Numerical Simulation on Deformation and Damage Mechanism of Existing Underground Structures Induced by Adjacent Construction of Super-Large-Diameter Tunnels

Abstract

The development of urban underground spaces has led to an increasing number of projects involving super-large-diameter shield tunnels, making research on their impact on existing structures particularly significant. This paper investigated the numerical simulation on deformation and damage mechanism of existing underground structures induced by adjacent construction of super-large-diameter tunnels. A 3D finite element model using ABAQUS (version 2022) software incorporating the Concrete Damaged Plasticity (CDP) constitutive model was established, and this paper was used to systematically analyze the deformation, internal force response, and damage evolution of existing tunnels. The results showed the following: (1) The double-line tunnel excavation intensified settlement superposition, increasing the maximum settlement from −19.70 mm (single-line) to −24.51 mm (double-line) and transforming the settlement trough from a V shape to a W shape. (2) The vertical bending moment evolved from a single peak to double peaks being the dominant loading mode, with the maximum horizontal moment only about 1/8 of the vertical value. (3) During the construction, the peak tensile stress at the tunnel bottom reached 2.655 MPa, exceeding the C50 concrete tensile strength, but later decreased to 2.097 MPa. Damage was primarily caused by bending-induced tension. (4) Tunnel damage was triggered by the historical peak stress and accumulated irreversibly, resulting in a final state of low-stress and high-damage, with a maximum tensile damage of 92.4%. This research can provide a theoretical basis for safety control in similar adjacent engineering projects.

1. Introduction

With the continuous advancement of urban underground space development, projects involving new tunnels undercrossing existing tunnels are increasing [1]. The shield method is widely adopted due to its relatively minor disturbance to the surrounding ground. However, the shield tunneling construction under the tunnel will disrupt the original stress balance, cause ground settlement, and result in the existing tunnel bearing additional loads and undergoing extra deformation, posing a threat to its normal operation and structural safety [2,3]. Reducing and controlling the mechanical response of existing tunnels to ensure their safety is a key objective in shield undercrossing construction control [4]. The issue is particularly pronounced in complex projects involving double-line undercrossing by super-large-diameter shields (with a diameter of 14.5 m); the construction disturbance has a wider range and higher intensity, making the influence mechanism and damage evolution on existing tunnels more complex. Therefore, systematically studying the deformation patterns, internal force responses, and damage mechanisms of super-large-diameter shield machines passing under existing tunnels has significant theoretical and practical significance.

Numerous scholars have conducted extensive research on the influence of adjacent construction on existing tunnels, employing methods such as theoretical analysis, model testing, and numerical simulation. Shield undercrossing construction disturbs the surrounding soil and redistributes its stress, causing soil displacement between the new and existing tunnels, which in turn drives the deformation of the existing tunnel. This process is essentially the result of the interaction between the shield, soil, and existing tunnel [5]. In terms of theoretical analysis, Han et al. [6] proposed a three-dimensional failure mechanism for the limit support pressure of tunnel faces in multi-layered cohesive–frictional soils based on the limit analysis method. In terms of numerical simulation, Lu et al. [7] established a full-process simulation method for shield construction that considers shield–soil interactions, lining–grouting and layer–soil coupling, as well as synchronous grout hardening and pressure dissipation. Peng [8] analyzed the deformation patterns of existing double-line tunnels and the ground surface under different grouting pressures and face support forces through numerical simulation. Liu et al. [9] validated model reliability using field measurements and further studied the mechanical response of existing tunnels during the crossing process of new tunnels and the effectiveness of micro-deformation control measures. Yang [10] employed a displacement-controlled finite element model to reveal the influence mechanisms of factors such as the entire shield undercrossing process, ground loss rate, and tunnel spatial positions on existing tunnels in soft soil strata. Lin et al. [11] found that the deformation mode of existing tunnels gradually evolves with shield advancement, with torsional deformation being closely related to the position of the cutterhead. Recent international studies have further refined the simulation and assessment of tunnel undercrossing effects. For instance, refined numerical analyses considering construction joints have been conducted to evaluate the impact of twin-tunnel excavation on operational high-speed railway tunnels [12], while empirical models based on extensive monitoring data have been proposed to predict the settlement of existing tunnels [13].

Regarding damage in existing tunnels, scholars both domestically and internationally have also conducted a series of studies. The Concrete Damaged Plasticity (CDP) model has been effectively applied to investigate reinforcement mechanisms for defective tunnel linings [14]. Nie [15] analyzed the mechanism of segment cracking in adjacent shield tunnels induced by foundation pit excavation and the laws of crack propagation through finite element simulation. Zhang et al. [16] investigated the influence of key segment positions on the mechanical performance of shield tunnel linings under seismic loads. Wei [17] established a damage model for adjacent tunnel linings under shield construction disturbance, exploring the effects of construction depth, concrete grade, and lining thickness on damage development. In terms of experimental research, Zhang et al. [18] studied the failure characteristics and crack propagation of key segments in super-large cross-section shield tunnels through full-scale tests. Liu et al. [19] analyzed the bearing capacity, deformation behavior, and crack evolution of linings under bias pressure using experiments and XFEM simulation. Chen et al. [20] developed a refined model to study the structural behavior of segmental rings and the interrelationships among concrete cracks, reinforcement stress, and bolt stress. Sharghi et al. [21] evaluated the impact of joint dislocation and incomplete contact on lining damage parameters through numerical analysis. Shi et al. [22] investigated the influence of longitudinal crack parameters on the bearing capacity and crack propagation laws of segments based on ABAQUS and XFEM. Cao et al. [23] explored the influence of different crack locations on the failure modes of segment linings through scaled model tests. Dong [24], starting from structural and geological factors, proposed a crack distribution pattern for the failure of subway tunnel linings. Xu et al. [25] combined numerical simulation and model tests to reveal the mechanical response of linings containing non-penetrating longitudinal cracks. Moreover, the concepts of resilience assessment and intelligent monitoring have been introduced to evaluate and enhance the safety of underground structures under disturbance [26,27]. Concurrently, advanced technologies, such as intelligent geological identification, structural health monitoring systems, and digital twin-based predictive frameworks, are increasingly being explored to address complex challenges in tunnel engineering [28,29,30].

Based on the studies mentioned above, scholars both domestically and internationally have achieved certain results in research on the deformation and lining cracks of existing tunnels induced by conventional diameter shield undercrossing. The deformation patterns and disturbance mechanisms of existing tunnels induced by super-large-diameter shield undercrossing remain unclear [31,32,33]. Moreover, comprehensive research on the deformation, internal force response, and damage evolution of existing tunnels under double-line super-large-diameter shield undercrossing remains relatively limited. The study established a numerical model of double-line super-large-diameter shield tunnels undercrossing existing tunnels, incorporates the CDP constitutive model, and systematically investigates the deformation patterns, internal force responses, and damage mechanisms of existing tunnels during super-large-diameter shield undercrossing construction. This research can provide a theoretical basis for the safety control of similar engineering projects.

2. Numerical Simulation and Constitutive Model

2.1. Numerical Model

Based on the project, a 3D finite element model was established using ABAQUS (version 2022) software to simulate double-line super-large-diameter shield tunnels undercrossing existing tunnels. A schematic diagram of the model is shown in Figure 1. The model’s dimensions are 150 m (length) × 90 m (width) × 70 m (height). The existing tunnels, OldT1 and OldT2, have a diameter of 6.5 m, a burial depth of 22.5 m, and a transverse spacing of 5.5 m. The new tunnels, NewT1 and NewT2, have a diameter D = 14.5 m, an overburden depth of 41 m, and a spacing of 15 m. The vertical separation between the new and existing tunnels is 12 m. The model’s length is 150 m, which is 10.3 times the tunnel excavation diameter. The distance from both sides of the new tunnels to the model boundary is 53 m (>3.5D), satisfying the requirement for the soil disturbance range induced by shield tunneling and allowing the boundary effects to be neglected [34].

During the simulation of tunnel construction, the new tunnel NewT1 was excavated first, and after its breakthrough, the new tunnel NewT2 began excavation, with the tunneling direction being the positive X-axis. The new tunnels were advanced for a total of 30 rings, with a ring width of 3 m. The ground strata were simplified as a homogeneous stratum and modeled using a silty clay formation. The upper 35 m corresponds to the zone where the existing tunnels are located, while the lower 35 m is the zone of the new tunnels.

The soil was simulated using the M-C constitutive model. The geometric dimensions of the new tunnel lining and the existing tunnels were adopted from the actual project values, and key parameters in the model were determined with reference to the relevant literature [35,36,37]. The soil, lining, and shield were modeled with C3D8R elements, resulting in a total of approximately 51,900 mesh elements in the model. The shield is generally simulated using shell or solid elements, and its self-weight was accounted for by increasing the density of the shell [38]. The interaction between the shield and the soil was represented by defining contact relationships between the outer surface of the shield shell and the inner surface of the soil after tunnel excavation. The contact formulation reasonably describes the normal and tangential interactions between the shield shell and the soil [39]. Both the new tunnels and the shield were simulated using a linear elastic constitutive model to characterize their mechanical behavior. The model’s parameters are listed in

Table 1. Model material parameters.

Material Name	Thickness (m)	Mass Density (kg·m−3)	E (MPa)	Poisson’ s Ratio	Cohesion (kPa)	Internal Friction Angle (°)
Silty clay	70	1900	140	0.23	25.4	22.4
New tunnel lining (C50)	0.55	2500	33,500	0.2	-	-
Shield shell	0.3	7850	210,000	0.25	-	-
Existing tunnel lining	0.5	2400	34,500	0.2	-	-

. It is acknowledged that the ground in this study is simplified as a homogeneous silty clay layer governed by the M-C criterion. This simplification is adopted to prioritize the investigation of the fundamental interaction mechanism between the super-large-diameter shield and the existing tunnel, based on the relatively uniform soil profile indicated in the project’s geotechnical report.

2.2. CDP Constitutive Model

The constitutive relationship of a material describes its stress–strain behavior under loading, predicts its performance under various loading conditions, and influences the accuracy of numerical simulation results. Compared to the traditional linear elastic model, the Concrete Damaged Plasticity model (hereinafter referred to as the CDP model) can more realistically simulate the nonlinear behavior of concrete under complex stress states. The CDP model quantifies the irreversible degradation of concrete stiffness through two key internal variables: the tensile damage factor (DAMAGET) and the compressive damage factor (DAMAGEC). The damage factor ranges from 0 (undamaged) to 1 (completely failed), intuitively characterizing the extent and location of material damage. The linear elastic model lacks this capability, as its stiffness remains constant.

Therefore, in order to accurately simulate the deformation and damage of the existing tunnel under the construction disturbance of the super-large-diameter shield tunneling, the constitutive relationship of the existing tunnel concrete segments in this study adopted the CDP model. The stress–strain relationship of the CDP model is stipulated in accordance with the Code for Design of Concrete Structures (GB 50010-2010) [40], as shown in Figure 2.

The elastic stage is defined by the Young’s modulus and Poisson’s ratio of the concrete, while the plastic stage is determined based on the concrete stress–strain relationship provided by the design code. The concrete grade for the segments of the existing tunnel is C50. The CDP model is adopted, with its parameters listed in

Table 2. C50 concrete CDP model parameter table.

Parameter	Unit	Value	Parameter	Unit	Value
Mass density	kg/m3	2400	Peak compressive strength	MPa	36.88
Elastic modulus	GPa	34.5	Peak tensile strength	MPa	2.64
Poisson’s ratio	-	0.2	Peak compressive strain	-	1.85 × 10−3
Dilation angle	-	38	Peak tensile strain	-	1.20 × 10−4
Invariant stress ratio K	-	0.667	fb0/fc0	-	1.16
Viscosity parameter	-	0.0005	Eccentricity	-	0.1

, and the constitutive curves for concrete under tension and compression are shown in Figure 3. The elastic stage corresponds to the region before reaching the peak strength, and the plastic stage refers to the region after the peak.

3. Deformation and Internal Force Variation Patterns of Existing Tunnels Induced by Super-Large-Diameter Shield Undercrossing

3.1. Settlement and Stress Contours

This section analyzes the settlement and stress contours of the existing tunnel when the single-line and double-line excavation of the new super-large-diameter shield tunnel are completed, investigating the macro impact of double-line super-large-diameter shield undercrossing construction on the existing tunnel. Although the existing tunnel consists of two lines, OldT1 is the first to be undercrossed and experiences the greatest influence. Therefore, this paper initially focuses on data analysis for OldT1 only. The settlement contours of existing tunnel OldT1 during the excavation are shown in Figure 4.

The settlement contours indicate that shield excavation has induced a distinct settlement trough in the existing tunnel. When the single-line excavation of NewT1 (the first line) is completed, the center of the settlement trough of the existing tunnel is roughly directly above NewT1’s axis, presenting a V shape, with the maximum settlement value being approximately −19.70 mm, as shown in Figure 4a. With the advancement of NewT2 (the after line), the shape of the settlement trough undergoes a significant change. After completion of the double-line excavation, the center of the settlement trough shifts towards the central axis of the double-line tunnel, and the shape transitions to a W shape. The influence range of settlement visibly expanded, and the maximum settlement value further increased to −24.51 mm, as shown in Figure 4b. The settlement development of the existing tunnel indicates the disturbance superposition effect induced by the sequential excavation of the double-line super-large-diameter shield tunnels.

The stress contours are shown in Figure 5. When the single-line excavation of NewT1 was completed, the maximum principal stress at the tunnel bottom of the existing tunnel reached 2.624 MPa, approaching the tensile strength standard value of C50 concrete (2.64 MPa). However, after the excavation of NewT2 was completed and the double-line tunnels were fully connected, the maximum principal stress did not continue to increase; instead, it decreased to 2.097 MPa. This indicates that during the excavation of NewT1, the stratum was subjected to intense disturbance for the first time, causing stress to concentrate directly above the new tunnel, specifically in the bottom region of the existing tunnel, thereby forming a relatively high stress peak. Subsequently, although the excavation of NewT2 introduced new disturbance, it also altered the overall stress field of the stratum, leading to partial release of the soil stress between the two tunnels due to the disturbance. More importantly, as described later in Section 3, during the excavation of NewT2, the historical peak stress exceeded the tensile strength of concrete, resulting in irreversible tensile damage in critical areas of the existing tunnel. The stiffness degradation induced by material damage and the stress redistribution effect led to a wider load distribution, which manifested as a reduction in the stress peak after double-line completion, while concurrently causing a significant expansion in the distribution range of high stress.

In summary, both the settlement and stress contours collectively illustrate the impact of double-line super-large-diameter shield undercrossing construction on the existing tunnels. The excavation of the after tunnel not only intensified the overall settlement of the existing tunnels but also altered the spatial distribution of settlement, transforming the settlement trough from a V-shaped profile during single-line excavation to a W-shaped profile after double-line excavation. Meanwhile, the peak reduction and range expansion of stress indicates that after experiencing the historical maximum load, the material of the existing tunnel structure underwent damage, leading to a redistribution of internal forces. Therefore, this research further underscores the necessity of studying the evolution of internal forces and damage in the existing tunnel throughout the entire construction process in the subsequent sections. Moreover, direct field monitoring data (e.g., lining strain) for this specific project were not available at the time of this study, as it focuses on the construction phase. To assess the reasonableness of the numerical results, the predicted maximum settlement of −24.51 mm and the development of a W-shaped settlement trough are compared with the range of values and patterns reported in the literature for similar shield undercrossing projects in soft soil conditions [7,23,30]. The consistency observed provides indirect support for the model’s capability in simulating key deformation characteristics.

3.2. Settlement Development Patterns of the Existing Tunnel

During shield tunneling, the excavation position of the new tunnel is an important factor influencing the deformation of the existing tunnel. Therefore, five key time points were extracted from the advancement of the left and right lines, respectively, as monitoring points for tracking the dynamic settlement changes in the existing tunnel. Here, the right line is the first-line NewT1, and the left line is the after line NewT2. Two surveying lines at the vault and the invert of the existing tunnel were selected for study; the specific locations of the surveying lines are detailed in Figure 1b. The dynamic curves of the top and bottom displacements of the existing tunnel OldT1 are shown in Figure 6.

The development patterns of the settlement curves for the tunnel top and tunnel bottom of the existing tunnel differ. The tunnel top’s settlement curve consistently maintains a symmetrical V-shaped distribution throughout the double-line tunneling process, with the center of the settlement trough gradually shifting from the axis of the first tunnel to the central axis of the double-line tunnels as construction progresses. In contrast, the tunnel bottom’s settlement curve is more complex. During the single-line tunneling stage, it presents a single V-shaped settlement trough. As the double-line reaches completion, the curve morphology gradually transitions from a V shape to a W shape, ultimately forming a broad and gentle settlement trough with a slight uplift in the area above NewT2 when the double-line tunnels are fully connected. However, the peak settlement of the tunnel bottom remains concentrated above NewT1, whereas the peak settlement area of the tunnel top is closer to the central axis of the double-line tunnels.

In terms of settlement magnitude, the settlement of the tunnel bottom was similar to that of the tunnel top. Upon completion of the single-line tunneling, the maximum settlement of the tunnel top was −17.86 mm and that of the tunnel bottom was −19.70 mm, with a difference of 1.84 mm. After the double-line connection, the maximum settlement of the tunnel top increased to −24.24 mm, while that of the tunnel bottom was −23.81 mm, resulting in a difference of 0.43 mm. The influence range of the settlement expanded significantly as construction progressed, extending from about 40 m upon single-line completion to about 60 m upon double-line completion. The pattern indicates that, during double-line super-large-diameter shield undercrossing construction, the overall settlement of the existing tunnel was characterized by continuous ovalization deformation. Furthermore, this settlement distribution pattern was noticeably modified by the construction of the after tunnel.

3.3. Internal Force Variation Pattern of Existing Tunnels

3.3.1. Axial Bending Moment

Based on the numerical simulation results, this subsection analyzes the dynamic response pattern of the axial bending moment in the existing tunnel OldT1 during the undercrossing construction of double-line super-large-diameter shield tunnels. Figure 7 illustrates the distribution of vertical and horizontal bending moments along the tunnel axis during the single-line (T1 to T5) and double-line (T6 to T10) excavation stages of the super-large-diameter shield tunneling process.

The distribution curves of axial bending moments indicate that the excavation of the super-large-diameter shield significantly affects the internal forces of the existing tunnel, and its distribution pattern evolves systematically with the construction progress. As shown in Figure 7a, when the single-line excavation is completed, the vertical bending moment along the tunnel axis approximates a normal distribution. Within the section from 75 m to 105 m, the bending moment is negative, indicating that the tunnel top of the existing tunnel is under compression and the tunnel bottom is under tension in this zone. Outside this section, the bending moment becomes positive, meaning that the tunnel top is under tension and the tunnel bottom is under compression. The extreme value point of the bending moment is located directly above the axis of the new tunnel NewT1, demonstrating that the tension at the tunnel bottom is greatest at this location, which represents the most critical structural area. After the double-line is completed, the distribution pattern of the vertical bending moment has undergone a significant change: it evolves from a single-peak distribution into a double-peak distribution, and the influence range expands substantially. In this stage, the bending moment remains negative within the section from 50 m to 105 m, where the tunnel top of the existing tunnel is compressed and the tunnel bottom is tensioned, while outside this section, the bending moment becomes positive, indicating tension at the tunnel top and compression at the tunnel bottom. It is worth noting that the negative bending moment at the tunnel bottom above the axis of the after tunnel NewT2 is smaller than the peak above the axis of the first tunnel NewT1, which implies that the region subjected to the most severe tension is still located above the axis of the new tunnel NewT1.

The development pattern of the horizontal bending moment in the existing tunnel is shown in Figure 7b. Its distribution shape gradually changes from an M-shaped pattern to a W-shaped pattern as the shield advances. The values of the horizontal bending moment near the critical positions at 50 m and 105 m remain close to zero and hardly vary with shield excavation. In terms of magnitude, the maximum horizontal bending moment is only about 1/8 of the maximum vertical bending moment, indicating that under these engineering conditions, vertical bending is the dominant mode of axial loading in the existing tunnel.

The evolution of the vertical bending moment visually reflects the superimposed disturbance caused by the double-line construction of super-large-diameter shield tunnels. The excavation of NewT1 caused the tunnel bottom of the existing tunnel directly above it to reach the maximum negative bending moment. Subsequently, the excavation of NewT2 altered the overall state of the stratum, causing the peak bending-moment zone to expand laterally from the axis of NewT1, ultimately forming a broad and gentle W-shaped distribution covering the entire span of the new double-line tunnels.

It is worth noting that during the double-line undercrossing by super-large-diameter shields, the vertical bending moment curve of the existing tunnel is not an ideally smooth curve; instead, it exhibits an abrupt change near the axis of the first tunnel. This differs somewhat from the simulation results induced by ordinary-diameter shield undercrossing [7,8,9]. There are two main reasons for this: firstly, the double-line super-large-diameter shield undercrossing construction leads to a complex internal force state in the existing tunnel; secondly, compared with simplifying the existing tunnel as a linear elastic model, the introduction of the CDP constitutive model enables a more accurate simulation of the internal force variations and damage in the existing tunnel.

3.3.2. Axial Shear Force

This subsection analyzes the dynamic evolution pattern of the axial shear force in the existing tunnel OldT1 during the double-line excavation of super-large-diameter shield tunnels. Figure 8 shows the distribution of vertical and horizontal shear forces along its axis at ten key construction stages. Unlike the distribution pattern of bending moments, the shear force curves exhibit a typical antisymmetric distribution characteristic. Throughout the entire construction process, the shear force values remain at zero directly above the new tunnel axis at 60.25 m and 89.75 m, as well as at the model boundaries, while the extreme shear force points occur in the regions between these zero value points.

As shown in Figure 8a, when the single-line excavation is completed, the vertical shear force curve exhibits an antisymmetric distribution about the tunnel center as the axis of symmetry. The positive and negative peaks occur on either side of the axis of the first-line NewT1, corresponding to axial distances of approximately 78 m and 103 m, respectively. This area is a high-risk zone for interring misalignment and shear cracking of tunnel lines. With the advancement of the after tunnel NewT2, the basic distribution pattern of the shear force remains unchanged, but the shear force values generally show an increasing trend, and the influence range widens, which clearly demonstrates the shear force superposition effect induced by the double-line super-large-diameter shield construction.

The horizontal shear force also exhibits an antisymmetric distribution, as shown in Figure 8b. During single-line excavation, the shear force in the region from 90 m to 120 m on the right side of the axis is positive, while the region from 60 m to 90 m on the left side is negative. However, after the double-line excavation is completed, the positive and negative zones of the horizontal shear force undergo a reversal: the originally negative region from 60 m to 90 m becomes positive, and the originally positive region is converted to negative. This phenomenon indicates that the double-line excavation process of super-large-diameter shield tunnels significantly alters the lateral shear state experienced by the existing tunnel. In terms of magnitude, the maximum horizontal shear force is much smaller than the vertical shear force, which is approximately 1/8 of the maximum vertical shear force.

4. Damage Development Mechanism of the Existing Tunnel

4.1. Damage of the Existing Tunnel

This section analyzes the final damage state of the existing tunnel OldT1 after it was completed by the double-line super-large-diameter shield. Damage of the existing tunnel is typically induced by tension and compression. With the introduction of the CDP model, the damage zones can be studied by extracting the output variables of tensile damage and compressive damage, combined with the stiffness degradation rate. The contours of DAMAGET, DAMAGEC, and SDEG after construction completion are shown in Figure 9, while the stress contours can be referred to in Figure 5 above.

Based on the contours from the numerical simulation results, it can be clearly observed that the damage is predominantly concentrated in the tunnel bottom region of the existing tunnel OldT1, located above the first line of the new tunnel NewT1. This area shows a high degree of coincidence with the location of the maximum bending moment illustrated in Figure 7, indicating that the damage is primarily dominated by bending tensile stress.

When the double-line of the super-large-diameter shield machine is completed, the segments of the existing tunnel exhibited severe tensile damage in the concrete, with the maximum value of DAMAGET reaching 92.4%, indicating that the concrete had been heavily cracked. The maximum stiffness degradation rate (SDEG) was 89.58%, and the area where the stiffness decreased most significantly coincided with the region of tensile damage. Although compressive damage occurred in some areas, its value was far lower than the tensile damage, with the maximum compressive damage being only 0.9368%, which is negligible. Therefore, it is shown that the failure mode of the existing tunnel is mainly concrete cracking caused by bending-induced tensile stress, and the damage in the existing tunnel is dominated by tensile damage, which is significantly greater than compressive damage both in magnitude and extent.

4.2. Damage Mechanism of the Existing Tunnel Based on the Entire Construction Process

The previous subsection analyzed the final state of the existing tunnel upon completion of construction, revealing severe tensile damage in the concrete segments of the existing tunnel, with the maximum value of DAMAGET reaching 92.4%, as shown in Figure 9a. However, the maximum stress after the completion of the double-line super-large-diameter shield was only 2.097 MPa, which did not reach the tensile strength of C50 concrete (2.64 MPa). This final state of high-damage forms a sharp contradiction with the final low-stress result shown in Figure 5.

To investigate the reasons for this, this subsection extracts the dynamic stress curves at the tunnel bottom of the existing tunnel T1 during both single-line and double-line excavation, as shown in Figure 10, along with the development patterns of the maximum principal stress contours and tensile damage contours, as listed in

Table 3. Tensile stress contours of OldT1 during the entire construction process.

Times	Tensile Stress Contours	Description
T1		Tensile stress initially appears, concentrated above the first line NewT1, but with a low magnitude.
T2		The stress continues to rise, showing a local high stress zone.
T3		The stress continues to increase, and the high stress zone expands slightly to both sides.
T4		The stress decreases slightly, but the distribution pattern remains centered above the NewT1 axis.
T5		Single-line completion: the high-stress zone remains stable, with a peak value of 2.624 MPa.
T6		After tunnel NewT2 advances six rings, stress reaches the process peak of 2.655 MPa, exceeding the tensile strength of C50 concrete.
T7		Stress begins to decrease but still maintains a wide distribution of high stress.
T8		With continued advancement of the after line, the influence of double-line tunneling superimposes, expanding the distribution range.
T9		Stress further decreases, and its distribution becomes more uniform.
T10		Double-line completion: the stress peak decreases to 2.097 MPa, but its influence range is the widest.
Legend

and

Table 4. Development Pattern of Tensile Damage of OldT1.

Times	DAMAGET Contours	Description
T1		No damage; the structure remains intact.
T2		Damage begins to appear locally at the tunnel bottom of the existing tunnel directly above NewT1.
T3		Damage rapidly expands, forming a relatively distinct high-damage band.
T4		Damage continues to accumulate, and its extent extends toward both sides.
T5		Single-line completion: damage approaches its maximum value and remains stable in distribution.
T6		After tunnel advances 6 rings: damage increases slightly.
T7		Damage reaches its peak of 92.4% and stabilizes, though its range expands.
T8		Damage remains stable, and its distribution pattern is essentially unchanged.
T9		No further propagation of damage; the structure enters a stable damage stage.
T10		Construction ends; damage stabilizes at 92.4%, exhibiting a high-damage and low-stress state.
Legend

. Based on these data, an analysis of the damage mechanism of the existing tunnel throughout the entire construction process is conducted. The contour plots only display the section from 30 m to 120 m along the length of the existing tunnel, which represents the main affected zone during the double-line excavation of the super-large-diameter shield.

By integrating the tensile stress curves and tensile stress contours of the existing tunnel throughout the entire construction process of super-large-diameter shield undercrossing, it can be demonstrated that the damage development of the existing tunnel is directly related to the historical peak stress it experienced. As shown in Figure 10b, when the after tunnel NewT2 advanced six rings (corresponding to time T6), the maximum stress at the tunnel bottom of the existing tunnel T1 reached the process peak of 2.655 MPa (in Table 3), which, for the first time, exceeded the tensile strength of C50 concrete (2.64 MPa). It is worth noting that the dynamic stress curve exhibited a sharp drop near the peak, in other words, after reaching the peak, the stress values at nearby points decreased rapidly, and both the stress peak and the steep-drop zone were located directly above the new tunnel NewT1. This is because the CDP constitutive model accurately reflects material softening behavior: when the local material stress exceeds the tensile strength, damage occurs at that point, stiffness degrades significantly, and its load-bearing capacity rapidly decreases. At the same time, the load carried by the peak zone is transferred to adjacent areas, resulting in a sudden change in the dynamic curve over the entire process, representing a sudden release and redistribution of stress from the peak point. Furthermore, it can be observed that compared with the single-line excavation process (times T1 to T5), the high-stress zone of the existing tunnel increased significantly as the after line advanced (times T6 to T10).

The development process of the tensile damage contours of OldT1 is shown in Table 4; it can be observed that the high-damage zones show a high degree of spatial coincidence with the peak stress zones. Tensile damage began when NewT2 advanced to the 6 rings (time T6), reaching 92.38%, and stabilized at the maximum value of 92.4% throughout the subsequent construction process, indicating severe irreversible cracking in the concrete. The abrupt change in the stress curve corresponds precisely in both time and space with the maximum damage, both occurring directly above the area of the new tunnel T1. After passing this critical point, although the maximum principal stress decreased to 2.097 MPa in the subsequent excavation due to stress redistribution (time T10, in Figure 5) and fell below the tensile strength, the damage remained stable at 92.4% (times T7 to T10, in Table 4) without recovery. The study clearly reveals that under the disturbance induced by double-line super-large-diameter shield undercrossing, the damage in the existing tunnel is triggered by the historical peak stress during the construction process, and due to its irreversibility, accumulates and persists.

In summary, through the analysis of the entire construction process of super-large-diameter shield undercrossing, the study investigated the damage mechanism of the existing tunnel: the damage is not determined by the final stress state but is governed by the historical maximum stress it experienced. The peak and abrupt change in the dynamic stress curves directly manifest the initiation and development of damage. This indicates that for double-line super-large-diameter shield undercrossing construction, traditional safety assessment methods based solely on the final state may be insufficient. This study demonstrates that by implementing whole-process dynamic monitoring and numerical simulation, structural damage can be predicted and controlled. Moreover, the parameters of the CDP model (Table 2), particularly the tensile strength (2.64 MPa), were adopted directly from the Chinese national design code GB 50010-2010 for C50 concrete [40]. This choice ensures consistency with the regulatory standards used in the project’s design and represents a conventional, widely accepted baseline for numerical studies of concrete structures in similar contexts.

5. Conclusions

The study established a three-dimensional finite element numerical model of a double-line super-large-diameter shield (14.5 m) undercrossing the existing tunnels (6.5 m). By introducing the Concrete Damaged Plasticity model and assigning the CDP constitutive relation to the existing tunnel, the deformation patterns, internal force responses, and damage mechanism of the existing tunnel throughout the entire construction process were systematically investigated. While the identified mechanism—that damage is governed by the historical peak stress rather than the final state—is of general significance, the quantitative values of deformation and damage are contingent upon the specific geometric, material, and geotechnical parameters of the project. The main conclusions are as follows:

• When the single-line excavation was completed, the maximum settlement of the existing tunnel was −19.70 mm and the settlement trough exhibited a V shape. After the double-line completion, the maximum settlement increased to −24.51 mm, and the settlement trough evolved into a W shape, with the influence range expanding from about 40 m during the single-line stage to about 60 m. The position of the peak settlement shifted from the axis of the first tunnel toward the central axis of the double-line tunnels.
• The vertical bending moment reached its maximum after the completion of single-line excavation, with the peak located above the axis of NewT1. Following the double-line completion, its distribution evolved from a single-peak pattern into a broad, gentle, double-peak pattern, and the influence range expanded significantly. The maximum horizontal bending moment was only 1/8 of the vertical bending moment, confirming that vertical bending is the dominant loading mode. The shear force analysis indicated that the construction of the after tunnel generally increased the vertical shear force values, widened the affected zone, and formed a high-shear risk area between approximately 78 m and 103 m on both sides of the axis of the first tunnel.
• The damage mechanism of the existing tunnel induced by the double-line super-large-diameter shield undercrossing is as follows: damage is triggered by the historical peak stress and accumulates irreversibly, rather than being determined by the final stress state. When the after tunnel advanced six rings (time T6), the stress peak at the tunnel bottom of the existing tunnel reached 2.655 MPa, briefly exceeding the tensile strength of the concrete, and irreversible tensile damage occurred at the tunnel bottom. Although the final stress peak decreased to 2.097 MPa, the final DAMAGET value reached 92.4%, representing a high-damage state. Moreover, the distribution area of the damage highly coincided with the location of the maximum bending moment, indicating that cracking was primarily caused by bending-induced tension.

Based on the releved mechanism in this paper, two main control recommendations can be proposed to guide the design and construction of actual engineering projects: (1) Optimize the first undercrossing by prioritizing precise control of construction parameters during the first new tunnel’s undercrossing to minimize initial peak disturbance to the existing tunnel. (2) Implement targeted monitoring by deploying real-time stress or strain monitoring at the most vulnerable location—the invert of the existing tunnel directly above the first new tunnel’s axis. Set an early warning threshold near the concrete’s tensile strength for timely intervention.