1 Introduction

Support optimization for soft rock tunnel is always the hot issue in engineering field (Choi and Shin 2004). As soft rock has significant expansibility and time-variant characteristics due to its low strength, easily weathered, large void ratio, small bulk density and large water permeability, it always results in a serious impact on the engineering safety and construction period (Sharifzadeh 2013). Results show that the most important, for the final surface subsidence, was the installation time and stiffness of the initial support (Bizjak and Petkovšek 2004). Authors can simulate the actual situations with the three dimensional finite difference code, and propose new optimal support patterns (Choi and Shin 2004). Alejano assessed the effects of the standard support and reinforcement in soft rock (Alejano et al. 2009). In order to control the deformation in soft rock tunnel and determine a reasonable support form, some other experts at home and abroad have done a lot of research on structural design of tunnel section, parameter settings of support structure, construction method and technology (Sakurai 1978; Cristescu et al. 1987; Indraratna 1993; Vermeer et al. 2002; Tomanovic 2006; Li 2007; Tsesarsky et al. 2013; Alexakis and Makris 2013; Zhang 2014; Vlachopoulos and Diederichs 2014), and have made some achievements. Although extensive research on support optimization for soft rock tunnel have been carried out, it is still short of systematically theoretical research. So far, the problem of large deformation in surrounding rock has not yet been included in the design specifications. In view of the particularity of the tunnel project, it needs a single theory or control measures to deal with the actual situation of complicated geology (He et al. 2002). Through analyzing the condition of the first lining surrounding rock pressure, secondary lining surrounding rock pressure and deformation for different supporting stiffness, the reasonable supporting parameters form is finally confirmed based on analyzing monitoring data (Li et al. 2013). To control excessive deformation of mica schist tunnel, the anchoring bolts of side wall and skewback foundation must be strengthened to improve the annular effect of primary support, the stiffness of steel frame should also be increased appropriately, and applied the secondary lining when necessary (Zuo et al. 2014).

Based on the study on the multiple occurrences of large deformation during construction in Youfangping tunnel of Gu-zhu highway and its deformation-failure rule, this paper has proposed the support scheme of “weakening anchor bolts while enhancing the stiffness and strength of initial lining”, and the other two support schemes are selected to do comparative study and analysis. By numerical simulation under different support schemes, the tunnel deformation characteristics, such as rock deformation, the distribution of anchor force, the development of plastic zone, the stress of sprayed concrete and secondary lining, are analyzed. And three groups of test sections on site are selected for real-time monitoring the deformation and pressure of surrounding rock in three kinds of support schemes. The numerical simulation and the field monitoring results both show that the support scheme of “weakening the anchor bolts while enhancing the stiffness and strength of initial lining” is feasible, and can provide a reference in support optimization for the similar projects along the highway.

2 Project Overview

Youfangping tunnel is very long, located in Youfangping Village, Zijin Town, Gucheng County, Xiangfan City, with the NE–SW trending. Tunnel width 10.25 m, height 5.0 m, with R = 550 cm single center circle curved wall lining. The tunnel site is located between the South Qinling tectonic belt and the Yangtze paraplatform, severely affected by the geological structure. The exposed formation is mainly composed of eluvial silty clay of Quaternary Holocene (Q el+dl4 ) and schist of Proterozoic Wudang Group (Pt2Wb). The maximum buried depth of tunnel is about 100 m. During the construction in tunnel, the exposed surrounding rock is moderate and strong weathered sericite quartz-schist, as shown in Fig. 1.

Fig. 1
figure 1

Sericite quartz-schist in tunnel

Full size image

As the surrounding rock is very broken with low strength, in high in situ stress and significant compression, the self-stability of cavern is very poor. The beyond-limit deformation of initial support produced by the squeezing action in the original design is shown in Fig. 2.

Fig. 2
figure 2

The beyond-limit deformation of initial support

Full size image

3 Numerical Simulation and Analysis

3.1 Support Scheme in Tunnel

The rock of Grade V accounts for a large scale in Youfangping tunnel, and the large deformation occurs basically in this type of rock. In view that the large deformation occurred frequently, the support scheme of “weakening anchor bolts while enhancing the stiffness and strength of initial lining” is proposed in this paper. Meanwhile, the two other alternative schemes are chosen to do comparative study. Therefore in the numerical simulation of support structure, the three different schemes of the original design of Z5 and the changed support form of Z5c, Z5b are respectively analyzed. The parameters of three support schemes are shown in Table 1. From the comparison among the above three support schemes, it can be found that the support types of secondary lining are the same, and the initial supports are slightly different except steel mesh reinforcement, which is disregarded in the process of simulation and regarded as the restorative effect of loose surrounding rock and preventing the cracking damage of spray layer. The difference between Z5c and Z5 is the anchor length and space in the circumferential direction, and the values are respectively 3 and 1.6 m in Z5c, while those are 3.5 and 1 m in Z5. The difference between Z5b and Z5c is the steel support type of initial lining and support thickness, which change from HW150 steel to HW175 and from 23 to 25 cm. In general, the support form of Z5c has weaker anchor compared with Z5, while Z5b has greater strength and stiffness of initial lining relative to Z5c.

Table 1 The support parameters
Full size table

3.2 The Calculation Model and Parameters

3.2.1 The Basic Assumption

  1. 1.

    Without considering the effects of time and space, the released load, support force and deformation are thought to be instantaneous. All material is supposed to be homogeneous and continuous in plane strain model (Song-hua et al. 2000).

  2. 2.

    Use Mohr–Coulomb elastic-perfectly plastic element to simulate the rock, and elastic solid element to simulate secondary lining. In initial lining, shell elements are used to simulate sprayed concrete and steel arch, while cable elements are used for the anchor. Each support structure is shown in Fig. 3 below.

    Fig. 3
    figure 3

    The support structure of initial lining and secondary lining

    Full size image
  3. 3.

    Only consider the gravity stress field, regardless of the tectonic stress field.

3.2.2 Determination of the Calculation Range and Constrains

The surface is taken as the upper boundary of the model in numerical simulation, and the actual depth of the tunnel is 100 m. Five times of the tunnel span is set as the left and right borders which are both 60 m. And five times of the tunnel height is taken as the lower boundary which is 50 m. The thickness of the model is about a construction cycle footage (80 cm). The grid model is shown in Fig. 4. The mesh model has a total of 3528 units and 5376 nodes.

Fig. 4
figure 4

Calculation grid model

Full size image

The constraint conditions of model are set as follows: the DOFs of X-direction at left and right sides are constrained; the DOF of Y-direction (parallel to the direction of the tunnel) in front and back of the model is fixed; the DOFs of X-direction, Y-direction and Z-direction in lower boundary are constrained; and the upper boundary is free.

3.2.3 The Selection of Calculation Parameters

The geotechnical parameters of surrounding rock in calculation model are determined according to preliminary survey data and laboratory tests, while the calculation parameters of sprayed concrete, secondary lining, steel arch and anchor are in accordance with Code for Design of Road Tunnel (JTG D70-2004) and other similar projects. The specific parameters are shown in Tables 2 and 3.

Table 2 Calculation model parameters (1)
Full size table
Table 3 Calculation model parameters (2)
Full size table

3.2.4 The Realization of Construction Step in Tunnel

The benching tunneling construction method is used in Youfangping tunnel. It is divided into several steps to simulate the construction process in FLAC3D. After the grid model of the tunnel is set up, different construction steps are simulated by deleting and adding units. And there is mainly five construction steps in the benching tunneling construction method, which are the basis of analyzing the stress field and displacement field of surrounding rock and internal force of supporting structure with the solution of numerical simulation.

3.3 Numerical Results and Analysis

The calculation results in this section are obtained under the conditions of three support schemes of Z5, Z5b and Z5c, and the steps of 1, 2, 3, 4 and 5 in the results respectively represent the excavation and supporting of upper bench, the excavation and supporting of lower bench and secondary lining.

3.3.1 Deformation Analysis of Surrounding Rock

The values of vault settlement and horizontal convergence under each construction step in three kinds of support schemes are shown in Table 4.

Table 4 The deformation values of surrounding rock in three support schemes (unit: mm)
Full size table

From Table 4, the values of vault settlement and horizontal convergence in three kinds of support schemes are in the same order of magnitude without big difference, indicating that weakening the anchor has little impact on controlling the rock deformation. In addition, the deformation in the support of Z5b is smaller than that in the other two support schemes, indicating that initial support structure has good effect on controlling the rock deformation after enhancing support structure. And the excavation of upper bench has greater impact on the vault settlement, and the values respectively account for 87.2, 87.4 and 87.2 % of total deformation after excavating the upper bench in three supports of Z5c, Z5b and Z5. While, the excavation of lower bench has greater impact on horizontal convergence, and the horizontal convergence between the step of installing initial support of upper bench and the excavation of lower bench respectively accounts for 63.6, 63.4 and 64.2 % of total deformation in the three kinds of support schemes.

3.3.2 Stress Analysis of Anchor

The axial stress distributions of anchor in three kinds of support schemes of Z5c, Z5b and Z5 are shown in Fig. 5 below (followed by the above sequence). From the figure, the axial stress distributions in three kinds of support schemes are basically the same, and the maximum axial stress are respectively 225.6, 227.2, 197.9 MPa. The maximum forces in the anchor are all at the skewback of upper bench, and the stress is small or in compression within the range of 45° at left and right sides of the vault, which indicates that the system anchor effect is not obvious in the range area. The anchor force is also small at the right and left sidewalls of lower bench. All this shows that the effects of the anchor have little difference in three kinds of support schemes.

Fig. 5
figure 5

The axial stress distribution of anchor in three kinds of support schemes

Full size image

3.3.3 Plastic Zone Analysis of Surrounding Rock

The distributions of plastic zones in the four kinds of support schemes Z5c, Z5b, Z5 and no support scheme are shown in Fig. 6 below (followed by the sequence of no support scheme, Z5c, Z5b and Z5). From the figure, the three support schemes of Z5c, Z5b, Z5 have a better effect on controlling the development of plastic zones in surrounding rock. Among these schemes, Z5c has relatively weaker system anchors compared with Z5 and they both almost have the same size of plastic zones. From the distribution of plastic zone in Z5b, it can reduce the development of plastic zone more by enhancing the stiffness and strength of initial support under the premise of weakening system anchor.

Fig. 6
figure 6

The distribution of plastic zone in four support schemes

Full size image

3.3.4 Stress Analysis of Sprayed Concrete

The bending stress and shear stress nephograms of sprayed concrete in three kinds of support schemes are shown in Fig. 7 below (followed by the sequence of Z5c, Z5b and Z5). The bending stress and shear stress values are both small, far less than the strength of sprayed concrete. Therefore, the initial support structures in three kinds of support schemes are all safe with the solution in the elastic plastic model. The maximum bending tensile stress in Z5c, Z5b and Z5 are respectively 0.0273, 0.0229 and 0.0370 MPa, which are mainly at the junction of upper bench and lower bench. This is also in conformity with the characteristic of large horizontal convergence deformation and large anchor force at the junction of upper bench and lower bench during excavation.

Fig. 7
figure 7

The bending stress and shear stress nephograms of sprayed concrete in three support forms (unit: Pa). a Bending stress. b Shear stress

Full size image

3.3.5 Stress Analysis of Secondary Lining

The principal stress nephograms of secondary lining in three kinds of support schemes are shown in Fig. 8 below (followed by the sequence of Z5c, Z5b and Z5). The maximum principal stress and minimum principal stress are both small. Therefore, the secondary lining structures in these schemes are all safe with the solution in the elastic–plastic model. The maximum stress of secondary lining are mainly distributed at the skewback of lower bench and at the left and right hances. The lateral part of secondary lining at the skewback is in tension, while the lateral part at the left and right hances is in compression. This situation is consistent with the deformation of surrounding rock in tunnel.

Fig. 8
figure 8

The principal stress nephogram of secondary lining in three support schemes (unit: Pa). a The minimum principal stress. b The maximum principal stress

Full size image

4 Support Parameters Optimization in Large Deformation of Surrounding Rock

To effectively control the large deformation of surrounding rock and determine reasonable support parameters, different support schemes are carried out for field test. By contrastive analysis of the structure force and the rock deformation in different forms of support parameters, the most appropriate form of support parameters is synthetically determined.

4.1 Field Monitoring Items

The main purpose of field monitoring is to comparatively analyze the monitoring data and to determine the actual effect in different kinds of support schemes by real-time monitoring the horizontal convergence, crown settlement, rock pressure of initial lining and secondary lining (Xian-min et al. 2011).

4.2 Selection of Test Area and Analysis of Monitoring Results

4.2.1 Selection of Test Area

Three regions of YK 41 + 720 ~ YK 41 + 740, YK 41 + 845 ~ YK 41 + 865, ZK41 + 560 ~ ZK41 + 580 are taken as test sections, and the support forms of Z5, Z5c, Z5b respectively adopted as the support schemes in the three regions for field test.

4.2.2 Analysis of Monitoring Results

After real-time monitoring for nearly 6 months and the systematic processing and analysis of monitoring results, the variation characteristics of the deformation values and deformation rate of surrounding rock and the rock pressure of initial lining and secondary lining are obtained under different support parameters. Now the monitoring data of typical sections in the regions have been done preliminary statistics, as shown in Table 5 below.

Table 5 Monitoring data of typical sections in three kinds of support schemes
Full size table
  1. (1)

    Monitoring results analysis of rock pressure in initial lining

From the monitoring results analysis of rock pressure in initial lining and field geological inspections, the following conclusions can be drawn: (1) the average rock pressure in the test section of Z5b is the largest, and the pressure in Z5c and Z5 are relatively small with little difference. From the analysis of rock pressure in compression, it can be obtained that the larger the stiffness of the support structure, the larger the rock pressure on the support structure, and vice versa; (2) from the analysis of a single monitoring section, the rock pressures at the skewback and the side walls are the largest, followed by the pressure at spring line, the minimum pressure at the vault. It can be obtained that the lateral pressure coefficient of the formation is large, indicating that the horizontal tectonic stress is large, which is consistent with the simulation results; (3) with the field inspections, it can be seen that the concretes of initial lining appear crack in partial sections in the support schemes of Z5 and Z5c, and some even fall off in local area, while the support structures in other local area of test sections are intact and unspoiled. Thus, the support scheme of Z5b has a good effect on controlling the deformation of surrounding rock.

  1. (2)

    Monitoring results analysis of rock pressure in secondary lining

From the monitoring results analysis of rock pressure in secondary lining and field geological inspections, the following conclusions can be drawn: (1) the values of rock pressure in three kinds of supporting parameters are all not large, and the trend of contact pressure in secondary lining is that the pressure at the sidewalls is the largest, followed by the pressure at the hance, the minimum pressure at the vault. The average rock pressure in test section of Z5b is the largest, and the rock pressure in Z5c and Z5 are relatively small with little difference; (2) from the changes of pressure, the values in Z5b need a period of time to stabilize gradually, while it needs more time to be stable in Z5c and Z5. Thus, it can be obtained that the large stiffness of initial lining is conducive to helping speed up the stability of surrounding rock and tunnel structure to a certain extent; (3) for soft rock tunnel, the secondary lining is often required to bear a greater proportion of the load, playing a role in bearing the late rock pressure. From the monitoring results in Table 5, the proportion of rock pressure born by secondary lining in the region of Z5b is as much as 79.31 %, and the rock pressure tends to stabilize quickly, which is consistent with the principles of construction in soft rock. The proportions born by secondary lining in Z5c and Z5 are respectively 43.05 and 43.58 %, and it needs a long time to be stable for the two support schemes; (4) from the suffered pressure, the values in the regions of Z5c and Z5 are smallest, indicating that the stress release rate of surrounding rock is so large that it easily leads to the adverse geological problems such as large deformation. While, the rock pressure in Z5b is the largest, which is conducive to protecting the integrality of surrounding rock and the stability of surrounding rock structure.

  1. (3)

    Monitoring results analysis of deformation in surrounding rock

Through the study on statistical result of rock deformation in three kinds of support schemes from Table 6, the following conclusions can be obtained: (1) the convergence and subsidence in the region of Z5b are significantly lower than that in Z5c and Z5, and the deformation in Z5c and Z5 have little difference from the monitoring data. With the augmentation of the support parameters, the deformation and the deformation rate of the vault subsidence and the horizontal convergence in upper bench and lower bench show the law of a significant decrease, and the horizontal change has an evident trend. All this indicates that strengthening the support can play an effective role in controlling the large deformation of surrounding rock, while weakening the anchor has little effect; (2) through the comparative analysis of the deformation in different parts, it can be obtained that the horizontal convergence value and rate are larger than that in the vault, indicating the surrounding rock is subjected to strong horizontal stress; (3) by comparing and analyzing the effect of controlling large deformation in three kinds of support parameters, the support form of Z5b, which has a better effect on controlling the deformation of surrounding rock, is obviously better than Z5c and Z5.

Table 6 Statistics for the deformation of tunnel in different support parameters
Full size table

5 Discussion and Analysis

By analyzing the deformation, the anchor force, the development of plastic zone, the stress of sprayed concrete and the force of secondary lining in three support forms of Z5c (weakening the anchor), Z5b (weakening the anchor while enhancing the stiffness and strength of initial lining) and Z5 (original design) with numerical simulation, it shows that:(1) the vault subsidence and horizontal convergence of surrounding rock in three kinds of support schemes are in the same order of magnitude with little difference, indicating that weakening the anchor has little impact on support effect. Also the overall deformation in the support form of Z5b is smaller than the other two support forms, indicating that the initial support structure has a significant effect on controlling the deformation of surrounding rock after strengthening the support and respectively has great influence on the vault subsidence during the excavation of upper bench and the horizontal convergence during the excavation of lower bench; (2) the stress distribution in anchor is basically the same in three kinds of support schemes, and the maximum values are all at the skewback of upper bench. The anchor force within the range of 45° at left and right sides of vault is very small or in a compressed state; (3) the three support forms of Z5c, Z5b, Z5 all have a good effect on controlling the development of plastic zone in surrounding rock, and under the premise of weakening the anchor system, enhancing the stiffness and strength of initial lining can reduce the development of plastic zone more. From the above analysis, it can be obtained that weakening the anchor has little impact on the deformation and plastic zone of surrounding rock in elastic–plastic model, indicating that this method is feasible in engineering applications, but the anchor below 45° at left and right sides of the vault should be strengthened; (4) with the solution in elastic–plastic model, the bending stress and shear stress in initial support structures are both small, and so is each principal stress in secondary lining. In short, the overall force in Z5b is relatively small. The above analysis shows that weakening the anchor and enhancing the strength and stiffness of initial lining have a better effect on controlling the deformation of tunnel.

Through long-term monitoring and the analysis of monitoring data, the conclusions can be obtained as follows: with comprehensive comparisons and analysis of rock pressure in initial lining and secondary lining, the deformation and deformation rate of each monitoring section in different support parameters, it can be drawn that the support form of Z5b can effectively bear the load of surrounding rock and keep the rock intact so as to stabilize the deformation and pressure of surrounding rock rapidly, playing an effective role in controlling large deformation of surrounding rock. Meanwhile, weakening the anchor has little impact on the overall support effect, which has been proved by the field monitoring results and the simulation results. Therefore, the support scheme of “weakening anchor bolts while enhancing the stiffness and strength of initial lining” is feasible in engineering applications.

6 Conclusions

  1. 1.

    After contrastive analysis of the deformation, the anchor force, the development of plastic zone, the stress of sprayed concrete and secondary lining in three support schemes of Z5c (weakening the anchor), Z5b (weakening the anchor, enhancing the stiffness and strength of initial support), and Z5 (original design) with numerical simulation, the support scheme of Z5b is considered to have a better support effect.

  2. 2.

    With comprehensive comparisons and analysis of rock pressure in initial lining and secondary lining, the deformation and deformation rate of each monitoring section in different support parameters, it can be drawn that the support form of Z5b can effectively bear the load of surrounding rock and keep the rock intact so as to stabilize the deformation and pressure of surrounding rock rapidly, playing an effective role in controlling large deformation of surrounding rock. Meanwhile, weakening the anchor has little impact on the overall support effect, which has been proved by the field monitoring results and the simulation results.

  3. 3.

    Through the verification of numerical simulation and field test results, weakening the anchor can not only have little impact on the overall support effect of support system, but also reduce the construction process and project cost.

  4. 4.

    As the buried depth of tunnel and in situ stress are both small, the support scheme of ‘‘enhancing the stiffness and strength of initial lining” can effectively control large deformation in this type of surrounding rock. Therefore, the proposed scheme of “weakening anchor bolts while enhancing the stiffness and strength of initial lining” is feasible and can provide a reference for similar projects.