You are on page 1of 9

ResearchGate

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/267936830

Numerical Analysis for Umbrella Arch Method in


Shallow Large Scale Excavation in Weak Rock
Article

CITATION READS
S
280
65
authors, including:

Ranjith P.G Jit Sharma

Monash University York University

(Australia) 256 PUBLICATIONS 25 PUBLICATIONS 272 CITATIONS

3,040 CITATIONS

SEE SEE
PROFILE PROFILE

Jiyun Zhao
Nanyang Technological University

582 PUBLICATIONS 6,984 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

i ANALYSIS OF THE BEHAVIOUR OF ORDINARY AND GEOSYNTHETIC ENCASED GRANULAR

COLUMNS
Project

SUBJECTED TO SHEAR LOADING View project project

Fundamental Mechanism on Rock Fractures View project


All content following this page was uploaded by Ranjith P.G on 25 May 2015.

The user has requested enhancement of the downloaded file.


Numerical Analysis for Umbrella Arch Method in Shallow Large
Scale Excavation in Weak Rock

A.M. Hefny1, W.L. Tan1, P. Ranjith2, J. Sharma3, J. Zhao1

'Nanyang Technological University, School of Civil & Environmental Engineering, Singapore


2
Monash University, Australia 3University of Saskatchewan, Canada

ABSTRACT

The steel pipe Umbrella Arch reinforcement is a form of forepoling method used to improve stability
of a tunnel prior to excavation. Steel pipes are inserted into drilled holes and then mortar is poured
into the holes and pipe shaft. Grouting enhances the stability of the tunnel crown by forming a
reinforced arch that controls permeability and improves the cohesion and other geotechnical
parameters of the reinforced soil mass. This method is widely used in large scale excavations in Italy
and Japan, especially with NATM operations. Conventionally, this method is modelled numerically
both in 2dimensional and 3-dimensional analyses by approximating the steel pipes, grout and rock
material as a single composite material. However, such crude approximations may yield inaccurate
results and to date, there are no accurate approximations to efficiently model the Umbrella Arch. This
paper presents the results of a numerical study performed to investigate the effect of different
approximations for the umbrella arch on the predicted ground behaviour. Two methods of
approximation for the umbrella arch supporting a large shallow excavation in weak rock are
compared. The first method is the conventional approximation method of composite zone of enhanced
properties and the second method is more accurate where the steel pipes are modelled individually in a
grouted zone.

1 INTRODUCTION
The steel pipe Umbrella Arch method is used most often at fractured zones in especially poor ground
conditions over a long work section or where surface settlement is restricted, but is also executable in
other soil condition from moraine to sand. It is commonly employed as an auxiliary reinforcing
method in large diameter NATM tunnels, especially in Italy and Japan. This method has provided
successful results in restricting surface settlement for shallow tunnel excavations as described for
example by Yang et al. (2001), Haruyama et al. (2001), Matsuo et al. (1996), and Barisone et al.
(1982). The typical construction method and specifications of the steel pipe Umbrella Arch is shown
in Figures 1 and 2 for the longitudinal and cross sectional profiles respectively.

H05 1
Grout + Rock

Fig. 1. Longitudinal profile of Umbrella Arch Fig. 2. Cross sectionSteel


profile of Umbrella Arch
pipe
Method method
2. CASE STUDIES

Although the Umbrella Arch method has been widely used in weak ground and shallow depth
excavations since 1980s, there are no simple and definite methods to model the actual behavior of the
reinforced arch that is formed with steel pipes and grout.

Numerous authors had described detailed construction procedures for case studies in Japan and Italy
(e.g., Yang et al. (2001); Haruyama et al. (2001); Matsuo et al. (1996) and Barisone et al. (1982)).
Numerical analyses had also been carried out to some of the field cases stated above. Nishimaki et al.
(1995) conducted a three dimensional finite element analysis to model the steel pipe umbrella. The
three dimensional solutions were used to conduct a back analysis whereby the umbrella arch is
simulated by improving the material properties of the ground in two dimensional finite element
analyses.

Similarly, using the improvement of material property method, Ohtsu et al (1995) also studied the
behaviour of the ground due to tunnel excavation using a three dimensional finite element analysis.
Ohtsu et al (1995) assumed that the pre-lining arch (forepoles and shotcrete) acted as a combined shell
element.

Sato and Ito (1993) used a three dimensional finite element method to analyse the effect of the
Umbrella Arch method on tunnel face stability.

Most numerical modelers of the Umbrella Arch method had crudely simulated the steel pipes and grout
material as a composite material. Hoek (2000) showed that the method of weighted averages can be
used to estimate the strength and deformation of the reinforced zone surrounding the tunnel periphery.
In addition, Hoek (2000) also reinforced the fact that this method provides crude approximations in the
study of the effect of the Umbrella Arch method. Hence, a more accurate approach should incorporate
modeling the steel pipes individually in a grouted zone directly above the excavated tunnel rather than
as a composite material.

3. SIMULATION OF STEEL PIPES AND GROUT MATERIAL IN THE UMBRELLA ARCH


METHOD

The objective of this study is to compare ground deformations generated by 2 different approaches of
simulating the Umbrella Arch method as a pre-reinforcement when tunnelling in weak and shallow
ground. The conventional approach of combining steel pipes and grout material to form a composite
material and a new approach whereby steel pipes are modeled singularly as a separate component from
the grout zone are used to simulate the Umbrella Arch method. 2D finite difference analyses are
conducted to compare the difference in surface and crown settlement, and vertical and horizontal
displacement at the tunnel periphery of a test model.

H05 2
The problem studied is shown in Figure 3, which is approximately similar to the case study of the
Egnatia Highway in Hoek (2000). A horseshoe shaped tunnel of 10m height and 6m width is to be
excavated as part of a top heading and bench method in weak rock. The tunnel is situated at a shallow
depth of 15 m from the ground surface in highly weathered and disturbed overburden rock mass. Tunnel
construction is separated into the excavation of the top heading which is a semi circular tunnel of 6m
diameter followed by the bottom bench of height 7m. Pre-support reinforcement is required to stabilise
the top heading construction in this weak ground and the Umbrella Arch method was chosen. The
Umbrella Arch spans 180° around the tunnel crown. It was anticipated that the umbrella arch will
minimise the surface settlement and hence allow for the excavation of the bottom bench. Rock mass
properties are assumed for this analysis based on Hoek and Brown (1997) and shown in Table 1.
Table 1. Rock mass properties
Geological strength index, GSI 20
Hoek-Brown constant, mi 8
Intact rock strength, oci 3 MPa
Friction angle, 0 21°
Cohesive strength, c 76 kPa
Deformation modulus, E 308 MPa
Tensile strength 0.001 MPa

The numerical modelling only considered the top heading which is the semi circular part of the horse
shoe tunnel. This is because the forepoling system is installed prior to excavation over the entire stretch
of the tunnel crown to control the collapse of the rock mass whilst tunnelling and to control the surface
settlement. The bottom bench is excavated after the top heading has been completed.

The first simulation (Method 1) approximated the Umbrella Arch as a composite material whereby steel
pipes, grout and rock material properties are combined using weighted averages and an equivalent rock
mass strength is derived as shown in Table 2. The other rock mass parameters shown in Table 3 are
calculated using the equivalent rock mass strength and original rock mass properties (GSI=20 and m i=8)
based on Hoek and Brown (1997). These parameters only apply to a strip of rock

H05 3
material (denoted as composite ‘beam’) of 600 mm around the tunnel crown as shown in Figure 4 after
accounting for pipe size and site conditions. In the analysis, steel pipes of 150 mm diameter with 7 mm
thickness are used as shown in Figure 5.
Table 2. Specifications of the composite ‘beam’
Component Area (m2) Strength Product
Spacing between pipes, s = 500mm(MPa)
Rock 0.6*1.0 = 0.6 0.16 0.096
Steel pipes 2*0.00314= 0.00628 200 1.256
Grout 2*0.0145=0.029 30 0.87
Sum 0.635 2.222
Rock mass strength = 2.222/0.635 = 3.50 MPa
◄ --------- ►

Table 3. Rock mass parameters for the composite


‘beam’
Geological strength index, GSI 20
Hoek-Brown constant, mi 8
Intact rock strength, oci 48 MPa
Friction angle, 0 20°
Cohesive strength, c 1.2 MPa
Deformation modulus, E 1230 MPa
Tensile strength 0.014 MPa

The second method (Method 2) uses beam elements to form the steel pipes as shown in Figure 5, and
these steel pipes will be inserted individually around the excavated tunnel and in the grout zone. The
beam elements have dimensions of thickness 7 mm and length 150 mm. Steel pipes have Young’s

Fig. 5.
Description and
dimension of a
steel pipe.

modulus of 200 GPa and second moment of area, I, of 2.86E-08 m4.

H05 4
In this new method, grout zones are simulated with an equivalent rock mass strength using the method
of weighted averages. The zone has a thickness of 600 mm spanning around the tunnel crown for 180°
and the steel pipes are embedded in these zones. Figure 6 shows the calculations of the equivalent rock
mass strength for the grout zone.
From the Hoek-Brown criterion, the equivalent rock mass strength translates into the parameters shown

in Table 4.

Table 4. Rock mass strength of grout zone


Geological strength index, GSI 20
Hoek-Brown constant, mi 8
Intact rock strength, oci 21 MPa
Friction angle, 0 20°
Cohesive strength, c 0.5 MPa
Deformation modulus, E 815 MPa
Tensile strength 0.006 MPa

4. RESULTS & DISCUSSIONS

Surface settlement and crown displacement for both methods are compared in Figures 7 and 8. The
surface settlement profile generated using the method of combining the steel pipes and grout material as
a composite material (Method 1) showed that the maximum value developed directly above the tunnel
crown is approximately 4.0 mm as compared to a value of 2.0 mm generated from singularly modelling
the steel pipes in grout zone (Method 2). The crown displacement generated by the two methods also
showed significant difference of approximately 4.0 mm. The analysis showed also that Method 1
produced about twice the crown displacement estimated from Method 2.

Table 5 shows the vertical and horizontal tunnel closures. Vertical closure is the difference in vertical
displacement between the tunnel crown and the base of the excavated tunnel. Horizontal closure is the
difference in the horizontal displacement between the left and right side of the tunnel wall at the tunnel
springline. Method 1 yielded a vertical tunnel closure that is about 30% larger than that yielded by
Method 2. It can be observed that different approximations produce different results and hence it is
important to compare such approximations with actual field cases and 3-D numerical analyses to

H05 5
obtain more comprehensive conclusion about the appropriate approximation for such problems. Field
case studies and 3D numerical analyses are currently under investigation.

It should be noted that in the composite ‘beam’ method, the mixture of grout with the rock material
outside the pipe diameter were not accounted for. In actual forepoling methods, holes are drilled larger
than the pipe size before the perforated pipes are inserted and subsequently, the pipe shaft and the space
between the pipe and drilled hole are grouted. Permeation or fracture grouting are commonly used and
thus the grout may also penetrate a larger area outside the hole due to fractures and cracks produced by
drilling operations. As such, the grout region is larger than the approximated area and deformations
around the excavated zone could be substantially lesser since the reinforced area is larger.

Fig. 7. Comparison of surface settlement using various methods of analysis.

Fig. 8. Comparison of crown settlement with various methods.

H05 6
Table 5. Vertical and horizontal tunnel closures
Method vertical closure (mm) horizontal closure (mm)
1 20 7
2 15 6.5

5. CONCLUSIONS

An investigation on the ground behaviour around tunnels supported by the Umbrella Arch method was
conducted. Two methods of approximation for the Umbrella Arch are used in this investigation. The
first method is based on the principles of improvement of geotechnical material properties. In the second
method, the steel pipes are modelled individually. The results of the analyses show that different
approximations produce significantly different values of surface settlement and crown displacement.
Hence, crude approximations may not be viable for design of the Umbrella Arch method. In order to
produce an accurate 2-D approximation for the Umbrella Arch, the 2-D numerical solutions should be
calibrated with field case histories and 3-D numerical analyses, which take into account of the 3-D
deformation at tunnel face.

6. REFERENCES

Barisone, G., Pigorini, B. and Pelizza, S., 1982. Umbrella Arch method for tunnelling in difficult
conditions - Analysis of Italian cases. Proceedings of the 4th Congress International Association of
Engineering Geology, New Delhi, Vol. 4, pp. IV 15- IV 27.
Haruyama, K., Teramoto, S., Harada, H. and Mori, M., 2001. Construction of urban expressway tunnel
with special large cross section by NATM-Metropolitan Inter-city Highway (ken-O-Do) Ome Tunnel.
Modern Tunnelling Science and Technology, Adachi et al. (eds.), pp. 693-698.
Hoek, E., 2000. Numerical Modelling for Shallow Tunnels in Weak Rock. Presented during the 5th GRC
Lecture, NTU, Singapore.
Hoek, E. and Brown, E.T. 1997. Practical estimates of rock mass strength. International Journal for
Rock Mech. & Mining Sci. & Geomechanics Abstracts. Vol. 34, pp. 1165-1186.
Matsuo, H., Yamamura, S., Amano, M. and Taira, K., 1996. New construction method for urban tunnels
in uncemented ground under high groundwater pressure. North American Tunnelling '96, Ozdemir (ed.),
pp. 345-352.
Nishimaki, A., Mitarashi, Y. and Uematsu, S., 1995. Study of the effects of the AGF method. South East
Asian Symposium on Tunnelling and Underground Space Development, Bangkok, pp. 125-132. Ohtsu,
H., Hakoishi, Y., Nago, M. and Taki, H., 1995. A prediction of ground behaviour due to tunnel
excavation under shallow overburden with long length forepilings. South East Asian Symposium on
Tunnelling and Underground Space Development, Bangkok, pp. 157-164.
Sato, J. and Ito, J., 1993. Numerical analysis of the umbrella method for tunnel excavation.
Infrastructures Souterraines de Transports, Reith (ed.), Balkema, Rotterdam, pp. 355-360.
Yang, T., Woo, J. and Lee, S., 2001. Ground reinforcement for a tunnel in weathered soil layer beneath
Han riverbed in Korea. Modern Tunnelling Science and Technology, Adachi et al. (eds.), pp. 493-496.

H05 7

View publication ■

You might also like