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

net/publication/303583232

Forepoling Design in Weak Medium based on the Convergence- Confinement

Method

Conference Paper · May 2016

DOI: 10.13140/RG.2.2.23716.68481

CITATIONS READS

3 7,331

3 authors, including:

Rouhollah Basirat Rahim Hassani

Tarbiat Modares University Amirkabir University of Technology
51 PUBLICATIONS 87 CITATIONS 32 PUBLICATIONS 42 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Rouhollah Basirat on 08 August 2016.

The user has requested enhancement of the downloaded file.

 Forepoling Design in Weak Medium based on the Convergence-
Confinement Method

Rouhollah Basirat *1, Rahim Hassani2, Navid Mahmoodian3

1. Iran Oston Consulting Engineers Company; Tarbiat Modares University, PhD Student ([email protected])
2. Iran Oston Consulting Engineers Company, MSc in Rock Mechanics Engineering
3. Iran Oston Consulting Engineers Company, Manager of Structural Engineering Department

ABSTRACT

Umbrella arches are tunnel heading reinforcements used for supporting the unlined portion of the tunnel
heading when excavating a tunnel in soft ground or loose soil conditions by open face excavation methods. They are
installed at the tunnel face at the top perimeter of the tunnel section. In this paper, the contribution of Convergence–
Confinement method for tunneling design, forepoling technique, and Ring-Cut method considered for Gilavand tunnel.
Three analysis are considered: top and bench excavation, ring-cut excavation, and top and bench excavation with
presence of forepoling condition. The results show that the pre-consolidation is need for top and bench excavation
method, however, the settlement is happened by installing supporting systems adjacent tunnel face. The settlement also
is occurred in Ring-cut method, however the tunnel is stable. The both stability and settlement is provided, so that
supporting system is installed in the close distance of tunnel face.

Keywords: Forepoling Method, Weak Medium, GRC, Settlement

1. INTRODUCTION
In general, poor ground prevails and because of the traffic and/or planning requirements tunnels tend to be laid at
a shallow depth. On the other hand, tunnels have to be constructed in ways that cause the least amount of detrimental
damage to surface structures, i.e., to minimize ground deformation as much as possible. Also in non-urban areas,
tunnels are sometimes excavated in difficult ground in which it is difficult to maintain stability and limit deformation of
the excavated opening. Hence, tunnel engineers have to be able to handle various problems in both urban and non-urban
areas[1].
In the case of low ground strength and/ or the need to restrict ground deformation, either shield tunnelling is used
or the face is divided into small sections. However, when the latter construction procedure is implemented large
construction machinery cannot be used, and as a result, the relative importance of the manual labor increases. Taking
into account the present circumstances in tunnel construction, i.e., a shortage of manpower and the increase in
mechanization, saving labor in tunnelling is necessary. Hence, a construction method that reinforces the surrounding
ground, and then allows large-sized machinery to excavate the tunnel rapidly is particularly appealing. Such action is
also necessary to ensure the safety of the construction work.
The umbrella method is a ground reinforcement technique in which all or part of the support of a tunnel section
is placed ahead of the cutting face before excavation is begun (Fig. 1).
The aim of this method is to preserve the face and perimeter of the excavation from disturbances as much as
possible.
Principles of the Umbrella Method in the umbrella method an "arch-like" shell is created ahead of the face prior
to excavation, enabling tunnel excavation to be safe and speedy under an umbrella-like structure. The outstanding
characteristics of the method are the support provided by this arch-like shell structure[1].
Given the existing shallow ground cover, implementation of forepoling was generally accepted to be vital but
was not readily approved initially. As there was no acceptable design guidelines for forepoling a series of detailed three-
dimensional finite element analyses on the construction process was commissioned that showed forpoling to reduce
ground surface settlements [2].
Kamata et al investigated tunnel heading collapse mechanisms where forepoles and face nails were excluded.
Excessive friction might be present between the aluminum plug and sand under high-g which could have affected the
collapse mechanisms[3].

www.C-Civil.ir 1
 Date et al. tested model tunnel headings in dry sand with face reinforcements or umbrella arch in the centrifuge.
Face reinforcements and umbrella arch were modelled with aluminum rods. The umbrella arch was found to isolate the
face wedge and spread the deformation field further forward of the tunnel heading[4].
In this paper, the ground-support interaction in ring-cut excavation technique has been considered. The method
has been adopted for Gilavand tunnel near Tehran-Iran. Also, results have been compared to forepoling technique based
on GRC method and by using numerical modeling.

Fig. 1: Schematic of the umbrella method [5]

2. SUPPORTING MECHANISMS OF THE UMBRELLA METHOD

The supporting mechanism of the umbrella method is to stabilize the excavation face in both the transverse and
longitudinal directions by an arch-like reinforced zone (Fig. 2).
As shown in Fig. 2 (a), regarding the stabilization in the transverse direction, an arch like reinforced zone carries
the ground load. Regarding the stabilization in the longitudinal direction, on the other hand as shown in Fig. 2 (b), the
ground load of the freshly excavated portion is supported by a beam with one end supported by the tunnel support and
the other end supported by the ground.
By supporting ground load in this way, the following effects on tunnel excavation occur[1].
- restraining ground settlement ahead of the face
- increasing face stability
- Reducing dimensions of tunnel supports (shotcrete, steel arch support, etc.)
- enlarging the cutting face, which makes it possible to use large machinery, which in tum, leads to rapid
excavation
In contrast, in the case in which the umbrella method is employed, the unsupported section of the tunnel is
covered with the umbrella structure and, as a result, the structure bears the overburden pressures. Similar to the relation
between the lining and the ground, vertical stresses acting on the umbrella structure depend on the relative stiffness
between the umbrella structure and the ground. It should be emphasized that, due to the umbrella structure, the stress
concentrations both ahead of the face and behind the face are smaller[1].

Fig. 2: Schematic of supporting mechanism of umbrella method [After 6]

www.C-Civil.ir 2
3. RING-CUT METHOD FOR SOFT GROUND
Ring-Cut method is a usual technique for excavating medium diameter tunnels in soft grounds[7]. This method
can be performed in different shape. If tunnel face is taken as Fig 3, one of the common performance sequences can be
as follows:
- Excavating parts UU, LU and RU by advancement of 0.5m to 1.0m in face.
- Performing shotcrete on excavated surfaces.
- Excavation of central part (NU).
- Installing temporary support for invert.
- Repetition if last items in next steps.
- Excavation of bench part by advance of 0.5m to 1.0m.
- Installing support system for walls and tunnel invert.
It should be added that sometimes in very soft ground it is necessary to perform ground improvement techniques
such as Forepoling before excavation of first part [8,9].

Fig. 3: Excavation sequences in Ring-Cut technique

Fig. 4: Excavation sequences in longitude view for Ring-Cut technique

In fact, the pressure of the central non-excavated part (NU and BU) controls large displacements spatially in
tunnel face and provide enough time to perform support system. In this situation, making a good observation about
ground behavior and interaction between ground and support near tunnel face is really important to access desirable
results.

4. THE CONVERGENCE–CONFINEMENT METHOD

The Convergence–Confinement method is a common procedure to observe ground- support behavior. In fact,
this method allows the load imposed on a support installed behind the face of a tunnel to be estimated. When a section
of support is installed in the immediate vicinity of the tunnel face, it does not carry the full load to which it will
eventually be subjected [10].
Carranza–Torres and Fairhurst [10] plotted the LDP, GRC and SCC on a single diagram (Figure 5). The
horizontal axis shows radial displacement in a given section of tunnel, The left vertical axis represents internal pressure
which acts on tunnel wall (Pi) or support (Ps) and the right vertical axis displays distance from a section of the tunnel to
the face (ahead or behind). In this reference system, the LDP is rotated counterclockwise 90 degrees.
Interpreting the relationships among the LDP, GRC and SCC makes it possible to find the final pressure that
rock mass will transmit to the support. To illustrate the procedure, consider a support is installed in the distance of
LDs,I from the face, which corresponds in the LDP of Figure 4 as point I. In the horizontal axis this point is u= u0s,I
and defines point K of the SCC. As long as the face does not move, stability is maintained solely by the ability of the
face to carry the load redistributed by excavation[11].

www.C-Civil.ir 3
 As the face advances, internal pressure decreases and pressure on the support increases then both the support and
excavation deform by the same amount. When the supporting effect of the face disappears completely, the system
reaches equilibrium at point D, which is the intersection of the GRC and the SCC. The pressure defined by this point
represents the final pressure and its corresponding displacement (time dependent weakening of the rock mass is not
considered)[10].
The hazard warning levels could be determined from critical strain (ɛc). The critical strain could successfully be
used for assessing displacement measurements in tunnels, such as crown settlements and convergence. Sakurai (1993)
obtained the relation of critical strain, compressive strength and Young‚s modulus as follows for level II[12]:
Log ɛc = - 0.25 LogE - 1.22
where E is Young‚s modulus in kg/cm2 .

Face
Wall and support pressures, Pi and Ps

LDPI
GRC I

Distance to the face

I D
Ls,I
0
SCC

N
D
ps,ID
k
u0s,I uDs,I
Displacement of the wall, u

Fig. 5: Relationships of LDP, GRC, and SCC [9]

5. THE CASE STUDY

The Gilavand way is one of the civil projects under construction in Iran. This way has two twin road tunnels,
varying from 340 m to 510 m in length named Gilavad 1&2 twin tunnels[13].
Based on geological longitudinal profile, there is only one type of lithology at the tunnel elevation including
young alluvium layers. In this paper, section for analysis is selected at the overburden 30 m. during underground
investigation, four boreholes have been drilled and laboratory tests have been done on soil samples. The ground
properties have been estimated by use of geological surveying and laboratory test results. The geotechnical parameters
in the selected section of the tunnel are provided in Table 1 [13].
Table 1: Rock mechanics parameters of the selected section
Overburden K C
φ
E Ρ
(m) (σh/σv) (KPa) (MPa) (kg/m3)

400 0.55 40 30 0.25 300 1800

6. NUMERICAL MODELING
1.1. Simulation of Ring-Cut Excavation Method
The Convergence-Confinement GRC and LDP curves are convoluted. As a result, in this research 2D and 3D
numerical modeling is used to define the curves and their effect on each other[14,15]. The geometry of tunnel in the
selected section is modeled by use of 2D finite difference element software (FLAC) and five stage excavations are
taking into account in modeling (Figure 6). Without considering the support effect, GRC of first excavation stage is
determined.

Fig. 6: a) Geometry of tunnel section and stages of excavation, b) 2D geometry of model in FLAC

www.C-Civil.ir 4
 In order to define LDP of first excavation stage the 3D finite difference element software (FLAC 3D) is used to
model the stages of excavation (Figure 7). In this 3D- modeling the effect of support is not considered. In Figure 9,
interpreting of GRC and LDP curve on a simple diagram is shown. As it can be seen, approximately 15 mm of
displacement is occurred by stress releasing of %35 at the distance of 0.5 m from tunnel face. In more value of stress
releasing, plastic displacement occur in tunnel which may cause abrupt failure. So, support system should be active in
this level of stress releasing. The installed support is a combination of a 25cm thick shotcrete and steel ribs of IPE18
spaced at 0.75 m. The combination support is resulted on the basis of simple assumption by Oreste[16].

Fig. 7: 3D model in FLAC3D to define LDP

1 45
40
0.9 GRC_I
35
0.8 LDP-FALD3D 30
25
0.7 20
1-(Relaxation Stress)

Distance from Face(m)

15
0.6
10
0.5 5
0
0.4
-5
0.3 -10
-15
0.2 -20
-25
0.1
-30
0 -35
0 0.01 0.02 0.03 0.04 0.05
Displacement(m)

Fig. 8: GRC and LDP diagrams resulted by 2D and 3D numerical modeling

After this stage, part II will be excavated and 6 m rock bolts spaced at 1.5 m x 1.5 m will be performed. In the
next stage, after excavation of part III, IV and V, final settlement in surface of ground was 3.2 cm. Therefore, the
induced settlement is more than allowable settlement (25 mm) and we should use pre-consolidation methods such as
forepoling technique.
1.2. Two Dimensional Modeling of Forepoling Technique
6.2.1 Analysis of face support
Analysis of the support provided by systems such as forepoles is even more difficult than the analysis of face
stability described earlier. A full solution requires the use of a program such as FLAC 3D but such programs are seldom
used for routine tunnel design[5].
Consequently, it is worth considering whether two-dimensional models such as FLAC2D can provide any
guidance on this complex issue.
There are no general rules currently available for the support provided by forepoles and, in the absence of such
rules, a crude equivalent model is used in this analysis. This assumes that a process of weighted averages can be used to
estimate the strength and deformation of the zone of ‘reinforced rock’. For example, the strength is estimated by
multiplying the strength of each component (rock, steel and grout) by the cross-sectional area of each component and
then dividing the sum of these products by the total area[5]. In this case, the steps in the tunnel roof required to install

www.C-Civil.ir 5
the forepoles are approximately 0.6 m deep and hence we will consider a rock beam 1 m wide and 0.6 m deep. The
forepoles have an outer diameter of 114 mm and an inner diameter of 100 mm and are spaced at 0.5 m. The quantities
involved are as follows:
Component Area (m2) Strength (MPa) Product
Rock 0.6 0.8 0.48
Forepoles 0.005 200 1
Grout 0.015 30 0.45
Sum 0.62 - 1.93

The resulting rock mass strength for this composite ‘beam’ is 1.93/0.62 = 3.1 MPa. The equivalent rock mass
properties can be estimated by iteration of the Hoek-Brown failure criterion (using the program RocLab) as follows:
Geological Strength Index GSI 50
Hoek-Brown constant mi 11
Intact rock strength σci 30 MPa
Rock mass strength σcm 3.1 MPa
Material constant mb 0.504
Material constant s 0.0002
Material constant a 0.531
Deformation modulus E 1100 MPa
Note that the Disturbance Factor D = 0 in this case since the forepoling is assumed to be undamaged. Figure 9
shows two dimensional geometry of forepoling modeling in FLAC 2D software.
JOB TITLE : Forepoling Analysis (*10^1)
1.600

FLAC (Version 7.00)

Pre-consolidation Area
LEGEND
(Forepoling)
1.200
16-Feb-16 0:05
step 21941
-1.177E+01 <x< 1.157E+01
-6.944E+00 <y< 1.640E+01
0.800
bulk_mod
2.000E+08
7.000E+08
shear_mod
1.200E+08
4.400E+08 0.400
cohesion
3.800E+04
7.000E+05

0.000

-0.400

D:\FLAC\Dynamic\Gilavand\Frequenc
-0.800 -0.400 0.000 0.400 0.800
(*10^1)

Fig. 9: 2D geometry of forepoling modeling in FLAC2D software

6.2.2. Design of the pipe umbrella by using statical (beam) method

As mentioned before, the pipe umbrella must be able to carry the loads from traffic and from the weight of the
overburden. As for the proof of stability of the tunnel face, the traffic load was increased by the factor of safety η t = 1.6
and the overburden pressure of the soil by the factor of safety ηγ = 1.4 for the design of the pipe umbrella[17].
The round length for the crown heading amounts to 1 m and the shotcrete support is to be installed and
temporarily closed at the invert after each round. For the design of the pipe umbrella it is conservatively assumed that
its maximum free span amounts to approx. 0.5 m. It is taken into account here that the green shotcrete does not have any
bearing capacity yet directly after its application. The calculation of designing the pipe umbrella, statical systems and
loading is as follow:

www.C-Civil.ir 6
 Traffic Load: Pt×ηt= 20×1.6=32 kN/m2
Overburden pressure: ×ηh= 1.4×24.5×4=137.2 kN/m2
Total Load: q=(32+137.2)×0.5=84.6 kN/m

Fig. 10: Design of the pipe umbrella, calculation of the section modulus of the pipes

The beam on two supports and the beam fixed at both ends are considered as statical systems for the design of
the pipe umbrella.
For the reasons given above, the maximum span of the beam is assumed as 2 m. From the superposition of the
traffic load and the overburden pressure, taking into account the spacing of the pipes (L=0.5 m), the loading of the beam
results to q = 84.6 kN/m.
The calculation of the section modulus W of the pipes, which is required for the stress proof is specified as
follows:
qL2
M  2.64  kN .m 
8

I   D 4  d 4   1.1106  mm 4  ,W   1.94 105  m3 
I
64 D/2
M 2.64
   140  MPa    all  240  MPa  
W 1.94 105
For the proof of safety, the computed stresses are compared to the yield stress of the pipes made from steel of the
grade St37 (σadm = 240 N/mm2). This is permissible, since the assumed loads were provided with factors of safety. The
deflection of the pipe umbrella is estimated at approx. 0.3 mm.
6.2.3. Numerical Method Results Analysis
Figure 11 shows the interpreting of GRC and LDP curves on a simple diagram for excavating with and without
forepoling technique, respectively.
1 40 1 40
35 GRC_I 35
0.9 0.9
GRC_I 30 30
LDP (with Forpoling)
0.8 25 0.8 25
LDP (without Forepoling)
20 20
0.7 0.7
15
1-(Relaxation Stress)

Distance from Face(m)

15
1-(Relaxation Stress)
Distance from Face(m)

0.6 10 0.6 10
5 5
0.5 0.5
0 0
0.4 -5 0.4 -5
-10 -10
0.3 0.3
-15 -15
0.2 -20 0.2 -20
-25 -25
0.1 0.1
-30 -30
0 -35 0 -35
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0 0.005 0.01 0.015 0.02 0.025
Displacement(m) Displacement(m)

Fig. 11: GRC and LDP curves for excavating a) with and b) without forepoling technique

When excavatin is done without forepoling (Figure 11a), supporting system is installed at 3.6 mm based on
Sakurai’s strain. According Figure 11a, preconsolidation should be performed before excavating of tunnel crown (x= -
2m from the face). In this condition, if supporting system attached to the tunnel face, the settlement is reached to 10 cm.
According to Figure 11b, although Sakurai’s strain represents the supporting system is installed at a farther distance

www.C-Civil.ir 7
from the face, but in this condition the settlement in created with the presence of forepoling. Therefore, the optimal
distance from the face in installation of forepoling condition is two meters.
Figure 12 shows the settlement curve in forepoling condition. The maximum settlement is 23 mm and it is less
than allowable settlement.
Distant from Tunnel Axis (m)
0
-50 -40 -30 -20 -10 0 10 20 30 40 50
-5
Settlement (mm)

-10

-15

-20

-25

Fig. 12: Ground settlement curve in forepoling condition

7. CONCLUSION
In this paper, the contribution of Convergence–Confinement method for tunneling design, forepoling technique,
and Ring-Cut method considered for Gilavand tunnel. Settlement and tunnel stability are two important issues for
excavating of Gilavand tunnel in overburden of 30m. Three analysis were performed for these issues: top and bench
excavation, ring-cut excavation, and top and bench excavation with presence of forpoling condition. The pre-
consolidation is need for top and bench excavation method, however, the settlement is happened by installing
supporting systems adjacent tunnel face. The settlement also is occurred in Ring-cut method, however the tunnel is
stable. The both stability and settlement is provided, so that supporting system is installed in the close distance of tunnel
face.

REFERENCES

[1] Muraki, Y. (1997). The Umbrella Method in Tunnelling, MSc Thesis, Massachusetts Institute of Technology.
[2] Yeo, C. H., Lee, F. H., Tan, S. C., Hasegawa, O., Suzuki, H. & Shinji, M. (2009). Three dimensional numerical
modelling of a NATM tunnel. International Journal of the Japanese Committee for Rock Mechanics. Vol. 5, No.
1, pp.33-38.
[3] Kamata, H. & Mashimo, H. (2003). Centrifuge model test of tunnel face reinforcement by bolting. Tunnelling
and Underground Space Technology. Vol. 18, pp. 205-212.
[4] Hoek, E. Numerical modelling for shallow tunnels, Discussion paper # 3, (2004).
https://www.rocscience.com/documents/pdfs.
[5] Date, K., Mair, R.J. & Soga, K. (2008). Reinforcing effects of forepoling and facebolts in tunnelling.
Proceedings of the 6th International Symposium of Geotechnical Aspects of Underground Construction in Soft
Ground (eds. Ng, C. W. W., Huang, H. W. & Liu G. B.), Shanghai. pp. 635-641.
[6] Geo-Fronte Research Association, (1994). Technical report on Rodin jet forepoling method.
[7] M. Hisatake, S. Ohno, T. Katayama, Y. Ohmae, S. Sano. (2009). Effect of Ring-Cut excavation Method on the
resistance of displacements ahead of tunnel face. Tunneling and Underground Space Technology. Vol. 24, Issue
5. pp 547-554.
[8] M. Hisatake, S. Ohno, T. Katayama, Y. Ohmae. (2012). Effect of Ring-Cut method as a settlement deterrent in a
soft ground tunnel. Tunneling and Underground Space Technology. Vol. 28. pp 90-97.
[9] Zhang Huashuang, (2009). Optimization of Construction Scheme for Xiangshan Railwa y Tunnel in Fully
Weathered Granitic Porphyry Ground. Tunnel Construction, Hebei, China.
[10] Carranza–Torres, C., and C. Fairhurst. (2000). Application of convergence-confinement method of tunnel
design to rock masses that satisfy Hoek-Brown failure criterion. Tunneling and Underground Space Technology.
Vol. 15, No 2. pp 187-213.
[11] Salari rad. H, Hassani. R, Aghchai. M.H. (2011). Multi Stage Excavation on the Basis of Convergence -
Confinement Method (Taloon Tunnel Tehran-Shomal Freeway of Iran). ARMA, Chicago.
[12] Sakurai, S. (1993). Direct Strain Evaluation Technique in Construction of Underground Openings. In Proc.22
U.S. Symp. Rock Mech. Boston. M A (Edited by H., H., Einstein). pp. 278-282.

www.C-Civil.ir 8
 [13] Geology Report of Gilavand way. (2011). I.O.C.E Consulting Engineers.
[14] Itasca Consulting Group, (2012). FLAC3D Fast Lagrangian Analysis of Continua in 3 Dimension. s. Version
4.0. Minneapolis.
[15] Itasca Consulting Group, (2012). FLAC2D Fast Lagrangian Analysis of Continua in 2 Dimensions. Version 7.0.
Minneapolis.
[16] Oreste, P.P. Analysis of structural interaction in tunnels using the convergence – confinement approach.
Tunneling and Underground Space Technology 18, pp 347–363, 2003.
[17] W. Wittke, C. Erichsen, J. Gattermann, (2006). Stability Analysis and Design for Mechanized Tunnelling, WBI,
Felsbau GmbH, Aachen.

www.C-Civil.ir 9

View publication stats