RETRACTED: Effectiveness of anchor pile protection for metro tunnels under construction disturbances

c′

effective cohesion

D

pile diameter

$E50ref$

reference secant modulus

$Eoedref$

reference tangent modulus

$Eurref$

reference unloading-reloading modulus

f_s

strength retention rate

G₀

initial shear modulus

$G0ref$

reference initial shear modulus

H

depth of pit excavation

H_u(cr)

depth of disturbed zone

h

equivalent thickness of plate element

K₀

coefficient of earth pressure at rest

m

power-law exponent

m_c

mass of cement

m_s

mass of original soil material

P_c

yield stress of natural soil

$Pc′$

yield stress of disturbed soil

p_i

remaining overburden load after i stages of unloading

p_max

initial overburden load

p^ref

reference stress

R

unloading ratio

R_f

failure ratio

SD

disturbance degree defined according to yield stress

S_i

shear strength at unloading point

S₀

shear strength at consolidation point

t

pile spacing

γ_0.7

shear strain at 70% of initial shear modulus

$σ3′$

lateral stress of soil

υ_ur

Poisson’s ratio

$υurσ′1$

vertical effective stress of soil

φ_c

percentage of cement content

φ′

effective internal friction angle

ψ

dilation angle

The unloading due to pit excavation can lead to the soil rebound within the pit and redistribution of the surrounding soil stresses, causing additional displacements in the underlying tunnels and the degradation of soil engineering properties due to construction disturbances, which also affect the structural safety of the tunnels (Burford, 1988; Huang et al., 2012; Marta, 2001; Zhang et al., 2011). Most operational metro tunnels, constructed using precast concrete segments joined with high-strength bolts, are particularly sensitive to deformations. Excessive uplift at the pit bottom may result in damage to these tunnel segments, such as cracking and water leakage, leading to further deformation and potential structural failure (Limited, 2001). In practical engineering applications, the use of anti-pull piles has been identified as an effective method to mitigate this uplift at the pit bottom (Wen, 2010; Zheng et al., 2012). However, the construction of anchor piles can also disturb the surrounding soil and adversely impact the tunnel. Previous studies have indicated that a primary cause of engineering problems in pits and tunnels is the inadequate recognition of the changes in soil properties due to construction disturbances (Zhu, 2011). Consequently, research into the effects of construction disturbances, specifically from pit excavations and anchor piles construction on the deformation of underlying metro tunnels holds significant engineering value. This research is crucial for developing a comprehensive understanding of the interaction between construction activities and tunnel stability, and for enhancing protective strategies in urban underground construction.

Zhejiang province is located in a developed coastal region, where coastal soft soils exhibit widespread sedimentation, high moisture content, high compressibility, low strength, and pronounced structure. The structural integrity of these soils is highly susceptible to disturbances, leading to irreversible structural damage and significant changes in mechanical properties (Leroueil and Vaughan, 1990). When coastal soft soils are disturbed, the equilibrium system composed of the cementing materials between soil particles and water molecules is disrupted. This leads to changes in their internal structure and stress state. The reduction in strength and increase in compressibility of coastal soft soils following disturbance are attributed to the inherent structural properties of the soil (Wang et al., 2020). The structural nature of soil encompasses the spatial arrangement of soil particles, the condition of the pores, and the characteristics of inter-particle contacts and bonds. Due to the structural properties of soft soils, which are attributed to their cementation and fabric, Zhang et al. (2013) and Eskisar (2015) suggested the inclusion of cement in clay to prepare structurally bonded soils. They demonstrated the validity of this sample preparation method through microscopic experiments and uniaxial tests. Liu and Shen (2007) simulated the structural properties of natural clay by incorporating small amounts of cement and salt into the native soil particles to induce inter-particle cementation and create a large-pore structure. Luo et al. (2013) and He et al. (2017) conducted laboratory uniaxial tests to verify the feasibility of this approach. Compared with other methods, this technique is notably straightforward and practical.

Numerous domestic and international scholars have studied the soil disturbances caused by construction through laboratory tests and field monitoring. Hardin and Drnevich (1972) found through resonant column tests that even minimal disturbances could alter the soil’s initial modulus. Wang et al. (2001), based on the Wenzhou coal yard project, discovered through vane tests that the installation of drainage boards disturbed the soft soil, reducing its strength by approximately 50%. Chen et al. (2014) conducted field tests and laboratory experiments on the disturbed soils at the bottom of the Hangzhou Xianghu Station pit, which collapsed, finding that the soil strength reduced by 40%–80%. Zuo et al. (2015) installed earth pressure sensors on bored piles to study the disturbance range during pile construction, noting significant effects on the soil within three diameters horizontally and vertically from the pile. Li et al. (2021), based on multifunctional piezocone penetration tests (CPTU), determined the range of disturbance caused by pit excavation unloading and proposed that the depth of disturbance could be defined as the location where cone resistance decreases to 20% of its original value.

While existing research primarily focuses on the impacts of construction disturbances on pit or tunnel engineering, there is limited research on the effects of pit construction disturbances on nearby tunnels. Wang et al. (2005) considered construction disturbances and proposed equations for the stability against heave of deep excavations. Lu et al. (2021), combining finite element numerical analysis, analysed the effects of disturbances on the heave at the bottom of pits. Zhang et al. (2019), based on the Terzaghi–Rendulic consolidation theory, studied the effects of shield tunnel disturbances on soil pore water pressures and surface subsidence. Zhu et al. (2018) used three-dimensional finite element analysis to study the impacts of diaphragm wall construction on soil disturbances and adjacent buildings. In research considering the impacts of pit excavation construction disturbances on existing tunnels, Hu et al. (2013) set the disturbed area at the bottom of the pit and analyzed the effects on adjacent tunnels through numerical simulation. Similarly, Wang et al. (2020), through numerical simulation, set the entire area as a disturbance zone and studied the effects of pit excavation on adjacent tunnels. From the research conducted by these scholars, studies on the impacts of deep pit excavation considering construction disturbances on nearby underlying tunnels are scarce, and existing research has not detailed the division of disturbed soil areas according to actual disturbance levels, nor considered the impact of disturbances from anchor pile construction on the protective effectiveness for metro tunnels.

This paper verifies the methods of adding small amounts of cement and salt grains to silty clay in Ningbo, simulating artificially disturbed structured soils, and obtains the parameters of different disturbance degrees of soil bodies based on uniaxial compression tests and triaxial tests. Disturbed zones are classified based on the unloading ratio and field disturbance tests conducted at the Gaotangqiao metro station excavation site in Ningbo. Based on a practical engineering project in Zhejiang province, the impacts of pit excavation and anchor pile construction on tunnel displacement are analysed using the finite element software PLAXIS 3D, considering the degradation of soil properties due to the disturbances caused by construction. The protective effects of these anchor pile measures on the metro tunnel are discussed.

A practical pit engineering project in Zhejiang province consists of two blocks, north and south (with an interconnected basement), covering an excavation area of approximately 34 000 m² and a perimeter of about 960 m. Except for a single basement layer set up in the southernmost 40 m range, the remaining areas are constructed with two basement layers. A dual-line metro shield tunnel passes beneath the pit, with an external diameter of 6.2 m and an internal diameter of 5.5 m. The tunnel lining utilizes assembled segments of C50 grade concrete, and the cover depth above the tunnel is approximately 22.7 m.

As shown in Figure 1, the primary support system employs large-diameter bored cast-in-place piles (ϕ1000@1400) in combination with two reinforced concrete braces, with the exception of the intersection area between the metro shield tunnel and retaining structures, extending approximately 30 m on either side. Triple-axis cement mixing piles are implemented externally to the bored piles for effective water and soil retention.

In soft soil foundations, the pre-construction of anchor piles (bored cast-in-place piles) along both sides of the tunnel prior to excavation is a common practice to control tunnel deformation during excavation. For this project, to mitigate potential damage and water infiltration in underlying tunnels caused by basal heave, three rows of anchor piles (ϕ1200@4500) are strategically arranged along both lateral boundaries and the central axis of the underlying dual-line tunnel within the excavation area.

Figure 2 presents the typical cross-section A-A of the northern foundation pit illustrated in Figure 1. As depicted in Figure 2, the soil strata at the project site are characteristic of the typical soft soil layers found in Zhejiang province. Apart from the mixed fill soil of Layer ①-0-1, the plain fill soil of Layer ①-0-2, and the clayey silt of Layer ②-1, the remaining strata consist primarily of silty clay. The northern foundation pit is supported by two reinforced concrete braces, each measuring 0.7 m by 0.9 m. The ground level is situated at an elevation of −0.8 m, while the top elevations of the two struts are positioned at −1.9 m and −4.9 m, respectively. The basement slab’s bottom elevation is at −11.1 m, with a 300 mm thick cushion layer. The excavation depth of the foundation pit reaches 10.6 m. The retaining piles extend to an elevation of −31.4 m, and the anchor piles, with a length of 30 m, extend to an elevation of −41.4 m.

The nature of the retaining piles and anchor piles are bored cast-in-place piles. The construction of pit excavation and bored cast-in-place piles induces significant soil disturbance through multiple mechanisms. The processes of pit excavation and drilling lead to stress release, resulting in changes in the effective stress within the soil. The relieved stress is replaced by the tension in the soil’s pore water, resulting in negative pressure in the pore water, which causes the gas dissolved in the water to escape, further altering the effective stress in the soil. During the construction of pit excavation and bored cast-in-place piles, construction disturbances cause extensive soil remodelling, disrupting the original soil structure and leading to particle rearrangement, fundamentally altering the natural soil structure. During the concrete pouring process of bored cast-in-place piles, the injection of fresh concrete significantly alters the stress state of the surrounding soil, causing changes in pore water pressure and effective stress distribution. These combined effects create a complex disturbance zone around the piles, significantly affecting the engineering behaviour of the soil (Hu, 1997; Wei, 1987).

During the construction of the retaining piles and pit excavation, soil disturbance inevitably occurs, which is particularly critical in the context of the high sensitivity of the soft soil surrounding the tunnel. This disturbance leads to a significant reduction in the bearing capacity of the soft soil, consequently inducing continuous deformation of the tunnel. Furthermore, the construction of anchor piles generates additional soil disturbance. This study primarily investigates the impact of disturbances caused by pit excavation and retaining pile construction on tunnel displacement, with particular emphasis on the deterioration of soil engineering properties caused by these disturbances. In addition, the influence of construction disturbance from anchor piles on the protective effect for the metro tunnel is systematically analysed.

Due to the complex conditions of actual engineering projects, it is difficult to obtain soil with the same degree of disturbance, and secondary disturbances to the soil can occur during sampling, transportation, and sample preparation. Regarding the indoor preparation of disturbed soil, Meng (2019), Li (2015), and Hu et al. (2013) used methods such as artificial squeezing, ring knife, and kneading to prepare disturbed soil; Liu (2018) used a vibration table to disturb the soil, controlling the vibration time to obtain soil samples with different degrees of disturbance. The soil samples produced through squeezing and vibration methods are unevenly disturbed, and the operations are challenging. These methods make it difficult to quantitatively control the degree of soil disturbance and difficult to obtain completely consistent disturbed soil samples, have poor repeatability, and are not conducive to application in indoor geotechnical tests.

Carter and Liu (1999) proposed that the difference between disturbed soil and undisturbed soil lies in the change in its structure, and Wang et al. (2020) verified the reliability and rationality of simulating disturbed soil using artificially structured soil.

As delineated in Section 2, the soil strata at the project site are characteristic of the typical soft soil layers found in the region, and with the exception of the surface layer, the remaining strata predominantly consist of silty clay. To facilitate the investigation of soil disturbance effects on tunnel displacement and the protective efficacy of anchor piles, subsequent analyses are conducted under the assumption that the foundation pit enclosure, tunnel structure, and anchor piles are situated within a homogeneous silty clay stratum. The soil samples used in this experimental study were extracted from the ④-2 silty clay layer at an approximate depth of 33 m. These samples exhibit a dark grey coloration, which is typical of deep soft soil formations in the coastal regions of Zhejiang. The basic physical properties of the soil (CL, according to ASTM D2487) are comprehensively presented in Table 1.

The preparation of artificially structured soil and subsequent experimental procedures were strictly conducted in accordance with the Chinese National Standard for Geotechnical Testing Methods (GB/T 50123-1999). After drying the original soil material, it was crushed and passed through a 0.5 mm sieve, with different amounts of 525# Portland cement and a small amount of salt grains added (the salt content was 8% relative to the total weight of silty clay and cement). The salt grains were made by crushing large-grain edible salt and passing it through a 0.5 mm sieve. After thoroughly mixing the required mass of the mixture, it was compacted in five layers using a standard compactor. After compaction, the saturator containing the soil sample was placed in a saturation infiltration system for vacuum saturation. After saturation, the saturator was placed in flowing water, ensuring the soil sample was completely immersed in water. The flow of water helped flush away the dissolved salt grains, forming a structured soil sample with a cementing effect and large pores. Following sample preparation, the soil samples were transferred to a controlled curing chamber to ensure that the cement reaction time was maintained at 3 days for a standardized curing period.

The cement content calculation is shown in the following equation:

where φ_c is the percentage of cement content, m_c is the mass of cement, and m_s is the mass of the original soil material.

Based on the research results of Wang et al. (2020), this paper uses artificially structured soil with different cement contents to simulate disturbed soil with different degrees of disturbance. Uniaxial compression tests and triaxial consolidated undrained tests were conducted for six types of soil samples: natural soil, artificially structured soil 1 (cement 2.0%), artificially structured soil 2 (cement 1.5%), artificially structured soil 3 (cement 1.0%), artificially structured soil 4 (cement 0.5%), and remoulded soil (Figure 3). For the one-dimensional compression tests, three replicate tests were conducted for each specimen configuration, with the mean value of these measurements being adopted as the representative result to ensure statistical reliability. For the triaxial testing programme, each soil specimen was subjected to a comprehensive series of tests under four distinct confining pressures to characterize its stress–strain behaviour.

The one-dimensional compression tests were conducted using a 16-cell pneumatic consolidation apparatus, with applied pressure levels of 12.5, 25, 50, 100, 200, and 400 kPa. The deformation measurements were systematically recorded using an automated data acquisition system to ensure accuracy and consistency. For the triaxial testing programme, consolidated undrained tests were performed using a GDS triaxial testing system under confining pressures of 25, 50, 100, and 200 kPa. The deviator stress and axial strain were continuously monitored and recorded through a computerized data acquisition system, enabling precise measurement of the stress–strain behaviour and pore pressure development throughout the testing process.

Based on the results of uniaxial compression tests, the compression indices of different structured soils were obtained, and the fitting curves of the compression indices and cement content were drawn, as shown in Figure 4. It was found that as the cement content increased, the compression modulus of the artificially structured soil increased, reducing its compressibility. The hydration reaction of cement can build bonds between clay particles, effectively enhancing the cementing action between soil particles. These results indicate that using artificially prepared structured soil can restore the compressibility of remoulded soil to its natural state.

Figure 5 shows the relationship curves between the shear strength indices of different structured soils and cement content. It can be seen from the figure that as the cement content in the remoulded soil increases, its cohesion c and c′ both show exponential growth; when the cement content reaches 2%, the sample’s cohesion is nearly consistent with that of the natural soil; the internal friction angles φ and φ′ do not change significantly with increasing cement content, remaining essentially consistent with the natural soil’s internal friction angle. The results of uniaxial compression tests and triaxial tests indicate that as the cement content increases, the structural integrity of the soil is enhanced, allowing artificially prepared structured soil to restore the natural soil, and the deformation and strength characteristics of the natural soil after disturbance can be simulated by controlling the cement content.

Many scholars have conducted some research on the disturbance degree of structured soft soils, proposing their own methods for determining the disturbance degree of soil based on changes in soil pore water pressure, e–lgp curves, and soil strength (Hvorslev, 1949; Ladd and Lambe, 1964; Raymond et al., 1971). Based on the research results of Wang et al. (2020), this paper adopts the evaluation method proposed by Nagaraj et al. (2003), evaluating the disturbance degree of artificially disturbed soil with different cement contents based on the structural yield stress obtained from uniaxial compression consolidation tests, as shown in Table 2. The disturbance degree is defined as shown in Equation 2:

where SD is the disturbance degree defined according to yield stress; P_c is the yield stress of the natural soil, $Pc′$ is the yield stress of the disturbed soil, and the structural yield stress corresponds to the axial pressure at the turning point of the e–lgp curve.

Based on Table 2, the fitting curve of soil disturbance degree and cement content is drawn. As shown in Figure 6, the disturbance degree of the soil decreases linearly with increasing cement content, reaching nearly 0% when the cement content is 2%. This also verifies the rationality and reliability of simulating the disturbance degree of natural soil by controlling the cement content.

To study the impact of construction disturbances on the underlying tunnel, the disturbed soil zones caused by pit excavation were classified based on existing research results and field tests. The unloading due to pit excavation leads to the release of soil stresses at the bottom of the pit, reducing the soil strength and causing rebound deformation. The maximum depth of soil affected by the engineering property changes due to pit excavation is called the unloading influence depth (Li et al., 2021).

Pan et al. (2001) used the unloading ratio R to measure the change in stress levels at the bottom of the pit (Equation 3). Through direct shear tests, the strength curves under loading and unloading states were obtained. The strength retention rate f_s was obtained (Equation 4).

where p_max is the initial overburden load, p_i is the remaining overburden load after i stages of unloading. S₀ and S_i are the shear strengths at the consolidation point and unloading point, respectively.

The depth of the disturbed zone H_u(cr) can be determined using Equation 5 (Pan et al., 2001), where H is the depth of the pit excavation:

Several studies on the disturbance depth of excavation areas are shown in Table 3.

It is reasonable to divide the bottom of the pit into strong and weak disturbance zones. Since many field measurements and laboratory tests determine the strong disturbance zone to be about 0.25H, this paper sets the depth of the strong disturbance zone as 0.25H; the depth of the weak disturbance zone is set as 1.3H, according to Pan et al. (2001).

Pit excavation not only disturbs the soil at the bottom of the pit but also affects the soil behind the retaining walls. This paper also divides the disturbance zones based on the field disturbance tests conducted at the Gaotangqiao metro station pit in Ningbo. The degree of soil disturbance was evaluated using cone penetration tests (CPT), vane shear, and shear wave velocity tests, based on which the soil was categorized into three disturbance zones, as shown in Figure 7.

To facilitate a clearer comparison of soil strength reductions across the different disturbance zones as described, the test results have been organized into Table 4. This paper divides the disturbance zones behind the retaining walls according to the field disturbance tests of the Ningbo metro pit project. Zone 1 is classified as a strong disturbance zone, and Zone 2 as a weak disturbance zone.

The division of the disturbance zones caused by pit excavation is shown in Figure 8. The disturbance is divided into bottom disturbance and outside wall disturbance: the bottom disturbance is divided into two areas, from the bottom of the pit to 1.25H excavation depth as a strong disturbance zone, and from 1.25H to 2.3H as a weak disturbance zone. The soil disturbance outside the diaphragm wall is divided into two areas, with the connection area between the buried depth of the underground continuous wall at 1.4H and outside the pit at 1H classified as a strong disturbance zone, and the connection area between the buried depth of 2.0H and outside 2.0H as a weak disturbance zone.

Lu et al. (2021) measured a disturbance degree of about 60% in the strong disturbance zone through field CPTU tests, with the disturbance degree of the soil in the weak disturbance zone gradually decreasing from 60% to 0% with increasing distance from the pit. Based on the parameters of disturbed soil measured in Section 3, the disturbance degree of the strong disturbance zone is set at 62%, and the weak disturbance zone at 40% for numerical analysis.

During the construction of bored piles, the soil compaction effect causes certain disturbances to the surrounding soil. According to research by Liu and Tang (2005), bored pile construction can cause significant disturbances to the soil within a 1.5 m radius around the piles. As shown in Figure 9, the central anchor pile is about 5.6 m away from the tunnel, and the side anchor piles are about 7.0 m away from the tunnel, with the anchor pile disturbance zone set as a 1.5 m radius around the piles. Based on the parameters of disturbed soil measured in Section 3, the disturbance degrees of the anchor pile construction zone are set at 0%, 18%, 40%, 62%, and 83% for numerical analysis.

In this paper, the soil is modelled using the hardening soil model with small-strain stiffness (HSS). The HSS model parameters include 11 HS model parameters and 2 small strain parameters. This paper obtained the soil’s reference tangent modulus $Eoedref$ from one-dimensional compression consolidation tests; the reference secant modulus $E50ref$⁠, failure ratio R_f, and dilation angle ψ from triaxial consolidation drained shear tests; effective cohesion c′ and effective internal friction angle φ′ from triaxial consolidation undrained shear tests; and the reference unloading-reloading modulus $Eurref$ from triaxial consolidation drained unloading-reloading tests. The other HS model parameters m, υ_ur, p^ref, and K₀ can refer to existing research results for specific values and literature sources as shown in Table 5. The small strain parameter $G0ref$ can be determined based on the conclusion by Wang et al. (2013), taking $G0ref=3.5:5.0Eurref$⁠, and γ_0.7 can be calculated according to the introduction by Brinkgreve and Broere (2006) using Equation 6.

where

In the equation, $σ′1$ is the vertical effective stress of the soil, which can be taken as the vertical effective stress at the middle point of the corresponding soil layer during calculation; $σ′3$ is the lateral stress of the soil.

The main parameters of the HSS model are detailed in Table 6.

The concrete strength grade of the retaining piles is C30, and the piles are simulated using plate elements, with the thickness determined according to Equation 8, where D is the pile diameter, t is the pile spacing, and h is the equivalent thickness of the plate element. The three-dimensional calculation parameters of the retaining structure plate element are shown in Table 7.

The northern block of the project is equipped with two supports, consisting of 0.7 m × 0.9 m reinforced concrete supports, using C30 concrete, and simulated using beam elements in the finite element model. The calculation parameters of the internal supports are shown in Table 8.

The metro tunnel constructed by the shield method has longitudinal joints between the segments, leading to reduced lateral bending stiffness of the lining. Ye (2017) used the modified customary method from the Japanese tunnel standards to account for the reduction in bending stiffness caused by longitudinal joints. This method considers the reduction in bending stiffness caused by longitudinal joints as an effective reduction in the overall bending stiffness of the lining, introducing an effective lateral bending stiffness ratio η to account for the impact of longitudinal joints on the tunnel lining, taking the lateral bending stiffness of the tunnel as ηEI. Huang et al. (2006, 2012) determined through tunnel model tests that the effective lateral bending stiffness ratio of shield tunnel misaligned joints can be taken as about 0.75, and the longitudinal bending stiffness can be taken as 1/6. The tunnel has an inner diameter of 5.5 m and an outer diameter of 6.2 m. The strength grade of the concrete for the tunnel lining structure is C50, with a weight of 15.5 kN/m³, and Poisson’s ratio is taken as 0.2. The tunnel lining parameters are shown in Table 9.

In numerical calculations, anchor piles are simulated as plates, with specific parameters as shown in Table 10.

This paper primarily analyses the impact of construction disturbances on the underlying tunnel, so only the northern pit is modelled for analysis. Based on the actual engineering project, the numerical simulation model used is shown in Figure 10. The pit area is 238 × 110 m, excavated in four steps to the bottom of the pit, with excavation depths of 2.5, 4.9, 8.1, and 10.6 m, and supports placed at −1.9 m and −4.9 m. The tunnel diameter is 6 m, a dual-line tunnel, with a burial depth of −22.7 m. To consider the influence of boundary effects, the model dimensions are set at 305 × 190 m. In the numerical analysis, both the disturbances caused by pit excavation and tunnel protection measures are considered. According to the disturbance zones in Section 4, the disturbed zones in the three-dimensional model are set as shown in Figure 11.

The numerical analysis process for the model is outlined as follows:

• generation of the initial geostress field

• resetting initial displacements and tunnel activation

• resetting initial displacements and activation of the support structure’s plate elements

• resetting initial displacements and activation of anchor pile’s plate elements

• excavation to 2.5 m below ground surface and construction of corresponding supports

• excavation to 4.9 m below ground surface and construction of corresponding supports

• excavation to 8.1 m below ground surface

• excavation to the bottom of the pit at 10.6 m below ground surface.

Since the two tunnels are closely spaced, their deformations and displacements are similar, so only the left line tunnel is analysed. The tunnel crown beneath the pit is most affected by excavation unloading, and the following analysis focuses on the deformation of the tunnel crown. As shown in Figure 12, the tunnel experiences uplift displacement within the range of pit excavation, and subsidence occurs outside the range of pit excavation due to dewatering. When the tunnel is near the boundary of the pit, it experiences significant horizontal displacement, while in the middle of the pit, the horizontal displacement of the tunnel is close to zero. Due to the small angle of intersection between the tunnel and the pit at the tunnel length of 223 m, the impact of pit unloading is greater, leading to larger horizontal displacement. Figure 11 shows the tunnel displacements under two conditions: with and without considering soil disturbance. When considering disturbance, the maximum horizontal displacement of the tunnel is 4.08 mm, and the maximum vertical displacement is 9.74 mm. Without considering disturbance, the tunnel’s maximum horizontal displacement is 3.92 mm, and the maximum vertical displacement is 8.23 mm. It is evident that considering soil disturbance has a significant impact on the vertical displacement of the tunnel, increasing the maximum vertical displacement by 18%; however, the change in horizontal displacement is minimal, with a maximum increase of 4.1%. The results considering disturbances are used for analysis in the following sections. In the text below, ‘unmitigated excavation’ refers to considering only the disturbances caused by pit construction.

Figure 13 shows the impact of different disturbance degrees of soil on tunnel displacement after anchor pile construction. As shown in the figure, after anchor pile construction followed by pit excavation, without considering anchor pile disturbance, the maximum vertical displacement of the tunnel is 5.14 mm, and the maximum horizontal displacement is 3.02 mm, which are reduced by 47% and 33%, respectively, compared with unmitigated excavation. Anchor piles provide good protective effects on the tunnel.

When considering anchor pile disturbance, both the vertical and horizontal displacements of the tunnel increase. Compared with the condition without considering anchor pile disturbance, with a soil disturbance degree of 83%, the maximum uplift displacement of the tunnel increases by 18%, and the maximum horizontal displacement increases by 17%. It is evident that soil disturbance caused during the construction process can significantly increase tunnel displacement. Compared with unmitigated excavation, even when considering anchor pile disturbance, anchor pile construction still provides good control over tunnel displacement. In addition, pile foundation construction between the dual-line tunnels, being close to the tunnels, is prone to pile hole wall collapse, causing loss of surrounding soil. In actual engineering, adopting construction methods such as long casing and rotary pile foundation soil extraction can significantly reduce soil disturbance, thereby reducing tunnel displacement.

Based on a practical engineering project in Zhejiang province, this paper conducted uniaxial compression tests and triaxial tests on natural silty clay and artificially structured soil, and studied the effectiveness of anchor pile protection measures for metro tunnels considering construction disturbances using PLAXIS 3D. The main conclusions are as follows:

• For the silty clay in the Ningbo area, mixing 2% cement into remoulded soil results in compression and strength characteristics that are essentially similar to those of undisturbed soil. Within the range of 0% to 2% cement content, as the cement content increases, the compression index of the artificially structured soil decreases linearly, the compression modulus increases linearly, cohesion c′ rises exponentially, and the internal friction angle φ′ remains largely unchanged. The artificially structured soil’s disturbance degree decreases linearly with increasing cement content, reaching nearly 0% when the cement content reaches 2%.

• In the area within the pit excavation, the tunnel experiences uplift displacement; near the boundary of the pit, the tunnel exhibits significant horizontal displacement, while in the centre of the pit, the horizontal displacement of the tunnel is close to zero. The soil disturbances caused by pit excavation significantly affect the vertical displacement of the tunnel, increasing the maximum vertical displacement by 18.3%; however, the change in horizontal displacement is minimal, with a maximum increase of 4.1%.

• After anchor pile construction followed by pit excavation, without considering anchor pile disturbance, the maximum vertical displacement of the tunnel is 5.14 mm, and the maximum horizontal displacement is 3.02 mm, which are reduced by 47% and 33%, respectively, compared with unmitigated excavation. Anchor piles provide good protective effects on the tunnel.

• When considering disturbances from anchor pile construction, both the vertical and horizontal displacements of the tunnel increase. Compared with the condition without considering anchor pile disturbance, with a soil disturbance degree of 83%, the maximum uplift displacement of the tunnel increases by 18%, and the maximum horizontal displacement increases by 17%. Compared with unmitigated excavation, even when considering anchor pile disturbance, anchor pile construction still effectively controls tunnel displacement.

Not applicable.