1 Study on bearing failure characteristics and

d
2 parameters of pipe roof support in super shallow

we
3 buried tunnel in soft surrounding rock
4 Lun Gonga#, Jiaqi Fenga, Ping Zhou b#*, Dishuang Suna, Feicong Zhoua#, Hui Hua, Lichuan Wanga
5 #These authors contributed to the work equllly and should be regarded as co-first authors
6 a School of Civil Engineering, Southwest Jiaotong University, Chengdu, Sichuan610031, China

vie
7 b Faculty of Geosciences and Environmental Engineering, Southwest Jiaotong University, Sichuan 611756,

8 China

9 Abstract: Through theoretical analysis, laboratory test and numerical simulation, this paper
studies the bearing failure characteristics and support parameters of pipe roof support in super

re
10

11 shallow buried double track railway tunnel crossing soft surrounding rock. Based on the
12 analysis of the load and load range of the pipe roof in the super shallow double track railway
13
er
tunnel, Pasternak double parameter model is selected as the mechanical analysis model of the
14 pipe roof. Based on this, 11 groups of laboratory tests of pipe roof grouting steel pipes are
15 proposed and designed, and the influence of pipe diameter, wall thickness, grouting fullness
pe
16 and reinforcement cage measures of pipe roof grouting steel pipes on the change law of bearing
17 capacity of pipe roof grouting steel pipes is analyzed. The bearing capacity of pipe roof
18 grouting steel pipes under different parameters is obtained, and the test results are compared
ot

19 and verified with ABAQUS numerical simulation. Based on the analysis and summary of the
20 bearing failure characteristics of pipe roof support and the test results, it is concluded that the
tn

21 grouting steel pipe with full grouting, wall thickness of 4mm and pipe diameter of 108mm
22 without reinforcement cage can meet the support capacity requirements of most super shallow
23 buried double track railway tunnels crossing soft surrounding rock. At the same time, in the
rin

24 actual engineering, the grouting fullness of the pipe roof should be focused on to prevent the
25 pipe roof steel pipe from separating from the internal grouting body, and the wall thickness of
ep

26 the steel pipe can be reduced and no reinforcement cage can be set up if unnecessary.
27 Keywords: Tunnel Engineering; Pipe roof support; Laboratory test; Grouting fullness;
—————————
28
*Corresponding author at: Faculty of Geosciences and Environmental
Support parameters
Pr

29 Engineering, Southwest Jiaotong University, Sichuan 611756, China

Email address：[email protected]

Tel/fax：86-028-66367910/86-028-66367910

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 30 1 Introduction

d
31 With the rapid development of urban rail transit in China, more and more large-span

we
32 double track railway tunnels have appeared, which are usually accompanied by complex
33 engineering conditions such as soft rock, super shallow burial, crossing shallow burial,
34 eccentric pressure, karst, and rich water, as well as crossing important buildings.

vie
35 For tunnels built in soft surrounding rock, Ping Zhou et al. [1-3] studied the disaster
36 mechanism and failure mechanism of tunnels built in sandy dolomite stratum, weak fractured
37 slate and high-content salt rock stratum; Xingzong Liu et al. [4] analyzed the collapse
mechanism of tunnel in a soft-hard interbedded surrounding rock; Hongliang Tu et al. [5]

re
38

39 conducted kinematic analysis on shallow buried large-span tunnel excavated in an up-soft/low-

40 hard rock stratum; Wenqiang Mu et al. [6] studied the long-term deformation and supporting
41
er
structure of soft rock tunnel based on numerical simulation and on-site monitoring. For tunnels
42 constructed at unfavorable locations such as super shallow buried tunnels and tunnels crossing
43 shallow buried tunnels, H.T. Wang et al. [7] proposed the roof collapse mechanism of shallow
pe
44 buried tunnels in two-layer rock stratum considering the influence of groundwater; Zheng
45 Wang et al. [8] analyzed the surface settlement induced by the construction of large-diameter
46 shallow buried twin-tunnel in soft ground; Feng Huang et al. [9] studied the instability
ot

47 mechanism of shallow tunnel in soft rock under overload.

48 In general, during tunnel excavation, 30%~40% of the total surface subsidence of the
tn

49 excavation section or 40%~50% of the total tunnel crown subsidence occurs before the initial
50 support plays its role. Therefore, the tunnel excavation requires the construction of advance
51 pre support structure to enhance the self-stability of surrounding rock. Oreste, P.P., et al. [10]
rin

52 proposed the design method of reinforcing shallow tunnel excavation face with fibreglass
53 dowels; Andreas Sjölander et al. [11] proposed a numerical model to simulate bolt-anchored
ep

54 and fiber-reinforced shotcrete to prevent failure to surrounding rock; R. Pan et al. [12] used
55 grouting bolt to reinforce the surrounding rock, and studied the parameters of grouting bolt
56 through experiment; Keqi Liu et al. [13] put forward a design method for pre-support of
Pr

57 subway tunnels under different water content conditions in silty clay.

58 Compared with other types of advanced pre support technologies, pipe roof support is

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 59 widely used in shallow buried, super shallow buried, portal section, and important buildings

d
60 under which surface settlement and surrounding rock deformation are strictly controlled in
61 tunnel and underground engineering construction [14-18] because of its mature technology,

we
62 large overall stiffness, and strong ability to limit surrounding rock deformation. Its structure
63 is shown in Fig. 1. At the same time, during the construction of the Gongbei tunnel of the
64 Hong Kong Zhuhai Macao Bridge, a new type of pre support technology combining the pipe

vie
65 roof method with the freezing method was innovatively used [19-22].
66 Because of the complexity and uncertainty of underground engineering, the empirical
67 design method is still the mainstream design method of underground engineering structures.

re
68 Xiao, H. et al. [23] analyzed the stability of shallow tunnel supported by pipe roof, and studied
69 the tunnel collapse case on this basis; Xiao, J.Z. et al. [24] analyzed the mechanical behavior
70 of pipe roof during the excavation of shallow tunnel in loose surrounding rock; Bo Lu et al.
er
71 [25] proposed a new pipe roof method to improve construction safety and utilization rate of
72 underground space.
pe
73 Up to now, domestic and international scholars have conducted a lot of research on the
74 pipe roof support system, and analyzed the mechanism, bearing capacity, and support effect
75 of the pipe roof support from theoretical analysis, numerical simulation, model test, on-site
ot

76 monitoring [14-27]. However, the quantitative analysis of the impact of various support
77 parameters of the pipe roof on the pipe roof bearing capacity by laboratory test is relatively
tn

78 small, and no system has been formed, As a result, the parameter design of pipe roof support
79 is still largely dependent on engineering experience analogy.
80 In view of the above problems, this paper quantitatively analyzes the influence of pipe
rin

81 roof grouting steel pipe diameter, wall thickness, grouting plumpness and reinforcement cage
82 measures on the change law of pipe roof bearing capacity through laboratory tests, analyzes
83 the bearing failure characteristics of pipe roof support and test results, and gives the
ep

84 recommended parameters of pipe roof grouting steel pipe for super shallow double track
85 railway tunnel. On the premise of ensuring the safety of the engineering, we hope to better
Pr

86 control the engineering cost, shorten the engineering duration, reduce the construction
87 difficulty, and promote the quantitative parameter design research of the pipe roof support
88 system.
3

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 89 2 Mechanical analysis of pipe roof support

d
90 2.1 Analysis of load and load range of pipe roof in super shallow buried

we
91 double track railway tunnel
92 （1）Calculation of overlying load
93 The calculation formula for surrounding rock pressure in China railway tunnel design

vie
94 code is an empirical formula based on the statistical analysis of a large number of actual
95 engineering cases, which can reflect the real situation of surrounding rock pressure to a certain
96 extent. According to the code for design of railway tunnel of China(TB 10003-2016), the depth

re
97 of tunnel can be divided by using the load equivalent height hq :

98 hq  0.45  2 s 1   （1）

99
er
Where: hq - load equivalent height ( m ); s - surrounding rock grade;  - width

100 influence coefficient, taken as   1  i ( B  5) ; B - tunnel width ( m ); i - increase/decrease

pe
101 coefficient of surrounding rock pressure: at that time B  5m , take i =0.2 ; at that time B  5m ,
102 take i =0.1 .

103 When the tunnel burial depth H  hq , it is a super shallow tunnel, and the calculation
ot

104 formula of the surrounding rock pressure on the tunnel is:

q H （2）
tn

105

106 Where:  - surrounding rock weight ( N / m 3 ); H - tunnel burial depth ( m ).

107 For the super shallow double track tunnel, on the one hand, because most of the overlying
rin

108 surrounding rock has poor physical and mechanical properties, it basically does not have the
109 ability to self stabilize; on the other hand, because of the large span of the double track tunnel,
ep

110 the stability of the overlying surrounding rock is further weakened. Therefore, for the pipe
111 roof support structure of super shallow double track tunnel, the load it bears can be seen as all
112 the self weight of all the surrounding rocks above it. Therefore, formula (2) is used to calculate
Pr

113 the overburden load on the pipe roof of super shallow double track tunnel.
114 （2）Longitudinal load range

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 115 As shown in Fig. 2, the pipe roof can be divided into four sections longitudinally, and the

d
116 OA section is the excavation support section. Because of the strong support effect of the initial
117 support, the A end can be regarded as the fixed end; Section AB is the excavated and

we
118 unsupported section, and the surrounding rock pressure is borne by the pipe roof in this section;
119 BD section is an unexcavated section, of which BC section is a relaxation section of
120 surrounding rock. The pipe roof in this section is subject to both the upper surrounding rock

vie
121 load and the elastic resistance of the lower foundation. CD section is an undisturbed section
122 of surrounding rock. The pipe roof in this section is only subject to the elastic resistance of the
123 foundation caused by the deformation of the pipe roof. According to the calculation theory of

re
124 tunnel surrounding rock pressure, it is assumed that the fracture surface of surrounding rock
125 in front of the excavation face starts from the tunnel excavation face, and the fracture angle is

45    / 2 , then the distance of section BC is d  h  tan(45    / 2) , and h is the tunnel

126
er
127 excavation height. In conclusion, the longitudinal load range of pipe roof is AD section.
pe
128 2.2 Pasternak double parameter mechanical model of pipe roof support

129 2.2.1 Pasternak elastic foundation beam model

130 At present, the elastic foundation beam theory is the most widely used in the research of
ot

131 pipe roof support mechanism, in which Winkler mechanical model and Pasternak double
132 parameter model are the most commonly used. Compared with Winkler model, Pasternak
tn

133 double parameter model is more widely used. It solves the problem that Winkler model does
134 not consider soil continuity. It can be used for calculation of more kinds of rock and soil mass
135 and better reflect the properties of rock and soil mass [28].
rin

136 Therefore, in order to make the simulation more consistent with the actual situation and
137 simplify the calculation, Pasternak double parameter elastic foundation beam model is adopted.
138 This model transfers the interaction between soil masses by setting a shear layer that can not
ep

139 be compressed but can only produce transverse deformation on the independent springs in
140 Winkler model, so that the simulated soil masses have continuity, as shown in Fig. 3.
Pr

141 In two-dimensional case, the relationship between load and displacement of the model
142 can be expressed by the following formula:

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 d 2 w( x)
143 P( x)  kw( x)  G p （3）

d
dx 2

Where: P( x) - foundation reaction ( kN / m 2 ); k - foundation reaction coefficient (

we
144

145 kN / m 3 ); w( x) - pipe roof deflection ( m ); G p - foundation shear stiffness ( kN / m ).

146 When G p  0 in equation (3), the model is Winkler model.

vie
147 When establishing the double parameter mechanical model of pipe roof, the following
148 basic assumptions shall be made:
149 （ 1） The pipe roof structure is regarded as an Euler Bernoulli beam acting on the

re
150 elastic foundation, that is, the influence of the horizontal shear force between the
151 pipe roof and the surrounding rock and the membrane effect are ignored;
152 （2） The friction between the pipe roof and the surrounding rock contact surface is not
er
153 considered;
154 （3） In the short section before and after the tunnel excavation face, the surrounding
pe
155 rock pressure borne by the pipe roof is regarded as a uniform load of equal size.
156 （4） During tunnel excavation, the front soil fracture surface starts from the excavation
157 surface, and the length of the stressed section of the pipe roof in front of the

158 excavation face is d  h  tan(45   / 2) , and h is the excavation height of the
ot

159 tunnel.
160 According to the above assumptions and the analysis of the longitudinal load range of the
tn

161 pipe roof, the following mechanical model of the pipe roof can be established:
162 When the tunnel excavation face is close to the front end of the pipe roof, the pipe roof
163 can be considered as a finite elastic foundation beam. The mechanical model of the pipe roof
rin

164 can be divided into the excavated and unsupported section AB and the unexcavated section
165 BC, of which the pipe roof in section AB is only subjected to the load of the overlying
ep

166 surrounding rock; The pipe roof in section BC is not only subjected to the load of the overlying
167 surrounding rock, but also subjected to the reaction of the lower foundation. Its mechanical
168 model is shown in Fig. 4.
Pr

169 2.2.2 Differential equation solution

170 In this mechanical model, the pipe roof can be regarded as a finite length elastic
6

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 171 foundation beam. It is necessary to establish differential equations of deflection curves for

d
172 section AB and section BC respectively for solution and calculation. The differential equations
173 for deflection control of pipe roof in each section are:

we
174 AB section:

d 4 ( x)
175 EI  bq ( x) （4）
dx 4

vie
176 BC section:

d 4 ( x) d 2 ( x)
177 EI  G p b  kb ( x)  bq ( x) （5）
dx 4 dx 2

re
178 Where: b - width of Pasternak elastic foundation beam ( m ); E - elastic modulus of pipe

179 roof ( Pa ); I - inertia moment of pipe roof ( m 4 ); q ( x) - load on pipe roof ( kN / m 2 ); w( x) -

180
er
pipe roof deflection ( m ); k - foundation reaction coefficient ( kN / m 3 ); G p - foundation

181 shear stiffness ( kN / m ).

pe
182 The general solution of differential equation in each section is:
183 AB section:

q( x)b 4
184  1 ( x)  x  C1 x 3  C 2 x 2  C 3 x  C 4 （6）
24EI
ot

185 BC section:

q( x)
tn

186  2 ( x)  e ax (C 5 cos  x  C 6 sin  x)  e  ax (C7 cos  x  C 8 sin  x)  （7）

k

187 Where: C 1 、 C 2 、 C 3 、 C 4 、 C 5 、 C 6 、 C7 、 C 8 are undetermined coefficients;

rin

kb
  1   G p  2  / k  ;   1   G p  2 / k   ;  4 
1/ 2 1/ 2
188 .
4EI

189 The boundary conditions of the model are:  1  0 ， 1  0 ， 1  2 ，

ep

x 0 x 0 xs xs

190  1(') xs

  2(')
xs
， 1
('')
xs
  2('')
xs
， 1
(''')
xs
  2(''')
xs
， 2
('')
xsd
 0 ，  2(''')
xsd
0.

191 According to the above, the deflection equation of each section of the pipe roof can be obtained
Pr

192 by substituting the boundary conditions. According to the Euler Bernoulli beam theory, the
193 rotation angle, bending moment and shear force of the pipe roof can be calculated by the

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 194 following methods, respectively:

d
195 Rotation angle:

d ( x)

we
196  ( x)  （8）
dx

197 Bending moment:

d 2 ( x)

vie
198 M ( x)   EI （9）
dx 2

199 Shear force:

dw 3 ( x) dw( x)
200 Q( x)   EI  G pb （10）

re
3
dx d ( x)

201 3 Test on bearing and failure characteristics of pipe roof

grouting steel pipe
202
er
203 The pipe roof support system is composed of grouting steel pipe and surrounding rock
pe
204 grouting reinforcement area. The stiffness of the grouting steel pipe is far greater than that of
205 the surrounding rock grouting reinforcement area. It is the most important bearing part in the
206 pipe roof support system, and its parameters have a greater impact on the bearing capacity of
207 the pipe roof. Therefore, referring to the mechanical model in Pasternak double parameter
ot

208 model of pipe roof in Section 2 and the actual working state of pipe roof, this section designs
209 laboratory tests to conduct quantitative research on the four support parameters of single pipe
tn

210 roof grouting steel pipe, such as pipe diameter, wall thickness, grouting plumpness, and
211 reinforcement cage measure, and analyze the change rule of its bearing capacity.
rin

212 3.1 Test conditions

213 A total of 4 groups of laboratory tests and 11 working conditions are designed for this
214 test. The pipe diameter, wall thickness, grouting plumpness and reinforcement cage of the pipe
ep

215 roof grouting steel pipe are studied respectively. See Table 1 for the detailed working
216 conditions design.
217 Q235 hot-rolled seamless steel pipe is selected for the test. Referring to the mechanical
Pr

218 model in the Section 2 of pipe roof and the stress state of the pipe roof structure in the actual
219 engineering, the length of the test steel pipe is selected as 3.6m, of which 1.2m is used to
8

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 220 simulate the excavation of supported section, 1.2m is used to simulate the excavation of

d
221 unsupported section, and 1.2m is used to simulate the unexcavated section; The grouting
222 material in the pipe is pure cement slurry with a water cement ratio of 1:0.4, and the cement

we
223 is 425 ordinary portland cement; The reinforcement cage adopts 4 HRB335 reinforcements
224 with a diameter of 16mm. The specimen form is shown in Fig. 5.

3.2 Specimen production

vie
225

226 Different types of specimens were made according to Table 1. All specimens were made
227 with the same materials, in the same way and under the same curing conditions. The
manufacturing process of different types of specimens is as follows:

re
228

229 For the specimen with 100% grouting fullness, in order to make the cement slurry fullness
230 in the steel pipe reach 100%, vertical grouting is adopted. For the specimen with insufficient
231
er
grouting, it shall be placed horizontally for grouting. Before grouting, thin steel sheets with
232 different heights shall be welded at both ends of the steel pipe, and the weld shall be sealed to
233 make the grouting fullness reach 50% and 75% respectively. The steel pipe grouting process
pe
234 is shown in Fig. 6.
235 For the specimen with reinforcement cage, the reinforcement cage is welded with 4 pieces
236 of 3.6m long, 16mm diameter HRB335 steel bars and a steel pipe of 42mm diameter and 5cm
ot

237 length every 50cm, as shown in Fig. 7. After the reinforcement cage is welded, vertically insert
238 the fixed steel pipe and conduct subsequent grouting. All specimens were tested 28 days after
tn

239 grouting and curing.

240 3.3 Material performance

rin

241 3.3.1 Mechanical properties of grouting body

242 Pour 3 cement slurry cubes ( 150mm  150mm  150mm ) and 6 cement slurry cubes (
243 150mm  150mm  300mm ) with the same pure cement slurry as the grouted steel pipe
ep

244 specimens, a total of 9 specimens and curing them for 28 days. Three cubes are used to
245 measure the compressive strength of the grouting body, as shown in Fig. 8; Three cubes are
Pr

246 used to measure the axial compressive strength of the grouting body, as shown in Fig. 9; The
247 elastic modulus of the grouting body is measured by sticking strain gauges on 4 longitudinal
248 faces of 3 cube specimens, as shown in Fig. 10. The mechanical property test of grouting body
9

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 249 were tested with reference to the compressive strength test and static compression elastic

d
250 modulus test methods in the standard for test methods of physical and mechanical properties
251 of concrete (GB/T 50081-2019) and were conducted on the microcomputer controlled electro-

we
252 hydraulic servo pressure testing machine in the laboratory.
253 Through the test, the average compressive strength of grouting body material is 42.7MPa,
254 the average axial compressive strength is 32.7MPa, and the average elastic modulus is

vie
255 15.1Gpa.

256 3.3.2 Mechanical properties of steel

Q235 steel is used for all steel pipes in this test. According to the relevant provisions in

re
257

258 metallic materials tensile testing part 1: room temperature test methods (GB/T 228.1-2010),
259 when the pipe diameter is less than 200 mm, three tensile specimens are cut in situ with a pipe
260
er
diameter of 108 mm and a wall thickness of 4 mm, and they are named Group A; When the
261 pipe diameter is greater than 200 mm, three tensile specimens are cut in situ by using a steel
262 pipe with a pipe diameter of 219 mm and a wall thickness of 6 mm, and they are named Group
pe
263 B. The tensile test were carried out according to the relevant provisions in the specification.
264 The process and results of steel tensile test are shown in Fig. 11.
265 Through the test, the average yield strength of group A specimens is 367.1MPa, the
ot

266 average tensile strength is 500.6MPa, and the average elastic modulus is 203.7MPa; The
267 average yield strength of group B specimens is 311.8 MPa, the average tensile strength is
tn

268 474.8 MPa, and the average elastic modulus is 201.3 MPa.

269 3.4 Test layout

rin

270 3.4.1 Test device

271 According to the boundary conditions of the pipe roof mechanical model analyzed in
272 Section 2, the stress state of the pipe roof grouting steel pipe is simulated, as shown in Fig. 12.
ep

273 The test device shown in Fig. 13 is adopted for this test. The 1.2m pipe roof above the
274 steel arch and concrete bearing is used to simulate the excavation of supported section, the
Pr

275 1.2m in the middle is used to simulate the excavated of unsupported section, and the 1.2m
276 above the rubber support is used to simulate the unexcavated section.
277 The device consists of a bearing system and a load dispersion system. The bearing system
10

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 278 consists of the rubber bearing simulating the soil, the steel arch and the concrete bearing

d
279 simulating the initial support; The load distribution system consists of 4 large fasteners at both
280 ends of the steel pipe and 2 small fasteners at the mid span flying section. The jack is placed

we
281 on the horizontal steel plate at the top of the fastener for loading.
282 The steel arch in this test is spliced by two 30cm long and 6cm wide channel steels. In
283 order to prevent the instability of the steel frame, several vertical large-diameter

vie
284 reinforcements are welded on the section of the channel steel to strengthen it before splicing;
285 Considering the conversion of reinforcement strength, the concrete bearing in this test is made
286 of C35 plain concrete, with a longitudinal length of 0.488m. After pouring, it will be cured for

re
287 28 days for testing. Considering different steel pipe diameters, four types of concrete supports
288 are made, corresponding to the grouting steel pipes with pipe diameters of 108mm, 159mm,
289 219mm and 299mm respectively, as shown in Fig. 14.
er
290 For super shallow tunnel, the surrounding rock is mostly Grade V. In order to better
291 simulate the stress of pipe roof steel pipe in Grade V surrounding rock, the test uses ordinary
pe
292 rectangular plate rubber bearing to simulate the surrounding rock. Plate rubber bearing is a
293 kind of bearing formed by inlaying, bonding and vulcanizing multi-layer thin steel plates and
294 multi-layer natural rubber. The bearing has sufficient vertical stiffness to bear vertical load,
ot

295 good elasticity to adapt to the rotation of beam ends, and large shear deformation to meet the
296 horizontal displacement of upper structure. Two plate rubber bearings (
tn

297 600mm  500mm  125mm ) are used in this test, as shown in Fig. 15. The compressive elastic
298 modulus of the rubber bearing is about 1.0MPa.
299 In this test, arc fasteners are made to disperse the pressure of the jack and make it evenly
rin

300 distributed on the surface of the grouting steel pipe as much as possible. The large fastener is
301 obtained by halving the seamless steel pipe, its inner diameter is the same as the outer diameter
302 of the grouting steel pipe, and the longitudinal length of each large fastener is 0.6m. According
ep

303 to the pipe diameter of the grouting steel pipe, 4 kinds of large fasteners with different pipe
304 diameters are made. Small fasteners are cut from 3cm thick steel plates, and their inner
Pr

305 diameters are consistent with the outer diameters of steel pipes. They are also made into 4
306 types. The fabricated fasteners are shown in Fig. 16.
307 In order to better bear the pressure of the jack, 6 pieces of steel plates (
11

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 308 300mm  300mm  3mm ) are welded or placed on the top of the fastener as the contact

d
309 surface of the jack. The assembled test device is shown in Fig. 17.

3.4.2 Loading Scheme

we
310

311 In this test, six 200t electric hydraulic jacks are used for loading. The two jacks in the
312 mid span flying section are located at the trisection point. To keep the load of each jack

vie
313 consistent, the six jacks are connected to an oil pump, and the specimen is loaded at the same
314 time. The jack applies load to the reaction beam, distributes the load to the grouting steel pipe
315 through large and small fasteners, and forms a pure bending section at the 1.2m section of the
steel pipe span. In this test, each specimen is loaded in stages. Before yielding, the specimen

re
316

317 is loaded in stages according to 1/10 of the estimated ultimate load; After that, carry out step
318 loading according to 1/15 of the estimated limit load, and stop loading for about 2min for each
319
er
step until the deformation of the specimen is stable. In order to avoid structural instability due
320 to excessive deformation, causing danger or damage to the instrument, stop loading when the
321 mid span deflection reaches about 4cm~5cm.
pe
322 3.5 Analysis of test results

323 3.5.1 Failure mode of specimen

ot

324 Fig. 18 shows the final failure patterns of the grouting steel pipe specimens under
325 different working conditions in different groups.
tn

326 It can be seen from the figure that the failure process and failure form of each specimen
327 are basically the same. The support sections at both ends are basically straight, with little
328 deformation. The mid span section is bent, showing a smooth concave curve.
rin

329 For the specimen with insufficient grouting, during the loading process, the cross section
330 at the supports at both ends of the specimen will have large transverse deformation, which will
331 flatten the specimen at the supports at both ends, and the shape of the cross section is elliptical,
ep

332 as shown in Fig. 19. After the test, the grouting body in the specimen has been separated from
333 the steel pipe.
Pr

334 3.5.2 Analysis of load deflection relationship curve

335 According to the mid span section deflection data of each specimen obtained from the

12

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 336 test, draw and analyze the load deflection diagram lines of each specimen of pipe diameter

d
337 group, wall thickness group, grouting plumpness group and reinforcement cage group, as
338 shown in Fig. 20.

we
339 According to the load mid span deflection curves of each specimen in different groups,
340 all curves can be roughly divided into three stages, namely elastic stage, elastic-plastic stage
341 and plastic stage. In the elastic stage, the load and the mid span deflection of the specimen are

vie
342 linearly increasing; In the elastic-plastic stage, with the increase of load, the deformation rate
343 of the specimen increases, and the slope of the load mid span deflection curve decreases
344 gradually; When the specimen enters the plastic stage, with the increase of load, the

re
345 deformation rate of the specimen is further accelerated compared with the elastic-plastic stage,
346 and the load displacement curve is still basically linear at this stage.
347 It can be seen from Fig. 20 that the pipe diameter of the specimen has a great influence
er
348 on the bearing capacity. With the increase of the pipe diameter, the slope of the elastic phase
349 of the load midspan deflection curve of the specimen also increases, and the ultimate bearing
pe
350 capacity of the specimen increases rapidly; With the increase of the wall thickness of the
351 specimen, the slope of the elastic stage of the load mid span deflection curve and the ultimate
352 bearing capacity of the specimen also increase; With the decrease of grouting plumpness, the
ot

353 slope of the elastic stage of the load mid span deflection curve of the specimen decreases
354 significantly, and the ultimate bearing capacity decreases significantly; After the
tn

355 reinforcement cage is added, the slope of the elastic stage of the load deflection curve and the
356 ultimate bearing capacity of the specimen increase to a certain extent.

357 3.5.3 Bearing capacity analysis

rin

358 （1）Fitting of bearing capacity of each specimen

359 According to the load mid span deflection curve of each group of test specimens, since
ep

360 there is no obvious yield point, the elastic stage and plastic stage of the load deflection curve
361 of each specimen are linearly fitted using the linear fitting function in Origin, and the
362 intersection point load of two fitting straight lines is taken as the ultimate bearing capacity of
Pr

363 the specimen, as shown in Fig. 21.

364 After fitting calculation, the ultimate bearing capacity curves of different groups of

13

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 365 specimens are drawn, as shown in Fig. 22.

d
366 （2）Calculation of specimen safety factor
367 Different from the general bending statically determinate beam test with free ends, the

we
368 load device used in this test is closer to the state simulated by the statically indeterminate beam
369 with restrained ends. Therefore, there will be negative bending moment at both ends of the
370 support of the specimen, and the mid span bending moment cannot directly reflect the bearing

vie
371 capacity of the pipe roof grouting steel pipe in the surrounding rock.
372 Therefore, according to the load imposed by the jack, this paper converts it into a uniform
373 load on the steel pipe. This load value as the ultimate bearing capacity of the specimen can be

re
374 more intuitively compared with the surrounding rock pressure obtained according to the
375 surrounding rock pressure calculation theory, and more clearly reflect the bearing capacity of
376 the pipe roof grouting steel pipe in the surrounding rock.
er
377 After investigating a large number of pipe roof projects in China, it is found that the
378 circumferential spacing of pipe roof is mostly 0.3~0.5m, of which 0.4m is the most widely
pe
379 used. Therefore, the circumferential spacing of pipe roof in this section is taken as 0.4m. Since
380 the length of the pipe roof under the load applied by each jack is 0.6m, the formula for
381 converting the load applied in this test into the uniform load on the pipe roof is:
ot

Test ultimate load

382 Uniform load  （11）
0.4m  0.6m
tn

383 According to the calculation formula of the load on the pipe roof of the super shallow
384 buried double track railway tunnel in Section 2, and referring to the standard section of the
385 350km/h double track high-speed railway tunnel in China, the tunnel excavation span is 14m,
rin

386 the surrounding rock is taken as Grade V surrounding rock, and its volume weight is

387 20kN / m 3 . It is calculated that the pipe roof bears the surrounding rock load when the critical
ep

388 burial depth of the super shallow buried double track railway tunnel is:

389 q   hq  20  0.45  2 4  [1  0.1  (14  5)]  274kPa （12）

390 According to formula (11), calculate the uniform load of grouting steel pipe for pipe roof
Pr

391 under different conditions in the test, and use the following formula to calculate the bearing
392 capacity safety factor n of grouting steel pipe with different conditions.
14

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 Uniform load
393 n （13）

d
Bearing surrounding rock load

394 After calculation, the safety factors of each condition are listed in Table 2, as follows:

we
395 According to Fig.22 and Table 2, the following conclusions can be drawn:
396 （1）The diameter of the pipe roof grouting steel pipe has the greatest influence on the

vie
397 bearing capacity of the pipe roof. Compared with the pipe roof with a pipe diameter of 108
398 mm, the bearing capacity increases by 183.58% when the pipe diameter is 159 mm, 398.62%
399 when the pipe diameter is 219 mm, and 874.78% when the pipe diameter is 299 mm.
Increasing the pipe diameter can significantly improve the bearing capacity of the pipe roof.

re
400

401 （2）The wall thickness of the pipe roof grouting steel pipe has little influence on the
402 bearing capacity of the pipe roof. The bearing capacity increases by 20.76% when the wall
403
er
thickness increases from 4.5mm to 6mm, 16.25% when the wall thickness increases from 6mm
404 to 8mm, and 11.94% when the wall thickness increases from 8mm to 10mm. With the
405 increasing of the wall thickness, the bearing capacity of the grouting steel pipe of the pipe roof
pe
406 increases slowly.
407 （3）The insufficient grouting defect of the pipe roof will greatly weaken the bearing
408 capacity. When the grouting fullness decreases from 100% to 75%, the bearing capacity
ot

409 decreases by 29.8%; when the grouting fullness of the pipe roof decreases to 50%, the bearing
410 capacity decreases by 34.5%; when the grouting fullness of the pipe roof decreases to 50%,
tn

411 the bearing capacity of the pipe roof is similar to that of the empty pipe; the relationship
412 between the grouting fullness and the bearing capacity of the steel pipe is non-linear.
413 （ 4） The reinforcement cage added to the pipe roof grouting steel pipe has little
rin

414 influence on the bearing capacity of the pipe roof. Compared with the specimen without
415 reinforcement cage, the bearing capacity of the specimen with reinforcement cage can be
ep

416 increased by 15.6%.

417 （5）The pipe roof made of steel pipes with full grouting, wall thickness of 4mm and
418 diameter of 108mm without reinforcement cage can meet the bearing capacity requirements
Pr

419 of most super shallow buried double track railway tunnels.

15

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 420 4 Numerical simulation of bearing and failure characteristics

d
421 of pipe roof grouting steel pipe

we
422 Based on the laboratory test of pipe roof grouting steel pipe, this section uses the large-
423 scale general finite element analysis software ABAQUS to conduct finite element numerical
424 simulation for 11 working conditions of the laboratory test of pipe roof grouting steel pipe,

vie
425 and compares the failure form, load deflection curve and bearing capacity of each specimen
426 obtained from the numerical simulation with the results of the laboratory test to further explore
427 the bearing and failure characteristics of pipe roof grouting steel pipe.

re
428 4.1 Numerical model and setting

429 4.1.1 Numerical model

430
er
The numerical model shall be consistent with the laboratory test as far as possible. In
431 order to increase the convergence of the model, the large fasteners used at both ends of the test
432 are simplified, and the equivalent uniform load is applied instead. The small fasteners are
pe
433 simulated by rigid bodies; For the end plates at both ends of the specimen, its effect is
434 simplified to constrain the longitudinal displacement of the grouting body along the steel pipe
435 to reduce the contact analysis. Except that D3T2 truss element is used for reinforcement cage,
ot

436 C3D8R solid element is used for simulation analysis of all structures. The numerical analysis
437 model is shown in Fig. 23.
tn

438 For the specimen with insufficient grouting, the non-grouting part is divided into grids,
439 but the material calculation parameters are reduced to close to 0, so as to achieve the effect
440 close to the actual test; The reinforcement cage is built by D3T2 truss element and embedded
rin

441 into the grouting body during assembly. The steel pipe and reinforcement cage with
442 insufficient grouting are shown in Fig. 24.
ep

443 4.1.2 Material constitutive model and parameters

444 For steel, the double broken line model is selected for this calculation, as shown in Fig.
445 25. The model is divided into linear elastic stage before yielding and steel strengthening stage
Pr

446 after yielding. The slope of the linear elastic stage is the elastic modulus of the steel itself, and
447 the slope of the yield stage can be roughly taken as 1/100 of the linear elastic stage.
16

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 448 As a homogeneous rock mass like material, pure cement grout has cohesion and friction

d
449 angle. Therefore, the grouting body of this model adopts the mohr coulomb constitutive model.
450 With reference to the compression test results of pure cement specimens mentioned above, its

we
451 parameters are taken as the parameters of class III surrounding rock in the code for design of
452 railway tunnel (TB 10003-2016). In this calculation, the rubber bearing adopts the elastic
453 constitutive model, and the concrete adopts the elastic model. The basic calculation parameters

vie
454 used for each component in the numerical calculation are shown in Table 3.

455 4.1.3 Contact type

In this calculation, surface to surface contact is used between grouting body and steel

re
456

457 pipe, steel pipe and rubber bearing, steel pipe and steel block, steel pipe and concrete bearing.
458 Tie contact is used between small fasteners and steel pipe surface, embedded region is used
459
er
for contact between reinforcement cage and grouting body inside steel pipe, and the
460 established reinforcement cage model can be placed into the grouting body.
pe
461 4.1.4 Boundary condition and loading setting
462 According to the actual situation of the laboratory test, this model limits the displacement
463 and rotation angle of rubber bearing, steel arch and concrete bearing bottom in three directions,

namely U x  U y  U z  UR x  UR y  UR z  0 . In the test, two large fasteners are used to

ot

464

465 disperse the jack pressure at 1.2m at both ends of the grouting steel pipe, which is equivalent
tn

466 to the uniform load on the steel pipe surface in the calculation; Two small fasteners are used
467 to disperse the pressure of the jack at 1.2m of the mid span, and concentrated force is applied
468 on the small fasteners during calculation. The model boundary conditions and loading
rin

469 conditions are shown in Fig. 26.

470 4.2 Comparison between numerical simulation and laboratory test

ep

471 4.2.1 Comparison of failure form

472 Four specimens calculated by finite element method (condition 1, condition 6, condition
473 9 and condition 11) are selected to compare their failure forms with the actual failure forms of
Pr

474 the laboratory test, as shown in Fig. 27. It can be seen from the figure that the failure form is
475 the same as that of the laboratory test specimens. When the grouting is full, the deformation
17

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 476 of the steel pipe is smooth at the rubber bearing end, and obvious dents appear on the surface

d
477 of the steel pipe at the steel frame concrete bearing end with large stiffness; When the grouting
478 is not full, the section where the specimen contacts the support will have large transverse

we
479 deformation, making the circular steel pipe section elliptical.
480 According to the comparison of the failure form between the numerical calculation
481 specimen and the laboratory test specimen, it is found that although there is some deviation

vie
482 between the deformation of the test specimen and the finite element simulation specimen, the
483 deformation characteristics and deformation trends are relatively consistent.

484 4.2.2 Comparison of load deflection relationship curve

re
485 The mid span deflection of each specimen obtained by numerical calculation is extracted,
486 and the load mid span deflection relationship curves of four test pieces (condition 1, condition
487
er
6, condition 9, condition 11) are also drawn, and compared with the load mid span deflection
488 curves obtained from laboratory tests, as shown in Fig. 28.
pe
489 4.2.3 Comparison of bearing capacity
490 The ultimate load of each specimen in the numerical calculation is obtained by fitting in
491 the same way, and the ultimate load of each specimen obtained from the laboratory test and
492 the ultimate load of each specimen obtained from the numerical calculation are listed in Table
ot

493 4. The error value calculation formula shown in the table is:
Numerical simulation load
Error = 1 （14）
tn

494
Test load
495 According to the comparison of load midspan deflection curves between laboratory test
496 and numerical simulation as shown in Fig. 28 and the comparison of ultimate bearing capacity
rin

497 of each specimen between laboratory test and numerical calculation as shown in Table 4, it
498 can be found that:
499 （1）The trend of the load deflection curve obtained from the finite element simulation
ep

500 is basically consistent with that from the laboratory test. It is feasible to use ABAQUS to
501 simulate and calculate the stress process of the grouting steel pipe specimen of the pipe roof,
Pr

502 and it also verifies the correctness of the test.

503 （2）The slope of the load midspan deflection curve of each specimen obtained by finite
504 element simulation at the elastic stage is slightly larger than that obtained by laboratory test,
18

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 505 mainly because the setting of material parameters and boundary conditions in finite element

d
506 simulation tends to be ideal, while there is a certain deviation in the production, installation,
507 loading, etc. of the specimen in the test.

we
508 （3）The influence of the parameters of the pipe roof grouting steel pipe on its bearing
509 capacity is basically the same. The pipe diameter and grouting fullness have a greater impact,
510 while the wall thickness and reinforcement cage have a relatively small impact.

vie
511 （ 4） The bearing capacity of pipe roof grouting steel pipe obtained by numerical
512 calculation is generally larger than that of laboratory test, but the difference is not significant,
513 the maximum is 12.4%, and most of them are within 10%, which verifies the correctness of

re
514 various settings of numerical simulation calculation.

515 5 Conclusion
516
er
Based on the engineering background of super shallow buried double track railway tunnel
517 in soft surrounding rock, combined with theoretical analysis, laboratory test and numerical
518 simulation, this paper studies the bearing and failure characteristics of pipe roof support and
pe
519 the influence of support parameters on the bearing capacity of pipe roof. The main conclusions
520 are as follows:
521 （1）Through the analysis of the load and load range of pipe roof support in the super
ot

522 shallow buried double track railway tunnel, Pasternak double parameter model is selected to
523 build a more practical mechanical model of the pipe roof support, and based on this, 11 groups
tn

524 of pipe roof grouting steel pipe laboratory loading tests are proposed and designed.
525 （2）In the test of the bearing and failure characteristics of the pipe roof grouting steel
526 pipe, the influence of the pipe diameter, wall thickness, grouting fullness and reinforcement
rin

527 cage measures of the pipe roof grouting steel pipe on the change law of the bearing capacity
528 of the pipe roof grouting steel pipe is analyzed, and the order of the influence degree is: pipe
ep

529 diameter > grouting fullness > wall thickness > steel cage. At the same time, consistent results
530 are obtained in the numerical simulation.
531 （3）The pipe diameter has the greatest influence on the bearing capacity of the pipe
Pr

532 roof grouting steel pipe. Compared with the pipe roof with a pipe diameter of 108 mm, the
533 bearing capacity increases by 183.58% when the pipe diameter is 159 mm, 398.62% when the

19

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 534 pipe diameter is 219 mm, and 874.78% when the pipe diameter is 299 mm. Increasing the pipe

d
535 diameter can significantly improve the bearing capacity of the pipe roof.
536 （4）The insufficient grouting defect of the pipe roof will greatly weaken the bearing

we
537 capacity of the pipe roof. When the grouting fullness decreases from 100% to 75%, the bearing
538 capacity decreases by 29.8%; when the grouting fullness decreases to 50%, the bearing
539 capacity decreases by 34.5%; when the grouting fullness decreases to 50%, the bearing

vie
540 capacity of the pipe roof is similar to the empty pipe; the relationship between the grouting
541 fullness and the bearing capacity of the steel pipe is non-linear.
542 （5）The wall thickness of grouting steel pipe and the reinforcement cage have little

re
543 influence on the bearing capacity of the pipe roof. When the wall thickness increases from
544 4.5mm to 6mm, the bearing capacity increases by 20.76%, and when the reinforcement cage
545 is added, the bearing capacity increases by 15.6%. The pipe roof made of steel pipes with full
er
546 grouting, wall thickness of 4mm and diameter of 108mm without reinforcement cage can meet
547 the bearing capacity requirements of most super shallow buried double track railway tunnels.
pe
548 6 Declaration of Competing Interest
549 The authors declare that they have no known competing financial interests or personal relationships

550 that could have appeared to influence the work reported in this paper.
ot

551 7 Acknowledgments
552 The work was supported by the High Speed Railway and Natural Science United Foundation of China
tn

553 [Grant number U1934213] and National Natural Science Foundation of China [Grant number 51678498].

554 8 Reference
rin

555 [1] Zhou, P., Jiang, Y.F., Zhou, F.C., Wu, F., Qi, Y.L., Wang, Z.J., 2022. Disaster mechanism
556 of tunnel face with large section in sandy dolomite stratum. Eng. Fail. Anal. 131, 105905.
557 [2] Zhou, P., Jiang, Y.F., Zhou, F.C., Gong, L., Qiu, W.G., Yu, J.W., 2022. Stability
ep

558 Evaluation Method and Support Structure Optimization of Weak and Fractured Slate Tunnel.
559 Rock Mech. Rock Eng. 55, 6425-6444.
Pr

560 [3] Zhou, P., Li, J.Y., Jiang, Y.F., Zhou, F.C., Lin, M., Lin, J.Y., Wang, Z.J., 2021. Damage
561 mechanism of tunnels in the high-content salt rock stratum. Bull. Eng. Geol. Environ. 80,
562 7633–7652.
20

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 563 [4] Liu, X.Z., Liu, F., Song, K.Z., 2022. Mechanism analysis of tunnel collapse in a soft-hard

d
564 interbedded surrounding rock mass: A case study of the Yangshan Tunnel in China. Eng. Fail.
565 Anal. 138, 106304.

we
566 [5] Tu, H.L., Zhou, H., Qiao, C.S., Gao, Y., 2020. Excavation and kinematic analysis of a
567 shallow large-span tunnel in an up-soft/low-hard rock stratum. Tunn. Undergr. Sp. Tech. 97,
568 103245.

vie
569 [6] Mu, W.Q., Li, L.C., Chen, D.Z., Wang, S.X., Xiao, F.K., 2020. Long-term deformation
570 and control structure of rheological tunnels based on numerical simulation and on-site
571 monitoring. Eng. Fail. Anal. 118, 104928.

re
572 [7] Wang, H.T., Wang, L.G., Li, S.C., Wang, Q., Liu, P., Li, X.J., 2019. Roof collapse
573 mechanisms for a shallow tunnel in two-layer rock strata incorporating the influence of
574 groundwater. Eng. Fail. Anal. 98, 215-227.
er
575 [8] Wang, Z., Yao, W.J., Cai, Y.Q., Xu, B., Fu, Y., Wei, G., 2019. Analysis of ground surface
576 settlement induced by the construction of a large-diameter shallow-buried twin-tunnel in soft
pe
577 ground. Tunn. Undergr. Sp. Tech. 83, 520-532.
578 [9] Huang, F., Wu, C.Z., Jang, B.A., Hong, Y., Guo, N., Guo, W., 2020. Instability mechanism
579 of shallow tunnel in soft rock subjected to surcharge loads. Tunn. Undergr. Sp. Tech. 99,
ot

580 103350.
581 [10] Oreste, P.P., Dias, D., 2012. Stabilisation of the Excavation Face in Shallow Tunnels
tn

582 Using Fibreglass Dowels. Rock Mech. Rock Eng. 45, 499–517.
583 [11] Sjolander, A., Hellgren, R., Malm, R., Ansell, A., 2020. Verification of failure
584 mechanisms and design philosophy for a bolt-anchored and fibre-reinforced shotcrete lining.
rin

585 Eng. Fail. Anal. 116, 104741.

586 [12] Pan, R., Wang, Q., Jiang, B., Li, S.C., Sun, H.B., Qin, Q., Yu, H.C., Lu, W., 2017. Failure
587 of bolt support and experimental study on the parameters of bolt-grouting for supporting the
ep

588 roadways in deep coal seam. Eng. Fail. Anal. 80, 218-233.
589 [13] Liu, K.Q., Li, S.C., Ding, W.T., Hou, M.L., Gong, Y.J., Li, H.L., 2020. Pre-supporting
Pr

590 mechanism and supporting scheme design for advanced small pipes in the silty clay layer.
591 Tunn. Undergr. Sp. Tech. 98, 103259.
592 [14] Ma, P., Shimada, H., Sasaoka, T., Hamanaka, A., Dintwe, T.K.M., Pan, D.J., 2021.
21

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 593 Investigation on the Performance of Pipe Roof Method Adjacent to the Underground

d
594 Construction. Geotech. Geol. Eng. 39, 4677–4687.
595 [15] Xie, X.Y., Zhao, M.R., Shahrour, I., 2019. Experimental Study of the Behavior of

we
596 Rectangular Excavations Supported by a Pipe Roof. Applied Sciences. 9, 2082.
597 [16] Hisatake, M., Ohno, S., 2008. Effects of pipe roof supports and the excavation method
598 on the displacements above a tunnel face. Tunn. Undergr. Sp. Tech. 23, 120-127.

vie
599 [17] Heng, C.Y., Sun, S., Zhang, J.T., Zhou, Z., 2021. Calculation Method of Underground
600 Passage Excavation on Interactive Effects among Pipe-Roof, Steel Bracing and Foundation
601 Soil. KSCE J. Civ. Eng. 26, 448–459.

re
602 [18] Yang, X., Li, Y.S., 2018. Research of surface settlement for a single arch long-span
603 subway station using the Pipe-roof Pre-construction Method. Tunn. Undergr. Sp. Tech. 72,
604 210-217. er
605 [19] Cai, H.B., Hong, R.B., Xu, L.X., Wang, C.B., Rong, C.X., 2022. Frost heave and thawing
606 settlement of the ground after using a freeze-sealing pipe-roof method in the construction of
pe
607 the Gongbei Tunnel. Tunn. Undergr. Sp. Tech. 125, 104503.
608 [20] Zhang, D.M., Chen, C.C., Zhang, D.M., 2022. Ground surface movement of Shallow-
609 Buried Large-Sectional tunnel under Full-Ring Pipe-Jacking roof and ground freezing. Tunn.
ot

610 Undergr. Sp. Tech. 127, 104600.

611 [21] Hu, X.D., Fang, T., Chen, J., Ren, H., Guo, W., 2018. A large-scale physical model test
tn

612 on frozen status in freeze-sealing pipe roof method for tunnel construction. Tunn. Undergr.
613 Sp. Tech. 72, 55-63.
614 [22] Hu, X.D., Deng, S.J., Wang, Y., 2018. Test investigation on mechanical behavior of steel
rin

615 pipe-frozen soil composite structure based on Freeze-Sealing Pipe Roof applied to Gongbei
616 tunnel. Tunn. Undergr. Sp. Tech. 79, 346-355.
617 [23] Xiao, H.J., Zhou, S.H., Sun, Y.Y., 2019. Stability Analysis and Case Study of Shallow
ep

618 Tunnel Using Pipe Roof Support. Geotech. Geol. Eng. 37, 1249–1260.
619 [24] Xiao, J.Z., Dai, F.C., Wei, Y.Q., Xing, Y.C., Cai, H., Xu, C., 2016. Analysis of
Pr

620 mechanical behavior in a pipe roof during excavation of a shallow bias tunnel in loose deposits.
621 Environ. Earth. Sci. 75, 293.
622 [25] Lu, B., Dong, J.C., Zhao, W., Du, X., Cheng, C., Bai, Q., Wang, Z.G., Zhao, M.C., Han,
22

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889
 623 J.Y., 2022. Novel pipe-roof method for a super shallow buried and large-span metro

d
624 underground station. Undergr. Space. 7, 134-150.
625 [26] Volkmann, G., Schubert, W., 2006. Optimization of excavation and support in pipe roof

we
626 supported tunnel sections. Tunn. Undergr. Sp. Tech. 21, 404.
627 [27] Zhao, C.Y., Lei, M.F., Shi, C.H., Cao, H.R., Yang, W.C., Deng, E., 2021. Function
628 mechanism and analytical method of a double layer pre-support system for tunnel underneath

vie
629 passing a large-scale underground pipe gallery in water-rich sandy strata: A case study. Tunn.
630 Undergr. Sp. Tech. 115, 104041.
631 [28] Wang, H.T., Jia, J.Q., Kang, H.G., 2009. Analytical approach and field monitoring for

re
632 mechanical behaviors of pipe roof reinforcement. J. Cent. South Univ. Technol. 16, 827–834.

er
pe
ot
tn
rin
ep
Pr

23

This preprint research paper has not been peer reviewed. Electronic copy available at: https://ssrn.com/abstract=4355889