INTERNATIONAL SOCIETY FOR

SOIL MECHANICS AND

GEOTECHNICAL ENGINEERING

This paper was downloaded from the Online Library of

the International Society for Soil Mechanics and
Geotechnical Engineering (ISSMGE). The library is
available here:

https://www.issmge.org/publications/online-library

This is an open-access database that archives thousands

of papers published under the Auspices of the ISSMGE and
maintained by the Innovation and Development
Committee of ISSMGE.
The reinforcing effects of Forepoling Umbrella System in soft soil tunnelling
Les effets de renforcement du système Forepoling Umbrella dans tunneling sol mou

Binh Thanh Le, R Neil Taylor

Department of Civil Engineering, City, University of London, UK, [email protected]

ABSTRACT: Tunnelling inevitably induces ground deformations which may lead to damage of nearby infrastructure. A
method of reducing tunnelling-induced ground movements for open-faced tunnelling is to use in-tunnel support methods
such as the Forepoling Umbrella System (FUS). The FUS comprises steel pipes that provide structural support to the
surrounding soil above and around the tunnel heading. A series of three-dimensional centrifuge tests modelling a
reinforced tunnel heading were carried out to assess the FUS reinforcing proficiency by varying the arrangement and
bending stiffness of the steel pipes at different tunnel depths. The inclusion of a FUS leads to significant reductions in
both the magnitude of surface and subsurface ground movements and the overall extent of the deformed area. The tunnel
depth was found to be a critical factor in designing optimum Forepoling Umbrella Systems.

RÉSUMÉ: Le creusement d’un tunnel induit inévitablement des déformations de terrain qui peuvent donner lieu a des
endommagements des infrastructures avoisinantes. Une technique qui permet de réduire les déformations induites par le creusement
de tunnel est d’utiliser des techniques de support au sein des tunnels comme la voute parapluie (Forepoling Umbrella System FUS).
La technique de la voute parapluie utilise des tubes en acier qui fournissent un support structurel aux sols environnants situés au
dessus et autour de la tête de tunnel. Une série d’essaies tridimensionnels sur centrifugeuse simulant une tête de tunnel renforcé ont
été effectué pour évaluer les capacités de renforcement de la technique de voute parapluie en faisant varier l’agencement des tuyaux
d'acier et la profondeur du tunnel. Les résultats montrent des réductions significatives à la fois sur l'ampleur des mouvements de
terrain en surface et sous-terre et l'étendue globale de la zone déformée. La profondeur du tunnel a été démontré comme étant un
facteur critique dans la conception optimale de la technique de la voute parapluie.
KEYWORDS: centrifuge modelling, tunnel, soil deformations, soil reinforcement.

1 INTRODUCTION.
The main parameters of a tunnel heading and a Forepoling
Tunnelling in soft ground inevitably induces ground Umbrella System are illustrated in Figure 2. D is the tunnel
deformations and is a critical issue relating to the safety of diameter, C is the cover above the tunnel crown, P is the
people and nearby structures especially in crowded urban areas unlined portion of the tunnel heading. S is the centre to centre
with congested underground space. Damage to existing spacing between the forepoles. L is the length of the forepoles
buildings and utilities due to tunnelling-induced ground which are installed from the tunnel face at an insertion angle of
deformations has been encountered world-wide. Therefore, β. EL is the embedded length of the forepoles into the soil in
systems that minimise the ground movements caused by the front of the tunnel face. The soil beneath the embedded length
tunnelling process will be beneficial. of the forepoles acts like a foundation to support the steel pipes
The Forepoling Umbrella System (FUS) has proved to be an as they bridge over the structurally unsupported tunnel heading
efficient soil reinforcement method in open face tunnelling. The and this is known as the foundation effect (depicted in Figure 2).
system comprises steel pipes installed from the tunnel face to A minimum EL is required to maintain adequate foundation
form a roof above the tunnel heading (Figure 1), thereby support for the steel pipes.
contributing to decreasing the deformations caused by
tunnelling and increasing the tunnel heading stability. One of
the noticeable advantages of FUS is the immediate support after S - Spacing Forepoles
installation of the steel pipes (also termed as forepoles) that C-C C
L 1
allows the excavation to be carried out with minimal waiting
time. 
 EL D Tunnel lining
Overlapped
section
Direction Foundation
of tunnel effect P
advance Cross section 1
1-1
Figure 2. FUS schematic diagram.

Typical dimensions of various parameters used in a FUS in

FUS
practice are presented in Table 1.
consists
steel Understanding the soil deformation mechanisms is essential
pipes to achieve optimal soil reinforcement design. Upper bound
collapse mechanisms proposed by Davis et al., (1980) indicate
Tunnel that for a shallow tunnel, soil movements tend to be
concentrated at the crown of the tunnel (roof mechanism). For
deeper tunnels, the soil mobilisation involves not only the
Figure 1. Forepoling Umbrella System (after Carrieri et al., 2002). crown but also the sides and invert of the tunnel.

- 1709 -
 Proceedings of the 19th International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

Table 1. Typical parameters of a FUS. variables including C/D, forepoles arrangement and bending
Parameter Unit Value stiffness are presented in Table 2 and illustrated in Figure 4. In
Steel pipe diameter and wall mm 70-80 the reinforced tests, the forepoles were modelled by fourteen
thickness mm 4-8 1mm rods (brass or steel).
By modelling half of the tunnel, the surface and subsurface
Steel pipe length, L m 12-18 ground deformations could be observed and measured during
Embedded length, EL m 3-6 the tests. The stiff tunnel lining was modelled by a half section
of a stainless steel tube. The model tunnel diameter, D, was
Insertion angle, β  5-7 50mm. The unlined portion P and the insertion angle β in all the
Filling angle, α  60-75 tests were 25mm and 5 respectively. The tunnel cavity was
supported by a compressed air pressure contained in a latex
membrane lining the tunnel. The air pressure was controlled to
2 BACKGROUND
balance the total overburden stress at the tunnel axis level. The
2 .1 Current understanding in Forepoling Umbrella System overburden stress σob for C/D=3 and C/D=1 tests were 360kPa
and 155kPa respectively. A pressure transducer was installed at
Calvello & Taylor (1999) assessed the performance of a soil the end of the latex membrane to monitor the tunnel support
reinforcement measure by considering the reduction in soil pressure.
settlement and increase in tunnel stability. The tunnel stability
ratio, N, was defined by Broms & Bennermark (1967) as the
difference between the overburden stress at the tunnel axis, σob,
and the tunnel support pressure σT expressed as a ratio of the
undrained shear strength Su as:

N = [σob – σT]/Su (1)

where σob = (C + D/2) γ,

γ: unit weight of soil.

Juneja et al., (2010) used centrifuge modelling to investigate

the effect of forepoling reinforcement on a tunnel face in clay.
Juneja et al., (2010) found that that the use of forepoles reduced
the extent of the settlement trough ahead of the tunnel face
while the width of the settlement trough remained unaffected.
Figure 3. Centrifuge test diagram (for C/D=3)
Results from centrifuge tests and an upper bound plasticity
analysis conducted by Yeo (2011) suggested a significant
improvement in stability of the tunnel heading in clay can be Table 2. Test variables
achieved by using long and stiff forepoles. The tunnel model Test reference
used by Yeo (2011) had C/D=1 and the forepoles were modelled C/D=3 C/D=1 L (mm) EL (mm) S (mm)  ()
by brass rods. Inspection of the model forepoles post test
showed that the top forepoles had large deformation whereas 2BL 8BL 100 25 1.7 – 3.4 75
the lower forepoles had negligible deformation. That implied (Fig. 4)
the majority of soil movement occurred at the tunnel crown as a 3BL 11BL 100 50 3 90
roof deformation mechanism anticipated by Davis et al., (1980). 12BL*
Volkmann & Schubert (2007) reported the site measurement
data at a tunnelling project (D=11m, C=15m) using FUS. The 4BL 10BL 100 25 3 90
geological conditions were mainly mudstone, clay stone and 13BL*
sandstone. The results suggested that the tunnel lining and the 5BL** 9BL** - - - -
soil underneath the steel pipes provides foundation effects for
the whole FUS system. Therefore, the reinforcing effects of a * Tests 12BL and 13BL used steel rods for model forepoles. **5BL &
Forepoling Umbrella System depends not only on the stiffness 9BL are reference tests (no FUS). Other reinforced tests used brass rods.
of the steel pipes but also the strength of the surrounding soil.
Despite the research carried out to investigate the effect of L=100
FUS, understanding of the influence of the tunnel depths and 2BL/8BL
the relative effects of the FUS parameters (bending stiffness,
EL/L=0.25
ELto its reinforcing performance are still limited and form
the objectives of this paper. The centrifuge modelling technique =75
is chosen as the research methodology due to its capability in EL/L=0.5
3BL/11BL
replicating the behaviour of soils (Taylor 1995). With careful
selection of dimensions and materials the structural behaviour
of steel pipes and their interaction with soil can also be
modelled. =90
EL/L=0.25
4BL/10BL
3 CENTRIFUGE MODELLING TEST SERIES

3 .1 Centrifuge model tests =90

Ten centrifuge tests have been conducted to investigate the FUS
effect at two different tunnel cover depths C/D=1 and C/D=3. A Figure 4. Forepoles arrangements in centrifuge tests
typical model test apparatus is described in Figure 3. The

- 1710 -
 Technical Committee 204 & Joint Technical Committee 2 / Comité technique 204 & Comité technique conjoint 2

The model clay (Speswhite kaolin) was one dimensionally Figure 7 presents the amount of ground settlement reduction
consolidated to a vertical effective stress, σ’vo, of 175kPa. The delivered by the FUS, (S0–Sr)/S0x100% (Sr and S0 are
tests were conducted at 125g. respectively the surface settlements in the reinforced and
A guide, produced by precision 3D printing, was used to unreinforced tests at the corresponding tunnel support pressure).
insert the model forepoles into the clay sample when the model The presence of the FUS reduces the surface settlement by
was constructed at 1g during the model preparation (Figure 5). approximately 5%-85% at different tunnel support pressure.
Increase in the bending stiffness of the steel pipe by 80% (steel
compared with brass) only yielded an improvement in
settlement reduction of an average of 25% for the same
arrangement of forepole (13BL compared with 10BL, 12BL
compared with 3BL). This implies the reinforcement effect of
the FUS also depends on the strength of the surrounding soil.

Figure 5. Insertion guide, high precision produced by 3D printing.

3 .2 Centrifuge model tests

The models were accelerated to 125g while simultaneously
increasing the tunnel support pressure, σT, to balance the
overburden stress at the tunnel axis σob. It was left running until
the excess pore pressure dissipated and the clay had reached
effective stress equilibrium. After the clay model reached
equilibrium, the tests were started by gradually reducing the
tunnel support pressure to zero to simulate the excavation
process. During the tests, the surface and subsurface soil
displacement and tunnel support pressure were recorded at one-
second intervals for later analysis.

4 CENTRIFUGE TEST RESULTS

4 .1 Soil displacements
Maximum surface settlement is of great concern because it Figure 7. Settlement reduction offered by FUS.
indicates the potential damage to the building. Figure 6
illustrates the maximum surface settlement above the tunnel 4 .2 Subsurface soil deformations
heading obtained from LVDT (marked x in Figure 3, which lies
above the mid-point of the heading). For clarity, only the
significant settlements at the latter stage of the tests are
presented rather than from the very start of tunnel pressure
reduction.
Displacement (mm)

a) Comparison between 8BL & 9BL (σT reduced from 55 to 20kPa)

Distance from tunnel tail (mm)
400 350 300 250 200 150 100
0

a) C/D=1 tests 50

100
Displacement (mm)

2BL 2
5BL** 2 8
4

150 4 4
FUS
Tunnel
8

200 **: reference test

a) Comparison between 2BL & 5BL (σT reduced from 180 to 126kPa)
b) C/D=3 tests Figure 8. Engineering shear strain developed in the ground.
Figure 6. Surface settlement above the tunnel face in centrifuge tests.

- 1711 -
 Proceedings of the 19th International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

The subsurface soil deformations obtained from image analysis A longer embedded length (EL) provides improved support
showed that the areas that benefit the most from the FUS are efficiency because the forepoles benefited from the greater
above the tunnel heading where the model rods were installed foundation effects provided by the surrounding ground (tests
to reinforce the surrounding soil. In contrast in the areas that are 3BL and 11BL in Figures 4, 6, 7). The increase in the forepoles
far from the tunnel face or below the FUS, the effect of the FUS stiffness also improved the reinforcement effect of the FUS
is less pronounced. The effect of Forepoling Umbrella System (tests 12BL and 13BL).
on the extent and magnitude of engineering shear strains is The tunnel depth was shown to be an important factor that
illustrated in Figure 8. Large engineering shear strains (> 8%) dictates the soil mobilisation mechanisms (Davis et al., 1980)
developed near the tunnel invert and tunnel heading near the which in turn indicates a beneficial forepole arrangement. For
stiff lining edge in the reference tests with no forepoles. relatively shallow tunnels, the soil mobilisation mechanism is
However, in tests with the FUS, large shear strains only concentrated at the tunnel crown. Therefore, the presence of
occurred near the invert of the tunnel and not at the tunnel forepoles above the tunnel crown is more effective. For deep
heading where model rods were present. tunnels, the soil mobilisation involves the areas at the tunnel
crown and the tunnel spring line. Thus, having forepoles near
4 .2 Tunnel stability ratio the tunnel spring line is beneficial to the reinforcement effect of
the FUS.
The tunnel support pressure at collapse and the undrained shear
For deeper tunnels, the increase in the tunnel stability ratio
strength of clay are used to calculate the tunnel stability at
was not as much as for shallow tunnels (Table 3). This can be
failure. The stage at which there is a significant increase in the
explained by the overburden pressure in C/D=1 test which was
rate of settlement with reduction in tunnel support pressure is
relatively small and can be supported by the structural capacity
used to define failure and thus the tunnel support pressure at
of the FUS. For the deep tunnels C/D=3, this overburden stress
collapse (Mair 1979).
was much larger than it in C/D=1 tests and exceeded the
Most of the elements of clay around and above the tunnel in
support capacity of the FUS hence the smaller increase in the
three-dimensional heading tests experience extension stress
tunnel stability. The implication is that for tunnels in clay, the
paths during the reduction of tunnel support pressure. Therefore,
strength and stiffness of forepoles may need to be increased for
the undrained shear strength of one-dimensionally consolidated
deep tunnels.
kaolin in triaxial extension is deemed the relevant strength for
these three-dimensional tunnel heading tests (Mair 1979). The
6 ACKNOWLEDGEMENT
relationship between the undrained shear strength and OCR
(Mair 1979) was used to calculate the following undrained The first author acknowledges the Vietnam government for
shear strengths of clay Su1 and Su2 for tests with C/D=3 and funding his doctoral scholarship. The authors are grateful to the
C/D=1; colleagues in the Research Centre for Multi-scale Geotechnical
Engineering at City, University of London for their support.
Su1 = 0.18σ’vo (2)
Su2 = 0.16σ’vo (3) 7 REFERENCES
Table 3 presents the tunnel support pressure at collapse, with Broms, B.B. & Bennermark, H. 1967. Stability of clay in vertical
NTC, calculated using Equation 1 and Su1=31.5kPa and openings. J. Soil Mechanics and Foundation Division. American
Su2=28kPa from the Equations 2 and 3. The increase in the Society of Civil Engineers 93, SM1: 71-94.
tunnel stability delivered by the FUS (NTCr–NTC0)/NTC0x100% Carrieri, G., Fiorotto, R., Grasso, P. & Pelizza, S. 2002. Twenty years
are also tabulated (NTCr and NTC0 are respectively the tunnel of experience in the use of the umbrella-arch method of support for
tunnelling. Proc. 4th Int. Workshop on Micropiles, Venice.
stability ratios at collapse in reinforced and unreinforced tests).
Calvello, M. & Taylor, R.N. 1999. Centrifuge modelling of a spile-
reinforced tunnel heading. Proc. 2nd Int. Symp. Geotechnical
Table 3 Tunnel stability ratio at collapse Aspects of Underground Construction in Soft Ground. Tokyo.
Series Test EL/L  σTC NTC Increase in Davis, E.H., Gunn, M.J., Mair, R.J & Seneviratne, H.N. 1980. The
stability of shallow tunnels and underground openings in cohesive
(o) (kPa) NTC (%)
material. Géotechnique 30, No. 4: 397-416.
C/D=3 2BL 0.25 75 112 7.9 6.0 Juneja, A., Hedge, A., Lee, F.H. & Yeo, C.H. 2010. Centrifuge
3BL 0.5 90 98 8.3 12.0 modelling of tunnel face reinforcement using forepoling.
4BL 0.25 90 107 8.0 8.1 Tunnelling and Underground Space Technology 25: 377-381.
Mair, R.J. 1979. Centrifugal modelling of tunnelling construction in
5BL - - 126 7.4 - soft clay. PhD Thesis, University of Cambridge
12BL 0.5 90 87 8.7 16.7 Mair, R.J. & Taylor, R.N. 1997. Theme lecture: Bored tunnelling in the
C/D=1 8BL 0.25 75 15 5.0 13.8 urban environment. Proc. 14th Int. Conf. Soil Mechanics and
Foundation Engineering, Hamburg. Balkema, Rotterdam.
9BL - - 32 4.4 -
Taylor, R.N. 1995. Geotechnical Centrifuge Technology. Blackie
10BL 0.25 90 22 4.8 8.1 Academic and Professional, Glasgow.
11BL 0.5 90 6 5.3 21.1 Volkmann, G.M. & W. Schubert 2007. Geotechnical Model for Pipe
13BL 0.25 90 10 5.2 17.9 Roof Supports in Tunneling. In Proc. of the 33rd ITA-AITES
World Tunneling Congress, Underground Space - the 4th
Dimension of Metropolises, Volume 1. Prague, Czech Republic.
5 DISCUSSION Yeo, C.H. 2011. Stability and collapse mechanisms of unreinforced and
forepole-reinforced tunnel headings. PhD Thesis, National
Significant reinforcing effects delivered by using a FUS were University of Singapore.
observed from centrifuge tests (Figures 6, 7, 8). Initially, the
overburden pressure was supported by tunnel support pressure
σT. When σT reduced, the induced stress difference (σob – σT)
was supported by the surrounding soil and the FUS. Thus, the
effects of FUS became more significant when the tunnel
support pressure σT reduced.

- 1712 -