Mechanical Characteristics of an Open-Buried Double-Arch Tunnel during Construction

: The open excavation and concealed construction method is widely adopted for the construction of bias double-arch tunnels. However, the mechanical behavior of the tunnel during the whole construction period by using the open excavation and concealed construction method is not well understood


Introduction
Because of its advantages of a smooth alignment, a small footprint, and excellent space utilization, the double-arch tunnel becomes one of most widely used types of tunnel structures, and it is frequently employed in highway engineering. Thus, the mechanical behavior of the tunnel is regarded by many scholars in terms of the deformation laws of the tunnel lining and surrounding rock, as well as the control measures to restrain the deformation of the tunnel and surrounding rock [1][2][3][4].
Generally, the engineering behaviors of different geotechnical structures can be well reflected by numerical simulation [5][6][7] and experimental investigation [8][9][10]. Zhao et al. [11] investigated the deformation characteristics of the crossing section of two municipal tunnels subjected to construction load by on-site experiments. Chang et al. [12] investigated the behavior of a tunnel caused by adjacent excavation and established a reference for excavation standard. Sharma et al. [13] discussed the influence of the stiffness of tunnel lining on the displacement and distortion by on-site investigation. Fahimifar et al. [14] predicted the deformation of the surrounding rock around the tunnel based on different calculation models or methods.
The construction method for the double-arch tunnel includes the double-side-drift method, the CD method, the benching method, and the CRD method [15,16]. The different methods are suited for different engineering geological conditions, and the suitability of different methods for the construction tunnels has been discussed by many scholars [17][18][19][20][21].

Tunnel on Xiajuan Road
The double-arch municipal tunnel with a two-way four-lane structure is located on Xiajuan Road, Changsha City, China. The mileage range of the tunnel is K2 + 060~K2 + 380 with a total length of 320 m, and the concealed section is about 280 m in length. The width and the height of tunnel structure are 14.75 m and 5.5 m, respectively. The cross-section area of internal profile of a single tunnel cave is 154.39 m 2 . The tunnel is oriented from north to south, and the overall topography of construction site shows a high elevation in the south and a low elevation in the north. The inlet and the cave section are high in the west and low in the east. A composite tunnel lining is adopted according to the principle of NATM, in which the primary lining adopts a 15 × 15 bar-mat reinforcement of Φ8, I22b steel arch and C25 shotcrete, and the secondary lining is supported by molded reinforced concrete with a thickness of 70 cm.

Engineering Geological Condition
The stratigraphic distribution in construction site mainly includes Plant layer, Quaternary sedimentary layer, Quaternary relict deposits, and underlying bedrock. The underlying bedrock is paleoproterozoic muddy and sandy slate. The lithological characteristics of stratigraphic layer are described in Table 1. The location, longitudinal and cross profiles, and geological conditions of the tunnel project are shown in Figure 1. Table 1. Summary of stratigraphic lithology characteristics.

Number
Stratigraphic Types Characteristic Description 1 Plant layer This layer is brown, tawny, and mainly composed of clayey soil interspersed with plant roots and mixed with a small amount of gravel in a slightly wet and loose state, which is mainly distributed on the surface of the mountain. The thickness of this layer is in the range of 0.1~0.30 m. 2

Quaternary freshly deposited silt
This layer is grey-black, with a wet-saturated and flow plastic state. Its dry strength and toughness are low, with a layer thickness of <3.00 m. 3 Quaternary residual deposit silty clay This layer is brownish-yellow or reddish-brown, locally interspersed with weathered rock blocks that have not weathered completely. It is in a slightly wet, hard plastic state and its dry strength and toughness are medium. Its layer thickness is 0.50-7.60 m, and its average thickness is 2.25 m. 4 Paleoproterozoic slate Within the exploration depth, this layer can be divided into muddy slate and sandy slate according to its mineral composition.
The cavern section is surrounded by soil, strongly weathered rock layers, and medium weathered rock layers, with large differences in rock hardness and integrity. Consequently, the grade of the surrounding rock is determined as V.

Construction Process of the Tunnel Entrance
The tunnel entrance section of Xiajuan Road in Changsha City, China (mileage range K2 + 060~K2 + 100) shows a great difference in burial depth between the two sides of the The cavern section is surrounded by soil, strongly weathered rock layers, and medium weathered rock layers, with large differences in rock hardness and integrity. Consequently, the grade of the surrounding rock is determined as V.

Construction Process of the Tunnel Entrance
The tunnel entrance section of Xiajuan Road in Changsha City, China (mileage range K2 + 060~K2 + 100) shows a great difference in burial depth between the two sides of the cavern. The left cavern is very shallow, with more than 1/2 of the cavern exposed to the surface, while the right cavern is located inside the mountain (burial depth of 2~6 m), forming a distinctive open-buried double-arch tunnel. Because of a serious topographic deviation in this section, a deflecting retaining wall was added on the left side of the left cavern to prevent tunnel instability or other disasters during the tunnel construction and service period. When the linings in both caverns were completed, soil backfill was performed above the left cavern to balance the topographic deviation of the mountain. The left cavern of the entrance section adopted an open excavation and a concealed construction method. Before the initial support of the left cavern, the excavated soil from other places was back-filled into the left cavern, so as to establish the up-step steel arch. The right cavern was constructed by the concealed excavation method. Both the left and the right caverns were excavated using the thee-step ring-cut method. The left and right caverns were constructed in order, keeping the staggered distance of the cavern palm surface more than twice that of the tunnel span, and the longitudinal distance of the upper and lower steps should be less than 5 m. Meanwhile, the longitudinal distance of each cavern was adjusted in time according to the monitoring results. To make the construction process clear, Figure 2 shows the construction schedule in detail, and the corresponding construction is shown in Table 2.
surface, while the right cavern is located inside the mountain (burial depth of 2~6 m), forming a distinctive open-buried double-arch tunnel. Because of a serious topographic deviation in this section, a deflecting retaining wall was added on the left side of the left cavern to prevent tunnel instability or other disasters during the tunnel construction and service period. When the linings in both caverns were completed, soil backfill was performed above the left cavern to balance the topographic deviation of the mountain. The left cavern of the entrance section adopted an open excavation and a concealed construction method. Before the initial support of the left cavern, the excavated soil from other places was back-filled into the left cavern, so as to establish the up-step steel arch. The right cavern was constructed by the concealed excavation method. Both the left and the right caverns were excavated using the thee-step ring-cut method. The left and right caverns were constructed in order, keeping the staggered distance of the cavern palm surface more than twice that of the tunnel span, and the longitudinal distance of the upper and lower steps should be less than 5 m. Meanwhile, the longitudinal distance of each cavern was adjusted in time according to the monitoring results. To make the construction process clear, Figure 2 shows the construction schedule in detail, and the corresponding construction is shown in Table 2.  Mid-step excavation in the right cavern and primary lining construction 16 Construction of the secondary lining of the right cavern Excavation and primary lining

Monitoring Objects and Content
According to the condition of construction site, a typical monitoring section is selected at the mileage of K2 + 097. The stress and strain of the supporting structure are investigated, including the strain of steel arch in primary lining, the stress of rebar and the strain of concrete in secondary lining. Meanwhile, the deformation of the surrounding rock is also observed, including the vault settlement and the tunnel perimeter convergence. The layout of measurement points and the point number at the tunnel cross-section are shown in Table 3 and Figure 3.

Monitoring Objects and Content
According to the condition of construction site, a typical monitoring lected at the mileage of K2 + 097. The stress and strain of the supporting investigated, including the strain of steel arch in primary lining, the stress o strain of concrete in secondary lining. Meanwhile, the deformation of the rock is also observed, including the vault settlement and the tunnel peri gence. The layout of measurement points and the point number at the tunne are shown in Table 3 and Figure 3.

Component Arrangements
According to existing reports [30][31][32], the measurement of stress or strain at the steel arch was mostly conducted by arranging steel stress gauges or strain gauges parallel to the inner and outer flange sides of the steel arch. Thereafter, the test results were converted into axial force or bending moment of the steel arch, and then the internal force of the steel arch could be analyzed. There were few experimental results related to the force or deformation investigation in any abdominal region of the steel arch. If the strain gauges are arranged on inner and outer flanges of the steel arch in parallel, the changing trend of the contact pressure between the liners cannot be reflected sensitively. The strain gauge arrangement on the web axis of the steel arch can not only reflect the deformation of the steel arch effectively, but also indirectly reflect the changing trend of the contact pressure between the liners.
A fixing device was utilized to weld the strain gauge to the center of web along the axis of steel arch, and a protective shell was welded to steel arch to ensure that the strain gauge would not be damaged during the shotcrete process, as shown in Figure 4. The stress gauge was welded to the reinforcement of secondary lining in the direction of tunnel circumference to measure the main circumferential axial force of reinforcement. The strain gauges were fixed to the reinforcement of secondary lining by using ties at both ends to measure the strain of the second liner concrete. The vault settlement was measured by a micrometer with a high precision. The tunnel perimeter convergence was measured by a digital convergence meter. 6 would not be damaged during the shotcrete process, as shown in Figure 4 gauge was welded to the reinforcement of secondary lining in the direction of t ference to measure the main circumferential axial force of reinforcement. The s were fixed to the reinforcement of secondary lining by using ties at both en ure the strain of the second liner concrete. The vault settlement was measured meter with a high precision. The tunnel perimeter convergence was measured l convergence meter. he on-site test data were collected corresponding to the construction progress ity was achieved. More than 240 days were spent on this on-site monitoring. To y illustrate the force and the deformation evolution of the supporting structur unding rock during the construction, the construction cases were numbered ac

Construction Cases
The on-site test data were collected corresponding to the construction progress until stability was achieved. More than 240 days were spent on this on-site monitoring. To more clearly illustrate the force and the deformation evolution of the supporting structure and surrounding rock during the construction, the construction cases were numbered according to the construction sequence and time, as shown in Table 4. There are mainly nine construction cases which are numbered as Cases 1 ~9 , and four construction stages marked by Stages I~IV. The ending of soil backfills 5 October Figure 5 shows the strain-time distribution of the steel arch in the monitoring section of the tunnel, in which tension strain is represented by a positive value, and compression strain is marked by a negative value. The strain is assumed as zero when the upper steel arch is erected along the upper step. Combined with the arrangement of the measurement points in the tunnel section (see Table 3 and Figure 3) and the division of the construction cases (see Table 4), the variation of the strain at the steel arch is analyzed corresponding to different construction stages, as follows. steel arch to increase rapidly. When the secondary-lining trolley is removed, there is a certain release of interaction force between the linings. Meanwhile, the secondary lining concrete at this time tends to be stable due to its solidification. It is worth noting that the strain of the steel arch at each measurement point of the right cavern did not increase repeatedly due to the soil excavation when the secondary lining of the left cavern was completed. The strain tends to be stable after it slightly decreases during the construction of the secondary lining of the right cavern. It is seen that the supporting strength at the open cavern side in the open-buried double-arch tunnel is effectively utilized, which is of great significance to guarantee the stability of the tunnel structure system. (4) Construction stage IV. When the secondary linings of the left and right caverns reach the design strength, the soil is back-filled on the left cavern (open cavern). It is seen that the backfill presents different effects on the strain at the steel arch in both caverns. Soil backfill increases the strain at the steel arch of the left cavern again, which is especially significant at the LW, while the strain at the steel arch of the right cavern decreases at the end of backfill. Soil backfill will increase the surrounding rock pressure around the left cavern, which indicates that it causes a transformation from an "open" cavern into a "buried" cavern. The bias pressure caused by the slope is balanced, which subsequently reduces the strain at the steel arch of the right cavern, and increases the strain at the left cavern. During the whole process, the strain at each measurement point is still within the safe range.

Vault Settlement
The vault settlement was continuously monitored after the steel arch support at the up-step was established in the left and right caverns. The value of the vault settlement and its sinking rate are shown in Figure 6.
(1) The process of vault settlement can be divided into three stages: the rapid development stage within 15 d after the construction of up-step, the continuous development stage before the application of the invert, and the gradual stabilization stage after the application of the invert. The vault settlement at each stage accounts for about 50%, 30%, and 20% of the total settlement. After the application of the invert, the sinking rate gradually tends to be stable.  (1) Construction stage I. The up-step of the left tunnel is located above the original hillside surface. Consequently, after the construction of the deflecting retaining wall, the primary lining at the up-step of the left tunnel is constructed (as shown in Figure 2). It is seen that after the completion of the primary lining, the strain at the steel arch shows a rapid growth trend at the early stage due to the effect of the concrete, formwork, and its own gravity. The strain development at each potion of the steel arch gradually flattens out with time. It is worth noting that the strain at the remaining measurement point continues to increase while the strain at the LW and the LS of the steel arch decreases because the right side of the steel arch that overlaps with the mid-partition wall shares part of the right side of the surrounding rock pressure. However, when the mid-step of the left cavern is excavated, the strain at the LW and the LS increases simultaneously, and the increasing trend is smaller than at other positions. This indicates that the soil displacement of the middle-lower parts of both sides of the left cavern are caused by the excavation of the mid-step, driving a stress adjustment of the surrounding rock and squeezing the steel arch to a certain extent. Under the deflecting effect of right hillside body, the mid-partition wall on the right side shows a tendency to deflect to the left cavern after the removal of surrounding rock at the left side. Consequently, the pulling and squeezing effect at the steel arch on the same side is more obvious. During the construction of the lower step of the cavern, the above phenomenon becomes more obvious, and the strain growth increases significantly. It is seen that after The steel arch strain increases rapidly when it begins to bear the surrounding rock pressure. As the up-step of the right cavern continues to be excavated, the strain of the steel arch at the RW and the RS is essentially stable, while the other points continue to increase until the construction of the invert begins. The reason for the above phenomenon is that the right side of the right cavern is on a higher elevation of mountain topography, while the left side is reduced in height and linked with the mid-partition wall. The caverns are in a right-to-left skewed topography. After the excavation of the up-step and the invert in the right cavern, the surrounding rock deforms towards the left side with a stress adjustment, and the left side of the steel arch bears the surrounding rock pressure. It is seen that the location of the joint area (near the mid-partition wall) is a sensitive area during the buried-cavern construction of the open-buried double-arch tunnel. (3) Construction stage III. The strain at the steel arch increases rapidly in both caverns within 3 days after the secondary lining is applied, and reaches the maximum value during the monitoring period. After 5 days, the strain at the steel arch of the left and right caverns decreases rapidly until they are essentially stable. The reason for the above phenomenon is that contact pressure arises between the primary and the secondary linings due to gravity and the solidification of the concrete when the secondary lining is completed [33]. While the secondary-lining trolley is still acting on the secondary lining, the contact pressure between the linings causes the strain at the steel arch to increase rapidly. When the secondary-lining trolley is removed, there is a certain release of interaction force between the linings. Meanwhile, the secondary lining concrete at this time tends to be stable due to its solidification. It is worth noting that the strain of the steel arch at each measurement point of the right cavern did not increase repeatedly due to the soil excavation when the secondary lining of the left cavern was completed. The strain tends to be stable after it slightly decreases during the construction of the secondary lining of the right cavern. It is seen that the supporting strength at the open cavern side in the open-buried double-arch tunnel is effectively utilized, which is of great significance to guarantee the stability of the tunnel structure system.

(4) Construction stage IV. When the secondary linings of the left and right caverns reach the design strength, the soil is back-filled on the left cavern (open cavern)
. It is seen that the backfill presents different effects on the strain at the steel arch in both caverns. Soil backfill increases the strain at the steel arch of the left cavern again, which is especially significant at the LW, while the strain at the steel arch of the right cavern decreases at the end of backfill. Soil backfill will increase the surrounding rock pressure around the left cavern, which indicates that it causes a transformation from an "open" cavern into a "buried" cavern. The bias pressure caused by the slope is balanced, which subsequently reduces the strain at the steel arch of the right cavern, and increases the strain at the left cavern. During the whole process, the strain at each measurement point is still within the safe range.

Vault Settlement
The vault settlement was continuously monitored after the steel arch support at the up-step was established in the left and right caverns. The value of the vault settlement and its sinking rate are shown in Figure 6.

Peripheral Displacement
Peripheral displacement refers to the displacement of measurement points in surrounding rock within the tunnel boundary after excavation. It is difficult to accurately measure the absolute value of displacement of spatial points. Consequently, it is common to determine the relative displacement of two points along the direction of inner wall of the tunnel. The peripheral displacement of the surrounding rock of the tunnel is reflected by monitoring the relative displacement of the points along the upper and lower horizontal lines. The peripheral displacement vs. time on the left and right caverns of the tunnel is shown in Figure 7, in which the negative value indicates an inward convergence, and the positive value refers to an outward extension deformation. The positions of the upper horizontal line (UH) and the lower horizontal line (LH) are shown in Table 3 and Figure  3.
(1) The peripheral displacement around the tunnel shows an inward convergence, and then experiences an outward extension. After the construction of the up-step of the tunnel, the peripheral displacement of the left and right caverns tends to move inwards. The main reason for this phenomenon is that the mid-partition wall moves towards the shallowly buried side (left cavern) subjected to the deflection of slope. Regarding the right cavern, after the construction of the up-step of the right cavern, (1) The process of vault settlement can be divided into three stages: the rapid development stage within 15 d after the construction of up-step, the continuous development stage before the application of the invert, and the gradual stabilization stage after the application of the invert. The vault settlement at each stage accounts for about 50%, 30%, and 20% of the total settlement. After the application of the invert, the sinking rate gradually tends to be stable. The open-excavation and concealed construction method is effective to control the settlement of the tunnel during the construction of the tunnel.

Peripheral Displacement
Peripheral displacement refers to the displacement of measurement points in surrounding rock within the tunnel boundary after excavation. It is difficult to accurately measure the absolute value of displacement of spatial points. Consequently, it is common to determine the relative displacement of two points along the direction of inner wall of the tunnel. The peripheral displacement of the surrounding rock of the tunnel is reflected by monitoring the relative displacement of the points along the upper and lower horizontal lines. The peripheral displacement vs. time on the left and right caverns of the tunnel is shown in Figure 7, in which the negative value indicates an inward convergence, and the positive value refers to an outward extension deformation. The positions of the upper horizontal line (UH) and the lower horizontal line (LH) are shown in Table 3 and Figure 3.
the mid-partition wall moves to the right because of the construction deflection pressure caused by tunneling construction between the caverns. A similar phenomenon is also discussed by Duan et al. [35] with regard to a bias-pressure double-arch tunnel, where the bias effect caused by the asymmetric construction is not considered. When the plastic zone of the surrounding rock is initially formed, the primary lining begins to work, and the relative displacement of the cavern gradually moves toward both sides.
(2) The displacement development pattern of the tunnel cavern is different when subjected to different construction stages. For the first half of the month, because of a rapid stress release of the surrounding rock, the maximum peripheral displacement rate occurs at the early stage, and it moves toward the inner side. Compared with the construction of the lower step, the mid-step construction shows a greater effect on the accumulated displacement and the displacement rate at the UH. The effect of the invert construction on the accumulated periphery displacement and the displacement rate at the UH is greater than the construction of the mid-step. The periphery displacement rate of the left and right caverns converges gradually after the invert construction, which indicates that the timely construction of the invert plays an important role in controlling the displacement of the primary lining and surrounding rock. (3) The accumulated peripheral displacement around the UH is larger than the LH because the deflecting retaining wall and the mid-partition wall restrict the horizontal displacement of the lower ling structure and surrounding rock. The accumulated peripheral displacement around the UH is smaller than the LH since there is no deflecting retaining wall on the right side of the right cave, which means that no restraint exists outside the right cave.

Correlation Analysis
The deformation and stress release of the surrounding rock leads to a settlement in the upper area of the primary lining. According to the test data of the sinking rate of the vault settlement and the steel arch strain, the correlation characteristics between the vault settlement rate and the steel arch strain are explored in Figure 8.
A rapid development of the vault settlement sinking rate and strain on the steel arch is observed within one week after the erection of the up-step steel arch because the shotcrete does not fully work yet. The surrounding rock pressure is mainly borne by the steel arch. When the shotcrete begins to work, the strength and stiffness of the primary lining (1) The peripheral displacement around the tunnel shows an inward convergence, and then experiences an outward extension. After the construction of the up-step of the tunnel, the peripheral displacement of the left and right caverns tends to move inwards. The main reason for this phenomenon is that the mid-partition wall moves towards the shallowly buried side (left cavern) subjected to the deflection of slope. Regarding the right cavern, after the construction of the up-step of the right cavern, the mid-partition wall moves to the right because of the construction deflection pressure caused by tunneling construction between the caverns. A similar phenomenon is also discussed by Duan et al. [35] with regard to a bias-pressure double-arch tunnel, where the bias effect caused by the asymmetric construction is not considered. When the plastic zone of the surrounding rock is initially formed, the primary lining begins to work, and the relative displacement of the cavern gradually moves toward both sides.
(2) The displacement development pattern of the tunnel cavern is different when subjected to different construction stages. For the first half of the month, because of a rapid stress release of the surrounding rock, the maximum peripheral displacement rate occurs at the early stage, and it moves toward the inner side. Compared with the construction of the lower step, the mid-step construction shows a greater effect on the accumulated displacement and the displacement rate at the UH. The effect of the invert construction on the accumulated periphery displacement and the displacement rate at the UH is greater than the construction of the mid-step. The periphery displacement rate of the left and right caverns converges gradually after the invert construction, which indicates that the timely construction of the invert plays an important role in controlling the displacement of the primary lining and surrounding rock. (3) The accumulated peripheral displacement around the UH is larger than the LH because the deflecting retaining wall and the mid-partition wall restrict the horizontal displacement of the lower ling structure and surrounding rock. The accumulated peripheral displacement around the UH is smaller than the LH since there is no deflecting retaining wall on the right side of the right cave, which means that no restraint exists outside the right cave.

Correlation Analysis
The deformation and stress release of the surrounding rock leads to a settlement in the upper area of the primary lining. According to the test data of the sinking rate of the vault settlement and the steel arch strain, the correlation characteristics between the vault settlement rate and the steel arch strain are explored in Figure 8. are greatly increased, and its proportion for the surrounding rock pressure is also greatly increased. The strain growth rate of the steel arch slows down until stabilization. The supporting function of the steel arch and shotcrete effectively restrains the deformation of the surrounding rock. Thus, the vault settlement rate is greatly reduced. It is seen that (1)

Reinforcement Stress of the Secondary Lining
The variation regulation of the reinforcement stress is consistent with that of the concrete strain as for the secondary lining in the open-buried double-arch tunnel. The reinforcement stress of the secondary lining is shown in Figure 9, in which the positive value indicates tension stress and the negative value refers to compressive stress. The initial data derive from the test data when the component is installed.
(1) The mechanical properties of the reinforcement in the secondary lining are measured throughout the process of concrete casting and maintenance. After the secondary lining concrete is cast in the left and right caverns of the tunnel, the secondary lining forms a temperature gradient from the inside to the outside due to a large amount of heat released from the hydration heat reaction of the concrete. When the concrete cools down and shrinks from a high temperature, this results in tensile stress (tem- A rapid development of the vault settlement sinking rate and strain on the steel arch is observed within one week after the erection of the up-step steel arch because the shotcrete does not fully work yet. The surrounding rock pressure is mainly borne by the steel arch. When the shotcrete begins to work, the strength and stiffness of the primary lining are greatly increased, and its proportion for the surrounding rock pressure is also greatly increased. The strain growth rate of the steel arch slows down until stabilization. The supporting function of the steel arch and shotcrete effectively restrains the deformation of the surrounding rock. Thus, the vault settlement rate is greatly reduced. It is seen that (1) the value of the vault settlement rate in the left cavern shows that LM > RM. Additionally, the strain of the steel arch in the left cave shows that RS > LS. (2) The value of the vault settlement rate in the right cavern shows that RM > LM, and the strain of the steel arch in the right cavern shows that LS > RS.
The reason for the above phenomenon can be concluded as follows. The values of the burial depth at the arch shoulders of each cavern are similar, although the burial depth for the left and right caverns is different (the left cavern is an open cavern where the burial depth of the arch shoulders is 0; the burial depth at the arch shoulders of the right cavern is about 6 m). Without considering any other influencing factors, the values of the surrounding rock pressure on both sides of the caverns are approximately equal. The settlement rate of the vault is related to the surrounding rock pressure. A larger value of the surrounding rock pressure will increase the deformation and settlement rate.
The stress and strain of the steel arch are mainly caused by the surrounding rock pressure. An increasing strain at the steel arch indicates that more surrounding rock pressure is undertaken by the steel arch. The steel arch plays a supporting role and limits the displacement trend and the deformation rate of the surrounding rock.

Reinforcement Stress of the Secondary Lining
The variation regulation of the reinforcement stress is consistent with that of the concrete strain as for the secondary lining in the open-buried double-arch tunnel. The reinforcement stress of the secondary lining is shown in Figure 9, in which the positive value indicates tension stress and the negative value refers to compressive stress. The initial data derive from the test data when the component is installed.
foot > inner arch foot > outer arch waist > inner arch waist. The reinforcement stress at the arch waist of the right cavern is smaller than that in left cavern, while the reinforcement stress at other measurement points of the right cavern is larger than those of the left cavern. Due to the supporting effect of the deflecting retaining wall and the mid-partition wall, the topographic bias will mainly affect the secondary lining of the buried cavern, which effectively controls the bias instability in the open-buried double-arch tunnel.
(3) Soil backfill presents different effects on the forces of the secondary lining of the left and right caverns. After soil backfill, the surrounding rock pressure above the left cavern increases, which causes the pressure at the V, LF, and RF of the left cavern to increase, especially for the points at the V. The pressure at the LW and the RW of the left cavern shows a decreasing trend. After soil backfill, the vault is firstly subjected to the surrounding rock pressure, which will be transferred to the lower part of the structure (i.e., the arch foot). The inner arch effect produced the waist of the cavern due to a supporting effect from the deflecting retaining wall and the mid-partition wall. Unlike the left cavern, the pressure of the right cavern decreases briefly, and then increases continuously at all measurement points after soil backfill (the soil backfill is above the left cavern), especially for the points at V and RF. The deflecting loading of slope is balanced to some extent after soil backfill above left cavern. A stress adjustment occurs in surrounding rock of right cavern near the right side of slope, and the direction and the size of secondary lining of right cavern change at the same time. When the secondary lining is located at the surrounding rock of right arch foot, the force at the RF and V is most concentrated. At the end of backfilling, the reinforcement stress in the secondary lining of the left and right caverns tends to stabilize and reduce simultaneously, indicating that the interaction between the surrounding rock (including the backfill soil) and the internal linings of the tunnel is gradually balanced and stabilized.  (1) The mechanical properties of the reinforcement in the secondary lining are measured throughout the process of concrete casting and maintenance. After the secondary lining concrete is cast in the left and right caverns of the tunnel, the secondary lining forms a temperature gradient from the inside to the outside due to a large amount of heat released from the hydration heat reaction of the concrete. When the concrete cools down and shrinks from a high temperature, this results in tensile stress (temperature stress) [36], causing a significant increase in the tensile stress in the secondary lining. When the maintenance of the secondary lining concrete is completed, it presents a dominated compressive stress. (2) Compared with the left cavern, the distribution of stress in the right cavern is more discrete, and a great difference is observed among different measurement points.
The left shoulder of the right cavern is dominated by tensile stress. The supporting effect of the mid-partition wall leads to a relatively concentrated reinforcement stress at the location of the arch waist near the side of the mid-partition wall in the left and right caverns. If the mid-partition wall is taken as a central axis, the value of the reinforcement stress in the left and right caverns can be sorted as follows: vault > outer arch foot > inner arch foot > outer arch waist > inner arch waist. The reinforcement stress at the arch waist of the right cavern is smaller than that in left cavern, while the reinforcement stress at other measurement points of the right cavern is larger than those of the left cavern. Due to the supporting effect of the deflecting retaining wall and the mid-partition wall, the topographic bias will mainly affect the secondary lining of the buried cavern, which effectively controls the bias instability in the open-buried double-arch tunnel. (3) Soil backfill presents different effects on the forces of the secondary lining of the left and right caverns. After soil backfill, the surrounding rock pressure above the left cavern increases, which causes the pressure at the V, LF, and RF of the left cavern to increase, especially for the points at the V. The pressure at the LW and the RW of the left cavern shows a decreasing trend. After soil backfill, the vault is firstly subjected to the surrounding rock pressure, which will be transferred to the lower part of the structure (i.e., the arch foot). The inner arch effect produced the waist of the cavern due to a supporting effect from the deflecting retaining wall and the mid-partition wall. Unlike the left cavern, the pressure of the right cavern decreases briefly, and then increases continuously at all measurement points after soil backfill (the soil backfill is above the left cavern), especially for the points at V and RF. The deflecting loading of slope is balanced to some extent after soil backfill above left cavern. A stress adjustment occurs in surrounding rock of right cavern near the right side of slope, and the direction and the size of secondary lining of right cavern change at the same time. When the secondary lining is located at the surrounding rock of right arch foot, the force at the RF and V is most concentrated. At the end of backfilling, the reinforcement stress in the secondary lining of the left and right caverns tends to stabilize and reduce simultaneously, indicating that the interaction between the surrounding rock (including the backfill soil) and the internal linings of the tunnel is gradually balanced and stabilized.

Conclusions
To study the mechanical behaviors of the supporting structure and surrounding rock of open-buried multi-arch tunnel, the strain and stress of linings and the deformation laws of the surrounding rock were studied by on-site investigation. The following conclusions are obtained: (1) The application of the secondary lining in the left cavern (open cavern) shows an inhibitory effect on the strain development of the steel arch in the right cavern (concealed cavern). The backfill balances the bias caused by the slope and reduces the strain at the steel arch of the concealed cavern. (2) The process of vault settlement can be divided into three stages: a rapid development stage within 15 d after the construction of the up-step, the continuous development stage before the application of the invert, and the stable convergence stage after the application of the invert. When the invert is applied, the settlement rate of the vault is gradually stabilized. The mid-partition wall and deflecting retaining wall effectively decrease the settlement of the primary lining and present a weakening effect on the bias effect. (3) The peripheral displacement of the tunnel experiences a process of convergence inward and extension outward. The peripheral displacement of the left and right caverns tends to move inward after the construction of the up-step. The deflecting retaining wall and mid-partition wall show a good restraining effect on the tunnel.