Research on Double-Layer Support Control for Large Deformation of Weak Surrounding Rock in Xiejiapo Tunnel

Abstract

Double-layer primary support is proposed to control the deformation of surrounding rock in tunnels within weak geological conditions, where engineering challenges such as large deformations, tunnel faces, and arch collapse are encountered. This approach is based on the principle of combined resistance and release. A combined approach of numerical modeling and on-site surveillance was utilized to analyze the displacement and stress state of the tunnel support structure at different construction stages of primary support for the second layer, using Xiejiapo Tunnel as an engineering case. The findings indicate that the implementation of two-layer primary support can mitigate the progression of large deformations effectively in weak surrounding rock; the sooner the primary support for the second layer is applied, the better the deformation control, and the later the application takes place, the more effectively the tension in the surrounding rock is diminished, whereby the self-supporting capacity of surrounding rock comes into its own. The force of the shotcrete is reduced. Considering the structural deformation and stress state, as well as combination of resistance and release, it is best to implement the primary support for the second layer 10 feet behind the primary support for the first layer.

1. Introduction

With the ongoing progress of highway construction in China, particularly in the western regions, more and more technical problems have emerged. These problems are prevalent in relation to the design and complex geological conditions in long tunnels, which the majority of tunnel builders have to face [1,2,3]. A significant concern for tunnel builders is the tendency of soft surrounding rock to cause issues such as twisting and deformation of steel arches, primary supports exceeding the clearance limit, and tunnel face sliding. These engineering problems frequently result in difficulties in controlling the safety, duration, cost and quality of the tunnel project, and hinder the development of China’s transportation to a certain degree [4]. As the complexity and difficulty of engineering conditions increase, the problem of tunnel support becomes more and more prominent. Especially under complex geological conditions, how to effectively design and construct tunnel supports to control the deformation of the tunnel and ensure the safety and stability of the tunnel is an important issue that needs to be solved urgently. In response to the significant deformation of soft surrounding rock, numerous domestic and non-domestic researchers have studied tunnel deformation mechanisms, construction methods and large deformation control measures [5,6]. As numerical simulation software has been used to analyze the deformation characteristics of the surrounding rock and the supporting structures in the tunnel under construction, Ref. [7] proposed the three-bench tunnelling + temporary inverted arch method. The method was combined with increasing the size of steel sections, reinforcing leg-lock anchor bolts and strengthening the longitudinal connection of the arch to enhance the form of support. Such improvements controlled the surrounding rock’s large deformation effectively. Using Yufangping Tunnel of Guzhu Expressway as a case, through monitoring and measuring and on-site tests, Ref. [8] analyzed the primary support pressure and surrounding rock’s second lining. They also analyzed the deformation under different support stiffnesses, and ultimately adopted the support scheme of improving the arch stiffness, as well as increasing the length of anchor bolts at the sidewalls, which controls the surrounding rock’s deformation effectively. Through statistically analyzing the surveillance data of Muzailing Tunnel, Ref. [9] classified the soft surrounding rock’s large deformation growth into four types and proposed targeted distortion control measures, focusing on stress control technology, tunnel excavation methods and the support form. Addressing the large deformation problem of the Maoxian Tunnel of Chenglan Railway, Ref. [10] analyzed the behaviors of different support counterforces and support forms in relation to the control of deformation through the numerical model. They recommended the adoption of multiple layers and multiple support systems to alleviate some of the stress on surrounding rock, which ensures the tunnel’s long-term stability. Ref. [11] proposed a dual-shell grouting deformation control technology involving low-pressure grouting and high-pressure splitting grouting to address the problem of large deformations in highly stressed soft rock. The results indicate that the whole surrounding rock’s mechanical properties have been improved, and the effect of injection can be employed to control the surrounding rock deformation. Ref. [12] adopted the support method of constant resistance large deformation (CRLD) anchor bolts, anchor cables with hollow grouting in the bottom plate, and steel-fiber-reinforced concrete. The approach effectively controlled the large deformations of surrounding rock in deep soft tunnels. Ref. [13] developed a method for controlling large deformations in the surrounding rock of Yuntunbao Tunnel on the Chenglan Railway by assembling long anchor bolts and double-layer primary supports, wherein the accumulated tunnel deformation is only just over half of those of the two phases, effectively solving the problem of encroachment and primary support failure in large surrounding rock deformation. Ref. [14] applied continuous and non-continuous numerical simulation methods to study the formation and evolution of large deformations in soft rock, as well as the effectiveness of joint support in high-stress tunnels. Their findings indicate that the standard and multi-layer support measures have a positive suppression effect on large deformations, providing guidance for related technical studies. Ref. [15] considered the distributed features of the surrounding rock joints as well as performing discrete unit simulations for different anchor bolt lengths, circumferential spacings and prestresses. The results show that the deformation of the surrounding rocks tended to decrease linearly with the increase in circumferential length and prestress, offering a theoretical foundation for the use of the prestressing anchorage system to control large soft rock deformations. Aiming at the problem whereby the supporting structure of the Minxian soft rock tunnel is prone to damage, Ref. [16] proposed a method based on constant resistance large deformation (CRLD) by exploring how the surrounding rock deforms and breaks through on-site monitoring and data analysis methods. The latest support method of employing anchor cables had a substantial impact on controlling the surrounding rock’s deformation. Ref. [17] controlled horizontal stress by improving the resistance and stability of the shotcrete support system, addressing the challenge of the mergence of large deformations in soft rock in “three-soft” coal tunnels, which effectively improved the stability of the “three-soft” coal tunnel. Ref. [18] set up high-strength anchor bolts around the tunnel and used them to fix the rock at the bottom of the tunnel. This strategy reduced the friction between the rock and surrounding environment, as well as its sliding resistance, effectively controlling the surrounding rock’s deformation. Ref. [19] used FLAC 3D software in simulating tunnel excavation and analyzing the mechanism of rock deformation and failure. They proposed a number of joint, multi-layer support measures that limited surrounding rock deformations effectively. To solve the problems of large soft rock deformation in tunnels under high-ground stress, Ref. [20] applied FLAC 3D to perform numerical simulations of various tunnel excavation schemes, based on Jianshan tunnel in the Gansu province. It showed that the three-bench tunnelling temporary inverted arch and three-step overhang method mostly controlled the surrounding rock’s deformation effectively. Ref. [21] used FLAC 3D to model tunnel excavation methods and address the large deformation of the specially shaped super-large Y-shaped soft rock tunnel; they proposed a deformation control scheme in which the tunnel-builders built a pilot tunnel, and then excavated in the reverse direction. Ref. [22] compared the “centre pilot tunnel method” and the “bilateral pilot tunnel method”, combined with numerical simulations and field measurement data, to study different tunnel excavation methods. They found that the latter significantly controlled deformation under asymmetric terrain conditions. Aiming at the large deformation problem in highly weathered carbonaceous mudstone tunnels, Ref. [23] took the highly weathered carbonaceous mudstone section of Muzhailing Tunnel as the background of the research, and proposed a three-step method and a heading expansion method combined with field tests and numerical simulation results. The deformation of the surrounding rock was effectively controlled in highly weathered carbonaceous mudstone tunnels, which provides a reference for constructing similar carbonaceous mudstone tunnel sections. Aiming at the problem of strong–moderate weathering deformation in Gelong quartzite tunnels, Ref. [24] proposed a method combined with theoretical analyses and numerical simulations to enhance the longitudinal stiffness of the supporting structure and improve the integrity of surrounding rocks. In the case of Jiaoweiqin tunnel, Ref. [25] proposed the temporary double-side wall method. The method, derived from studies on the deformation problem of shallow-buried large sections and supported by numerical modeling and field survey data, effectively limited deformation. Ref. [26] constructed a ring structure composed of 15 adjustment units. By controlling and adjusting the position and diameter of the units, they optimized the main power system of the tunnel lining, demonstrating that the deformation of surrounding rock around primary supports could be controlled. Ref. [27] conducted a study whereby the deformation of surrounding rock was limited in a deep soft-rock tunnel. The deformation and damage were visually analyzed using finite element software. According to the findings, the surrounding rock’s vertical stress exhibited concentration at both the top and underneath, while the horizontal stress was focused on the same areas. The study found that stress was concentrated on the sides and bottom of the tunnel, revealing that plastic deformation and damage typically started on the side of the tunnel before extending to other parts. These insights are crucial for understanding the rock-related challenges in deep-buried soft-rock tunnels, and offer valuable guidance for addressing such issues in practice. Ref. [28] assessed the stress and displacement conditions of rocks around Dongsong Hydropower Station in Sichuan Province through on-site monitoring. They studied the stress field propagation and deformation principles during large deformation tunnel excavation in soft rock; they concluded that a short-step construction method controlled the expansion of rock deformation and the plasticity range of the surrounding area effectively. Aiming at the problem of instability and failing deep soft rock in Shiyaokou Coal Mine, Ref. [29] used on-site monitoring and borehole peeking methods to study tunnels’ deformation and damage characteristics. They proposed a method based on constant resistance large deformation (CRLD). The high-pressure-compensation method of using bolts had a positive influence on rock deformation. By examining the impacts of different parameters on tunnel deformation through a combination of numerical models and field tests, Ref. [30] optimized the parameters of the steel arch support structure to control rock deformation, thereby proposing a strength reduction method that enhanced the surrounding rock’s stability. Utilizing numerical calculations, Ref. [31] analyzed the stress deformation and distribution patterns of the surrounding rocks of large sections of whole coal cavern groups (WCCG). Following their analysis, they proposed a new layered reinforcement scheme of “long cable–anchor–grouting” (LBG), specifically designed to control surrounding rock deformation. Ref. [32] studied the large deformation mechanism and support methods, taking Maoxian tunnel as an example. Compared with single-layer primary support, double-layered primary support was more effective in controlling fractures under high-pressure geostress and phyllite rheological effects. Taking Muzhailing tunnel as an example, Ref. [33] evaluated various factors affecting the tunnel and proposed the use of high-constant-resistance bolts to limit the large deformations. In addition, anchor bolts also make a great contribution in controlling the deformation of rocks surrounding tunnels. Ref. [34] discovered that Swellex bolts significantly enhance support performance across various rock types, particularly exhibiting notable benefits in safety applications in tunnels and mines. Ref. [35] tackled the issue of conventional rock bolts’ inadequate energy absorption and limited stretching capabilities by introducing an innovative energy-absorbing rock bolt. This bolt was rigorously tested using static pull and dynamic drop tests, which assessed its load-displacement behavior and its capacity to absorb energy. The results indicate that the new rock bolt provides consistent resistance and remarkable elongation, substantially improving the rock support systems’ ability to absorb energy. Ref. [36] tackled the challenge of identifying optimal parameters for a newly developed non-adhesive anchoring mix that expands during the curing process. By conducting statistical and correlation analyses, they investigated the influence of the encapsulation depth and the geometric configuration of the mix–rock–anchor interface on anchoring efficacy. The study concluded that these parameters significantly impact the ultimate load-bearing capacity, which is vital for reinforcing the strength and stability of support structures in underground settings. Ref. [37] conducted laboratory testing on expansion shell rock bolts with deformable components to investigate their stress–strain and displacement characteristics. They quantified the contributions of each component to the overall displacement and provided essential design data that enhance the performance of rock bolt support systems in underground excavations under static loads.

The research into the deformation mechanism, construction methods and deformation control skills of soft surrounding rock tunnels has achieved specific outcomes. These results have provided a theoretical basis and technical guidance, and laid a solid foundation for the safe construction of tunnels with soft surrounding rock. However, the investigation of the large-scale deformations found in tunnels with soft surrounding rock, including the deformation characteristics, construction technology and other issues, still needs to be improved. This paper focused on the design and construction methods of tunnel supports, especially tunnel supports under soft surrounding rock and large deformation conditions, and found the optimal tunnel support solution. However, this study could not consider other types of underground works, such as underground warehouses or subway stations, to maintain the focus and depth of the study. This paper presents an analysis of the large deformation found in soft surrounding rock in the Xiejiapo tunnel. Utilizing on-site construction methods, field monitoring and numerical simulation and measurement comparative analysis methods, large deformation control measures in Xiejiapo Tunnel were proposed, which may provide references for related soft rock tunnel engineering, and also provide some inspiration and guidance for future research.

2. Project Overview

Xiejiapo Tunnel belongs to the section of the National Highway Yinbai Line (G69) from Ankang to Langgao (Shaanxi–Chongqing border), located near Xiejiapo in Jihe Town, Hanbin District, Ankang City, Shaanxi Province, with a tunnel length of 2870 m and a designed longitudinal slope of 18%. The tunnel site is a tectonically eroded low mountainous geomorphological unit with relief undulating and a relative height of 279.0 m. The formation lithology is complex, with phyllites of Meiziya formation in the lower Silurian system (S1m). We see here the growth of rock joints and fissures, and the rock is exceptionally fragmented, has a low strength, and is easy to disturb. It is, therefore, known that the surrounding rock is grade IV~V, as shown in Figure 1a.

The ground stress seen in the deeply buried section of Xiejiapo Tunnel is complex; the folds of surrounding rock around the tunnel face are noticeable; joints and fissures develop, and the phyllites’ uniaxial compressive strength is below 5 MPa. The tunnel belongs is a large soft-rock deformation tunnel, where the rock is very soft. The three-step reserved core soil construction method was here applied, and the support factors are shown in

Table 1. Support parameters of Xiejiapo Tunnel deeply buried large-deformation section (unit: cm).

Advanced Support Structure Parameters			Primary Support Structure Parameters					Inverted Arch
Small duct	Circumferential spacing	External angle	Steel arch	Reinforcement fabric	Anchor bolt	Leg-lock anchor bolts	Shotcrete	Concrete
Φ42 L = 450	40	12°	I20b @15 × 15	Φ8 @15 × 15	Φ22 L = 400 120 × 800	Φ420 L = 400 α = 12°	C30 26	C30 130

.

During the construction period, in accordance with the “Technical Specifications for Construction of Highway Tunnel” (JTG/T 3660-2020) [38], specific measuring devices were adopted for various construction stages. For the measurement of surrounding convergence, the JTM-J7100 steel ruler convergence meter from Changzhou Gold Civil Engineering Instrument Co., Ltd. was employed for data acquisition. For the measurement of arch top subsidence, the PENTAXR-422NM total station from PENTAX Instrument Co., Ltd. in Japan was used to obtain accurate readings. As for cross-sectional scanning, the SMTD-4 laser tunnel cross-section detection meter from Beijing Smartot Technology Co., Ltd. in China was utilized to ensure the precision of the data. And it was determined through field testing that the maximum value of vault sinking was 602.3 mm, the maximum surrounding convergence value was 900.3 mm, and the surrounding rock deformation rate was between 5 mm/d and 30 mm/d. Primary support deformation limitation, steel frame twisting, shotcrete spalling and cracking occurred in places, according to Figure 1b,c. The original support program fails to take on the surrounding rock’s deformation effectively, which hampers the project’s structural safety, threatening construction personnel and equipment.

3. Large Deformation Control Technology in Xiejiapo Tunnel

The high primary geostress of the deeply buried section of Xiejiapo Tunnel, as well as the rock crushing, strong weathering, and stiffness, all soften the original support scheme, causing large deformation in the construction. The maximum value of surrounding convergence is 900.3 mm, which is much larger than the 80~120 mm recommended for two-lane highway tunnel primary deformation in reserved class V surrounding rock, as described in the “Technical Specifications for Construction of Highway Tunnel (JTG/T 3660-2020)” [38]. The ratio of the maximum value of surrounding convergence to the width of the tunnel is 3.2%, which places it in class II (moderately large deformation). The substantial deformation of the soft surrounding rock in Xiejiapo Tunnel is beyond the effective limitation of the support parameters suggested in the current specification.

The primary support of the tunnel employed a single-layer initial support, and the specific support parameters are shown in Table 1. The single-layer support is shown in Figure 2. The support method is also shown in Scheme 1 in the following section. Double-layer initial support is added to the layer of support after the application of the single-layer initial support. This entails applying a second layer to the support shown in Figure 2. Currently, the main materials used in the primary tunnel support system are steel and concrete, which are rigid materials with minor ultimate strains compared to the soft and highly deformed surrounding rock. Given the complex stress conditions, the deformation resistance is much lower than its value under a single stress condition when supporting the tunnel structure. Therefore, when increasing the support strength (thickening shotcrete, increasing the density of arches and anchor bolts) to address the large deformation, the rigidity of the support is increased, which leads to an increase in supporting capacity and force, and limits the ability to solve the issue of large deformation stability effectively.

According to the principle of combining resistance and release, the flexibility of the first layer of the support plays a role in coordinating deformation and stress release such that the force of the support can be reduced; the rigidity of the second layer of the support is used for deformation control and safety protection in the structure. However, the difficulty of applying layered support lies in determining the first layer of the support’s strength and the timing of implementation of the second layer of the support.

Xiejiapo Tunnel’s large-deformation section was excavated using the three-step reserved core soil method, with a footage 0.6 m; the method of employing double-layer primary support plus increasing the deformation allowance was proposed to limit the large deformation. The first layer of support remained as described in the original design scheme, and the specific parameters are shown in Table 1. The primary support for the second layer utilized I20a steel arches spaced 80 cm apart, with C25 shotcrete of 26 cm thickness; four kinds of primary support were used for the second layer (the second layer of the primary support immediately followed the primary support of the first layer, and the second layer of the primary support was distanced behind the primary support for the first layer by 5 feet, 10 feet, and 15 feet) m. Using only the flexible support concept and increasing the deformation allowance, which fully releases the surrounding rock stress, not only does the amount of excavation and spray-mixing increase during construction, but the secondary lining also bears greater pressure, and this reduces the support structure’s safety. If the concept of rigid support is accepted, the stiffness of the support structure must be greatly increased in order to limit the release of surrounding rock stress, and then to control the deformation. Therefore, by comparing the normal- and large-deformation conditions of Xiejiapo Tunnel and the results of many on-site measurements, the deformation allowance for the section with large deformation was adjusted to 200 mm. The optimal positioning of the primary support for the second layer in the large-deformation section of Xiejiapo Tunnel was determined by numerical modeling.

4. Numerical Calculations

4.1. Modeling

To simulate the proposed supporting scheme, the Midas GTS finite element software 2022 was utilized. The model consisted of a solid element representing the tunnel’s surrounding rock, a plate unit representing the shotcrete, and implanted truss units representing the leg-lock anchor bolts. The model selected Xiejiapo Tunnel’s large-deformation section, with dimensions of 40 m × 80 m × 100 m (length × width × height), as depicted in Figure 3. Key points of the model were selected from the ZK 17+404 section points A, B, and C, as illustrated in Figure 4. The initial geostress considered only the self-weight, and the surrounding rock was modeled as an ideal elastic–plastic model, following Mohr–Coulomb yielding criteria. The surrounding rock’s physical–mechanical parameters were altered to simulate the advanced reinforcement area, with a reinforcement thickness of 0.5 m. Shotcrete and leg-lock anchor bolts were assessed in the elasticity model. The steel arch’s elastic modulus was converted to shotcrete by equivalent conversion [39], and only the leg-lock anchor bolts were separately set. Based on the geological investigation report, the design document, and field tests of the Xiejiapo Tunnel, the parameters of the rock mass and primary support structure are shown in

Table 2. Material parameters used in the model’s calculation.

Materials	Deformation Modulus (MPa)	Poisson Ratio	Bulk Density (KN/m3)	Internal Friction Angle (°)	Cohesive Force (KPa)
Phyllite	255	0.4	20	20	110
Advanced reinforcement area	1280	0.4	20	22	130
First layer of the primary support	30,000	0.25	24	/	/
Second layer of the primary support	30,000	0.25	24	/	/
Leg-lock anchor bolts	200,000	0.3	78.5	/	/
System bolts	210,000	0.3	78.5	/	/

.

4.2. Simulation Schemes

The Xiejiapo Tunnel was excavated through the three-step reserved core soil method, with 1 m of excavation per footage and first-layer primary support applied, and the lengths of the upper, middle, and lower steps were 6 m, 4 m, and 6 m, respectively. The left middle step and right middle step of the second-layer primary support were applied at the same time, and so were the left lower step and right lower step. The middle-inverted arch in the primary support of the double layer was applied at the same time. In Schemes 1–5, a total of five construction plans all adopted the method of initial support, the most important step of which was to be carried out near to the excavation surface on the upper, middle, and lower steps. Although all these planned designs were the same, they each entailed different implementations of the second layer of initial support. Specifically, Scheme 1 did not implement the second layer of initial support. The second layer of initial support in Scheme 2 was set up simultaneously with the first layer of the initial support. The application of the second layer of the initial support in Schemes 3–5 lagged behind the implementation of the first layer of initial support, where Scheme 3 lagged 3 m (5 feet), Scheme 4 lagged 6 m (10 feet), and Scheme 5 lagged 9 m (15 feet) behind. Each supporting scheme is shown in

Table 3. Numerical simulation schemes.

Support Schemes	Timing of the First-Layer Primary Support	Timing of the Second-Layer Primary Support
1	The upper, middle, and lower steps follow the excavation surface	/
2	The upper, middle, and lower steps follow the excavation surface	Applied at the same time as the first layer of primary support
3	The upper, middle, and lower steps follow the excavation surface	Applied 3 m behind the first layer of primary support (5 ft)
4	The upper, middle, and lower steps follow the excavation surface	Applied 6 m behind the first layer of primary support (10 ft)
5	The upper, middle, and lower steps follow the excavation surface	Applied 9 m behind the first layer of primary support (15 ft)

.

4.3. Analysis of Surrounding Rock Deformation and Supporting Structure

The five support schemes were simulated, and the surrounding rock’s vertical and horizontal displacements in a typical section are shown in Figure 5.

Figure 5 illustrates the distribution of horizontal and vertical displacements for five different support schemes. An analysis of the horizontal displacement distribution maps has revealed that the maximum displacement values typically arose at the left arch foot, whereas, in the case of vertical displacements, the largest movements were detected at the crown. This phenomenon is attributable to the arch foot being subject to the most significant geostatic pressure. Within the crust, rock layers impose pressure on the tunnel, with the pressure on both sides of the arch foot being notably pronounced. The geological strata at the crown, being comparatively softer, are more susceptible to subsidence. During the tunnel excavation process, the lower rock layers beneath the crown tend to move downwards due to gravitational forces. Given the relative softness of these layers, they are prone to sinking under the combined influences of crustal pressure and gravity; hence, they yield the maximum values of vertical displacement. A comparative analysis of the horizontal displacements across the five schemes showed that Scheme 4 presented the smallest proportion of maximum horizontal displacement at 9.5%, while Scheme 1 exhibited the largest at 13.8%. Similarly, a comparison of vertical displacements revealed that Scheme 5 had the smallest proportion of maximum vertical displacement at 11.51%, with Scheme 2 possessing the highest at 12.5%. The continuous and discontinuous numerical simulation methods used in reference [14] allowed for a comparison of the simulation results of the surrounding rocks under combined support with no support. The findings were in good agreement with those of this paper, demonstrating that the use of support has a beneficial effect in controlling large deformations in soft rocks.

According to the data presented in Figure 5, the horizontal displacement was focused on the arch waist on both sides of the tunnel. On the other hand, the vertical displacement was primarily focused on the tunnel vault and inverted arch bottom.

Table 4. Comparison of displacements of key points in various support schemes.

Support Schemes	Vertical Displacement of Point A (mm)	Horizontal Displacement of Point B (mm)	Horizontal Displacement of Point C (mm)	Horizontal Convergence (mm)
1	176.4	168.4	180.4	348.8
2	126.6	117.2	125.3	242.5
3	134.1	134.2	143.4	277.6
4	151.3	143.7	158.3	302.0
5	155.3	160.7	170.0	330.7

provides a comparison of the maximum displacements at critical points for a standard section, while Figure 6 illustrates the displacement dynamic curve that developed as excavation advances.

Figure 6 reflects the temporal curve of displacement at key points in the tunnel. It can be seen from the figure that, in general, the displacement first increases and then becomes constant as the excavation advance increases, which is consistent with the actual situation. The difference between the horizontal displacement and vertical displacement of the five support schemes was not obvious when the excavation footage reached 15 m. However, as the excavation advance increased, it can be seen that the horizontal displacement and vertical displacement of Scheme 1 were greater than those of the other schemes. The one with the smallest displacement was Scheme 2. With the completion of the support, the horizontal and vertical displacements of the tunnel tended to become stable, so the figure shows a relatively gentle shape.

Table 4 and Figure 6 show the following:

(1) The surrounding rock’s vertical and horizontal displacement patterns in all five schemes are consistent. Notably, the traditional single-layer primary support (Scheme 1) shows the highest surrounding rock displacement, with a total value of 176.4 mm for vault sedimentation and 348.8 mm for horizontal convergence, indicating that the Xiejiapo Tunnel is a tunnel with soft surrounding rock and large deformation;

(2) Adding a second layer of primary support (Schemes 2–5) in soft surrounding rock tunnels reduces the cumulative displacement value of the surrounding rock. With the lag of the implementation of the second layer of the initial support, the ranges of vault decline in Schemes 2–5 were 28.23%, 23.81%, 14.23%, and 11.96%, respectively, while the decreased horizontal convergence values were 30.48%, 20.41%, 13.42%, and 5.19%, respectively. The addition of a second layer of primary support is a common method for controlling the surrounding rock’s deformation in tunnels with significant deformation. It is best to add this layer as early as possible to achieve the smallest possible cumulative deformation. However, to prevent interference between the two layers of primary support, Scheme 4 is recommended, which involves delaying the construction of the second layer by 6 m (10 ft).

4.4. Support Structure Stress Analysis

The stress cloud diagrams of shotcrete, system anchor bolts and leg-lock anchor bolts for the five support schemes are shown in Figure 7 and Figure 8:

Figure 7a depicts the stress cloud map for Scheme 1, while Figure 7b,c depict the stress cloud maps for the first and second layers of support for Scheme 2, respectively. Figure 7d,e displays the stress cloud maps for the first and second layers of support for Scheme 3, respectively, and Figure 7f,g represents the first and second layers of support for Scheme 4. Furthermore, Figure 7h,i shows the stress cloud maps for the first and second layers of support for Scheme 5. An overall examination of Figure 7 reveals that the maximum stress values within the initial tunnel support were primarily concentrated at the vault. Comparative assessments of the supporting Schemes 2–5 with Scheme 1 reveal that Schemes 2 and 4 exhibited stresses that were not significantly different from those of Scheme 1, whereas Schemes 3 and 5 exhibited notable differences in stress levels when compared with Scheme 1.

Figure 8 illustrates the axial stress cloud diagrams for the system anchor bolts and the leg-rock anchor bolts across the five shoring scenarios, indicating the effects of different support timings on the axial stress of the anchors. In conjunction with the data in

Table 5. Stress comparison analysis of each support scheme.

Support Schemes	Maximum Stress of Spray-Mixing (MPa)		System Anchor Bolts Axial Stress (MPa)	Leg-Lock Anchor Bolts Axial Stress (MPa)
	First Layer Shotcrete	Second Layer Shotcrete
1	2.088	/	16.56	6.457
2	1.038	1.020	13.67	5.577
3	1.113	0.940	14.56	5.812
4	1.132	0.830	14.89	5.898
5	1.194	0.763	15.39	6.123

, the timing of secondary initial support application was observed to influence the stress on the system anchor bolts; the more delayed the support, the greater the axial stress sustained by the system anchor bolts, in accordance with the stress patterns evidenced in the figure.

The maximum stress values of shotcrete, system anchor bolts and leg-lock anchor bolts for the five support schemes are listed in Table 5.

From Figure 7 and Figure 8 and Table 5, we can infer the following:

(1) When employing the traditional single-layer primary support (Scheme 1), the shotcrete experiences a maximum stress of 2.088 MPa. On the other hand, when utilizing the double-layer primary support (Schemes 2–5), the total stress of double-layer shotcrete support is lower compared to single-layer support. Moreover, with the delay in construction of the second primary support layer, the total stress of the double-layer shotcrete gradually decreases. These results indicate that the double-layer primary support can be used to optimize the structural stress and enhance the structural safety. Additionally, implementing the primary support in the second layer at a later point leads to lower structural stress;

(2) Under Schemes 2–5, with the lag in the application time of the initial support in the second layer, stress borne by the first layer of shotcrete gradually increases, while the proportion borne by the second layer of shotcrete gradually decreases. This indicates that stress is continuously released during excavation. A delay in applying the primary support in the second layer allows for the full release of stress from the surrounding rock, thereby enabling the surrounding rock to exhibit a greater self-supporting capacity. Under constrained conditions, in the first layer of shotcrete, the surrounding rock can slowly release its stress, effectively achieving the effect of “combining resistance and release”. The second layer of shotcrete can optimally share the surrounding rock’s stress, optimize the stress state, and provide greater safety reserves;

(3) Under Scheme 1, where there is only one layer of primary support, the system anchor bolts and leg-lock anchor bolts experience the highest axial stress. However, under Schemes 2–5, where a primary support for the second layer was added, the axial stress on these bolts was reduced. Since the system and leg-lock anchor bolts were installed along with the primary support for the first layer, delaying the implementation of the second layer meant that the surrounding rock stress was not fully relieved, leading to a greater burden on these bolts. Therefore, it is advisable to implement the second layer of primary support as early as possible.

4.5. Comparative Analysis of Calculation Results

By numerically calculating five support schemes and analyzing the displacement and stress of the support structure, we can determine the following:

(1) From the perspective of deformation control, the earlier the primary second layer support is implemented, the better the deformation control will be. However, the premature implementation of the primary second layer support can lead to the in situ stress in the surrounding rock not being fully released, meaning the support structure must bear excessive force. At the same time, the primary first- and second-layer support structures are too close to each other, which is not conducive to construction operations;

(2) From the perspective of stress control, when the second layer is applied, the relief of the surrounding rock will be more complete, and the total stress that the double-layered pre-stressed concrete must withstand will be lower. In contrast, the axial stress endured by the system anchor bolts and leg-lock anchor bolts will be large. However, applying the second layer of primary support too late can lead to the excessive deformation of the support structure, or even relaxation and collapse.

Scheme 4 was chosen for application in the Xiejiapo Tunnel based on a thorough assessment of the design and careful consideration of the deformation and stress state of the support structure. This approach involves implementing the second layer of primary support at a 6 m (10 ft) lag, which relieves the stress in the surrounding rock effectively, minimizes load, and does not cause excessive deformation (with a maximum vertical displacement of 151.3 mm and maximum horizontal displacement of 158.3 mm, both well below the reserve deformation of 200 mm).

5. Analysis of Site Monitoring Results

To validate the efficacy of double-layer primary support, Scheme 4 (wherein the primary second-layer support lags behind the first layer by 6 m, 10 ft) was implemented in the Xiejiapo Tunnel through on-site construction. The vault sinking and dynamic monitoring of peripheral convergence were conducted in two sections, namely, ZK 17+398 and ZK 17+404, and temporal curves were charted and compared with the numerical simulation results, as can be seen in Figure 9 and Figure 10.

Figure 9 and Figure 10 provide field data and the results of numerical simulations under Scheme 4. The records show that both horizontal and vertical displacements increased with excavation depth. The most significant rise occurred within the first 20 m; afterwards, the rate of increase leveled off. This pattern is due to the reduction in spatiotemporal effects at the tunnel face as the tunnel progresses, allowing the initial support to absorb part of the deformation through its designed flexibility so as to adjust to ground pressure and reduce deformation.

Figure 9 and Figure 10 show the following:

(1) Based on on-site monitoring data, it has been observed that the supporting structure’s vaults accumulate sedimentation values of 166.4 mm and 163.4 mm, respectively. Additionally, the cumulative peripheral convergence values are slightly larger than the numerical simulation results, measuring 339.3 mm and 345.4 mm, respectively. However, these values fall within the reserved deformation range of the tunnel, ensuring both structural safety and normal construction;

(2) The temporal curve clearly demonstrates conformity between the trends in the development of surrounding rock deformation according to numerical modeling and field monitoring. This attests to the dependability of numerical simulation, and reinforces the reliability of the construction scheme that involves lagging the primary support for the second layer behind the first layer by 6 m (10 ft). Such a scheme can be used to manage the surrounding rock’s deformation effectively, alleviate the stress placed on it, and reduce the load applied on the supporting structure.

Comparing Figure 9 and Figure 10, it is evident that the approach adopted in this paper (Scheme 4) can effectively limit large deformations in soft rocks. This finding aligns well with the conclusions drawn in reference [32]. These comparisons, as shown in Figure 9 and Figure 10, aim to provide a comprehensive analysis of the practical effects of the double-layer initial support strategy, offering a quantitative, dynamic, and visual understanding of its efficiency when used in managing rock deformation. This contributes to the broader study of tunnel construction strategies.

6. Conclusions

The challenge of surrounding rock deformation in Xiejiapo Tunnel was addressed by utilizing a double-layer primary support system, and combining resistance and release methods. Numerical simulations and on-site measurements have allowed us to determine the optimal timing for the application of the second layer of the primary support. The key findings are:

(1) Based on the principle of combining resistance and release, a double-layer primary support is suggested to manage large deformations in soft surrounding rock. In the double-layer primary support, the first layer is a flexible support, which plays a role in coordinating deformation and releasing stress, while the second layer is a rigid support, which is used to control deformation and ensure structural safety;

(2) Through numerical simulations of five support schemes, it has been determined that implementing the primary support in the second layer at an earlier stage is more advantageous for deformation control. Conversely, delaying implementation of the second layer in primary support results in the full release of stress in the surrounding rock, thereby reducing the overall stress that the shotcrete endures;

(3) Considering the structural deformation and stress state comprehensively, it is suggested that the second layer of the primary support for Xiejiapo Tunnel be constructed 6 m (10 ft) behind the primary support for the first layer. This conclusion has been confirmed through measurements taken at the construction site.

When dealing with the issue of large deformation in soft surrounding rock, the method of double-layer initial support can be employed. In this approach, the first layer of the support is applied initially, followed by the second layer implemented 6 m (10ft) later. This progressive implementation strategy aids in more effectively controlling the deformation of the surrounding rock, thereby ensuring the safety and stability of the project. Moreover, this method has a wide range of applicability. It can be applied not only in the current project, but also as a reference for other similar projects.