1 Introduction
With the progress of urban construction in China, a large number of underground facilities, such as basements for high-rise apartment buildings [
1], underground commercial centers [
2], tunnels and buried pipelines [
3], are being constructed. Stress redistribution induced by excavations can cause the occurrence of soil movements; if there is no proper control, excessive ground deformation can lead to severe damages in adjacent buildings and other underground facilities. Thereby, it is of importance for designer and contractor to control the ground deformation during excavation.
At present, many studies have been conducted to understand the mechanisms associated with excavations, including the deformation characteristics [
4–
8], calculation of stress in deep soils [
9,
10], and development of deep excavation methods [
11,
12]. Different research methods have been adopted, such as theoretical analysis [
13,
14], numerical simulation [
15–
17], physical model testing [
18], and analysis of field measurements [
19–
22]. It is confirmed that the use of excavation method can have a significant impact on the deformation of retaining structures and existing underground structures.
Based on the findings from previous studies [
5,
11], the semi-top-down (STD) excavation method performed better than the bottom−up and the conventional TD excavation techniques, and the STD excavation approach resulted in a lower chance for stability failure incidence. Goh et al. [
23] analyzed a series of two-dimensional and three-dimensional (3D) finite element simulation models, through which a maximum wall deflection chart for various excavation length/excavation width (
L/
B) ratios was derived, along with an estimation equation. Tan et al. [
24] studied a deep excavation case that was adjacent to a metro line through field monitoring and numerical simulations, and they summarized the characteristics of oversized deep excavation on metro line. Similar studies focusing on the deformations of excavations and adjacent buildings or underground structures in urban areas have been reported in Refs. [
25–
27]. Based on a data set of 30 deep excavation cases in Nanjing silty soil, Li et al. [
28] investigated the influence of 3D effects on wall deflection and ground settlement, and then they proposed three normalized models for depicting the ground settlement. To forecast the construction risks, based on several risk evaluation models, Lin et al. [
29] identified the potential risk factors for a deep excavation project in Tianjin, China. Circular and pit-in-pit excavation are also getting more and more attention due to the booming scale of underground development. Chen et al. [
30] reported an oversized excavation in Shenzhen (Guangdong), based on numerical simulations and long-term field monitoring, and they investigated the deformation characteristics of structures and adjacent buildings, and the features of ground settlements. Wang et al. [
31] investigated the
in situ measurement data of a 56 m deep excavation which was circular in Shanghai downtown, where a novel double-wall system (an internally circular diaphragm wall and a rectangular cut-off wall on the outside) promoted the success of the project with several advantages, such as the impact of safety and greening. Most published works are related with the behavior of excavations in soft clayey soils [
23,
32], since the magnitude of lateral deflection of retaining structure is often hard to control for such cases. However, the impact of temperature is usually neglected for deep excavations considering the duration of construction or the relatively stable local ambient conditions.
Seasonal frozen soil is widely distributed worldwide [
33], and there are many factors that can affect the freezing-thawing process of such soil, in terms of permeability, clay mineral composition, and particle size distribution, etc [
34]. Freezing temperature also has a great influence on the soil behavior. Christ et al. [
35] found that the freezing-thawing process could reduce frozen soil’s compressive strength, and its compression and shear strength could increase with the increase of freezing temperature. To sum up, little attention has been paid to deep excavations under multi-condition situations, especially for the degradation of sandy soil through winter, where the low temperature can cause soil freezing and a long duration of downtime, followed by soil thawing after winter and continued excavation.
This study analyzes the performance of a deep pit-in-pit excavation in Shenyang, China using the cut-and-cover technique, which is in a very close proximity to Shenyang Metro Line 1. Field measurements were collected, including the deformation of retaining structures, the ground settlements and the responses of adjacent metro tunnel. Furthermore, the deformation characteristics of retaining structures and influence zones of the ground, which have experienced freezing and thawing of sandy soil, were revealed. Finally, the deformation laws of retaining structures, the ground, and metro line through winter excavation were obtained.
2 Project overview
2.1 Site conditions
The investigated excavation was located in a commercial area of downtown Shenyang, China. As the largest commercial and financial center in northeast China, Shenyang is characterized by the sub-humid continental climate, and its annual precipitation mainly occurs from June to August. Shenyang’s winter usually lasts for nearly 6 months with an average temperature of −11 to 0 °C. Moreover, the average depth of frozen ground is 1.2 m, and the maximum frozen ground depth is 1.48 m.
There were different nearby buildings that could be affected by the pit-in-pit excavation of the project, i.e., the building for Industrial & Commercial Bank of China on the north, a residential area on the west, a commercial shopping center on the east, and the building for New World Commercial Center on the south. Therefore, the surface environments around the excavation were exceedingly complex, as illustrated in Fig.1. Meanwhile, Shenyang Metro Line 1 (linking Shenyang Railway Station and Taiyuanjie Station) was in a very close proximity to the deep excavation on the north side. The existing two metro lines were 15 m below the ground surface (BGS), and the distance between the tunnel centerlines was 13 m. Furthermore, the outside diameter and thickness of tunnel segment were 6.0 m and 300 mm, respectively. The minimum distance between tunnel and northern retaining structure was 1.29 m. In addition, Langqin Underpass was directly above Shenyang Metro Line 1 with a depth of 1.5 m below the ground surface, which was at a distance of 10.2 m from the north side of retaining piles.
2.2 Geology and soil parameters
The deep excavation project lied in Hunhe River alluvial-proluvial fan plain with a gentle landscape. According to the geological investigation report, the excavation site existed quaternary pore phreatic water, and the depth of groundwater was 23.2 m. Furthermore, the topsoil was mainly consisted of clay and a slight amount of bricks and bituminous pavement. Subsoil was mainly consisted of medium coarse sand, gravelly sand, and round gravel. The properties of soil are shown in Fig.2. There were 6 layers in this deep excavation. The first layer was plain fill with a 3.5 m thickness, and it contained some remnants of bricks and bituminous pavement. The second layer was slightly wet medium coarse sand and its thickness was 5.5 m, and it contained a small amount of cohesive soil. The third layer was formed by round gravel and its thickness was 9.3 m, in which the upper layer was slightly dense, while the lower layer was moderately dense. The fourth layer was a 3.5 m gravelly sand layer, and it was mingled with thin layers of round gravel and coarse sand. The fifth layer was medium coarse sand with a thickness of 2.6 m. The last layer was formed by round gravel, and it was generally hard with a particle size from 2 to 22 mm. The last layer was relatively thick, so that the drill did not penetrate into this layer.
2.3 Supporting scheme and in situ measurement
In this case, the area of above-ground structure was 70440 m
2, and the area of underground structure was 25698 m
2. Hence, the deep excavation included 4 floors of basement, and the size was 78.6 m × 94.6 m in plan view. The excavated depth was 19.65 m, and a pit-in-pit excavation with a depth of 22.15 m was 12 m to the north side. As pointed out by Tan et al. [
24], oversized excavations commonly incurred significant greater deformations and displacements than regular-sized ones with the same engineering conditions. Considering the large size of this excavation, it was a big challenge to control the deformations and displacements of existing structures within an allowable range.
The retaining structures of the excavation included cast-in-place concrete piles, combined with prestressed cables and steel pipe supports. The typical excavation section with retaining structures is presented in Fig.3. As the north side of the excavation was adjacent to the metro lines and Langqin Underpass, retaining structures on this side were designed to be double-row piles with 5-layer prestressed anchor cables. The diameter of double-row piles was 1 m, and the distance between two rows was 1.3 m. The depth of double-row piles was 15 m, rendering an embedment ratio of 0.76.
The first three layers of prestressed cables on the north side were adjacent to the metro lines, and the minimum distance to the south metro line was 2.12, 3.26, and 1.73 m, respectively. The south side of the excavation was extremely close to the New World Commercial Center, and the depth of the southern capping beams was 15.55 m. The diameter of cast-in-place concrete piles was 0.8 m, and the distance between adjacent piles was 1.2 m. There was one layer of prestressed anchor cables on the south side. The diameter of cast-in-place concrete piles was 0.8 m on the west side, and the distance between adjacent piles was 1.3 m. A total of five layers of prestressed anchor cables were conducted on the west side.
The retaining form of the east side was similar to that of the west side, where cast-in-place concrete piles with a 0.8 m diameter and a 1.3 m pile spacing were designed. In addition, a total of five layers of prestressed anchor cables were designed on the east side. Since the northeast side of the excavation was adjacent to the underpass exit, three steel pipe bracings with an outside diameter of 609 mm and a thickness of 12 mm were set at the northeast corner of the excavation at a depth of 4 m.
As noted in Fig.4, the excavation schedule was divided into 7 stages: S-1 (from October 27, 2014 to December 16, 2014, for 50 d), S-2 (from December 17, 2014 to January 6, 2015, for 21 d), S-3 (from January 7, 2015 to February 10, 2015, for 35 d), S-4 (from February 11, 2015 to March 10, 2015, for 28 d), S-5 (from March 11, 2015 to April 14, 2015, for 35 d), S-6 (from April 15, 2015 to May 19, 2015, for 35 d), and S-7 (from May 20, 2015 to June 30, 2015, for 42 d). In total, the duration of this deep excavation was 246 d, from October 27, 2014 to June 30, 2015. Shenyang was in winter for 170 d from October 26, 2014 to April 14, 2015. According to the records, the average temperature in the whole winter was −3.9 °C, of which the average temperature at stage S-4 was −5.3 °C. We defined the “Winter Construction Stage” which includes construction stages of S-1, S-2 and S-3. It can be seen that the deep excavation case was exposed to a low temperature environment for quite a long time, and the impact of significant temperature variation could be evident. Essentially, the ground had experienced a whole freezing-thawing process through winter, adding more difficulty in the control of excavation-induced deformations.
The instrumentation layout for both excavation project and metro lines are shown in Fig.5. The excavation was divided into 7 zones: W-1, W-2, N-1, N-2, E-1, E-2, and S. All monitoring items are summarized as follows. 1) Horizontal displacement of capping beams (designated as C1 to C18). 2) Lateral displacement of piles (designated J31 to J243). 3) Vertical displacement of piles (designated as C1 to C18). 4) Ground settlement behind retaining piles (designated as L1 to L11). 5) Track settlements of 15 sections of the south tunnel (designated as R1 to R15). 6) Convergences of 7 tunnel rings (designated as R4 to R10).
3 Excavation-related responses
3.1 Retaining structures
3.1.1 Lateral displacement of capping beam
The inward horizontal displacement of capping beam (δhc) toward the excavation is defined as positive; otherwise, it is the negative direction.
Fig.6(a) shows the horizontal displacements of capping beam for monitoring C1 to C6 on the west side. It can be seen that in stage S-2, the N-1 zone on the west side of the excavation was excavated to 7 m BGS, and the displacement of capping beam was greatly disturbed. The maximum δhc of C1 and C2 reached a level of more than 5 mm. After the construction of stage S-2, the west side of the excavation was left standing for 63d until the end of stage S-4 (through winter). From December 2014 to March 2015, the displacements of C1 to C6 fluctuated slightly; during the warming period (from March to May 2015), the displacements of C1 to C6 increased by different degrees. It can be seen from Fig.6(a) that the displacement trends of C1 to C6 were significantly different from other monitoring points on the west side. From May to July 2015, the displacements of C1 and C2 were significantly larger than other monitoring points. The specific reasons will be discussed in detail in the analysis of Fig.7.
Fig.6(b) shows the horizontal displacements of capping beam for monitoring C7 to C12 on the north side of the excavation. It can be seen that the monitoring curves of δhc had a more significant fluctuation after the Winter Construction Stage (S-1 to S-3). For example, the maximum variation of δhc at C8 was 8.3 mm. The monitoring curves of δhc on the north side showed a trend of “sudden drop” at the beginning of stage S-5. Referring to the temperature records in Shenyang, the average temperature from March to April 2015 increased from −3 to 5 °C. By the end of stage S-5, the north excavation was excavated to 17.5 m BGS. In fact, this period of “sudden drop” lasted for 30 d, after which the displacements increased with the ambient temperature rapidly. Based on the temperature recorded during the deep excavation in Shenyang, it can be found that the temperature increased by nearly 10.5 °C during the stage S-5. Because the burial depth of the capping beam is within 600 mm, it is in the depth of frozen soil, when thawing occurred in the ground to result in liquid pore water, then the soil strength decreased, leading to more displacement for capping beams. With the continued construction of the north excavation, the earth pressure acting on the retaining piles gradually increased, as a result, the displacements of retaining piles grew.
Fig.6(c) shows the horizontal displacements of capping beam for monitoring C13 to C18 on the east side of the excavation. The monitoring data of capping beams on the east side showed that the displacement changed greatly from the initial stage of the excavation. The maximum displacement of capping beam on the east side reached 9.7 mm (C12), but the maximum displacement of capping beam on the north and west sides was 6.5 mm and 6.8 mm, respectively. The maximum displacement of capping beam on the north and west sides had negligible difference before the Winter Construction Stage.
From March to May 2015, as the ambient temperature increased, the displacement of monitoring C7 and C8 of capping beam on the east side fluctuated slightly, while the displacement data of other monitoring points showed a downward trend. According to the zoned excavation schedule in Fig.4, after the excavation on the east side reached −17 m below the ground surface in March, the east side had not been disturbed till May 2015. Hence, the displacement variation of capping beam was in a reasonable fluctuation range.
From May to July 2015, it can be seen from Fig.6(c) that the displacements of C17 and C18 of capping beam gradually increased to 6.8 mm, and the displacements of capping beam on the east side changed from 5.1 to 6.1 mm slowly. From the monitoring data in this stage, the displacements of capping beam had an obvious traction phenomenon. When the east side was excavated to −18.7 m, the displacement of C13 near the north side was relatively large (8.6 mm), and the displacements of C17 and C18 near the south side were smaller (2.2 and 2.0 mm, respectively). In the next excavation process, due to the coordinated deformation of capping beam on the east side, the displacements of monitoring C13 to C18 were gradually distributed around 5.1 to 6.1 mm.
Fig.7 shows the final δhc,max (the maximum horizontal displacement of capping beam) of capping beams. At C1 and C2 on the west side, δhc,max reached 12.50 and 9.40 mm, respectively, which were significantly greater than δhc,max of other monitoring points. The average δhc,max on the west side was 7.34 mm. The δhc,max on the north side was 6.60 mm, and their average δhc,max was 5.93 mm. The δhc,max on the east side was 6.10 mm, and their average δhc,max was 5.82 mm. The reason for the large displacements near C1 and C2 on the west side was analyzed. The monitoring points C1 and C2 are adjacent to a 6-floor residential building for about 18 m. In contrast, the north side was supported by double row piles while the east side was far from the existing commercial shopping center. Due to the additional stress of the residential buildings acting on the ground outside the foundation pit, the deformation of its capping beam on the west side was larger than the other two sides. The capping beam of this section finally had a large displacement, which was still much less than the warning threshold, without causing serious consequences.
3.1.2 Lateral displacement of piles
The patterns of lateral displacement of typical retaining piles, including piles J41, J64, J144, J185, J216, and J231, are shown in Fig.8 (the inward lateral displacement of retaining pile (δh) toward the inside of the excavation is the positive direction). All lateral displacement curves at different stages were basically identical, showing limited variations in deep soils during the excavation.
In Fig.8(a), J41 and J64 had an approximate interval of 30 m, and the embedment depth of these two piles was 6.5 m. Nevertheless, the lateral displacements of two measuring points showed different trends. The top pile displacement of J41 was 3.35 mm, and the maximum lateral displacement (δh,max) was 5.10 mm (at a depth of 5.09 m BGS). Below the maximum displacement section, the pile displacement fluctuated obviously. Compared with the lateral displacement of Pile J41, the curve of Pile J64 followed an obvious “bulge” displacement form. The maximum displacement of Pile J64 was 5.85 mm (at a depth of 14.00 m BGS). Compared with the top displacements of Pile J41 and Pile J64, the top displacement of Pile J41 (3.35 mm) was significantly greater than that of Pile J64 (0.48 mm). The displacement of the south–west capping beam was greater than that of the north–west capping beam, being also confirmed from the measuring data in Fig.6(a) and Fig.7. For the north side of the excavation which was adjacent to the metro lines, the displacement trends of Pile J144 and Pile J185 were consistent, as shown in Fig.8(b). It needs to be emphasized that the outer row of piles presented a slight “bulge”, because the tops of double-row piles were connected by a capping beam, enabling pile rows to work as a system to effectively control the displacement of pile top.
Around April 14, 2015, δh,max of Pile J144 was 8.13 mm (at a depth of 7.49 m BGS), while δh,max of Pile J185 was 6.17 mm (at a depth of 4.04 m BGS). Compared with the displacements of piles on the other three sides, δh,max of the north side adjacent to the metro lines was greater. The north side of the excavation was close to Zhonghua Road, which was in heavy traffic. The north side was in a very close proximity to Langqin Underpass, and Metro Line 1 also crossed through the north side of the excavation. Frequent disturbance was the main reason for the slightly greater displacement of retaining piles on the north side than other sides.
To further study the correlation between excavation depth
H and maximum lateral displacement
δh,max, different prediction models from the literature are compared in Fig.9. It can be seen that the distribution trend of
δh,max for this case varied from 0.01%
H to 0.06%
H. Due to the poor properties of clay in Shanghai [
20] and Singapore [
36], the
δh,max was greater than this case, which varied from
δh,max = 0.22%
H to
δh,max = 0.56%
H and from
δh,max = 0.44%
H to
δh,max = 0.86%
H, respectively. The engineering of deep excavation is a complex system affected by multiple factors, the factors which affect the displacement of retaining piles, in addition to soil properties, also include the excavation scheme, the retaining scheme and the plane size of the foundation pit.
3.1.3 Vertical displacement of piles
The vertical displacements of retaining piles during the excavation are illustrated in Fig.10 (the colored area in this figure represents the period in which the ambient temperature rapidly increases). During the Winter Construction Stage, the monitoring points had upward displacements with different magnitude except for C2, C3, and C4 on the west side. It needs to be emphasized that the temperature increased significantly after the end of the Winter Construction Stage. The temperature had risen by 16 °C in the beginning of April, resulting in thawing of the frozen ground, and pore water in surface soils gradually became the liquid phase. Consequently, the strength and the volume of soils slightly reduced, which caused significant fluctuation in vertical displacement of retaining piles. For example, the piles at C1 (1.9 mm, west side), C2 (3.4 mm, west side), C7 (3.9 mm, north side), C8 (3.7 mm, north side), C9 (3.6 mm, north side), C10 (2.8 mm, north side), C11 (2.8 mm, north side), C12 (4.1 mm, east side), and C14 (3.8 mm, east side) all showed the upward movements. The excavation depth of the north side was deeper than that of other areas (14 m BGS) in winter, resulting in a larger range of fluctuations in vertical displacement. At the end of the excavation, the retaining piles on the north side mainly settled (the minimum displacement was −0.8 mm at C10), while the other areas mainly showed the uplift behavior (the maximum displacement was 2.5 mm at C18). It should be stated that all monitoring data were considerably lower than the warning threshold for vertical displacement (20 mm).
3.2 Ground settlement
The six measuring points (L1 to L6) of ground settlement on the west were divided into three monitoring sections in this case, and each section was arranged with two measuring points. Only the monitoring data before April 17, 2015 were recorded in this case, at which the west side had been excavated to 17 m BGS (the designed excavation depth was 19.65 m). The variations of ground settlement δv during the excavation process are shown in Fig.11 (the colored area in this figure represents the period in which the extent of the excavation was large). It can be seen from that the trend of ground settlement at L1 to L6 was relatively consistent. In the stage S-1 (construction of retaining piles and capping beams), δv changed slowly. At the end of stage S-1, the maximum ground settlement δv,max was −0.91 mm (L2). In the construction stage of S-2 and the early stage of S-3 (3 m BGS), the ground surface had a slight variation, and δv fluctuated in a small range (1.40 mm).
After January 16, 2015, with the increase of excavation depth, δv began to change significantly. By February 3, δv was raised to −2.32 mm (L2). In the following week, δv had a significant change in the selected area highlighted by a red frame, as shown in Fig.11. On February 10, the excavation depth reached 12 m. The settlement of each monitoring point changed to −2.89 mm (L1), −1.10 mm (L2), −2.89 mm (L3), −2.48 mm (L4), −0.74 mm (L5), −2.05 mm (L6) within one week. During the Winter Construction Stage, the excavation construction was stopped. However, the western ground surface still maintained an increasing trend, but the trend was very small. During this period, the maximum change in settlement Δδv,max of six monitoring points was −1.04 mm (L5). By the end of stage S-5 (the west side was 17 m BGS), δv,max was −7.68 mm, and the ground settlement curves of L1 to L6 tended to be stable.
Combining Fig.6(a) and Fig.11, we can see that the monitoring data of C1 (capping beam) and L1 to L6 (ground settlement) in stage S-5 have a similarly increased trend. This shows that the adjacent residential buildings have a significant impact on the deformation of the retaining structures. The deformation of retaining piles move toward the foundation pit leading to the increase of the ground settlement, the monitoring data of capping beam and ground settlement are mutually verified.
To investigate the influence area of ground settlement, the relationship between
δv/
δv,max and
d/
He (
d represents the distance between the monitoring point and the edge of the deep excavation,
He represent the final depth of the excavation) is presented in Fig.12. With the increase of excavation depth, the development of
δv/
δv,max presented an increasing trend. The maximum value of
δv,max occurred at L1, and the location was 0.28
He from the excavation edge. Since the monitoring points of the same section were arranged at a distance of 8 m, it can be speculated that the maximum ground settlement should be within the range from 0.28
He to 0.71
He in practice. The influence area of ground settlement in this case was obviously smaller than that of the soft clay [
5,
37] or medium hard clay [
38] within 1.5
He. There are two main reasons. 1) As the ground in Shenyang mainly contained coarse-grained soil with high strength, the deformation caused by excavation could be far less than that in soft soil areas. 2) As the excavation was in winter, the average frozen ground depth in Shenyang was 1.2 m, which also reduced the settlement and the range of influence area.
Comparison curves between the maximum ground settlement (
δv,max) and the maximum lateral displacement of pile (
δh,max) are shown in Fig.13. In this case, the monitoring data of
δv,max at L3, L4, L5, L6 and
δh,max at J41, J64 were analyzed. It can be seen that
δv,max was greater than
δh,max in stage S-5, and
δv,max/
δh,max fell within the range from 1 to 3.95, which was larger than the monitoring data of excavations in Jinan silty clay [
39]. The main reason was that in the construction stage S-5, it experienced a period of temperature growing, where the temperature in Shenyang increased significantly within 35d from March 10, 2015 to April 14, 2015. In this period, the maximum temperature variation was from −1 to 20 °C. Hence, the surface soil was gradually thawed, which caused the occurrence of ground settlement. The findings suggested that soil freezing and thawing could have a great impact on the performance of excavation construction in Shenyang. As illustrated in Fig.14, the maximum ground settlement
δv,max at each excavation stage varied between 0.02%
H and 0.045%
H in this case. The variation range of
δv,max under different excavation depths in Shanghai soft clay was between 0.03%
H and 1.10%
H [
11,
12], and the range was between 0.04%
H and 0.30%
H in Suzhou stiff clay [
40] and less than 0.50%
H in other hard clays [
41].
4 Responses of adjacent metro line
4.1 Settlement of rail track and tunnel lining
Changes in vertical displacements of the rail track are shown in Fig.15 (the colored area in this figure represents the period in which the extent of the excavation was large), in which the vertical displacements of the track were the average value of S1 and S4 measuring points. It is found that the adjacent tunnel was moving toward the excavation, which was in agreement with several researchers [
24,
42]. During the excavation from S-2 to S-3, there was a small fluctuation in vertical displacement. The track at R14 and R15 sections on the east side mainly showed a trend of settlement, and other sections showed uplift of less than 1 mm. When the Winter Construction Stage began, construction of deep excavation was stopped, and the track displacement changed into a stable stage. After the Winter Construction Stage, continued excavation was implemented along with the temperature increase, and as such disturbance to the tunnel increased significantly. The temperature had raised by 16 °C during the excavation stage S-5. As previously illustrated, pore water became the liquid phase due to the increase of ambient temperature; therefore, the soil strength decreased, such that deep soils caused retaining piles to deform, further influencing the deformation of rail tracks (showing upward movement).
It can be seen that the track had a large upward deformation when the excavation stage S-5 began (the excavation depth was 14 m). The upward deformation value of section R4 was the largest (i.e., 2.31 mm, being less than the displacement warning threshold of 4 mm), and the increase in tunnel’s vertical displacement reached 2.36 mm. In S-6 (the excavation depth at the beginning of this stage was 17 m), the vertical track displacement tended to be stable and had a slight decrease, after the excavation speed was slowed down. The largest track deformation of the south metro line was 2.1 mm at section R2. In contrast, the values of upward track deformation at sections R14 and R15 were 0.995 and 0.825 mm, respectively.
According to Fig.3, the buried depth of metro lines was 15 m. According to the excavation schedule (Fig.4), the excavation depth of the north side was smaller than the buried depth of metro lines before the end of stage S-5, the tunnel had an upward displacement trend toward the excavation because of the unloading effect of soil. In stages S-6 and S-7, the excavation depth of north side was larger than the buried depth of metro lines, which resulting in a decrease trend of tunnel displacement. It can also be seen from the lateral displacement of the retaining piles (Fig.8), depth of the maximum lateral displacement of retaining piles on the north side gradually increases along with the increase of excavation depth.
The development of displacement of the south metro line at the end of different excavation stages is shown in Fig.16. The overall smoothness of the tunnel was well controlled, and the metro line displacement was less than the warning threshold during the whole excavation stages (the settlement was 2.1 mm, while the upward and horizontal warning thresholds were 3.5 and 3 mm, respectively).
In fact, the data in Fig.16 showed the final displacement of metro line at each excavation stage, and they were consistent with the trend in Fig.16. The vertical displacement of metro line could be divided into Stage A and Stage B. The first stage included the stages S-1, S-2, S-3 before the Winter Break (S-4). In this stage, the construction of retaining piles and capping beams had been completed, and in the process of subsequent excavation to the depth of 14 m, the upward displacement of metro line showed a certain decline. Furthermore, enough attention should be paid to the displacement monitoring of the metro line to prevent excessive deformation during the construction of retaining piles. In the first stage, the displacement was slight, and the maximum upward displacement and settlement were 0.78 (section R2 at the end of stage S-1) and −0.36 mm (section R15 at the end of stage S-4), respectively. Moreover, the maximum differential settlement of the metro line was 0.85 mm at the end of stage S-1, and the value was less than the warning threshold of 2 mm.
The second stage included the stages S-5, S-6, and S-7. The maximum vertical displacement of the metro line increased from −0.36 to 2.55 mm at the end of stage S-5 (maximum value appeared in section R2). Then, in the construction stage for basement floor, the vertical displacement of the metro line did not increase anymore. On the contrary, the metro line started to subside. At the end of stage S-7, the maximum vertical displacement of the metro line had been reduced to 2.06 mm (in section R2). The metro line’s maximum differential settlement occurred at the end of stage S-6, being 1.45 mm (less than the warning threshold of 2 mm).
As shown in the instrumentation layout of Fig.5, sections R10 to R13 had a distance of 16.5 m to sections R1 to R7, and steel pipe supports were adopted at the north–eastern corner. However, the difference in displacement curves between Pile J185 (north–eastern corner) and Pile J144 (northern corner) was not significant (as shown in Fig.8(b), on May 16, 2015, the lateral displacement of Pile J185 was 0.85 mm, which was greater than that of Pile J144). On the contrary, according to the vertical displacement of the metro line in Fig.16(a), the vertical displacement of sections R8 to R15 was significant smaller than sections R1 to R7.
The horizontal displacement curves of the metro line at different excavation stages are plotted in Fig.16(b). The horizontal displacement of metro line was also divided into Stage A and Stage B. The horizontal displacements of the metro line varied between 0.11 and 0.81 mm at Stage A. At Stage B, the maximum horizontal displacement reached 2.42 mm (at section R8). In addition, inward horizontal displacement toward the excavation was observed from section R3 to section R13 showing an obvious spatial effect. Although the excavation depth was 5.65 m in the last 3 months after winter, the displacement of metro line had a relatively huge variation between Stage A and Stage B. It is, thus, clear that the period of thawing of seasonal frozen soil had a significant influence on the adjacent metro line, although the excavation depth was smaller than that in former construction stages. Redistribution of soil pressure around the metro line induced unloading of deep soils; therefore, the metro line showed a hogging (concave down deformed shape) displacement pattern.
4.2 Convergences of tunnel rings
The variations of radial convergence of the south metro line are illustrated in Fig.17. Specifically, monitoring sections R4 to R10 were analyzed due to greater influence from the excavation. It can be seen that the convergences of 7 monitoring sections were relatively consistent. In stage S-1 (construction of retaining piles and capping beams), the convergence was small at monitoring sections R4 to R10, and the displacement continued until it was excavated to 14 m BGS. From the end of stage S-3 to the beginning of stage S-5, the convergence of metro line remained small all the time. Combined with Fig.15 and Fig.16, the displacement of metro line was gradually decreasing from stage S-1 to S-3, especially for monitoring sections R4 to R12 within the projection area of retaining piles on the north side of the excavation, and the radial convergence of this section was also maintained within −0.3 mm. Through the Winter Construction Stage, the numerical values of radial convergence decreased. At this time, the deformation of metro line appeared to rebound, but the overall smoothness of the tunnel was still well controlled. In addition, water in the soil behind the retaining piles turned from the solid into the liquid phase within a certain depth, along with the increase of temperature, resulting in slightly reduced soil strength and great variations in displacement of retaining piles. Therefore, the stress in deep soil redistributed, and the metro lines were forced to deform to meet a new state of equilibrium. After the values of radial convergence reached a stable state, the maximum convergence was −0.41 mm at section R6 (less than the warning threshold of −2.8 mm).
4.3 Displacement relationship between retaining piles and tunnel sections
To further investigate the relationship between excavation and displacement of metro line, the displacements of retaining piles and the metro line were compared at a same section from March to June 2015 in Fig.18. It should be emphasized that the negative direction of pile displacement was toward the inside of the excavation, and the positive direction of the metro line was away from the excavation.
In Fig.18(a), the maximum lateral displacement of pile J112 was −6.11 mm (at a depth of 7.122 m BGS), and the final horizontal and vertical displacements of the tunnel were −1.74 and 2.28 mm, respectively. Section R5 had an obliquely upward displacement toward the excavation, but the convergence of this section was a negative value. It indicates that section R5 was in a compressed status, resulting in an upward displacement. It should be emphasized that the displacement and convergence were less than the warning threshold.
In Fig.18(b), the δh,max of pile J144 was −6.59 mm (at a depth of 7.489 m BGS), and the final horizontal and vertical displacements of the tunnel were −2.16 and 2.34 mm, respectively. The displacement of section R7 was slightly different from that of section R5. It can be seen that during the construction from May to June 2015, the horizontal displacement decreased from 1.97 to 1.32 mm. Accordingly, the displacement of sections R5 and R7 had reached the stability in stage S-6 of the excavation.
As shown in Fig.18(c), the δh,max of pile J185 was −6.48 mm (at a depth of 7.497 m BGS), and the final horizontal and vertical displacements were −1.89 and 1.78 mm, respectively. The displacement of section R12 was also stable in stage S-6 of the excavation.
It can be seen from Fig.18 that the variation of vertical displacement was almost similar to the pattern of horizontal displacement. From March 23 to May 5, 2015, the horizontal displacement increment of section R5 was 1.07 mm, while the vertical displacement increment was 1.65 mm. From March 18 to April 17, 2015, the horizontal displacement increment of section R7 was 1.07 mm, while the vertical displacement increment was 2.13 mm. From March 18 to May 8, 2015, the horizontal displacement increment of section R12 was 1.61 mm, while the vertical displacement increment was 1.55 mm.
As is illustrated in Fig.18, the depth, at which the δh,max of retaining piles occurred, was always smaller than the burial depth of the tunnel (i.e., 15 m). As the excavation depth became lager, the depth of maximum horizontal displacement of retaining piles moved downwards, and the influence area of the excavation in deep soil also expanded. The imbalance of earth pressure drove the tunnel to move, showing a phenomenon of increasing first and then decreasing. The soils behind retaining piles were under the at-rest state before the implementation of the excavation. With the process of deep excavation, retaining piles started to deflect, and the deep soils between the metro line and the excavation were moving toward the retaining piles. Then, the stress in the soil was under the active state. Therefore, the lateral earth pressure and pressure at the tunnel crown became smaller, then the tunnel started to deform until they met a new state of equilibrium. Hence, tunnel sections began to move toward the excavation and uplift.
5 Conclusions
Though a comprehensive in situ measurement program, the performance of a deep pit-in-pit excavation constructed by the TD method in seasonal frozen soils in Shenyang was extensively examined. Based on the analyses of field data, the major findings are summarized as follows.
1) Using the TD method with the supporting scheme of combined retaining piles and anchor cables, the construction process of this deep excavation in seasonal frozen soils did not induce any adverse impact on adjacent infrastructure in downtown Shenyang.
2) The maximum displacement of capping beams occurred on the west side, showing an increase of 145.1% compared to the end of the Winter Construction Stage. The vertical displacements of double row piles on the north side changed from uplift to settlement, and the displacement of retaining structures was sensitive to changes in temperature.
3) The maximum ground settlement of deep excavation was within the range from 0.28He to 0.71He, and the ratio of maximum ground settlement δv,max to maximum lateral displacement of piles δh,max varied between 1 and 3.95. The seasonal frozen soil reduced the settlement and the range of influence area through the Winter Construction Stage.
4) After the Winter Construction Stage, the ratio of maximum horizontal displacement of the metro line to maximum lateral displacement of retaining piles was between 28% and 33%. In addition, the displacement of metro line showed a significant spatial effect, and the sections which were adjacent to the excavation had a hogging (concave down deformed shape) displacement pattern, being the main targets for implementing protective measures.
5) Earth pressure redistribution occurred around the retaining structures and metro line due to the combined effects of excavation and thawing of seasonal frozen soil, the stress in deep soil redistributed, and the metro lines were forced to deform to meet a new state of equilibrium. Consequently, enough attention should be paid to the excavation scheme and in situ measurements through winter.
PDF
(5506KB)
