You are on page 1of 12

DESIGN PROCEDURES FOR GROUNDWATER LOADS ON TUNNEL LININGS

Werner Bilfinger
1


Abstract:

Linings of tunnels below groundwater level are loaded by water pressure. These loads depend on a number of variables.
Relatively watertight linings will be loaded by almost full hydrostatic pressure, but in the case of relatively permeable
linings, a number of other factors affect the loads, like the relative stiffness between soil/ rock massif and tunnel lining, the
depth below groundwater and the shape of the flownet. A closed form solution for simplified problems, that include these
considerations, is presented and, for more complex situations, a design sequence, to be used with numerical methods, is
proposed.

Keywords: tunnel lining, groundwater flow, loads, stiffness, permeability

1. INTRODUCTION

Loads due to groundwater on buried structures are an issue that raises uncertainties, especially during the design phase,
when the main decisions are taken about the geometry and other main properties of the structure. For tunnels, lining
geometry, thickness, reinforcement and type of waterproofing are chosen and, in the case of open excavations, type,
structural dimensions and wall embedment of the retaining structure are defined. In several occasions, due to these
uncertainties, conservative assumptions are made, for example, adopting full hydrostatic pressure acting on the structure
(Guilloux, 1994).

Tunnel linings are almost never completely impermeable, as can be seen in a number of different publications and almost all
tunnels in operation. Therefore, in many cases, Owner, Designer and Contractor are interested in evaluating the benefits of
considering partially or fully drained tunnel linings.

Interesting comparisons can be made with available bibliography about deep excavations in Frankfurt clay (Moorman,
2001) (Moorman and Katzenbach, 2001) (Moorman and Katzenbach, 2002): studies with numerical models and in situ
instrumentation showed that using localized groundwater lowering measures, loads on the retaining structures were reduced
significantly, without affecting superficial water bearing strata. The consideration of groundwater flow instead of a
hydrostatic approach allows for an economic and safe design.

Two main topics are faced by the tunnel Designer when analyzing groundwater loads:

- availability of theoretical approaches that allow simulation of water flow and its effects on the tunnel lining this
issue can be considered solved, as theories regarding stress-strain behaviour of soils and linings, and on the other
hand, water flow, are common knowledge. For simplified conditions some closed form solutions are available, as
well as for more complex problems, a number of numerical methods through several commercial software;
- availability of reliable input data for representative numerical simulation, which is complex, as the uncertainties of
soil parameters, lining parameters and boundary conditions are significant.

In this paper, some considerations regarding lining loads due to groundwater are presented, focusing on the comparison
between impermeable and drained linings. An approach for safe and economic design is proposed.

2. BACKGROUND

Dimensioning a tunnel lining can be performed by several procedures. Common practice for NATM tunnels is the
sequential modeling of tunnel construction by numerical methods, simulating all important constructive phases and final
operating condition.

Normally, for the construction of tunnels below groundwater level, during construction the soil mass is drained by
groundwater lowering measures, using external deep wells or internal drainage with longitudinal sub-horizontal vacuum

1
Vecttor Projetos Ltda. Rua Cardeal Arcoverde, 1749, cj. 103 CEP 05407-002 SP - Brasil
Email: werner@vecttor.com.br
drains and/or well points. Only after completion of the final lining, the groundwater lowering measures are disrupted and
the groundwater table will re-establish its original position.

The dimensioning sequence of the lining normally includes this important phase and, as the lining is usually considered
impermeable, lining loads due to groundwater level will be full hydrostatic pressure. This load case, normally, leads to
higher axial loads and, especially, high bending moments, specifically in the tunnel invert. In several cases, lining thickness
and steel reinforcement are defined by this load case, as in this phase the highest loads are acting on the lining. Thus a
reduction of these loads can lead to more economic design.

According to Atkinson and Mair (1983), using a simplified approach, loads on tunnel linings are the same, considering
impermeable or fully drained linings. A simplified model can be used to visualize the proposed approach: consider two
containers full of water, where the first one is impermeable, and the second one, has a perforated bottom. If the second
container is continuously maintained full of water, both containers will have the same weight and the same loads will be
applied to the bottom of the container. This behaviour is assumed to be representative of the tunnel lining - soil mass
interaction in the case of no-flow or flow condition, meaning that a tunnel will support the same loads, with or without an
impervious lining. Szchy (1973) discusses, on the other hand, that using drainage the lining is subject to smaller forces.

The approach proposed by Atkinson and Mair cannot be considered universally valid, because it does not consider soil and
lining stiffness, neither the flowlines towards a tunnel are parallel, as in the simplified container model. Finally, the
simplified model does not consider the influence of different permeabilities, between different soil layers or between soil
and tunnel lining.

Franzn and Celestino (2002) discussed impermeable and drained linings and different design philosophies, including
examples, and concluded that more research and, mainly, interaction and exchange of ideas is necessary to improve safe and
economic design and construction. The same issue is discussed by Tunnels & Tunneling (2005), concluding that
construction industry needs to focus on sharing experience and to make more of the good quality materials and tools
available, by ensuring concerted and informed job-specific effort to satisfy desired long-term water targets.

Interesting results regarding tunnel water inflow rates were published by Fiztpatrick et al (1981). These authors evaluated
the influence of variable depth below the groundwater level, of different tunnel diameters and the distance to an
impermeable soil layer below the tunnel.

3. FACTORS AFFECTING LINING LOADS DUE TO WATER FLOW

The main factors that affect loads on tunnel linings due to water flow are:

- relative soil-lining stiffness;
- relative soil-lining permeabilities;
- geometric factors, like the depth below the groundwater level, distance to an impervious layer, location of water
source, etc.

If a no-flow condition is assumed, lining load due to water is full hydrostatic pressure acting on the lining or, in the case of
use of a water proof membrane, acting directly on it. The loads will not be affected by the factors listed above.

3.1 Relative Lining Soil Stiffness

The influence of the relative lining soil stiffness can be easily visualized for tunnels in rock, where often the groundwater
table lays high above tunnel crown, but linings are not dimensioned to support full water pressure. Instead, massif drains
into the tunnel or toward a drainage system, and the seepage forces of the water are equilibrated within the mass. This
behavior occurs because the massif is stiff, and the lining, relatively, soft. For a tunnel immersed in soft soil, normally
linings are dimensioned to withstand full hydrostatic loads and high loads due to the soil massif itself.

Using the potential function for radial flow (Verruijt, 1970) and an elastic approach (Timoshenko, and Goodier, 1970), to
model the lining and the soil massif in an idealized situation of a tunnel under radial flow, a closed form solution, presented
below, can be obtained, which shows that, for a flow condition, the lining loads are not equal to the no-flow condition, but
depend of soil and lining stiffness (Bilfinger, 1997).

For an isotropic linear elastic medium and for radial flow towards a circular tunnel, the following equilibrium equation is
valid:

( )( )
( )
r C
E
u
dr
du
r
dr
u d
r
w r
r r


1
2
2
2
1
2 1 1

+
= +

The general solution of this equation is:

( )( )
( )
(

+
+ = r r C
r
C
C
E
u
w r
ln
1 2
2 1 1 1
1
4
3




and the radial stress is:

( )( ) ( ) ( )
( ) ( )

+
+
+
+
= 1 ln
1 2 1 2 1 1
1
2
4 3
r
C
r
C C
w
r


C
1
is defined in (Verruijt, 1970), C
3
and C
4
are integration constants,
w
= unit weight of the water, E = soil Elasticity
Modulus, = Poisson coefficient, and r = radius.

Similar results were presented in (Wood, 1975).

Adopting, as a boundary condition, at the internal boundary of the radial flownet, the axial compression of a tunnel lining
with axial stiffness E
lining
x thickness, and, at the external boundary,
r
= 0, it can be seen in Figure 1 that a significant
decrease in lining loads occur for soft linings, when compared to the hydrostatic no flow condition (Bilfinger, 2005).

0
0,2
0,4
0,6
0,8
1
1,2
0 1 2 3 4 5 6 7 8 9

N

f
l
o
w

/

N

h
y
d
r
o
s
t
a
t
i
c

Figure 1 Influence of relative stiffness on non-dimensional axial force in tunnel lining for circular tunnel and radial
flownet.

To evaluate the relative stiffness, the dimensionless axial stiffness parameter , as proposed by Duddeck and Erdmann
(Duddeck, and Erdmann, 1982), can be used:

= (E x R) / (E
lining
x A)

where E is the soil Elasticity modulus, R, the tunnel radius, E
lining
the Elasticity modulus of the lining and A the thickness of
the lining.

Parametric analyses using a bidimensional finite differences computer software (Bilfinger, 1997) showed that, for fully
drained linings, tunnels can be assumed as being stiff for values of < 0,1. Higher values of , in drained conditions may
lead to significant reductions of lining loads due to seepage forces. For example, loads due to water on a typical 6 m
diameter tunnel, with a 30 cm thick concrete lining, will be influenced by drainage only if the surrounding massif has an
Elasticity modulus higher than 200 MPa.

Analogous to , the dimensionless parameter (Duddeck, and Erdmann, 1982) can be used to evaluate lining stiffness in
terms of bending moments:

= (E x R
3
) / (E
lining
x J)

where E is the soil Elasticity modulus, R, the tunnel radius, E
lining
the Elasticity modulus of the lining and J the inertia of the
lining.

Parametric analyses (Bilfinger, 1997) showed that bending moments are almost reduced to zero for values > 500. The
analyses showed also that bending moments are very sensitive to lining geometry.

3.2 Relative Permeabilities

The loads on the linining also depend of the relative permeability (K) between lining and its surrounding massif. For
example, for a tunnel where the linining has a relatively low permeability when compared to the surrounding soil mass, its
lining will behave almost as impermeable. A relatively permeable lining, on the other hand, will behave as a drain. Figure 2
shows the two situations described above.



Figure 2 Comparison between relatively drained and impermeable linings

For the case where K
lining
>> K
soil
, almost no head is lost when the water flows through the lining and no direct loads will
act on it. The loads due to the groundwater will only act on the lining indirectly, through the loads applied by the seepage
force onto the massif.

For the case where K
lining
<< K
soil
, almost no head is lost in the massif, and the seepage force will be concentrated in the
lining, as if the water pressure were applied directly on the lining.

Parametric analyses (Bilfinger, 1997) showed that the loads are affected significantly by the parameter K
soil/
K
lining
only for
values between 1 (fully drained lining) and 10, with some influence up to values in the range of 50. For higher values, the
lining loads will be almost equivalent to a no flow condition.

Similar results are obtained for water inflow rates, with high influence on the inflow rates for values K
soil
/K
lining
up to 50,
and very low influence for higher values of K
soil
/K
lining
(Fitzpatrick et al, 1981)(Bilfinger, 1997).

From a practical point of view, a permeable lining means that a) the soil has a low permeability, comparable to the lining, or
b) lining permeability is increased by the use of drains.

For the second case, parametric evaluations (Bilfinger, 1997) showed that a significant number of drains are necessary. The
use of, for example, 3 drains, modeled as slots in a bidimensional model, located in the centerline of the invert, and in the
springline of the tunnel lead to a condition comparable to a relative permeability of K
lining
/K
soil
= 50.

Results published by Celestino et al. (2001) showed that average lining permeabilities range between 1x10
-8
m/s and 1x10
-11

m/s. These lining permeabilities are significantly higher than the material permeabilities (concrete, cast iron, etc) and
represent an average value that includes imperfections. For the cases analyzed, soil permeabilities vary between 1x10
-6
m/s
and 1x10
-8
m/s. Franzn and Celestino (4) also show that, with adequate technology, including the use of microsilica and
steel fibers, well designed and applied linings could result in a permeability of 1 x 10
-12
m/s.

The relative permeability between soil layers can also influence lining loads significantly: for example, if the tunnel is
located below a relatively impervious soil layer, a significant part of the seepage forces will be dissipated inside this layer,
as schematically presented in Figure 3: the permeability K
1
is higher than the permeability K
2
, and, therefore, almost all
head is lost in layer 2, and layer 1 behaves almost as a reservoir. The permeability in layer 3 is higher than the
permeability in layer 2 and, therefore, head loss will be also concentrated in layer 2.



Figure 3 flownet for layered soil with variable permeabilities

Quantitative evaluations of this type of situation should be performed case by case by means of numerical methods, as
parametric analysis for all possible geometries are not viable.

3.3 Depth Below Groundwater Level

The influence of the depth below groundwater level on water inflow rates is significant (Franzn and Celestino, 2002), with
an almost direct linear relation between depth and water inflow rates.

Lining loads are also affected by the relative depth below groundwater level: for shallow tunnels, less than 1 tunnel
diameter below groundwater level, parametric analysis showed significant reductions in lining loads, especially in the tunnel
springline. In the tunnel invert, reductions are less important and almost negligible in tunnel crown.

In deeper tunnels (depth below groundwater level greater than 1 tunnel diameter) the influence of this factor becomes less
important.

3.4 Other Factors

Some other factors affect the lining loads, but are almost impossible to quantify for a generic solution.

One of the most important factors is the long term effect of the tunnel on the groundwater level, leading to a water balance
different from the natural original conditions. If tunnel water inflow rates are high and the water source is not recharging
the groundwater level, it will be lowered, reducing lining loads as a consequence and causing environmental changes, which
have to be evaluated case by case. In some countries it is not allowed to extract water permanently from the soil massif and,
therefore, the construction of a drained tunnel may not be possible from a legal point of view.

4. IMPERMEABLE TUNNEL LININGS

Tunnel linings are in general built of concrete. For NATM tunnels, primary lining consists of shotcrete and the secondary
lining can be constructed either with shotcrete or conventional concrete. Shield tunnel linings are normally constituted of
concrete segments and, in some cases, completed by a secondary lining.

The construction method, especially for NATM tunnels, leads to higher risk of occurrence of defects in the primary lining,
due to precarious installation, more difficult quality control, etc. As for the final use of the tunnel normally an imperfect
lining, with water inflow, is not suitable, in several cases a waterproof membraneis included as impermeabilization detail
between the primary and the secondary lining.

Segmental linings of shield tunnels are often provided with special gaskets to achieve an impervious lining and, in some
cases, a secondary lining is also installed.

But impermeable tunnel linings, in practice, are almost impossible to achieve. Even for full impermeabilization with a
geomembrane around the tunnel section, defects in the geomembrane are possible and will generate some water inflow. A
number of publications show that even for very tight specifications, some water inflow will always occur (Celestino et al,
2001)(Negro, 1994). Given that some water inflow will occur, it is important to know if and how this inflow can be used to
reduce lining loads.

The results presented in item 3 show that, to achieve some reduction in the lining loads, the ratio K
lining
/K
soil
must be smaller
than 50. This situation occurs only in tunnels immersed in very low permeability clays or when artificially drained linings
are used. Results from 16 cases where lining permeabilities were back-analyzed (Celestino et al, 2001) showed that the
factor K
lining
/K
soil
is normally higher than 100.

From a practical point of view, to design the tunnel lining assuming it to be drained, one has to be sure that:

a) soil permeability is low. This point is particularly critical, as an error in a magnitude of 10, which is common for
permeability laboratory and field tests, can change dramatically lining loads;
b) the lining is designed with drains that will work properly.

On the other hand, if the permeability ratio K
lining
/K
soil
is higher than 50, which is apparently common (12), lining can be
considered impermeable for the calculation of lining loads.

5. PROPOSED DESIGN SEQUENCE

The final use of tunnels is mandatory for establishing their design concepts in every aspect. The discussion of real
necessities and requirements of the lining and the environment should be clearly defined. In many situations water dropping
is not allowed on a road, railway or equipment, but there is no restriction regarding water inflow, for example, in the invert
of the tunnel.

In this case, a combination of an impermeable lining in tunnel crown, to avoid damage or restraints to final use of the
tunnel, associated to a drained invert, to reduce lining loads and, therefore, resulting in a more economic design, seems to be
the best solution. British Tunneling Society (2004) recently presented guidelines for lining design, and for the case where
the lining is assumed to be drained at tunnel invert and impervious in tunnel crown, states that This type of system limits
the development of hydrostatic pressures on the permanent lining and for most practical purpose these need not be taken
into account in the design.

The statement above is theoretically wrong, but maybe valid for the most common situations: with a drained invert, a
significant part of the flowlines that end at the tunnel invert are not flowing in its direction, as can be seen in Figure 3.
Therefore, a significant part of the seepage forces will not act along the tunnel invert direction, reliefing it if compared to
radial flow or a hydrostatic condition. The seepage forces will load the crown, as the integration of the seepage forces along
a flowline (almost parallel toward the tunnel) will result in an equivalent load of full hydrostatic pressure. As the tunnel
normally is dimensioned for a significant percentage of the overburden stress (with exception to deep and/or rock tunnels),
this condition is not very different of a dry condition load, as the unit weight of soil is close to 20 KN/m
3
, for the dry
condition, and the buoyant unit weight, 10 KN/m
3
, must be completed by the hydrostatic pressure due to the water (10
KN/m
3
), for the flow condition.

A more complete evaluation of the idealized situation of an impermeable lining in tunnel crown, and a drained tunnel invert,
could be run following the diagram presented in Figure 4. It is important to realize that the flow analysis that has to be
performed to evaluate the flownet around the tunnel shall be carried out for steady state condition and can be fully
uncoupled. Therefore, it is relatively easy to perform. After obtaining the pore pressure distribution inside the soil mass, this
pore pressure distribution is the input for the static analysis of the lining.


Tunnel final use requirements
Drained invert possible?
Evaluate water inflow rates through simplified model
- Impact on permanent pumping
- Impact on permanent groundwater lowering
Drained invert possible?
Initial design (geometry, lining thickness, etc)
Prepare numerical model (flow condition with drained invert)
Simulation, including different possible scenarios
Analysis of results - stresses in lining, settlements, water inflow rates.
Re-evaluate geometry, lining thickness, etc.
Final Design
Yes
No
No
Yes
C
o
n
v
e
n
t
i
o
n
a
l

C
a
l
c
u
l
a
t
i
o
n

Figure 4 proposed design sequence

One of the key aspects of the proposed design approach is to provide the tunnel invert with a drainage system that
guarantees high permeability during the lifetime of the tunnel and that allows safe and easy maintenance. If unpredicted
water pressure develops directly on a tunnel invert, certainly structural problems will damage it.

Figure 5 presents the results for a particular example: a two-lane subway tunnel, located in sedimentary tertiary clays and
sands. Three different load cases are presented: Case 1 shows the axial forces and bending moments for a dry soil massif.
Case 2 represents the long term situation, assuming an impermeable tunnel lining. Case 3 shows axial forces and bending
moments for a situation where tunnel crown is impermeable and the invert fully drained. The results show that, specifically
in the tunnel invert, the different considerations regarding loads due to water lead to significantly different results and,
therefore, to different costs. Clearly, Case 2 will lead to a more expensive final lining. On the other hand, water inflow rates
should be estimated, to evaluate long-term costs of pumping, for example. An interesting discussion about this issue is
presented in (13) and concludes that more comprehensive studies should be performed when defining



Figure 5 example of two lane subway tunnel

A recent publication (Insam et al, 2005) presents a cost comparison between drained and watertight linings for a railway
tunnel in Austria. The cost comparison included the maintenance and pumping costs during 150 years and the conclusion
was that the drained system resulted more economic, showing the importance of comparative analysis, considering all
relevant aspects.

6. CONCLUSIONS

Water pressures acting on tunnel linings are often a significant part of the overall load, especially if simplified approaches
are used, loading the lining with full hydrostatic pressure.

Alternatively, assuming a design procedure that considers the real flow conditions and provides tunnel lining with an
effective drainage system, significant reductions in lining loads due to the water table can be obtained.

For example, for stiff soil massifs and relatively soft linings, with < 0,1, assuming that the average tunnel lining
permeability (including drains) is similar to the soil permeability, lining loads will be significantly lower than for the
hydrostatic load case.

The impact of continuous drainage on the soil massif shall be evaluated for its potential environmental consequences.

7. ACKNOWLEDGEMENTS

The assistance of Prof. Dr. Waldemar C. Hachich during the development of the theoretical background and the parametric
analyses and of Prof. Dr. Tarcsio B.Celestino, for his valuable suggestions is gratefully acknowledged.

References

Atkinson, J. H.; Mair, R. J. Loads on leaking and watertight tunnel linings, sewers and buried pipes due to groundwater.
Gotechnique, v. 33, n.3, p. 341-4, 1983.

Bilfinger, W. Influence of Water-tightness on the Design of Tunnels. MSc. Thesis University of So Paulo, 1997 in
Portuguese

Bilfinger, W. Impermeabilization Versus Drainage Some Considerations Regarding Lining Loads. Felsbau, v. 23 n.3, p.
55-61, 2005.

British Tunneling Society. Tunnel Lining Design Guide. Thomas Telford Publishing, London, 2004.

Celestino, T. B.; Giambastiani, M.; Bortolucci, A. A. Water Inflows in Tunnels: Back-Analysis and Role of Different
Lining Systems.AITES-ITA World Tunnel Congress, 2001.

Duddeck, H.; Erdmann, J. Structural design models for tunnels. In Tunneling82, London, 1982, Proceedings. London,
IMM, p. 83-91, 1982.

Fitzpatrick, L.; Kulhawy, F. H.; ORourke, T. D. Flow patterns around tunnels and their use in evaluating construction
problems. In: Soft Ground Tunneling, Rotterdam, A.A. Balkema, p. 95-104, 1981.

Franzn, T.; Celestino, T.B. Lining of tunnels under groundwater pressure. In AITES-ITA World Tunnel Congress, Sydney,
2002.

Guilloux, A. French National Report on Tunneling in Soft Ground. In: International Symposium on Underground
Construction in Soft Ground, New Delhi, 1994.

Insam, R., Stacherl, B., Kusterle, W. Requirements for tunnel Waterproofing and Drainage Systems. Felsbau, v. 23 n.3, p.
8-13, 2005, in german.

Moormann, C. Design optimization of deep watertight excavations in clay. PhD Thesis Darmstadt University of
Technology, 2001, in german.

Moormann, C. Katzenbach, R. Performance of deep watertight excavations in urban area and their impact on adjacent high
rise buildings. XV ICSMGE, Istambul, 2001.

Moormann, C. Katzenbach, R. Wirtschaftlicher Verbauwandentwurf fr tiefe (teil-) wasserdichte Baugruben. 9.
Darmstdter Kolloquium, 2002.

Negro, A. Soil Tunnels and their support. X Cobramsef, Foz do Iguau, pp. 33-60, 1994

Szchy, K. The Art of Tunneling. Akadmiai Kiad, Budapest, 1973

Timoshenko, S. P., Goodier, J. N. Theory of Elasticity, 3 ed. Mc Graw Hill Book Company, 1970

Tunnels & Tunneling International. The way ahead for water inflow control. Vol. 37, n.3 p.43-45, 2005.

Verruijt, A. Theory of Groundwater Flow. London, MacMillan, 1970.

Wood, A. M. M. The circular tunnel in elastic ground. Gotechnique,. v.25, n. 1, p. 115-27, 1975


APPENDIX

Closed Form Solution for Lining Loads due to Groundwater Flow Radial Flownet


The seepage force acting in flow direction is defined as being:

i
dS
d
w

=
'
(A1)

S= distance along a flowline

dS
dH
i = (A2)

H = hydraulic potential

Therefore:

dS
dH
dS
d
w

=
'
(A3)

For radial flow, the potential function is (Verruijt, 1970):

2 1
) ln( C r C H + = (A4)

and the gradient is defined as:

r
C
r
H
i
1
=

= (A5)

substituting (A5) in (A3):

r
C
dS
d
w
1
'

= (A6)

For problems in polar coordinates, following equilibrium equations are valid (Timoshenko and Goodier, 1970):

0
1
= +

R
r r r
r r r


(A7)

0
2 1
= + +

S
r r r
r r

(A8)

For radial flow, the body forces R and S result:

r
C
R
w

1
= (A9)

0 = S

Due to symmetry, 0 =

r
, and
r
and

are not dependent of . Therefore, the first equilibrium equation can be rewritten
as below and second equation is identically zero.

( ) 0
1
=
w r
r
C r
d
d


(A10)

For an linear isotropic elastic medium:

dr
du
r
r
= (A11)

r
u
r
=

(A12)

where u
r
= radial displacement

( )( )
( ) ( )



+
+
=
r r
E
1
2 1 1
(A13)

( )( )
( ) ( )
r
E


+
+
= 1
2 1 1
(A14)

substituting (A11) and (A12) in (A13) and (A14):

( )( )
( ) |
.
|

\
|
+
+
=
r
u
dr
du E
r r
r


1
2 1 1
(A15)

( )( )
( ) |
.
|

\
|
+
+
=
dr
du
r
u E
r r

1
2 1 1
(A16)

Substituting in (A10):

( )( )
( )
r C
E
u
dr
du
r
dr
u d
r
w r
r r


1
2
2
2
1
2 1 1

+
= + (A17)

The solution of this equation is (Bilfinger, 1997):

( )( )
( )
( )
(

+
+ = r C
r
C
r C
E
u
w r
ln
1 2
2 1 1 1
1
4
3


(A18)

substituting u
r
in
r
:

( )( ) ( ) ( )
( ) | |

+
+
+
+
= 1 ln
1 2 1 2 1 1
1
2
4 3
r
C
r
C C
w
r
(A19)

For a flownet where the hydraulic potential at his outer boundary is C, and the inner and outer radii are defined as being,
respectively, r
i
and r
e
:

|
|
.
|

\
|
=
i
e
r
r
C
C
ln
1
(A20)

For the radial flownet, the boundary conditions are (Bilfinger, 2005):

at the outer boundary, r
e
:

0 =
r


at the internal boundary, r
i
:

l
i ri
ri
tE
r
u
2

= (A21)

where t = thickness of the lining and E
l
= elastic modulus of the lining.

( )( ) 2 1 1 + = A (A22)

( )
( )
( ) ( ) ( ) ( ) | |
( )
( )

(
(

|
|
.
|

\
|
|
.
|

\
|
(


+
(

+
+

(
(

|
|
.
|

\
|
|
.
|

\
|
+
(

|
.
|

\
|
+

=
A
r
E
r tE
r Er
tE
A
r
E
r tE
A r r r
E
r AtE
r
C
C
i i l
e i
l
i i l
e i i
i l
i
w
2
2
2
1
4
2 1
1
1
1 ln 1 ln
ln
1 2

(A23)

( )
( ) ( )A r
C
r
C C
e
w
e


+
|
|
.
|

\
|

|
|
.
|

\
|

= 1 ln
1 2
2 1
1
2
4 3
(A24)

You might also like