- Email: [email protected]
Investigation on the tail brush induced loads upon segmental lining of a shield tunnel with small overburden
Tunnelling and Underground Space Technology 97 (2020) 103283
Contents lists available at ScienceDirect
Tunnelling and Underground Space Technology journal homepage: www.elsevier.com/locate/tust
Investigation on the tail brush induced loads upon segmental lining of a shield tunnel with small overburden
T
⁎
Guan-lin Yea, Lei Hana, Santosh Kumar Yadava, Xiao-hua Baob, , Chen-cong Liaoa a b
Department of Civil Engineering and State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai 200240, China Department of Civil Engineering, Shenzhen University, Shenzhen 518060, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Construction loads Shield tunnel Tail brush Pressure on segments Monitoring data
Damage to the tunnel segmental lining induced by the construction loads has become an important issue in recent years. To investigate the influence of tail brush induced loads on linings during construction, in-situ monitoring was carried out to measure the pressure on the lining of a large cross-river shield tunnel with a small overburden in silty sandy ground in China. Time history of pressure distribution on the lining during tunnel construction was obtained using PAD type earth pressure gauges. Distribution of bending moments and axial forces and related crack width of segments were also examined. The in-situ monitoring results showed that the pressure induced by tail brush (steel plate brush and wire brush) had an important influence on the segments while the shield machine was passing through. Moreover, results from laboratory tests carried out on steel plate brush and wire brush to investigate the extent of loading, revealed that the load by the steel plate brush was more significant than that by the wire brush. Based on the data analysis and field observation, it can be concluded that the non-uniform distribution of pressure induced by steel plate brush might have caused cracking and water leakages in the segments, which can provide valuable references for damage prevention from construction loads in practice.
1. Introduction
Ishimura, 2006; Bakker & Bezuijen, 2008; Chen & Mo, 2009; Ramoni et al., 2011; Oh & Ziegler, 2014; Yu et al., 2014; Ye et al., 2015; Wang et al., 2016; Gall et al., 2018), and some of them pointed out that the tail brush and the tail grease may impose large pressure upon segments. A figure of the structural damage produced during the construction stage was summarized by the technical committee of Japanese Society of Civil Engineers (JSCE) (Japan Society of Civil Engineering, 2006; Sugimoto, 2006). It was found that the longitudinal cracks in the segments, which are the most common types of damages, mainly occurred in three construction scenarios: (a) the invasion of shield tail and segment (see in Fig. 1a) that usually occurs at curved tunnels with small curvature radius; (b) the jack thrusting at segments contacting simultaneously with two segments of the prior ring (see in Fig. 1b) that usually occurs at large shield tunnel (Cavalaro et al., 2011); (c) the overlarge unsymmetrical pressure (see in Fig. 1c) that is usually caused by the backfill grouting pressure (Mashimo & Ishimura, 2006). Currently, seals made of steel wire brushes are the state-of-the art (David et al., 2010), and they are positioned in several consecutive rows as shown in Fig. 2. The gaps between the brushes are filled with pressurized grease to seal against the backfill grouting liquid and water. Insitu monitoring data showed that the wire brushes pressure acting upon
In recent years, as the lining segments of shield tunnels become thinner and wider, construction loads (e.g. jack thrust force, backfill grouting pressure and tail brush pressure) on the segments have attracted the attentions of tunnelling engineers, and they found that these loads have induced damages to the segments (Wang et al., 2016; Zhang et al., 2017). Among the damages, cracking of the tunnel lining is one of the most common and serious problems occurring in tunnels, leading to the reduction of the durability of the concrete segments and ultimately, the durability and structural performance of the whole tunnel (Zhu & Li, 2011). Investigations on design, measurement and reinforcement method on the tunnel segments were performed (Yang et al., 2007; Cheung et al., 2010; Zhu & Li, 2011; Cavalaro et al., 2012; Liu et al., 2018). However, the construction loads are difficult to predict in the conventional design processes. Thus, identifying the potential risk of construction loads and evaluating their impacts on tunnelling linings has significant importance. Many investigations on tunnelling construction loads including insitu-monitor and theoretical analysis can be found in the literatures (Sramoon et al., 2002; Koyama, 2003; Bezuijen et al., 2004; Mashimo &
⁎
Corresponding author. E-mail addresses: [email protected] (G.-l. Ye), [email protected] (X.-h. Bao), [email protected] (C.-c. Liao).
https://doi.org/10.1016/j.tust.2020.103283 Received 18 June 2018; Received in revised form 24 April 2019; Accepted 3 January 2020 0886-7798/ © 2020 Elsevier Ltd. All rights reserved.
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 1. Typical construction loads in shield tunneling that induce longitudinal cracks in segments.
Fig. 2. Sketch of typical shield tail sealing system with tail brushes and grease.
tail bush in large shield machines has been changed from steel wire brush to steel plates in recent years. Because the steel plates have good wearing resistance and can provide a better sealing effect. However, the influence of such kind of new tail brush on segmental linings is still unknown. Based on the previous study (Han et al., 2017), the aim of this study is to further investigate the characteristic of the steel plate brush and wire brush induced loads and its impact on segments of a large shield tunnel with small overburden. In-situ monitoring was carried out to measure the lining pressure. The measured pressure data and the induced inner force on segments lining were analysed carefully. Apart from in-situ monitoring, laboratory tests were also performed on the steel plate brush and wire brush to study their mechanical characteristics. The results revealed that the non-uniform distribution of pressures due to steel plate brush might have caused cracking and water leakages in the segments. Therefore, the influence of tail brush on tunnel lining cannot be neglected, not only in the case of deep tunnels, but also in the case of shallow tunnels with small overburden.
the segments was much larger than the earth pressure (Arizumi et al., 1999; Koyama, 2003; Mashimo & Ishimura, 2006). Moreover, a numerical study indicated that the wire brush force even could change the shield behaviour, especially the yawing of the shield machine (Sramoon et al., 2002). Recently, the flow resistance of tail grease (Sugimoto et al., 2014) was also studied through element tests. It should be noticed that in almost all researches focused on the influence of the backfill grouting and tail brush pressures on the deep tunnels, the axial force is the dominant factor in the lining design. With the rapid development of urban traffic, cross-river tunnels in Yangtze River Delta, China (Shen et al., 2014, 2016; Wu et al., 2015; Cheng et al., 2017) with length about 1 km and diameter larger than 10 m are needed. For these tunnels, the depth of overburden has to be relatively small to satisfy the slope requirement of the road in such short distance tunnels. In this case, the construction loads are required to be treated carefully although the soil and water pressures are also relatively small. The previous study has evaluated pressures on the linings of a large shield tunnel with a small overburden (Han et al., 2017). However, the properties of the tail brush and its influence on the tunnelling line during construction were not clearly clarified. In another hand, the material of the outermost row (the 3rd row in Fig. 2) of 2
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 3. Steel plate brush (left) and wire brush (right) used for the laboratory tests.
2. Laboratory tests on two types of tail brushes
design (see in Fig. 5a). Table 1 lists the details of plate brush used in the laboratory tests. In the tests, the thickness of the plates are 0.5 t, 1.0 t, 1.5 t and 2.0 t respectively. The tests were conducted in a universal testing machine. As shown in Fig. 6, the brush was attached to a metal plate by bolts, just the same manner as that in the shield machine, and a piece of metal plate with a load cell was used to measure the pressure from the brush when the metal plate pressed the brush.
2.1. Test scheme for steel plate brush and wire brush As mentioned in previous session, in recent years, the material of the outermost row of tail bush has been changed from steel wire brush to steel plate brush in China. It is necessary to know their mechanical characteristics, especially the pressure and deformation relationships of different types of tail brushes. The commercially available steel plate brush and wire brush used in the tests are shown in Fig. 3, and their design drawings are shown in Figs. 4 and 5 respectively. Both types are designed for a void height of 240 mm. The geometries of wire brushes used in the large shield tunnel are almost the same, while the thickness of steel plate brush may change according to the advancing distance and depth of the tunnel. The site engineers found that the steel plate brush could be wore out so that the actual thickness had reduced gradually during the shield advancement. Therefore, in the present study, the steel plate with various thicknesses will be tested. As shown in Figs. 3 and 5, the plate brush is made of many pieces of thin steel sheets that have two types with the thickness of 0.5 mm and 0.75 mm respectively. These sheets are setup in a staggering manner one over other to make an 8-steel-sheet plate brush. The plate with a thickness t = 18 mm is considered as a reference standard unit in the
2.2. Laboratory testing results The pressure on wire brushes of the three similar samples with height of 240 mm, showed the consistent results as shown in Fig. 7. When the deformation was less than 135 mm, the pressure linearly increased with the deformation. Although the value of pressure was only 50 kPa at the deformation of 135 mm, when the deformation increased from 135 mm to 210 mm, the pressure increased exponentially with a final (maximum) value of 350 kPa. The test results of steel plate brushes presented a similar trend with those of wire brushes, as shown in Fig. 8. Thus, the deformation of 135 mm can be taken as the split line between linear growth and the exponential growth. In another hand, the thickness of the plate was found to have a large effect on the maximum pressure. The maximum pressure of the plate with thickness of 2.0 t was about 1000 kPa, which was two times larger than that of
Fig. 4. Design drawing of wire brush. 3
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 5. Design drawing of steel plate brush.
400
Table 1 Type and thickness of plate brush for laboratory tests. Thickness of tail brush
1 2 3 4 5 6 7 8
Plate brush No.1 (made of 0.5 mm steel sheet)
0.5 1.0 0.5 2.0 0.5 1.0 0.5 2.0
Pressure (kPa)
Type of tail brush
Plate brush No.2 (made of 0.75 mm steel sheet)
Wire brush-1 Wire brush-2 Wire brush-3
350
Test No.
t t t t t t t t
300 250 200 150 100 50
Note: t = 18 mm, a reference standard thickness of steel plate brush used in large shield machine.
0 0
50
100
150
200
250
Deformation (mm)
the plate with thickness of 1.0 t, and four times larger than that of the plate with thickness of 0.5 t. Meanwhile, the thickness of a single steel sheet had a minor effect on the pressure, although the pressure on the plate with the sheet of 0.75 mm thickness was a little larger than that with the sheet of 0.5 mm thickness. On the whole, the stiffness (=pressure/deformation) of the steel plate brush was approximately twice that of the wire brush.
Fig. 7. Pressure deformation curves for wire brush.
3. In-situ monitoring of tail brush induced loads 3.1. Tunnel profile As introduced in the previous study (Han et al., 2017), the West Chengjiang Road tunnel, a twin-tube shield tunnel consisting of the north and south lines build in Jiangyin City in China, is a cross-river
Load cell
Metal plate Tail brush
Clamp
Fig. 6. Compression tests of tail brush in a universal testing machine. 4
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
1600
Plate-0.5mm-0.5t Plate-0.5mm-1.0t Plate-0.5mm-1.5t Plate-0.5mm-2.0t Plate-0.75mm-0.5t Plate-0.75mm-1.0t Plate-0.75mm-1.5t Plate-0.75mm-2.0t
Pressure (kPa)
1400 1200 1000 800 600 400 200 0 0
50
100
150
200
250
Fig. 10. PAD type earth pressure gauge in tunnel lining.
Deformation (mm) rivers were selected for the in-situ monitoring. The soil profiles of the two sections are presented in Fig. 9. The depth of water and overburden soil of section A are 7 m and 7.69 m respectively, while those of section B are 3 m and 7.77 m respectively. Thus, the ratios of overburden depth of soil to tunnel diameter (H/D) of the two sections are approximately 0.7, which is classified as shallow overburden. In-situ monitoring was carried out to investigate the tail brush induced loads on the segments. Installed positions of pressure gauges in section A and B are shown in Fig. 9. Fig. 10 presented the installation of PAD type earth pressure gauge, a very thin pressure cell with 750 mm length and 450 mm width (Hashimoto et al., 1993, 2002), in segmental lining to measure the pressure on tunnel segments. A metal frame was used to mounted the pressure gauge on the segment, and the installation positions were determined when the segments were manufactured. The pressure cell of the earth pressure gauge was flexible and thin (< 6 mm), so that it can be easily fixed onto the segment extrados without modifying the external soil pressure. The filled liquid inside the pressure cell transferred the earth pressure to the pressure sensor precisely. The real-time data recording in Section A (Ring No. 118) commenced after the assembly of Ring No. 119 due to the construction
Fig. 8. Pressure deformation curves for plate brush.
road tunnel with small over burden. The total length of the tunnel is 1.27 km, of which 660 m was constructed by the shield tunnelling method. Most part of the tunnel is located in the silty sandy and silty clay strata. A slurry shield machine was used to excavate the north line first with the tunnel out diameter of 11.58 m and then the south line with the tunnel outer diameters of 11.36 m. Eight pieces of reinforced concrete (RC) segments with the thickness of 0.5 m and width of 1.5 m were used to form the tunnel ring by staggered assembly method. As the shield tunnel runs across two rivers. For the tunnel in under-river area, the thickness of overburden soil was less than one diameter (D) of the tunnel, and the minimum thickness of overburden soil was only 7 m. In the shield machine, the first and second rows of the tail brush were wire brushes, and the third row was a steel plate brush.
3.2. Outline of in-situ monitoring Section A and section B in the south line of the tunnel located under
SPT-N
10 20 30 40
SPT-N
10.77m
silty-sandy 11.36m
-20
-25 silty clay
-15
11.36m
-20 silty clay -25
-30
PAD earth pressure gauge Pore-water pressure gauge
PAD earth pressure gauge Pore-water pressure gauge
(a) Section A
(b) Section B
Fig. 9. Soil profiles of the selected sections and installed pressure gauges in tunnel. 5
3.13m 2.73m 1.46m 3m
silt
8.16m
-10
silty clay
6.52m
silty-sandy
river bottom soil 7.77m
silt
-5
11.26m
-15
7.69m
river bottom soil
10 20 30 40
river 4.39m 1.1m
14.69m
-5
-10
0
7m
river
6.25m
0
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 11. Comparison of time history of pressures upon segments between the measured and calculated values.
6
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
showed small fluctuation after the third row of tail brushes passed through. The reason of this fluctuation was considered to be the grouting pressure and the operation of the shield machine (Hashimoto et al., 2006). The maximum observed pressure in both sections was nearly 400 kPa, which was almost twice the theoretical earth pressure. Finally, the pressure on the tunnel linings decreased to a steady state, which is close to the theoretical earth pressure after 5 rings away from the shield tail. Through the measured results, it can be concluded that the most critical loads on the segments during construction appeared when the third row of tail brushes was passing through.
conditions. At that time, the first row of tail brushes had passed beyond the earth pressure gauges and the second row of tail brushes was acting upon the earth pressure gauges. For Section B (Ring No. 340), the realtime data recording commenced immediately after the assembly of this ring. In both sections, a ‘Handy Data logger’ measured the pressure data of the gauges just after the assembly of the segments, and these data were taken as the initial loading values.
4. Monitoring results and discussion 4.1. Overall monitoring results
4.2. Discussion on the influence of tail brush and tail grease The time histories of pressures on the segments of the two monitored sections are shown in Fig. 11 that the angle was taken from the starting point of the right spring line in an anticlockwise direction. The theoretical earth pressure, calculated using the Japanese conventional design method (Japan Society of Civil Engineering, 2010), is also shown in Fig. 11. In Section A (No.118), as mentioned above, the real-time data recording was started after the assembly of the next ring (No. 119), and at that time, the pressure gauges had already been passed by the first row of tail brushes and touched the second row of tail brushes. Therefore, the pressure induced by the first row and a part of that induced by the second row of tail brushes is missing in Fig. 11(a). In Section B (Ring No.340), the recording data commenced immediately after the assembly of the monitored ring. Accordingly, the pressure before the first row of tail brushes contacted with the ring was recorded, as shown in Fig. 11(b). By cross checking with the shield advancing records and the distances between tail brushes (Fig. 12), four pivotal moments are identified in Fig. 11. Time “a” indicates the time that the third row of tail brushes began acting upon the earth pressure gauges; Time “b” indicates the time that the third row of tail brushes began passing away from the earth pressure gauges. Time “c” indicates the time that the third row of tail brushes finished passing through the earth pressure gauges. Time “d” indicates the time that the shield advancing through the current ring finished. From Fig. 11, it can be seen that, the values of the pressures on segmental lining began to response when contacting with the tail brushes. The pressures were less than 200 kPa when the first and second rows of tail brushes were passing the ring, and the pressures at the crown were less than 50 kPa, while the pressures at the spring line were about 150 kPa. However, the pressures fluctuated dramatically when the third row of tail brushes passed through, and the maximum pressure was almost 900 kPa at the left spring line of Section B. The pressures
The results in Fig. 11 indicated that when the first and second rows of tail brushes passed through, the measured pressures on the lining were lower than the theoretical earth pressures. Apart from the left spring line of Section B, all measured values were less than half of the theoretical values. Some previous studies revealed that the tail grease might increase pressure on the lining (e.g., Koyama 2003; Mashimo & Ishimura 2006). The pressure of the tail grease depend on both the discharge pressure of grease pump and the pressure loss in the injection pipe. In current study, these relative small measured pressures in first and second row brushes indicated that the influence from the tail grease was quite limited. A small discharge pressure of grease pump perhaps was the reason. When the third row of the tail brushes passing through, most of the measured pressures increased and became larger than the theoretical values. The maximum value was about 700 kPa at right spring line of Section A, and about 900 kPa at left spring line of section B, that both were almost 3–4 times larger than the design earth pressure. The large discrepancy between the pressures induced by the first two rows and those by the third row of tail brushes was obviously due to the different materials used in the tail brushes. As mentioned in session 3.1, the first and second rows were conventional wire brushes, while the third row was steel plate brush as shown in Fig. 12(a). Similar characteristic of large discrepancy was also found during the laboratory tests on the wire brush and plate brush. The maximum loads induced by the steel plate brush was 700–900 kPa, very close to the maximum value measured in laboratory tests. The outmost steel plate brush is much stiffer than the wire brush and is considered to have provided better sealing effect than the wire type. However, it can impose large pressure upon segments. In addition, after construction, the engineers found that the gaps between steel sheets were filled with solidified backfill grouting, as shown in Fig. 12(b), which may have significantly
Steel plate brush
Wire brushes
450mm
Wire brushes
Steel plate brush
450mm
(a) Before construction
(b) After construction
Fig.12. Photograph of tail brushes before and after construction. 7
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
90 60
150
800
30
180
210 240
0 100 200 300 400 500 600 700 800
270
0
330
Tail brush pressure (kPa)
120
600 400 200 0
300 Unit: kPa
0 90 180 270
0
Time "a"
6
Theoretical earth pressure Tail brush pressure of 9 minutes after Time "a"
12
18
24
30
Time (min)
Fig. 13. Maximum non-uniform pressure in section A induced by 3rd-row steel plate tail brush.
increased the stiffness of the steel plate brush, resulting in unexpected large pressures upon tunnel lining. Fig. 13 presented the distribution of tail brush indueced lining pressure of Section A during passing through the third row of tail brush. The pressure at the right spring was about 700 kPa and that at the crown was less than 100 kPa when 9 min after time “a”. Such a nonuniform pressure distribution is prone to produce excessive bending moment in the tunnel lining. In the case of a shallow tunnel such as the project of West Chengjiang Road tunnel, since the design value of the maximum moment is relative small, a small amount of excessive bending moment can easily lead to cracking in the RC (reinforced concrete) segments. This phenomenon will be analysed in the next session. It is noted that besides the tail brushes induced pressure, other sources of loads, for example, backfill grouting (Bezuijen et al., 2004), tail grease (Koyama 2003; Mashimo & Ishimura 2006), jacking force and tolerances during ring assembly as well as small misalignments (Cavalaro et al., 2011, 2012) also produce high stresses in tunnel segments. These may be higher than the effect of the passing tail sealing.
Fig. 15. Relationship between bending moment and segment joint rotation.
circumferential direction, are represented by rotational spring elements with a rotational stiffness coefficient kθ. The relationship between kθ and the segment joint rotation θ obtained from the lining tests of a similar tunnel is shown in Fig. 15. A fictitious continuous spring supported with a coefficient of subgrade reaction kr in the normal direction, which has no stiffness on the tension side, is used to model the effect of supporting a tunnel ring inside of the shield machine (Mashimo & Ishimura 2006). Since the subgrade reaction spring is fictitious, whose values cannot be determined directly, four different values are given to kr, i.e. 10 MN/m3, 20 MN/m3, 50 MN/m3 and 100 MN/m3, which varies between the values of soft ground and stiff ground. In the calculation, the steel plate tail brush induced forces and the dead weight of the segments were considered as loading upon the lining. Table 2 lists the values of the parameters used in the calculation. An assumed distribution of the tail brush pressure extended from the
5. Analysis of inner force of lining by beam-spring model The beam-spring model as shown in Fig. 14 is used to analyse the inner force of the lining due to the steel plate tail brush. In the model, the RC (reinforced concrete) segments are modelled as beam elements, and the segment joints, which connect the segments in the
Table 2 Parameters used in the beam-spring model.
Fig. 14. Beam-spring model for one ring of segments. 8
Parameter
Value
Tunnel radius (m) Thickness of segment (m) Width of segment (m) Elastic modulus of segment (kN/m2) Poisson’s ratio of segment Unit weight of segment (kN/m3)
5.43 0.5 1.5 3.5 × 107 0.25 25
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 16. Tail pressure distribution in section A.
measured pressures in section A, was adopted in the calculation, as shown in Fig. 16. It is corresponding to the most non-uniform pressure distribution at the moment of 9 min after time “a” in Fig. 13. Fig. 17 shows the calculation results of bending moments and the axial forces with different values of kr. It can be seen that the maximum positive bending moment appears at the location of 0° (right spring line) and the maximum negative bending moment appeared at the location near 60° or 300° in all calculations. Based on the maximum bending moments (both positive and negative) and the corresponding axial forces with different values of kr, the crack widths of segments can be obtained, as shown in Fig. 18. It can be seen from the figure that all crack widths exceed the design limitation (wlim = 0.2 mm) of the crack for tunnel lining. In other words, the calculated crack widths far exceed the design value. In the current tunnel, water leakage from cracks in segments were observed here and there around the monitor sections right after passing through the tail voids, as shown in Fig. 19. These cracks were considered to be caused by large bending moments from non-uniform construction loads. Thus, this phenomenon of water leakage from cracks in turn proved that the non-uniform distribution of pressure induced by the steel plate brush might be one of the main factors that lead to cracking. The cracks will reduce the durability and
Fig. 18. Crack width of segments under different bending moments and corresponding axial forces.
service performance of RC segments (Wu et al., 2014). Therefore, the influence of tail brush induced loads (especial the steel plate type) on tunnel lining cannot be neglected, not only in the case of deep tunnel, but also in the case of shallow tunnel with small overburden. 6. Conclusions Laboratory tests, in-situ monitoring and theoretical calculation on the tail brush induced loads upon shield tunnel lining during construction were carried out in this study. The time history of lining pressure distribution and inner force were analysed. The following conclusions can be obtained. (1) Laboratory tests showed that the maximum load induced by the steel plate brush was twice that induced by the wire brush; however, insitu monitoring showed a much larger ratio. The reason was considered to be the solidification of backfill grouting inside of the steel plates, which significantly increased the stiffness of the steel plates brush. (2) The largest pressure acting upon the segments was observed
Fig. 17. Bending moments and axial forces induced by tail brush pressure. 9
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
the production tolerances of concrete tunnel segments. J. Perform. Constr. Facil. 26 (6), 824–834. Chen, J.S., Mo, H.H., 2009. Numerical study on crack problems in segments of shield tunnel using finite element method. Tunn. Undergr. Space Technol. 24 (1), 91–102. Cheng, W.C., Ni, J.C., Shen, S.L., 2017. Experimental and analytical modeling of shield segment under cyclic loading. International Journal of Geomechnics ASCE (2017), 04016146(1–17). Cheung, L.L.K., Soga, K., Bennett, P.J., Kobayashi, Y., Amatya, B., Wright, P., 2010. Optical fibre strain measurement for tunnel lining monitoring. Proceedings of the Institution of Civil Engineers - Geotechnical Engineering 163(3). 119–130. David, C., Nicole, M., Alfred, S., 2010. Introduction to Tunnel Construction. CRC Press, ISBN-13: 978-0415468428. Gall, V.E., Marwan, A., Smarslik, M., Obel, M., Mark, P., Meschke, G., 2018. A holistic approach for the investigation of lining response to mechanized tunnelling induced construction loadings. Undergr. Space 3 (1), 45–60. Han, L., Ye, G.L., Chen, J.J., Xia, X.H., Wang, J.H., 2017. Pressures on the lining of a large shield tunnel with a small overburden: a case study. Tunn. Undergr. Space Technol. 64 (2017), 1–9. Hashimoto, T., Konda, T., Brinkman, J., Feddema, A., Kano, Y., 2006. Simultaneous backfill grouting, pressure development in construction phase and in the long-term. Tunnelling-A Decade of Progress, GeoDelft 1995-2005, Taylor & Francis 2006, pp. 101-107. Hashimoto, T., Nagaya, J., Konda, T., Tamura, T., 2002. Observation of lining pressure due to shield tunnelling. Geotech. Aspects Undergr Constr Soft Ground IS-Toulouse 2002, 119–124. Hashimoto, T., Yabe, K., Yamane, S., Ito, H., 1993. Development of pad type earth pressure cell for shield segment. Proceedings of 28th Annual Report of JGS, Tokyo: Japan Geotechnical Society (1993), 2055–2058. Japan Society of Civil Engineering (JSCE), 2006. Tunnel library17: Construction loads during shield tunnelling. Tokyo Japan: JSCE (in Japanese). Japan Society of Civil Engineering, 2010. Japanese standard for shield tunnelling. China Architecture & Building Press, Beijing (in Chinese). Koyama, Y., 2003. Present status and technology of shield tunnelling method in Japan. Tunn. Undergr. Space Technol. 18 (2), 145–159. Liu, X., Jiang, Z., Yuan, Y., Mang, H., 2018. Numerical investigation of the mechanical behavior of deformed segmental tunnel linings, strengthened by epoxy-bonded filament wound profiles. Tunn. Undergr. Space Technol. 78, 231–244. Mashimo, H., Ishimura, T., 2006. Numerical modelling of the behaviour of shield tunnel lining during assembly of a tunnel ring. Geotechnical Aspects of Underground Construction in Soft Ground-Proceedings of the 5th International Symposium TC28, 587–593. Oh, J.Y., Ziegler, M., 2014. Investigation on influence of tail void grouting on the surface settlements during shield tunnelling using a stress-pore pressure coupled analysis. KSCE J. Civ. Eng. 18 (3), 803–811. Ramoni, M., Lavdas, N., Anagnostou, G., 2011. Squeezing loading of segmental linings and the effect of backfilling. Tunn. Undergr. Space Technol. 26 (6), 692–717. Shen, S.L., Cui, Q.L., Ho, E.C., Xu, Y.S., 2016. Ground response to multiple parallel microtunneling operations in cemented silty clay and sand. J. Geotech. Geoenviron. Eng. 142 (5), 1–11. Shen, S.L., Wu, H.N., Cui, Y.J., Yin, Z.Y., 2014. Long-term settlement behavior of the metro tunnel in Shanghai. Tunn. Undergr. Space Technol. 40, 309–323. Sugimoto, M., 2006. Causes of shield segment damages during construction. In: International Symposium on Underground Excavation and Tunnelling, Bangkok, 2-4 Feb. Bangkok Thailand: EIT67–74. Sugimoto, M., Hirai, Y., Huynh, T.N., Tamai, T., 2014. Study on characteristics of grease flow resistance by element test. Geotechnical Aspects of Underground Construction in Soft Ground. CRC Press, pp. 167–171. Sramoon, A., Sugimoto, M., Kayukawa, K., 2002. Theoretical model of shield behavior during excavation. II: Application. J. Geotech. Geoenviron. Eng. 128 (2), 156–165. Wang, Z., Wang, L., Wang, J., Li, L., 2016. Case Study of Rehabilitation of a Damaged Underwater Tunnel in the Construction Phase. J. Perform. Constr. Facil. 30 (1), C4014003. Wu, H.N., Huang, R.Q., Sun, W.J., Shen, S.L., Xu, Y.S., Liu, Y.B., Du, S.J., 2014. Leaking behavior of shield tunnels under the Huangpu River of Shanghai with induced hazards. Nat. Hazards 70 (2), 1115–1132. Wu, H.N., Shen, S.L., Liao, S.M., Yin, Z.Y., 2015. Longitudinal structural modelling of shield tunnels considering shearing dislocation between segmental rings. Tunn. Undergr. Space Technol. 50 (2015), 317–323. Yang, C.Y., Xu, M.X., Chen, W.F., 2007. Reliability analysis of shotcrete lining during tunnel construction. J. Constr. Eng. Manage. 133 (12), 975–981. Yu, H.T., Yuan, Y., Yu, G.X., Liu, X., 2014. Evaluation of influence of vibrations generated by blasting construction on an existing tunnel in soft soils. Tunn. Undergr. Space Technol. 43, 59–66. Ye, G.L., Hashimoto, T., Shen, S.L., Zhu, H.H., Bai, T.H., 2015. Lessons learnt from unusual ground settlement during Double-O-Tube tunnelling in soft ground. Tunn. Undergr. Space Technol. 49 (2015), 79–91. Zhang, Y., Shi, Y., Zhao, Y., Yang, J., 2017. Damage in concrete lining of an operational tunnel. J. Perform. Constr. Facil. 31 (4), 06017002. Zhu, J.H., Li, Y., 2011. Quality Testing and Evaluation of the Damaged Lining of Zhoulichong Railway Tunnel. Tunnel Management, Emerging Technologies, and Innovation, Geotechnical Special Publication No. 221.
Fig. 19. Water leakage due to cracks of segment after the installation of tail brushes.
during passing through the steel plate tail brush. The measured value was more than four times of the theoretical earth pressure value, and the distribution was quite non-uniform. Analysis results using beamspring model revealed that the non-uniform distribution of pressures induced by the steel plate brush resulted in a crack width of the segments far exceeded the design limitation. This indicated that the tail brush induced load might be the main reason for the observed longitudinal cracks in the segments. (3) The change of the third row of tail brush from wire type to the steel plate type by the engineers can prevent the leakage more effectively. However, the steel plate type has induced unexpected large loads upon segments. Especially in the case of shallow tunnel that the axial forces are relatively small, the large bending moment caused by the non-uniform distribution of tail brush pressure may damage the segments. Thus, the consideration of tail brush pressure is extremely important for tunnels with a shallow overburden. Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant Nos. 41630633 and 51678369), Technical Innovation Foundation of Shenzhen (Grant No. JCYJ20170302143610976). References Arizumi, T., Okadome, K., Igarashi, H., Nagaya, J., 1999. Investigation of the load acting on shield segments during tunnelling. Proc. Tunnel Eng., JSCE 9, 271–276 (in Japanese). Bakker, K.J., Bezuijen, A., 2008. Ten years of bored tunnels in the Netherlands: Part II structural issues. In: Geotechnical Aspects of Underground Construction in Soft Ground: Proceedings of the 6th International Symposium (IS-Shanghai 2008) pp. 249. Bezuijen, A., Talmon, A.M., Kaalberg, F.J., Plugge, R., 2004. Field measurements of grout pressures during tunnelling of the Sophia rail tunnel. Soils Found. 44 (1), 39–48. Cavalaro, S.H.P., Blom, C.B.M., Walraven, J.C., Aguado, A., 2011. Structural analysis of contact deficiencies in segmented lining. Tunn. Undergr. Space Technol. 26 (6), 734–749. Cavalaro, S.H.P., Blom, C.B.M., Aguado, A., Walraven, J.C., 2012. New design method for
10
Contents lists available at ScienceDirect
Tunnelling and Underground Space Technology journal homepage: www.elsevier.com/locate/tust
Investigation on the tail brush induced loads upon segmental lining of a shield tunnel with small overburden
T
⁎
Guan-lin Yea, Lei Hana, Santosh Kumar Yadava, Xiao-hua Baob, , Chen-cong Liaoa a b
Department of Civil Engineering and State Key Laboratory of Ocean Engineering, Shanghai Jiao Tong University, Shanghai 200240, China Department of Civil Engineering, Shenzhen University, Shenzhen 518060, China
A R T I C LE I N FO
A B S T R A C T
Keywords: Construction loads Shield tunnel Tail brush Pressure on segments Monitoring data
Damage to the tunnel segmental lining induced by the construction loads has become an important issue in recent years. To investigate the influence of tail brush induced loads on linings during construction, in-situ monitoring was carried out to measure the pressure on the lining of a large cross-river shield tunnel with a small overburden in silty sandy ground in China. Time history of pressure distribution on the lining during tunnel construction was obtained using PAD type earth pressure gauges. Distribution of bending moments and axial forces and related crack width of segments were also examined. The in-situ monitoring results showed that the pressure induced by tail brush (steel plate brush and wire brush) had an important influence on the segments while the shield machine was passing through. Moreover, results from laboratory tests carried out on steel plate brush and wire brush to investigate the extent of loading, revealed that the load by the steel plate brush was more significant than that by the wire brush. Based on the data analysis and field observation, it can be concluded that the non-uniform distribution of pressure induced by steel plate brush might have caused cracking and water leakages in the segments, which can provide valuable references for damage prevention from construction loads in practice.
1. Introduction
Ishimura, 2006; Bakker & Bezuijen, 2008; Chen & Mo, 2009; Ramoni et al., 2011; Oh & Ziegler, 2014; Yu et al., 2014; Ye et al., 2015; Wang et al., 2016; Gall et al., 2018), and some of them pointed out that the tail brush and the tail grease may impose large pressure upon segments. A figure of the structural damage produced during the construction stage was summarized by the technical committee of Japanese Society of Civil Engineers (JSCE) (Japan Society of Civil Engineering, 2006; Sugimoto, 2006). It was found that the longitudinal cracks in the segments, which are the most common types of damages, mainly occurred in three construction scenarios: (a) the invasion of shield tail and segment (see in Fig. 1a) that usually occurs at curved tunnels with small curvature radius; (b) the jack thrusting at segments contacting simultaneously with two segments of the prior ring (see in Fig. 1b) that usually occurs at large shield tunnel (Cavalaro et al., 2011); (c) the overlarge unsymmetrical pressure (see in Fig. 1c) that is usually caused by the backfill grouting pressure (Mashimo & Ishimura, 2006). Currently, seals made of steel wire brushes are the state-of-the art (David et al., 2010), and they are positioned in several consecutive rows as shown in Fig. 2. The gaps between the brushes are filled with pressurized grease to seal against the backfill grouting liquid and water. Insitu monitoring data showed that the wire brushes pressure acting upon
In recent years, as the lining segments of shield tunnels become thinner and wider, construction loads (e.g. jack thrust force, backfill grouting pressure and tail brush pressure) on the segments have attracted the attentions of tunnelling engineers, and they found that these loads have induced damages to the segments (Wang et al., 2016; Zhang et al., 2017). Among the damages, cracking of the tunnel lining is one of the most common and serious problems occurring in tunnels, leading to the reduction of the durability of the concrete segments and ultimately, the durability and structural performance of the whole tunnel (Zhu & Li, 2011). Investigations on design, measurement and reinforcement method on the tunnel segments were performed (Yang et al., 2007; Cheung et al., 2010; Zhu & Li, 2011; Cavalaro et al., 2012; Liu et al., 2018). However, the construction loads are difficult to predict in the conventional design processes. Thus, identifying the potential risk of construction loads and evaluating their impacts on tunnelling linings has significant importance. Many investigations on tunnelling construction loads including insitu-monitor and theoretical analysis can be found in the literatures (Sramoon et al., 2002; Koyama, 2003; Bezuijen et al., 2004; Mashimo &
⁎
Corresponding author. E-mail addresses: [email protected] (G.-l. Ye), [email protected] (X.-h. Bao), [email protected] (C.-c. Liao).
https://doi.org/10.1016/j.tust.2020.103283 Received 18 June 2018; Received in revised form 24 April 2019; Accepted 3 January 2020 0886-7798/ © 2020 Elsevier Ltd. All rights reserved.
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 1. Typical construction loads in shield tunneling that induce longitudinal cracks in segments.
Fig. 2. Sketch of typical shield tail sealing system with tail brushes and grease.
tail bush in large shield machines has been changed from steel wire brush to steel plates in recent years. Because the steel plates have good wearing resistance and can provide a better sealing effect. However, the influence of such kind of new tail brush on segmental linings is still unknown. Based on the previous study (Han et al., 2017), the aim of this study is to further investigate the characteristic of the steel plate brush and wire brush induced loads and its impact on segments of a large shield tunnel with small overburden. In-situ monitoring was carried out to measure the lining pressure. The measured pressure data and the induced inner force on segments lining were analysed carefully. Apart from in-situ monitoring, laboratory tests were also performed on the steel plate brush and wire brush to study their mechanical characteristics. The results revealed that the non-uniform distribution of pressures due to steel plate brush might have caused cracking and water leakages in the segments. Therefore, the influence of tail brush on tunnel lining cannot be neglected, not only in the case of deep tunnels, but also in the case of shallow tunnels with small overburden.
the segments was much larger than the earth pressure (Arizumi et al., 1999; Koyama, 2003; Mashimo & Ishimura, 2006). Moreover, a numerical study indicated that the wire brush force even could change the shield behaviour, especially the yawing of the shield machine (Sramoon et al., 2002). Recently, the flow resistance of tail grease (Sugimoto et al., 2014) was also studied through element tests. It should be noticed that in almost all researches focused on the influence of the backfill grouting and tail brush pressures on the deep tunnels, the axial force is the dominant factor in the lining design. With the rapid development of urban traffic, cross-river tunnels in Yangtze River Delta, China (Shen et al., 2014, 2016; Wu et al., 2015; Cheng et al., 2017) with length about 1 km and diameter larger than 10 m are needed. For these tunnels, the depth of overburden has to be relatively small to satisfy the slope requirement of the road in such short distance tunnels. In this case, the construction loads are required to be treated carefully although the soil and water pressures are also relatively small. The previous study has evaluated pressures on the linings of a large shield tunnel with a small overburden (Han et al., 2017). However, the properties of the tail brush and its influence on the tunnelling line during construction were not clearly clarified. In another hand, the material of the outermost row (the 3rd row in Fig. 2) of 2
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 3. Steel plate brush (left) and wire brush (right) used for the laboratory tests.
2. Laboratory tests on two types of tail brushes
design (see in Fig. 5a). Table 1 lists the details of plate brush used in the laboratory tests. In the tests, the thickness of the plates are 0.5 t, 1.0 t, 1.5 t and 2.0 t respectively. The tests were conducted in a universal testing machine. As shown in Fig. 6, the brush was attached to a metal plate by bolts, just the same manner as that in the shield machine, and a piece of metal plate with a load cell was used to measure the pressure from the brush when the metal plate pressed the brush.
2.1. Test scheme for steel plate brush and wire brush As mentioned in previous session, in recent years, the material of the outermost row of tail bush has been changed from steel wire brush to steel plate brush in China. It is necessary to know their mechanical characteristics, especially the pressure and deformation relationships of different types of tail brushes. The commercially available steel plate brush and wire brush used in the tests are shown in Fig. 3, and their design drawings are shown in Figs. 4 and 5 respectively. Both types are designed for a void height of 240 mm. The geometries of wire brushes used in the large shield tunnel are almost the same, while the thickness of steel plate brush may change according to the advancing distance and depth of the tunnel. The site engineers found that the steel plate brush could be wore out so that the actual thickness had reduced gradually during the shield advancement. Therefore, in the present study, the steel plate with various thicknesses will be tested. As shown in Figs. 3 and 5, the plate brush is made of many pieces of thin steel sheets that have two types with the thickness of 0.5 mm and 0.75 mm respectively. These sheets are setup in a staggering manner one over other to make an 8-steel-sheet plate brush. The plate with a thickness t = 18 mm is considered as a reference standard unit in the
2.2. Laboratory testing results The pressure on wire brushes of the three similar samples with height of 240 mm, showed the consistent results as shown in Fig. 7. When the deformation was less than 135 mm, the pressure linearly increased with the deformation. Although the value of pressure was only 50 kPa at the deformation of 135 mm, when the deformation increased from 135 mm to 210 mm, the pressure increased exponentially with a final (maximum) value of 350 kPa. The test results of steel plate brushes presented a similar trend with those of wire brushes, as shown in Fig. 8. Thus, the deformation of 135 mm can be taken as the split line between linear growth and the exponential growth. In another hand, the thickness of the plate was found to have a large effect on the maximum pressure. The maximum pressure of the plate with thickness of 2.0 t was about 1000 kPa, which was two times larger than that of
Fig. 4. Design drawing of wire brush. 3
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 5. Design drawing of steel plate brush.
400
Table 1 Type and thickness of plate brush for laboratory tests. Thickness of tail brush
1 2 3 4 5 6 7 8
Plate brush No.1 (made of 0.5 mm steel sheet)
0.5 1.0 0.5 2.0 0.5 1.0 0.5 2.0
Pressure (kPa)
Type of tail brush
Plate brush No.2 (made of 0.75 mm steel sheet)
Wire brush-1 Wire brush-2 Wire brush-3
350
Test No.
t t t t t t t t
300 250 200 150 100 50
Note: t = 18 mm, a reference standard thickness of steel plate brush used in large shield machine.
0 0
50
100
150
200
250
Deformation (mm)
the plate with thickness of 1.0 t, and four times larger than that of the plate with thickness of 0.5 t. Meanwhile, the thickness of a single steel sheet had a minor effect on the pressure, although the pressure on the plate with the sheet of 0.75 mm thickness was a little larger than that with the sheet of 0.5 mm thickness. On the whole, the stiffness (=pressure/deformation) of the steel plate brush was approximately twice that of the wire brush.
Fig. 7. Pressure deformation curves for wire brush.
3. In-situ monitoring of tail brush induced loads 3.1. Tunnel profile As introduced in the previous study (Han et al., 2017), the West Chengjiang Road tunnel, a twin-tube shield tunnel consisting of the north and south lines build in Jiangyin City in China, is a cross-river
Load cell
Metal plate Tail brush
Clamp
Fig. 6. Compression tests of tail brush in a universal testing machine. 4
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
1600
Plate-0.5mm-0.5t Plate-0.5mm-1.0t Plate-0.5mm-1.5t Plate-0.5mm-2.0t Plate-0.75mm-0.5t Plate-0.75mm-1.0t Plate-0.75mm-1.5t Plate-0.75mm-2.0t
Pressure (kPa)
1400 1200 1000 800 600 400 200 0 0
50
100
150
200
250
Fig. 10. PAD type earth pressure gauge in tunnel lining.
Deformation (mm) rivers were selected for the in-situ monitoring. The soil profiles of the two sections are presented in Fig. 9. The depth of water and overburden soil of section A are 7 m and 7.69 m respectively, while those of section B are 3 m and 7.77 m respectively. Thus, the ratios of overburden depth of soil to tunnel diameter (H/D) of the two sections are approximately 0.7, which is classified as shallow overburden. In-situ monitoring was carried out to investigate the tail brush induced loads on the segments. Installed positions of pressure gauges in section A and B are shown in Fig. 9. Fig. 10 presented the installation of PAD type earth pressure gauge, a very thin pressure cell with 750 mm length and 450 mm width (Hashimoto et al., 1993, 2002), in segmental lining to measure the pressure on tunnel segments. A metal frame was used to mounted the pressure gauge on the segment, and the installation positions were determined when the segments were manufactured. The pressure cell of the earth pressure gauge was flexible and thin (< 6 mm), so that it can be easily fixed onto the segment extrados without modifying the external soil pressure. The filled liquid inside the pressure cell transferred the earth pressure to the pressure sensor precisely. The real-time data recording in Section A (Ring No. 118) commenced after the assembly of Ring No. 119 due to the construction
Fig. 8. Pressure deformation curves for plate brush.
road tunnel with small over burden. The total length of the tunnel is 1.27 km, of which 660 m was constructed by the shield tunnelling method. Most part of the tunnel is located in the silty sandy and silty clay strata. A slurry shield machine was used to excavate the north line first with the tunnel out diameter of 11.58 m and then the south line with the tunnel outer diameters of 11.36 m. Eight pieces of reinforced concrete (RC) segments with the thickness of 0.5 m and width of 1.5 m were used to form the tunnel ring by staggered assembly method. As the shield tunnel runs across two rivers. For the tunnel in under-river area, the thickness of overburden soil was less than one diameter (D) of the tunnel, and the minimum thickness of overburden soil was only 7 m. In the shield machine, the first and second rows of the tail brush were wire brushes, and the third row was a steel plate brush.
3.2. Outline of in-situ monitoring Section A and section B in the south line of the tunnel located under
SPT-N
10 20 30 40
SPT-N
10.77m
silty-sandy 11.36m
-20
-25 silty clay
-15
11.36m
-20 silty clay -25
-30
PAD earth pressure gauge Pore-water pressure gauge
PAD earth pressure gauge Pore-water pressure gauge
(a) Section A
(b) Section B
Fig. 9. Soil profiles of the selected sections and installed pressure gauges in tunnel. 5
3.13m 2.73m 1.46m 3m
silt
8.16m
-10
silty clay
6.52m
silty-sandy
river bottom soil 7.77m
silt
-5
11.26m
-15
7.69m
river bottom soil
10 20 30 40
river 4.39m 1.1m
14.69m
-5
-10
0
7m
river
6.25m
0
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 11. Comparison of time history of pressures upon segments between the measured and calculated values.
6
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
showed small fluctuation after the third row of tail brushes passed through. The reason of this fluctuation was considered to be the grouting pressure and the operation of the shield machine (Hashimoto et al., 2006). The maximum observed pressure in both sections was nearly 400 kPa, which was almost twice the theoretical earth pressure. Finally, the pressure on the tunnel linings decreased to a steady state, which is close to the theoretical earth pressure after 5 rings away from the shield tail. Through the measured results, it can be concluded that the most critical loads on the segments during construction appeared when the third row of tail brushes was passing through.
conditions. At that time, the first row of tail brushes had passed beyond the earth pressure gauges and the second row of tail brushes was acting upon the earth pressure gauges. For Section B (Ring No. 340), the realtime data recording commenced immediately after the assembly of this ring. In both sections, a ‘Handy Data logger’ measured the pressure data of the gauges just after the assembly of the segments, and these data were taken as the initial loading values.
4. Monitoring results and discussion 4.1. Overall monitoring results
4.2. Discussion on the influence of tail brush and tail grease The time histories of pressures on the segments of the two monitored sections are shown in Fig. 11 that the angle was taken from the starting point of the right spring line in an anticlockwise direction. The theoretical earth pressure, calculated using the Japanese conventional design method (Japan Society of Civil Engineering, 2010), is also shown in Fig. 11. In Section A (No.118), as mentioned above, the real-time data recording was started after the assembly of the next ring (No. 119), and at that time, the pressure gauges had already been passed by the first row of tail brushes and touched the second row of tail brushes. Therefore, the pressure induced by the first row and a part of that induced by the second row of tail brushes is missing in Fig. 11(a). In Section B (Ring No.340), the recording data commenced immediately after the assembly of the monitored ring. Accordingly, the pressure before the first row of tail brushes contacted with the ring was recorded, as shown in Fig. 11(b). By cross checking with the shield advancing records and the distances between tail brushes (Fig. 12), four pivotal moments are identified in Fig. 11. Time “a” indicates the time that the third row of tail brushes began acting upon the earth pressure gauges; Time “b” indicates the time that the third row of tail brushes began passing away from the earth pressure gauges. Time “c” indicates the time that the third row of tail brushes finished passing through the earth pressure gauges. Time “d” indicates the time that the shield advancing through the current ring finished. From Fig. 11, it can be seen that, the values of the pressures on segmental lining began to response when contacting with the tail brushes. The pressures were less than 200 kPa when the first and second rows of tail brushes were passing the ring, and the pressures at the crown were less than 50 kPa, while the pressures at the spring line were about 150 kPa. However, the pressures fluctuated dramatically when the third row of tail brushes passed through, and the maximum pressure was almost 900 kPa at the left spring line of Section B. The pressures
The results in Fig. 11 indicated that when the first and second rows of tail brushes passed through, the measured pressures on the lining were lower than the theoretical earth pressures. Apart from the left spring line of Section B, all measured values were less than half of the theoretical values. Some previous studies revealed that the tail grease might increase pressure on the lining (e.g., Koyama 2003; Mashimo & Ishimura 2006). The pressure of the tail grease depend on both the discharge pressure of grease pump and the pressure loss in the injection pipe. In current study, these relative small measured pressures in first and second row brushes indicated that the influence from the tail grease was quite limited. A small discharge pressure of grease pump perhaps was the reason. When the third row of the tail brushes passing through, most of the measured pressures increased and became larger than the theoretical values. The maximum value was about 700 kPa at right spring line of Section A, and about 900 kPa at left spring line of section B, that both were almost 3–4 times larger than the design earth pressure. The large discrepancy between the pressures induced by the first two rows and those by the third row of tail brushes was obviously due to the different materials used in the tail brushes. As mentioned in session 3.1, the first and second rows were conventional wire brushes, while the third row was steel plate brush as shown in Fig. 12(a). Similar characteristic of large discrepancy was also found during the laboratory tests on the wire brush and plate brush. The maximum loads induced by the steel plate brush was 700–900 kPa, very close to the maximum value measured in laboratory tests. The outmost steel plate brush is much stiffer than the wire brush and is considered to have provided better sealing effect than the wire type. However, it can impose large pressure upon segments. In addition, after construction, the engineers found that the gaps between steel sheets were filled with solidified backfill grouting, as shown in Fig. 12(b), which may have significantly
Steel plate brush
Wire brushes
450mm
Wire brushes
Steel plate brush
450mm
(a) Before construction
(b) After construction
Fig.12. Photograph of tail brushes before and after construction. 7
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
90 60
150
800
30
180
210 240
0 100 200 300 400 500 600 700 800
270
0
330
Tail brush pressure (kPa)
120
600 400 200 0
300 Unit: kPa
0 90 180 270
0
Time "a"
6
Theoretical earth pressure Tail brush pressure of 9 minutes after Time "a"
12
18
24
30
Time (min)
Fig. 13. Maximum non-uniform pressure in section A induced by 3rd-row steel plate tail brush.
increased the stiffness of the steel plate brush, resulting in unexpected large pressures upon tunnel lining. Fig. 13 presented the distribution of tail brush indueced lining pressure of Section A during passing through the third row of tail brush. The pressure at the right spring was about 700 kPa and that at the crown was less than 100 kPa when 9 min after time “a”. Such a nonuniform pressure distribution is prone to produce excessive bending moment in the tunnel lining. In the case of a shallow tunnel such as the project of West Chengjiang Road tunnel, since the design value of the maximum moment is relative small, a small amount of excessive bending moment can easily lead to cracking in the RC (reinforced concrete) segments. This phenomenon will be analysed in the next session. It is noted that besides the tail brushes induced pressure, other sources of loads, for example, backfill grouting (Bezuijen et al., 2004), tail grease (Koyama 2003; Mashimo & Ishimura 2006), jacking force and tolerances during ring assembly as well as small misalignments (Cavalaro et al., 2011, 2012) also produce high stresses in tunnel segments. These may be higher than the effect of the passing tail sealing.
Fig. 15. Relationship between bending moment and segment joint rotation.
circumferential direction, are represented by rotational spring elements with a rotational stiffness coefficient kθ. The relationship between kθ and the segment joint rotation θ obtained from the lining tests of a similar tunnel is shown in Fig. 15. A fictitious continuous spring supported with a coefficient of subgrade reaction kr in the normal direction, which has no stiffness on the tension side, is used to model the effect of supporting a tunnel ring inside of the shield machine (Mashimo & Ishimura 2006). Since the subgrade reaction spring is fictitious, whose values cannot be determined directly, four different values are given to kr, i.e. 10 MN/m3, 20 MN/m3, 50 MN/m3 and 100 MN/m3, which varies between the values of soft ground and stiff ground. In the calculation, the steel plate tail brush induced forces and the dead weight of the segments were considered as loading upon the lining. Table 2 lists the values of the parameters used in the calculation. An assumed distribution of the tail brush pressure extended from the
5. Analysis of inner force of lining by beam-spring model The beam-spring model as shown in Fig. 14 is used to analyse the inner force of the lining due to the steel plate tail brush. In the model, the RC (reinforced concrete) segments are modelled as beam elements, and the segment joints, which connect the segments in the
Table 2 Parameters used in the beam-spring model.
Fig. 14. Beam-spring model for one ring of segments. 8
Parameter
Value
Tunnel radius (m) Thickness of segment (m) Width of segment (m) Elastic modulus of segment (kN/m2) Poisson’s ratio of segment Unit weight of segment (kN/m3)
5.43 0.5 1.5 3.5 × 107 0.25 25
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
Fig. 16. Tail pressure distribution in section A.
measured pressures in section A, was adopted in the calculation, as shown in Fig. 16. It is corresponding to the most non-uniform pressure distribution at the moment of 9 min after time “a” in Fig. 13. Fig. 17 shows the calculation results of bending moments and the axial forces with different values of kr. It can be seen that the maximum positive bending moment appears at the location of 0° (right spring line) and the maximum negative bending moment appeared at the location near 60° or 300° in all calculations. Based on the maximum bending moments (both positive and negative) and the corresponding axial forces with different values of kr, the crack widths of segments can be obtained, as shown in Fig. 18. It can be seen from the figure that all crack widths exceed the design limitation (wlim = 0.2 mm) of the crack for tunnel lining. In other words, the calculated crack widths far exceed the design value. In the current tunnel, water leakage from cracks in segments were observed here and there around the monitor sections right after passing through the tail voids, as shown in Fig. 19. These cracks were considered to be caused by large bending moments from non-uniform construction loads. Thus, this phenomenon of water leakage from cracks in turn proved that the non-uniform distribution of pressure induced by the steel plate brush might be one of the main factors that lead to cracking. The cracks will reduce the durability and
Fig. 18. Crack width of segments under different bending moments and corresponding axial forces.
service performance of RC segments (Wu et al., 2014). Therefore, the influence of tail brush induced loads (especial the steel plate type) on tunnel lining cannot be neglected, not only in the case of deep tunnel, but also in the case of shallow tunnel with small overburden. 6. Conclusions Laboratory tests, in-situ monitoring and theoretical calculation on the tail brush induced loads upon shield tunnel lining during construction were carried out in this study. The time history of lining pressure distribution and inner force were analysed. The following conclusions can be obtained. (1) Laboratory tests showed that the maximum load induced by the steel plate brush was twice that induced by the wire brush; however, insitu monitoring showed a much larger ratio. The reason was considered to be the solidification of backfill grouting inside of the steel plates, which significantly increased the stiffness of the steel plates brush. (2) The largest pressure acting upon the segments was observed
Fig. 17. Bending moments and axial forces induced by tail brush pressure. 9
Tunnelling and Underground Space Technology 97 (2020) 103283
G.-l. Ye, et al.
the production tolerances of concrete tunnel segments. J. Perform. Constr. Facil. 26 (6), 824–834. Chen, J.S., Mo, H.H., 2009. Numerical study on crack problems in segments of shield tunnel using finite element method. Tunn. Undergr. Space Technol. 24 (1), 91–102. Cheng, W.C., Ni, J.C., Shen, S.L., 2017. Experimental and analytical modeling of shield segment under cyclic loading. International Journal of Geomechnics ASCE (2017), 04016146(1–17). Cheung, L.L.K., Soga, K., Bennett, P.J., Kobayashi, Y., Amatya, B., Wright, P., 2010. Optical fibre strain measurement for tunnel lining monitoring. Proceedings of the Institution of Civil Engineers - Geotechnical Engineering 163(3). 119–130. David, C., Nicole, M., Alfred, S., 2010. Introduction to Tunnel Construction. CRC Press, ISBN-13: 978-0415468428. Gall, V.E., Marwan, A., Smarslik, M., Obel, M., Mark, P., Meschke, G., 2018. A holistic approach for the investigation of lining response to mechanized tunnelling induced construction loadings. Undergr. Space 3 (1), 45–60. Han, L., Ye, G.L., Chen, J.J., Xia, X.H., Wang, J.H., 2017. Pressures on the lining of a large shield tunnel with a small overburden: a case study. Tunn. Undergr. Space Technol. 64 (2017), 1–9. Hashimoto, T., Konda, T., Brinkman, J., Feddema, A., Kano, Y., 2006. Simultaneous backfill grouting, pressure development in construction phase and in the long-term. Tunnelling-A Decade of Progress, GeoDelft 1995-2005, Taylor & Francis 2006, pp. 101-107. Hashimoto, T., Nagaya, J., Konda, T., Tamura, T., 2002. Observation of lining pressure due to shield tunnelling. Geotech. Aspects Undergr Constr Soft Ground IS-Toulouse 2002, 119–124. Hashimoto, T., Yabe, K., Yamane, S., Ito, H., 1993. Development of pad type earth pressure cell for shield segment. Proceedings of 28th Annual Report of JGS, Tokyo: Japan Geotechnical Society (1993), 2055–2058. Japan Society of Civil Engineering (JSCE), 2006. Tunnel library17: Construction loads during shield tunnelling. Tokyo Japan: JSCE (in Japanese). Japan Society of Civil Engineering, 2010. Japanese standard for shield tunnelling. China Architecture & Building Press, Beijing (in Chinese). Koyama, Y., 2003. Present status and technology of shield tunnelling method in Japan. Tunn. Undergr. Space Technol. 18 (2), 145–159. Liu, X., Jiang, Z., Yuan, Y., Mang, H., 2018. Numerical investigation of the mechanical behavior of deformed segmental tunnel linings, strengthened by epoxy-bonded filament wound profiles. Tunn. Undergr. Space Technol. 78, 231–244. Mashimo, H., Ishimura, T., 2006. Numerical modelling of the behaviour of shield tunnel lining during assembly of a tunnel ring. Geotechnical Aspects of Underground Construction in Soft Ground-Proceedings of the 5th International Symposium TC28, 587–593. Oh, J.Y., Ziegler, M., 2014. Investigation on influence of tail void grouting on the surface settlements during shield tunnelling using a stress-pore pressure coupled analysis. KSCE J. Civ. Eng. 18 (3), 803–811. Ramoni, M., Lavdas, N., Anagnostou, G., 2011. Squeezing loading of segmental linings and the effect of backfilling. Tunn. Undergr. Space Technol. 26 (6), 692–717. Shen, S.L., Cui, Q.L., Ho, E.C., Xu, Y.S., 2016. Ground response to multiple parallel microtunneling operations in cemented silty clay and sand. J. Geotech. Geoenviron. Eng. 142 (5), 1–11. Shen, S.L., Wu, H.N., Cui, Y.J., Yin, Z.Y., 2014. Long-term settlement behavior of the metro tunnel in Shanghai. Tunn. Undergr. Space Technol. 40, 309–323. Sugimoto, M., 2006. Causes of shield segment damages during construction. In: International Symposium on Underground Excavation and Tunnelling, Bangkok, 2-4 Feb. Bangkok Thailand: EIT67–74. Sugimoto, M., Hirai, Y., Huynh, T.N., Tamai, T., 2014. Study on characteristics of grease flow resistance by element test. Geotechnical Aspects of Underground Construction in Soft Ground. CRC Press, pp. 167–171. Sramoon, A., Sugimoto, M., Kayukawa, K., 2002. Theoretical model of shield behavior during excavation. II: Application. J. Geotech. Geoenviron. Eng. 128 (2), 156–165. Wang, Z., Wang, L., Wang, J., Li, L., 2016. Case Study of Rehabilitation of a Damaged Underwater Tunnel in the Construction Phase. J. Perform. Constr. Facil. 30 (1), C4014003. Wu, H.N., Huang, R.Q., Sun, W.J., Shen, S.L., Xu, Y.S., Liu, Y.B., Du, S.J., 2014. Leaking behavior of shield tunnels under the Huangpu River of Shanghai with induced hazards. Nat. Hazards 70 (2), 1115–1132. Wu, H.N., Shen, S.L., Liao, S.M., Yin, Z.Y., 2015. Longitudinal structural modelling of shield tunnels considering shearing dislocation between segmental rings. Tunn. Undergr. Space Technol. 50 (2015), 317–323. Yang, C.Y., Xu, M.X., Chen, W.F., 2007. Reliability analysis of shotcrete lining during tunnel construction. J. Constr. Eng. Manage. 133 (12), 975–981. Yu, H.T., Yuan, Y., Yu, G.X., Liu, X., 2014. Evaluation of influence of vibrations generated by blasting construction on an existing tunnel in soft soils. Tunn. Undergr. Space Technol. 43, 59–66. Ye, G.L., Hashimoto, T., Shen, S.L., Zhu, H.H., Bai, T.H., 2015. Lessons learnt from unusual ground settlement during Double-O-Tube tunnelling in soft ground. Tunn. Undergr. Space Technol. 49 (2015), 79–91. Zhang, Y., Shi, Y., Zhao, Y., Yang, J., 2017. Damage in concrete lining of an operational tunnel. J. Perform. Constr. Facil. 31 (4), 06017002. Zhu, J.H., Li, Y., 2011. Quality Testing and Evaluation of the Damaged Lining of Zhoulichong Railway Tunnel. Tunnel Management, Emerging Technologies, and Innovation, Geotechnical Special Publication No. 221.
Fig. 19. Water leakage due to cracks of segment after the installation of tail brushes.
during passing through the steel plate tail brush. The measured value was more than four times of the theoretical earth pressure value, and the distribution was quite non-uniform. Analysis results using beamspring model revealed that the non-uniform distribution of pressures induced by the steel plate brush resulted in a crack width of the segments far exceeded the design limitation. This indicated that the tail brush induced load might be the main reason for the observed longitudinal cracks in the segments. (3) The change of the third row of tail brush from wire type to the steel plate type by the engineers can prevent the leakage more effectively. However, the steel plate type has induced unexpected large loads upon segments. Especially in the case of shallow tunnel that the axial forces are relatively small, the large bending moment caused by the non-uniform distribution of tail brush pressure may damage the segments. Thus, the consideration of tail brush pressure is extremely important for tunnels with a shallow overburden. Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant Nos. 41630633 and 51678369), Technical Innovation Foundation of Shenzhen (Grant No. JCYJ20170302143610976). References Arizumi, T., Okadome, K., Igarashi, H., Nagaya, J., 1999. Investigation of the load acting on shield segments during tunnelling. Proc. Tunnel Eng., JSCE 9, 271–276 (in Japanese). Bakker, K.J., Bezuijen, A., 2008. Ten years of bored tunnels in the Netherlands: Part II structural issues. In: Geotechnical Aspects of Underground Construction in Soft Ground: Proceedings of the 6th International Symposium (IS-Shanghai 2008) pp. 249. Bezuijen, A., Talmon, A.M., Kaalberg, F.J., Plugge, R., 2004. Field measurements of grout pressures during tunnelling of the Sophia rail tunnel. Soils Found. 44 (1), 39–48. Cavalaro, S.H.P., Blom, C.B.M., Walraven, J.C., Aguado, A., 2011. Structural analysis of contact deficiencies in segmented lining. Tunn. Undergr. Space Technol. 26 (6), 734–749. Cavalaro, S.H.P., Blom, C.B.M., Aguado, A., Walraven, J.C., 2012. New design method for
10