrounding rock in cross-section before and after the new tunnel construction. Sharifzadeh et al. [13] utilized a three-dimensional numerical simulation to analyze the stability of surrounding rock in cross-sections of large-span city tunnels with soft rock in the process of tunnel excavation with the CD method and the SD method.

Building a tunnel in an urban area, especially in a Central Business District (CBD), would involve the tunnel through or near existing ground buildings. In order to research the influence of tunnel construction on the stability of adjacent structures, scholars have analyzed the deformation characteristics of adjacent structures by numerical simulation, on-site measurement, and monitoring [14,15,16]. Shi et al. [17] applied the random medium theory to analyze surface movement and deformation along the tunnel’s vertical and horizontal direction caused by the excavation of a shallow buried section in Tong You Mountain multi-arch tunnel engineering and evaluated the influence of tunnel excavation on upper surface buildings. Taking the existing independent foundation frame structure on the upper part of the tunnel as the research object, Liu [18] analyzed the formation disturbance mechanism, calculated the additional internal force of the frame structure under uneven settlement, and obtained the variation curve of the formation settlement with time through similar model testing. Then, a series of protection measures for rock strata and buildings were proposed. Mroueh and Shahrour [19] utilized three-dimensional finite element and finite difference methods to perform numerical simulation on shield tunnel construction and surrounding buildings, studied the influence of weight and stiffness of buildings on surface displacement and believed that subsidence would occur in the area where buildings exist. Richard et al. [20] proposed a rolling beam model to analyze the deformation of ground three-dimension frame structures caused by deep excavation and evaluated damage to the buildings by considering the types of buildings and positions of doors and windows, beams, and columns.

Existing studies and fieldwork indicate that appropriate construction methods can mitigate stress concentrations and rock strata deformation, making them essential for addressing these challenges [21,22]. The New Austrian Tunneling Method (NATM), known for its minimal construction disturbance and rapid closure, is frequently employed in tunnel projects characterized by complex geological and construction conditions [23]. NATM encompasses various excavation techniques, including the full-section method, step method, and center cross diagram (CRD) method, each of which exerts a varying degree of impact on the surrounding rock. Consequently, investigating the influence of construction methods on the stability of surrounding rock can serve as a foundation for selecting the appropriate construction method. In this study, we investigate an underground circular road engineering project located in the CBD of Chongqing, China. Our objective is to assess the impact of different construction methods on the safety and stability of urban intersection tunnels through numerical simulations. Subsequently, we validate these numerical simulation results by conducting field monitoring measurements. Lastly, we discuss the effects of the three construction methods on the safety and stability of urban intersection tunnel construction. Our study not only confirms the minimal disturbance to the surrounding rock achieved through NATM but also explores the variations in stress and deformation within the rock strata under different construction methods. Ultimately, we derive the most suitable construction method for the engineering project. These findings offer valuable insights for the selection of construction methods for urban intersection tunnels and present a practical approach for evaluating such methods.

This paper comprises six interrelated sections, each providing a comprehensive perspective on the research. Section 1 offers an overview of the research’s significance and its current status. Section 2 describes the NATM theory and common urban tunnel construction methods. The project overview of the urban intersection tunnel is presented in Section 3. Information on the numerical model and simulation results are shown in Section 4. Section 5 provides an assessment of the three construction methods, along with a description of the field monitoring process. Finally, Section 6 wraps up our study with a comprehensive conclusion.

2. Urban Tunnel Construction Methods

2.1. The New Austrian Tunneling Method

In 1948, the Austrian scholar L.V. Rabcewicz proposed a novel construction method based on extensive experience in tunnel engineering and the principles of rock mass mechanics. This innovative approach centered around utilizing a combination of bolts and shotcrete as the primary means of support, with a crucial emphasis on harnessing the inherent bearing capacity of the surrounding rock itself. After undergoing rigorous engineering practice and theoretical research, this method was patented and officially christened as the NATM in 1963 [24].

Diverging from the traditional theories of tunnel design and construction, NATM theory treated rock mass as a continuous medium and placed a significant focus on factors such as viscosity, elasticity, plasticity, and other physical and mechanical properties of the rock mass. This theoretical framework effectively harnessed the time-dependent characteristics of stress redistribution within excavated rock strata. It also closely integrated a thin-wall, flexible support structure with the surrounding rock, thereby mobilizing the inherent bearing capacity of the natural rock to achieve the desired level of surrounding rock stability [25]. The NATM, which emphasizes the self-supporting ability of the surrounding rock and primarily employs shotcrete and bolts for support, offers advantages such as “minimal disturbance, early anchoring and shotcreting, frequent measurements, and rapid closure” [26]. As a result of these benefits, the NATM has found widespread application in tunnel engineering projects dealing with various complex strata.

2.2. Construction Methods

Within primary construction methods of NATM, the full-section method, step method, and CRD method are considered essential for urban intersection tunnel engineering.

The full-section method involves one-time blasting to form the tunnel design outline, followed by support and lining, as illustrated in Figure 2a. The construction sequence of this method is straightforward and can be divided into three main steps. Firstly, explosives are loaded, and fuses are connected after drilling the entire section using a jumbo drill. Secondly, explosives are detonated after the jumbo drill exits, blasting the tunnel section into shape. Lastly, slag is removed, and section support is added before proceeding to the next cycle. The full-section method simplifies construction organization and management due to fewer working procedures and reduced interference between them. Additionally, it promotes surrounding rock stability since the tunnel section is excavated at one time. However, the method demands a high rock strata grade due to the large excavation section. Field experience [27,28] indicates that the full-section method is suitable for rock strata with grades I to III. Furthermore, the full-section method requires ample working space and relies on large machinery. Consequently, it is best suited for urban tunnel engineering projects with high surrounding rock grades and favorable construction conditions, such as the metro tunnel project in Xi′an [29] and the Georgia No. 3 tunnel [30].

The step method, a widely applicable construction method within the NATM, divides the tunnel section into upper and lower steps, with the two steps staggered at a certain distance. The lower step is excavated after the upper step has been excavated to a certain distance and supported. The upper step and lower step can be excavated simultaneously on separate working faces, as shown in Figure 2b. Based on step length, the method is divided into three types: long step, short step, and super short step. The step method has rapid construction progress due to the two excavation faces not interfering with each other. Additionally, it has a small excavation area that is conducive to the stability of the palm surface. Consequently, the step method is suitable for weak rock strata with grades I to III and developed joints. However, the construction of two excavation faces increases disturbance to surrounding rock, negatively impacting surrounding rock stability. Thus, the step method can be used in urban tunnel engineering projects with weak surrounding rock and short construction periods, such as the Eling tunnel [31] and the Feng’an tunnel of the Zihui freeway project [32].

The CRD method, suitable for soft surrounding rock or long-span tunnels, involves partially excavating one side of the tunnel first, then partially excavating the other side after completing the median lamella and diaphragm, as shown in Figure 2c. The detailed construction process comprises four steps. In the first step, pilot-tunnel ➀ is excavated after surrounding rock is strengthened and initial support is completed. In the second step, pilot-tunnel ➁ is excavated, and corresponding initial support is completed. In the third step, after surrounding rock strengthening, pilot-tunnel ➂ is excavated, and initial support is completed. In the fourth step, pilot-tunnel ➃ is excavated, and corresponding initial support is completed. The CRD method adheres to the principle of “small blocks, short steps, multiple cycles and quick closure”. The support system of span reduction, temporary invert, and timely closed ring enhances the rigidity of the tunnel structure and effectively controls the deformation of the palm face. As a result, it is particularly suitable for double-line or multi-line tunnels in rock strata with grades V and VI. However, the CRD method has a complex construction process, high costs, and demanding construction organization requirements. Therefore, the CRD method can be used in urban tunnel engineering projects with weak surrounding rock and multiple construction influencing factors, such as the Re Shuitang No. 3 tunnel [33] and the Xi’an Metro Line 6 section tunnel [34].

3. Project Overview

The CBD, established in 1997 in Chongqing, China, faced several traffic challenges, including insufficient parking spaces, scattered parking garages, numerous garage entrances and exits, and difficult driving conditions. To address these issues and ensure smooth traffic flow, an underground circular roads project was constructed in the CBD, utilizing the principles of utilization, transformation, and local new construction. The project included a 2580 m long excavation tunnel, comprised of one main tunnel and five branch tunnels, several of which intersected. The mileage segment K1+000~K1+160 of the main tunnel and mileage segment K0+380~K0+420 of the branch tunnel 1# intersected vertically at 40 m underground, as shown in Figure 3. The hydrogeological conditions of the project were simple, and the groundwater was mainly bedrock fissure water. The surrounding rock of the intersection tunnel was mainly sandstone with two groups of fractures, classified as soft rock of grade IV. The surface plain fill was 4 m thick; the physical and mechanical parameters of each soil layer are shown in

Table 1. Stratigraphic physical and mechanical parameters.

	Internal Friction Angle (º)	Cohesion (kPa)	Volumetric Weight (kN/m3)	Elasticity Modulus (MPa)	Poisson’s Ratio
Plain fill	11.5	23.3	18.5	1.7	0.16
Sandstone	32.9	744.0	25.9	3500.0	0.23

. The main tunnel was 9.5 m in width and 5.5 m in height, the branch tunnel #1 was 7.0 m in width and 5.5 m in height, and the cross-sections of both were semicircular arch. The main tunnel was excavated after completion of the branch tunnel #1. There was an affected ground building with a height of 35 m and depth of 10 m, 7 floors on the ground, and 2 floors underground north of the intersection section tunnel, and the minimum distance between it and the ground projection of the intersecting section was 5 m.

In constructing the main tunnel, the safety of the tunnel itself, the disturbance effect on branch tunnel #1, and the stability of the nearby ground building had to be taken into account. The choice of tunnel construction method was crucial in ensuring the safety and stability of the tunnel structure and the ground building. To provide a basis for selecting a suitable construction method, a numerical simulation work was conducted to analyze the impact of different construction methods on the stability of tunnel construction. The simulation work would consider factors such as rock conditions, tunnel dimensions, and proximity to ground structures, ultimately providing valuable insights into the most appropriate construction method for the intersection tunnel engineering project. By selecting the optimal construction method, the project can ensure the safety of the tunnel structure, minimize disturbance to adjacent tunnels, and maintain the stability of nearby ground buildings.

4. Numerical Simulation

4.1. Structure Simplification

To account for the stiffness of the affected ground building in the numerical simulation, the building was simplified using Burland’s approach [35]. It was considered as a thin beam composed of plates of 7 layers and filling materials, with each plate having the same thickness and width, as shown in Figure 4. The distance between the plates represented the height between floors, while the floor thickness was considered negligible compared to the building’s height.

Assuming that each floor is stressed per unit size and the distance from the neutral axis of the beam to the bottom plate is λH,

$$\lambda H=\frac{{\displaystyle {\sum }_{i=0}^7{A}_i}{h}_i}{{\displaystyle {\sum }_{i=0}^7{A}_i}}$$

(1)

where i is the number of floors, the bottom floor is 0, A_i is the floor area, and h_i is the height of each floor to the bottom floor.

So the inertia moment I of the beam is,

$$I={\displaystyle {\sum }_{i=0}^7{A}_i}{\left(h_i-\lambda H\right)}^2$$

(2)

The stiffness W of the building is,

$$W=EI$$

(3)

where E is the equivalent elastic modulus of the building.

In addition, a surface uniform load of 210 kN/m² was applied on the surface of the affected ground building to replace the superstructure load of the simplified thin beam based on relevant survey data.

4.2. Numerical Model

A numerical model of the urban intersection tunnel was constructed by ANSYS simulation software, as shown in Figure 5. The model had dimensions of 240 m (length), 160 m (width), and 70 m (height), with the upper boundary representing the ground surface. The main tunnel and branch tunnel #1 intersected 40 m underground in a “T” shape. The length of the main tunnel was 160 m, and the width and height of the cross-section were 9.5 m and 5.5 m, respectively. The length of the branch tunnel 1# was 40 m, and the width and height of the cross-section were 7.0 m and 5.5 m, respectively. The affected ground building was simplified into a thin beam structure with a minimum distance of 5 m from the ground projection of the main tunnel. The distances between the affected building and other boundaries were set at 27 m, 30 m, and 35 m, respectively. The main tunnel was situated 40 m away from the right boundary of the model, while the distance between the branch tunnel 1# and the upper boundary was 45 m. Within the numerical model, the X-axis was perpendicular to the centerline of the main tunnel and pointed to the branch tunnel 1#, the Y-axis was opposite to gravity, and the Z-axis was aligned with the centerline of the main tunnel. Displacement constraints were applied in the Z direction to the four side interfaces of the model and in the negative Y direction to the bottom surface of the model.

The numerical model was composed of surrounding rock, ground building, and support materials (the anchor bolts and concrete material). It utilized three types of elements: solid92 element (ground building and surrounding rock), link8 element (anchor bolts), and shell93 element (concrete material). Given that the plain fill had a thickness of only 4 m, which was significantly less than the depth of the tunnels, and the foundation depth of the ground building exceeded 4 m, the plain fill was excluded from the model to simplify calculations. Consequently, the surrounding rock was modeled as sandstone with a cohesion of 744.0 kPa and an elastic modulus of 3.5 GPa. It was considered isotropic, homogeneous, and a continuous elastoplastic medium. Both the surrounding rock and the ground building were represented using the solid92 element. Additionally, the complex support process and type were simplified as the anchor bolt support and concrete support. The concrete support material, set as C30 concrete, was configured as a shell structure with a volume weight of 25 kN/m³, an elasticity modulus of 30 GPa, and a Poisson’s ratio of 0.20. Anchor bolts with a volume weight of 79 kN/m³, an elasticity modulus of 200 GPa, and a Poisson’s ratio of 0.20 were embedded within the model.

Table 2. Elements and parameters of the materials.

	Element Type	Internal Friction Angle (º)	Cohesion (kPa)	Volumetric Weight (kN/m3)	Elasticity Modulus (GPa)	Poisson’s Ratio
Surrounding rock	Solid92	32.9	744.0	25.9	3.5	0.23
Ground building	Solid92	-	-	25.0	33.0	0.20
Anchor bolt	Link8	-	-	79.0	200.0	0.20
C30 concrete	Shell93	-	-	25.0	30.0	0.20

shows the detailed physical and mechanical parameters of the materials. The Drucker-Prager yield criterion was adopted for the constitutive relationship of materials in the model. The volumes in the model were glued together by the “VGLUE” command. The model was meshed manually due to the complex lines of intersection tunnels. As a result, the model generated a total of 32,479 elements and 44,920 nodes.

4.3. Results Analysis

The construction of an urban intersection tunnel should not only ensure the stability of the existing tunnel but also guarantee the safety of the adjacent ground building. The full-section method, step method, and CRD method were used to excavate the tunnels in the intersection section engineering, respectively, and the construction characteristics caused by the three methods were analyzed from the displacement and stress of rock strata, the convergence of existing tunnel and the deformation of ground adjacent building.

4.3.1. Displacement and Stress Characteristics of Rock Strata

The construction methods did not change the vertical displacement variation trend of rock strata but affected the deformation degree. After the main tunnel was excavated, the rock strata above the tunnel settled down while the rock strata within 22 m below the tunnel uplifted upward. The deformations expanded outwards in a “flame” shape from the centerline. Specifically, the closer to the centerline, the higher the deformation value, as shown in Figure 6. The step method had the least influence, causing the maximum vault settlement amount of the main tunnel to be 2.717 mm and the maximum floor uplift amount was 2.254 mm. The full-section method had a great influence, resulting in the maximum vault settlement amount of the main tunnel being 2.754 mm and the maximum floor uplift amount being 2.823 mm. The CRD method had the greatest influence, causing the maximum vault settlement amount of the main tunnel to be 2.798 mm and the maximum floor uplift amount of 2.848 mm, as shown in

Table 3. Deformation amount of rock strata.

Construction Methods	The Maximum Vault Settlement Amount (mm)	The Maximum Floor Uplift Amount (mm)	The Maximum Compressive Stress (MPa)	The Maximum Tensile Stress (MPa)
Full-section method	2.754	2.823	13.80	1.31
Step method	2.717	2.254	11.90	1.14
CRD method	2.798	2.848	3.56	0.76

.

The construction methods did not affect the stress variation trend of rock strata but changed the stress degree. After the main tunnel was excavated, the rock strata near the vault were in a compression state, while the rock strata near the floor were in a tension state. The influence order of the three methods on stress degree was full-section method > step method > CRD method, as shown in Figure 7 and Figure 8. In detail, with the full-section method, the maximum compressive stress and tensile stress of the main tunnel rock strata were 13.80 MPa and 1.31 MPa, respectively. With the step method, the maximum compressive stress and tensile stress were 11.90 MPa and 1.14 MPa, respectively. With the CRD method, the maximum compressive stress and tensile stress were 3.56 MPa and 0.76 MPa, respectively, as shown in Table 3.

4.3.2. Deformation Characteristics of the GROUND Building

The construction methods did not alter the deformation variation trend of the ground building but influenced the deformation degree, as shown in Figure 9. The main tunnel excavation caused the rock strata to sink, which spread to the ground surface and resulted in the foundation on the right side of the ground building to settle. The settlement amount decreased gradually from the upper right side of the building to the left in the “water wave” type, and the upper right part had the largest value, While the rest of the foundation moved up. The building tilts slightly towards the right side (direction of the main tunnel). The influence order of the three methods on the deformation degree was CRD method > full-section method > step method. In detail, with the full-section method, the maximum settlement amount and the maximum gradient were 0.523 mm and 0.00338‰, respectively. With the step method, the maximum settlement amount and the maximum gradient were 0.507 mm and 0.00323‰, respectively. With the CRD method, the maximum settlement amount and the maximum gradient were 0.556 mm and 0.00354‰, respectively, as shown in

Table 4. Deformation amount of the ground building.

Construction Methods	The Maximum Settlement Amount (Mm)	The Maximum Gradient Value (‰)
Full-section method	0.523	0.00338
Step method	0.507	0.00323
CRD method	0.556	0.00354

.

4.3.3. Deformation Characteristics of the Branch Tunnel #1

The construction methods did not alter the vault subsidence variation of branch tunnel #1 but altered the deformation degree, as shown in Figure 10a–c. The vault area of branch tunnel #1 moved down while the area from waist to floor moved up after the main tunnel was excavated. The main tunnel excavation intensified vault subsidence of branch tunnel #1, generating three stages, as shown in Figure 10d. In the first stage, the rapid reduction stage, the vault subsidence affected amount of branch tunnel #1 decreased rapidly with the increase of distance within the range of 0~10 m from the intersecting section, and the affected amount was the largest at the intersecting section. In the second stage, the slow reduction stage, the vault subsidence affected amount decreased slowly with the increase of distance within the range of 10~30 m from the intersecting section. In the third stage, the stable stage, the vault subsidence affected amount remained stable at 0.215 mm within the range of 30~40 m from the intersecting section. Within the range of 0~16 m from the intersecting section, the influence order of the three methods on the vault subsidence was: full-section method > step method > CRD method. In detail, the maximum vault subsidence affected amounts were 1.881, 1.588, and 1.428 mm, respectively, under the full-section method, step method, and CRD method.

The construction methods did not alter the horizontal convergence variation trend of branch tunnel #1 but altered the deformation degree, as shown in Figure 11a–c. The main tunnel excavation intensified the horizontal convergence, presenting four stages, as shown in Figure 11d. In the first stage, the dramatic increase stage, horizontal convergence affected the amount of branch tunnel #1 increased rapidly with the increase of distance within the range of 0~3 m from the intersecting section. In the second stage, the dramatic decrease stage, the variation trend showed an inflection point at a 3 m distance from the intersection section. Horizontal convergence affected amount decreased rapidly with the increase of distance within the range of 3~10 m from the intersecting section. In the third stage, the slow decrease stage, the horizontal convergence affected amount decreased slowly with the increase of distance within the range of 10~30 m from the intersecting section. In the fourth stage, the stable stage, the horizontal convergence affected amount remained stable at 0.0571 mm within the range of 30~40 m from the intersecting section. Within the range of 0~16 m from the intersecting section, the influence order of the three methods on horizontal convergence was: full-section method > step method > CRD method. In detail, the maximum horizontal convergence affected amounts were 1.264, 1.157, and 0.931 mm, respectively, under the full-section method, step method, and CRD method.

5. Discussion

In the construction of urban intersection tunnels, the choice of appropriate excavation methods is crucial for ensuring the safety and stability of the excavating tunnel, the existing tunnel, and ground buildings. Thus, it was necessary to evaluate the construction methods from the perspective of engineering safety and stability.

5.1. Evaluation Criterion

Excavating tunnel safety could be evaluated by comparing the extreme stress of surrounding rock with the safety strength of the lining materials, such as concrete C30. This comparison would determine the safety degree of the excavating tunnel under different construction methods. The safety strength of concrete C30 is shown in

Table 5. Evaluation criteria for the main tunnel safety.

Lining Materials	Safe Compressive Strength (MPa)	Safe Tensile Strength (MPa)
Concrete C30	14.30	1.43

. For instance, when the maximum compressive stress of the surrounding rock exceeds 14.30 MPa, the excavating tunnel would be crushed, resulting in vault collapse.

In the matter of structural stability of the existing tunnel, which followed the guidelines of China’s national standard, “GB50911-2013 Code for monitoring measurement of urban rail transit engineering”, as shown in

Table 6. Deformation allowable value of existing tunnel during construction.

Types	Accumulated Value (mm)
Settlement of tunnel structure	3~10
Horizontal displacement of tunnel structure	3~5

. Considering the importance of the underground circular roads project, the stability of branch tunnel 1# was evaluated by comparing whether the vault settlement and horizontal convergence of the tunnel exceeded 3 mm. Once the deformation of the existing tunnel (either vault settlement or horizontal convergence) exceeded 3 mm, the tunnel was in a state of instability, and collapse occurred.

In addition, the safety of the affected ground building followed China’s national standard, “GB50007-2011 Code for design of building foundation”, and analogical engineering subway project in Beijing [36], as shown in

Table 7. Deformation allowable value of affected ground building.

Height of Building (m)	Allowable Gradient Value	Allowable Foundation Settlement Gradient (mm)	Safety Coefficient
24 < H ≤ 60	0.003	20	0.5

. It could be evaluated by comparing the foundation settlement value and overall gradient value with the allowable deformation values of 10 mm and 0.0015, respectively. If the deformation of the ground building reached any of these indicators, it signified an extremely unstable condition. Disasters would occur without reinforcement measures.

Lastly, we established a comprehensive safety evaluation criterion encompassing surrounding rock stress, lining displacement, and ground building deformation for underground circular road engineering, as summarized in

Table 8. Safety allowable criterion containing surrounding rock stress, lining displacement, and ground building deformation for the underground circular roads engineering.

	Safe Compressive Strength (MPa)	Safe Tensile Strength (MPa)	Allowable Gradient Value	Allowable Foundation Settlement (mm)	Allowable Vault Settlement (mm)	Allowable Horizontal Convergence (mm)
Surrounding rock	14.30	1.43	-	-	-
Branch tunnel 1#	-	-	-	-	3	3
Ground building	-	-	0.0015	10	-	-

.

5.2. Construction Methods Analysis

The surrounding rock stress and branch tunnel 1# deformation and ground building deformation under the three construction methods are shown in Figure 12. By analyzing surrounding rock stress, the CRD method was found to be more suitable for urban intersection tunnel construction, as it resulted in lower maximum compressive and tensile stresses on surrounding rock compared to the full-section method and step method. Although all three methods were within the allowable criterion of surrounding rock stress, the CRD method offered a higher degree of safety. By analyzing ground building deformation, all three methods were considered suitable for urban intersection tunnel construction, as the overall gradient values and maximum foundation settlement values were within the allowable criterion for each method. The differences in deformation amounts among the methods were minimal. By analyzing branch tunnel 1# deformation, the CRD method was found to be more suitable for urban intersection tunnel construction, as it resulted in the smallest maximum vault subsidence and horizontal convergence values. Although none of the methods exceeded the safety allowable value of 3 mm, the CRD method provided a higher degree of stability for the branch tunnel 1#.

The comprehensive analysis showed that while all three methods (full-section method, step method, and CRD method) did not negatively affect the safety of the urban intersection tunnel engineering, the CRD excavation method had less influence on the structural stability of both the excavating tunnel and the existing tunnel. As a result, the CRD method was considered more suitable for urban intersection tunnel engineering.

5.3. Field Monitoring

The CRD method, as indicated by the comprehensive analysis, provided greater construction stability for urban intersection tunnels and was subsequently utilized in urban intersection tunnel engineering. The deformation of tunnel sections was monitored on-site during construction. Section K0+415 of branch tunnel 1# was selected for monitoring, as it was 5 meters away from the intersecting section and could represent the maximum deformation of branch tunnel 1#. The vault subsidence and horizontal convergence of section K0+415 within 30 days were monitored and displayed in Figure 13.

The cumulative vault subsidence amount and cumulative horizontal convergence value both displayed a similar variation trend, increasing gradually with time. The variation curves were divided into three stages. In the rapid deformation stage, the cumulative deformation amounts increased rapidly from the 1st day to the 12th day, with an average daily vault subsidence amount of 0.19 mm and an average daily horizontal convergence value of 0.15 mm. In the slow deformation stage, the cumulative deformation amounts increased slowly from the 13th day to the 25th day, with both the average daily vault subsidence amount and average daily horizontal convergence value being 0.03 mm. In the stable stage, the cumulative deformation amounts remained stable from the 26th day to the 30th day, with both the daily vault subsidence amount and daily horizontal convergence value being close to 0 mm.

Based on these monitoring observations, it was determined that the maximum vault subsidence amount and maximum horizontal convergence value of section K0+415 were 2.480 mm and 2.110 mm. The numerical simulation showed that the corresponding deformation values were 2.361 mm and 2.157 mm, respectively, as shown in Figure 14. The simulated results were in close agreement with the monitoring data. The error between simulation results and monitoring data was less than 5%, as shown in

Table 9. Results comparison of numerical simulation and field monitoring.

	The Maximum Vault Settlement	The Maximum Horizontal Convergence
Field monitoring	2.480 mm	2.110 mm
Numerical simulation	2.361 mm	2.157 mm
Results error	4.7%	2.2%

. This consistency validated the accuracy of the numerical simulation.

6. Conclusions

In this study, we established a numerical model of an urban intersection tunnel based on an underground circular road project and conducted simulations of the excavation process using three construction methods (full-section method, step method, and CRD method). We analyzed the stability of tunnels and ground building and verified the simulation results with field monitoring data. It was essential to clarify that the foundation of our study was the underground circular roads project, and all numerical simulation parameters were tailored to this specific project. Consequently, the conclusions were primarily applicable to underground circular road engineering. While these research findings might not be universally applicable, they served as valuable references and a foundation for similar projects. To enhance the research comprehensiveness, it was suggested to conduct research on the construction methods of urban intersection tunnels with diverse geological conditions, varying burial depths, and different intersection angles. The main conclusions were as follows:

(1)

The CRD method proved highly effective in averting engineering disasters and demonstrated superior suitability for urban intersection tunnel projects from a safety perspective.

(2)

Construction methods did not alter the stress variation trend of surrounding rock but affected the variation degree. The influence order on the stress degree was full-section method > step method > CRD method. Specifically, the maximum compressive stress was 13.80 MPa, 11.90 MPa, and 3.56 MPa, and the maximum tensile stress was 1.31 MPa, 1.14 MPa, and 0.76 MPa, respectively, caused by the full-section method, step method, and CRD method.

(3)

Construction methods did not alter the deformation variation trend of the ground building but affected the deformation degree. The influence order on the deformation degree was CRD method > full-section method > step method. In detail, the overall tilt values were 0.00338‰, 0.00323‰, and 0.00354‰, and the maximum foundation settlement values were 0.532 mm, 0.507 mm, and 0.556 mm, respectively, caused by the full-section method, step method, and CRD method.

(4)

Construction methods did not change the deformation variation trend of branch tunnel #1 but affected the deformation degree. The influence order on the deformation was full-section method > step method > CRD method. Specifically, the maximum vault subsidence-affected values were 1.881 mm, 1.588 mm, and 1.428 mm, and the maximum horizontal convergence-affected values were 1.264 mm, 1.157 mm, and 0.931 mm, respectively, caused by the full-section method, step method, and CRD method.