Introduction

As there is continuous development of high-rise buildings and subway stations in urban areas, the scale of urban underground space development increases, leading to the increasing number of deep foundation pit projects as well (Li et al. 2012). Among engineering support structures, the pile-anchor retaining structure is more and more widely used due to its convenient construction and excellent supporting ability (Wang et al. 2013, 2014; Johnson and Sandeep 2016). Furthermore, prestressed anchor cable is an important part of the pile-anchor supporting system. It utilizes the self-stability of the rock and/or soil mass of the cables anchored in to share the supporting load and thus reduce the cross-section area of the supporting structure, which brings significant economic and social benefits.

The limitation of the prestressed anchor cable lies in that the loss of prestress is often founded in practices (Han et al. 2008; Chen et al. 2013; Sokolic and Szavits-Nossan2011; Jia 2011). The prestress loss rate of single tensioned locking cables is generally up to between 24.5 and 53.4% (Tang 2007). Besides, the ultimate stress state of the cables is affected by many factors such as its structural properties, partial excavation of the foundation pit, the deformation of the pile, and construction quality in anchor cable locking (Yang et al. 2016; Han et al. 2008; Chen and Wang 2012; Shan et al. 2013), all of which have made it the challenge in the design and construction of the pile-anchor retaining to ensure the favorable anchoring force state after the excavation.

The causes of prestress loss in prestressed anchor cables are diverse. Notably, combined with in-site monitoring and numerical simulation, it is concluded that the property of strata the cable anchored in is recognized as an essential factor responsible for the loss of prestress, which draws wide attention in current researches (Wang 2011; Wang et al. 2012). However, existing literature mainly focuses on the anchor cables anchored in a rock mass (Liu et al. 2014), whereas limited research concerning the prestress loss of the ones is anchored in the soil. Also, existing studies generally focus on the final magnitude of the prestress in anchor cable as the excavation is completed while lack in-depth exploration of the prestress variation during the whole construction of the foundation pit (Xing 2014; Baazouzi et al. 2017; Yang and Guo 2012). Besides, the loss of prestress in anchor cables could vary as the different positions of the anchor cable on the plane of the foundation pit, which is also worth investigating.

Therefore, to enable a better understanding of the effect of anchoring strata and installation position on the prestress in anchor cable, as well as the prestress variations during the construction, a field monitoring was carried out in this study. The monitored anchor cables were installed in a deep excavation foundation pit which passes through sand-cobble stone strata and anchored in clay or sandy mudstone. Based on the field monitoring and relevant literature, the objectives of this trial are to, during the entire construction process from anchor cable tensioning and locking to the completion of foundation pit excavation, (1) analyze the distribution and variation of the tension in anchor cables positioned differently on a same supporting pile, (2) reveal the variations of tension in the anchor cables anchored in different strata, and (3) investigate the variations of tension in anchor cables installed in different plane locations of the foundation pit.

Study site and field monitoring scheme

Engineering situation

The project is a foundation pit of a high-rise building in southwest China. The site is located on the Jialing River terrace; the terrain is geographically high in the south and low in the north. The designed basement is three floors, and the maximum excavation depth is about 22.5 m. The prestressed anchorage pile is adopted as the foundation pit supporting system. The main strata on the north and east sides of the foundation pit are plain soil and silty clay. The upper strata of the south side are mainly filled with plain soil and silty clay, while the lower is the bedrock composed of the sandy mudstone.

Subject to construction conditions, the project adopted an anchorage pile retaining structure in the north and east, anchor pile and anchor rod retaining wall combination structure on the south side, and a prestressed anchor cable structure of double-row piles support system near to bridge pier.

Comprehensively considering the influence of all aspects, 4φ15.2 steel strands were selected as the prestressed anchor cable and installed at an angle of 15° to the horizontal surface. The lengths of anchor cables are between 24.5 and 32.5 m and with the prestress magnitudes between 200 and 350 kN. Cement mortar is used to bond the anchor cable with the soil or rock mass. The distribution of the strata and supporting structure on the north side of the foundation pit is shown in Fig. 1.

Fig. 1
figure 1

Sectional view of the supporting structure and strata on the north side of the foundation pit

Full size image

Field monitoring scheme

According to the technical standard for monitoring of building excavation engineering (MOHURD 2019), the prestressed anchor cable dynamometer should be placed at the representative position where the foundation pit is under great stress, namely the middle part of each side, the corner, and the sections with complex geological conditions of the foundation pit.

The plane layout of the field monitored anchor cables is shown in Fig. 2. The monitoring points are distributed in the north, east, and south sides of the foundation pit with an average distance of 50 m. A total number of 14 anchor cables of 10 anchorage piles were selected to be monitored. The anchor cables in the upper three rows were monitored due to the anchor tension of the anchor cables in the upper part being more affected by foundation excavation than those in the lower part. Among them, M1, M2, and M3 are the monitored anchor cables for pile 25# and M6, M7, and M8 for pile 89#. The monitoring results of them were used for exploring the variation of multi-row anchor cable tension on a single retaining pile. To study the variation of anchor tension in different strata, the anchor tension of anchorage cables M4 (pile 51#) and M5 (pile 69#) anchored in silty clay and M14 (pile224#) and M15 (pile 244#) anchored in sandy mudstone were monitored, respectively. Additionally, anchor cables M9 (pile 110#), M10 (pile 130#), M11 (pile 153#), and M13 (pile 205#) were monitored for examining the variation of anchor cable tension in different zones of the foundation pit. All these 8 anchor cables are the first-row cables.

Fig. 2
figure 2

Plane layout schematic diagram of the field monitored anchor cables

Full size image

The field monitoring started on January 4, 2012, and ended when the completion of pit soleplates poured on January 1, 2013. Normally, the anchor tension was monitored once a week. During the excavation of a layer or the installation and locking of the prestressed anchor cables, according to the design requirements, monitoring was conducted once a day for 3 to 5 days. Subsequently, the monitoring frequency changed to once every 3 days. After the excavation to the bottom of the pit, the anchor tension was continuously monitored once a day for a week and then once a week. BGK-4900 vibrating-string anchor cable dynamometer manufactured by the Beijing Jikang Instrument Company was adopted in this project. Its major technical specifications are shown in Table 1.

Table 1 Major technical specifications for BGK-4900 vibrating-string anchor cable dynamometer
Full size table

Results

The prestress of cables set at different positions on the same supporting pile

To analyze the variation of anchor cable tension, the monitoring of multi-row anchor cable tension on the single retaining pile was carried out on the north side of the foundation pit. The cable tensions of M1, M2, and M3 on pile 25# and M6, M7, and M8 on pile 89# are shown in Fig. 3.

Fig. 3
figure 3

The tensions of anchor cables set in a the first, b the second, and c the third row on anchor pile 25# and 89#

Full size image

The initial prestress of these six anchor cables was all 350 kN. After the prestress was applied, a large loss occurred immediately at the first stage. One day, after the anchoring, the anchor tension of anchor cables M1, M2, and M3 on pile 25# reduced to 224 kN, 250 kN, and 268 kN, respectively, and those of anchor cables M6, M7, and M8 on pile 89# were 235 kN, 248 kN, and 265 kN, respectively. The average loss rates of anchor cables tension of piles 25# and 89# were 29.3% and 28.7%, respectively. As the anchor locked stably, with the progress of the foundation pit excavation, the anchor tension of each anchor cable increased in varying degrees. The tension of cables M1, M2, and M3 raised to 234 kN, 301 kN, and 294 kN, respectively, and those of cables M6, M7, and M8 elevated to 237 kN, 279 kN, and 277 kN, respectively. The anchor tension of cables in this row sharply dropped as the next row of anchor cables was locked and followed by a certain level of prestress recovery due to the redistribution of the soil stress caused by the foundation pit excavation. Subsequently, it is notable that under the action of lower anchor cable anchoring and foundation pit excavation repeatedly, the tension of cables fluctuated. And as a repeated excavation of the lower soil and locking of the lower anchor cable, the fluctuating amplitude gradually dwindled and eventually stabilized. Finally, anchor tension of anchor cables in the top three rows of piles 25# and 89# stabilized at 260 kN (M1), 300 kN (M2), 317 kN (M3), 220 kN (M6), 252 kN (M7), and 275 kN (M8), respectively.

The monitoring conducted by former researchers on anchor tension of pile-anchor retaining also showed similar variation characteristics that the tension of anchor cables installed on the retaining piles suggested a three-stage variation. Zhang (2011) monitored two rows of anchor cables installed on an anchor pile in a foundation pit (Fig. 4), and the monitoring data is shown in Fig. 5a. The altitude difference between the two rows of anchor cables is 6 m. The initial prestress of the upper cables was 300 kN, which decreased to 218 kN after the anchor cable stretching and locking. As the lower foundation pit was excavated, the anchor tension increased to 242 kN and followed by a slight decline after the prestress of the lower anchor cable was applied. When the whole excavation was completed, the tension of the upper cable stabilized at 244 kN. The initial prestress of the lower anchor cable was 400 kN and dropped to 311 kN after the locking. During the subsequent excavation, the anchor tension exhibited a slight increase and tended to level off which finally stabilized at 317 kN.

Fig. 4
figure 4

Sectional view of the supporting structure and strata in the literature of Zhang (2011)

Full size image
Fig. 5
figure 5

The tensions of anchor cables set at different positions on a single supporting pile reported by a Zhang (2011) and b Li et al. (2013)

Full size image

In another research (Li et al. 2013), four rows of cables in a foundation pit were monitored and the monitoring data is shown in Fig. 5b. The result shows that the prestress loss of anchor cables due to factors such as soil relaxation creep was completed in about 20 days after locking. During the excavation of the foundation pit in layers, the change of the tension in anchor cables was pronounced. It indicates that the excavation of the lower layer foundation soil and the application of the prestress would affect the tension of the locked anchor cables. Monitoring data revealed three fluctuations for the anchor tension of the first row of anchor cables, while two for those in the second and one for those in the third. This verified that the fluctuation of anchor cable tension is indeed caused by the lower construction.

The prestress of cables anchored in different strata

Figure 6 shows the variation curves of the anchor tension of anchor cables M4 (pile 51#) and M5 (pile 69#) anchored in silty clay (Fig. 6a) and anchor cables M14 (pile 224#) and M15 (pile 244#) anchored in sandy mudstone (Fig. 6b) with a function of time. The initial prestress of anchor cables M4 and M5 was set to 350 kN, and that of anchor cables M14 and M15 was set to 250 kN. Curves illustrated that the prestress loss of the cables anchored in the sandy mudstone was lower than that of cables anchored in the silty clay at the initial stage of locking. The corresponding average loss rates of the anchor tension of the cables anchored in the silty clay and sandy mudstone were 33.3% and 17.8%, respectively. In the subsequent construction, due to the excavation of the subsoil and the application of prestress in the lower-row anchors, significantly different fluctuation amplitudes of the anchor tension were shown between the cables anchored in the two strata. The fluctuation amplitude of anchor tension of the cables anchored in sandy mudstone was much smaller than that in the silty clay strata. A tension fluctuation of 10~20 kN in the cables anchored in sandy mudstone was found, whereas that in silty clay was more than 50 kN. When the foundation pit construction was completed and the anchor tension stabilized, the anchor tension of the cables anchored in the silty clay strata and that in the sandy mudstone decreased from 350 kN and 250 kN to 233 kN and 210 kN and with the average loss rates of 33.4% and 16.0%, respectively.

Fig. 6
figure 6

The tensions of anchor cables anchored in different strata as a function of time. a M4 and M5 anchored in silty clay. b M14 and M15 anchored in sandy mudstone

Full size image

Further investigation was conducted to enable a better understanding of the various characteristics based on the present studies on anchor tension in other different strata. Liu et al. (2014) reported that the variation law of the prestress of anchor cables anchored in the saturated fine sand. Among the 3 rows of anchor cables arranged in the foundation pit, the lower two rows of anchor cables were monitored. Part of the tension variation curves of the anchor cables is shown in Fig. 7.

Fig. 7
figure 7

The tensions of anchor cables anchored in saturated fine sand and set at different positions on a single supporting pile reported by Liu et al. (2014). a The second-row cables. b The third-row cables

Full size image

The initial prestress in the second (Fig. 7a) and third (Fig. 7b) rows of anchor cables was set to 380 kN and 430 kN, respectively. After locking, the prestress losses were high as the average loss rates of 66.8% (the second row) and 47.7% (the third row) for the two rows of anchor cables, respectively. In contrast, when it comes to the excavation of the subsoil, the anchor tension increased rapidly in a few days. Especially, the average increment of the tension in the second row was 51.4 kN after the first step excavation. The maximum increment of 69.1 kN was found in cable M2-N2. Notably, compared with silty clay and sandy mudstone, the prestress of cables anchored in the saturated fine sand layer had no obvious fluctuations, indicating no obvious effect of the application of prestress in the lower anchor cables on the tension in the upper ones. For cable M2-E1, it is on the artificial fill (low density) slope on the east side of the foundation pit where there were problems like shifting sand and seepage during the excavation, leading to a large settlement of the slope. The settlement of the soil under the anchor pier is one of the main reasons for the short-term loss of anchor cable prestress (Zhou et al. 2006). Therefore, the tension of cable M2-E1 then suffered a sharp decrease (104.5 kN). When the excavation was completed, the soil stress redistributed to a state of equilibrium; thus, the anchor tension in M2-E1 suggested a certain rebound and tends to be stable.

The prestress of cables installed in different plane locations of the foundation pit

The foundation pit in this study covers a wide range of land areas, which has a complicated engineering boundary. Field monitoring shows that the prestress of the anchor cables in the middle part of each side and the corner where the boundary shape of the foundation pit changes drastically differs greatly. Figure 8 shows the anchor tension of cable M9 and M10, the first-row anchor cables on the anchor pile 110# and 130#, respectively. Generally, the fluctuation of anchor tension near to the corner (M9) of the foundation pit was smaller than that near to the middle (M10) in the following stage of foundation pit excavation and anchor cable installation (Fig. 9). The initial prestress of the two cables was both set to 350 kN. One day after the installation, the tension reduced to 239 kN (M9) and 230 kN (M10) with initial loss rates of 31.7% and 34.3%, respectively. Then, they decreased to 200 kN (M9) and 197 kN (M10), respectively, at the end of the locking stage (first stage). The final tension of anchor cables M9 and M10 stabled at 225 kN and 242 kN, respectively. Notably, the monitored anchor tension of the first-row anchors, namely M1 (pile 25#) and M6 (pile 89#), showed a similar variation characteristic. The tension in the key construction stage of some anchor cables is shown in Table 2. Besides, based on the tension of anchor cables on the anchor piles 89# (M6) and 110# (M9), it could be conducted that the various characteristics of the anchor tension in anchor cables near the corners on the north and east sides of the foundation pit are basically the same.

Fig. 8
figure 8

The tensions of anchor cables installed in different plane locations of the foundation pit

Full size image
Fig. 9
figure 9

The tensions of anchor cables installed in different plane locations of a foundation pit reported by Liu (2014)

Full size image
Table 2 Partial field monitoring data of anchor cable tension on the north side of the foundation pit
Full size table

Additionally, as the east boundary of the pit is close to the existing municipal road and the bridge piers of Jiahua Bridge, the tension variation of anchor cables there suggested to be affected by traffic and other vibration loads alike. The first-row anchor cable M11 there was monitored during the construction, and its tension is shown in Fig. 9. Compared with cable M13 near the external corner with the same initial prestress (300 kN), the tension of M11 was found to be smaller than that of M13 during almost the entire construction process. M11 showed larger initial prestress loss (18.0% of M13 and 19.3% of M11) at the end of the first stage. In the second stage, the tension in M11 showed generally smaller fluctuation amplitudes. Most notably, although the whole construction was completed, the anchor tension of M11 still showed a more obvious fluctuation and continued to decrease in the third stage which finally stabilized with a residual rate of 78.0% (234 kN). This could be attributed to the difficulties for the cables to stabilize as well as the contact fatigue effect between the anchoring pastes and the slop soil caused by the traffic and other vibration loads alike.

Fig. 10
figure 10

The tensions of anchor cables near the external corner (M13) and under the influence of traffic load (M11)

Full size image

Field monitoring conducted by Liu (2014) on the variation of anchor tension in the pile-anchor retaining of a deep foundation pit in the west of Changchun uncovered a similar feature. The stratum of the monitored foundation pit is mainly silty clay, and the distance between monitoring points is 30 m. Fig. 10 shows the tension change curves of monitored anchor cables N1, N2, N3, and N4 on anchor piles 75#, 98#, 111#, and 149#, during the whole process of foundation pit construction. Among them, anchor piles 98# and 111# are close to the middle of the foundation pit boundary and anchor piles 75# and 149# near the corner of the foundation pit. The initial prestress in the four anchor cables was all 500 kN and reduced by 118 kN (N1), 146 kN (N2), 135 kN (N3), and 124 kN (N4) at a loss rate of 23.6%, 29.2%, 27.0%, and 24.8% during the locking stage (the first stage), respectively. Subsequently, the anchor cable tension rebounded to a certain extent after the excavation of the soil under the cables and reached 375 kN (N1), 368 kN (N2), 363 kN (N3), and 372 kN (N4), respectively, when the foundation pit construction was completed.

As the curves illustrated, during construction, the tension in the anchor cable set at similar positions on the boundary of the foundation pit (i.e., piles 98# and 111# near the middle, piles 75# and 149# near the corner) has similar fluctuation amplitudes. It is noted that the tension losses of the anchor cables near the middle were greater than those near the sides during the first stage, while no marked differences between the increases of the tension caused by the excavation of the soil under the anchor cables. Besides, since only two rows of anchor cables were installed, the anchor cable tension has no obvious fluctuation characteristics during the entire construction process.

Discussion

The variations of the tension in anchor cables on an individual supporting pile

According to the monitoring data of cables set at different positions on the same supporting pile, some discoveries can be summarized as follows:

  1. (1)

    The anchor cable tension fluctuates during the construction and gradually stabilizes after the construction is completed. After the initial prestress is applied, the anchor tension would have a significant loss in the subsequent days. This could be attributed to the deformation of the anchor end, as well as the creep deformation between the anchorage grouting body and soil (Yang et al. 2016). However, with the further excavation of the foundation pit, the change of earth pressure would lead to the lateral deformation of the pile and soil toward the foundation pit, increasing the anchor tension. Subsequently, when the later rows of anchor cables are installed, the anchor tension of the formerly rows of anchor cables gradually lost as the tighter contact between the piles and soil caused by their mutual squeeze.

  2. (2)

    During the excavation process, the number and amplitude of the fluctuations of the tension in the anchor cable vary with its construction sequence. Generally, more fluctuations and smaller fluctuant amplitudes of the tension were found in the first-row cables compared with those in the later rows. Besides, as the excavation of the foundation pit is completed, the final anchor tension in the same pile shows sequential growth from the top to the bottom. The anchor tension in the later constructed cables is less affected by the subsequent construction than the formerly constructed cables, which only show increases due to the excavation of the lower foundation pit and the lateral displacement of the piles and soil. The excavation of deep foundation pit presents mainly a disturbance effect on the excavating layer and its upper supporting system. Meanwhile, the “drum belly” deformation phenomenon of the piles would occur to the piles, namely the lateral displacement of the middle of the pile is relatively larger than the ends.

The variations of the tension in anchor cables anchored in different strata

Based on the monitoring results and literature, some discoveries can be summarized as follows:

  1. (1)

    The lowest prestress loss during the initial locking period was found in the ones anchored in the sandy mudstone. Inversely, the prestress loss of the ones anchored in the saturated fine sand was found the biggest and significantly greater than those of the others. The prestress loss of the ones anchored in the clay ranked second among the three.

  2. (2)

    Obviously, differences appeared between the prestress fluctuation amplitudes of the anchor cables anchored in the three different strata. Due to the excavation of the soil and installation of the anchor cables underneath, the tension of the former cables would fluctuate. The fluctuation in the tension of the cables anchored in sandy mudstone was smaller and more stable than that anchored in the clay strata. Especially, the prestress of anchor cables anchored in the saturated fine sand showed no such fluctuation, indicating no obvious effect of the later excavation and installation of cables on the tension of the formerly installed anchor cables in saturated fine sand.

  3. (3)

    Compared with silty clay and sandy mudstone, saturated fine sand is more susceptible to the shifting sand and seepage problems caused by rainfall, which would lead to a large settlement of the soil. The settlement of the soil under the anchor pier is one of the main reasons for the short-term loss of anchor cable prestress (Zhou et al. 2006). Thus, the anchor cables set in those places would suffer a large loss of prestress (such as 104.5 kN for M2-E1). Precautions should be taken during construction in such places.

The variations of the tension in anchor cables anchored different plane locations of the foundation pit

According to the monitoring data of cables installed in different plane locations of the foundation pit, some discoveries can be summarized as follows:

  1. (1)

    The stress in surrounding slopes of the foundation pit would redistribute to balance after excavation, leading to the lateral deformation of the pile and soil. The deformation is maximum in the middle and gradually decreases towards the corners of the pit. The compactness of the soil is related to the magnitude of the lateral displacement of the slope, which is generally reflected in the larger the lateral displacement, the greater the reduction amplitude, that is, the greater the creep deformation of the soil. Therefore, stress concentration would occur at the corners. The initial loss of prestress, the stress fluctuations caused by the subsequent construction, and the final tension in the anchor cables near the corner are all smaller than those near the middle of the boundary.

  2. (2)

    Under the influence of traffic load and other factors, the anchor cable near the tight pier still showed a more obvious fluctuation after the whole construction was completed. It indicates that the final anchor tension of the anchor cable is obviously affected by the surrounding environment of the foundation pit.

Conclusions

Based on a whole construction process field monitoring on the prestressed anchor cables used in a deep foundation pit construction, this paper investigates the variation characteristics of prestress in the anchor cables set at different positions on the same supporting pile, anchored in different strata, and installed in different plane locations of the foundation pit. The main conclusions are as follows:

  1. (1)

    The prestress of anchor cables shows a three-stage variation overall (Fig. 3a): the prestress loss stage after the anchor cables was locked (the first stage), the fluctuating growth stage of anchor cable prestress during subsequent excavation (the second stage), and the stabilization stage after the whole construction is completed (the third stage).

  2. (2)

    In general, on the same supporting pile, the prestress of the anchor cables constructed first (the first-row anchor cables) is the most affected by subsequent constructions, and the impact of construction on the later installed anchor cables gradually decreases.

  3. (3)

    Due to the drum belly phenomenon, the lateral displacement of the middle of the anchor pile is larger than that at its both ends, leading to the larger prestress fluctuation amplitudes of the anchor cables installed in the middle of the anchor pile. The final stable values of the anchor cable prestress on the same anchor pile increase from top to bottom of the pile.

  4. (4)

    The initial prestress loss of anchor cables anchored in the sandy mudstone is the smallest, whereas those in the saturated fine sand are significantly higher than those in the silty clay and sandy mudstone. The fluctuation amplitude of anchor tension of anchor cables anchored in the sandy mudstone is smaller than that of anchor cables anchored in the silty clay. No obvious fluctuation was found in the prestress of the cables anchored in the saturated fine sand.

  5. (5)

    The prestress loss after locking, the prestress fluctuation amplitude during the excavation, as well as the final prestress in cables are all getting smaller from the middle to the corner of the foundation pit slope. Traffic and other vibration loads alike would affect the prestress variation of anchor cables on the retaining piles nearby.