ADVANCED MATHEMATICS FOR ENGINEERS (CE 669)

CIVIL ENGINEERING DEPARTMENT, IIT GUWAHATI

Subject: Applications of Conformal Mapping

1 Seepage Analysis
In civil engineering, hydrogeology or soil mechanics, the flownet is an important tool to solve the two-dimensional
irrotational flow problems in presence of hydraulic structures like dams or sheet pile walls. Initially the construction
of the flownet was evolved as a graphical method considering some physical constraints of representing the flow lines
and the equipotential (piezometric) lines. Later analytical functions of complex variables, satisfying steady-state
flow condition by following the Laplace equation, were used to find some closed form solution for flow domains with
simple boundary conditions. But the process becomes extremely complicated when the flow domain is irregular in
shape or unknown altogether. In this scenario, methods of conformal mapping can be used successfully to transform
an unconfined flow domain (through an earth dam) or confined flow domain (underneath a concrete dam) into a
simple rectangular domain or a half-space plane where the closed form solutions can be obtained easily. Further, the
solutions can be mapped inversely to the original flow domain. The Laplace invariant property of conformal mapping
ensures the correctness of the solutions after inverse mapping from the simple domain to the original domain.

1.1 Earth dam

Let us define two spatial coordinates - a velocity potential function φ(x, y) and a stream function ψ(x, y), such that
they satisfy the governing Laplace equation of the steady-state flow, i.e., ∇2 φ(x, y) = 0 and ∇2 ψ(x, y) = 0. The
two functions φ(x, y) and ψ(x, y) are conjugate harmonic (i.e., they follow the Cauchy-Riemann equation), meaning
that the two families of curves given by φ(x, y) = constant and ψ(x, y) = constant represent mutually perpendicular
trajectories. Let us consider two complex plane z and w, respectively, such that

z = x + iy (1)
w = φ + iψ (2)

Now the Kozeny’s solution [1] for flow through the earth dams, as shown in Figure 1b can be obtained using the
following simple transformation

z = Cw2 (3)
⇒ x + iy = C(φ + iψ)2
⇒ x + iy = C(φ2 + i2φψ − ψ 2 )
∴ x = C(φ2 − ψ 2 ) and y = 2Cφψ (4)

where, C is a constant, depending on the boundary conditions of the problem. Now, along the free surface CDE
(commonly known as the phreatic line or line of seepage), ψ = q and φ = −ky, where k is the coefficient of hydraulic
conductivity (or the coefficient of conductivity) of the soil. Therefore, equation (4) yields
1
C=− (5)
2kq
The equation for the line of seepage, referred as Kozeny’s basic parabola can be obtained as:
 
1 1 1 q k
x=− (φ2 − ψ 2 ) = − (k 2 y 2 − q 2 ) = − y2 (6)
2kq 2kq 2 k q

It should be noted that the real upstream slope of the dam or face BCH (as shown in Figure 1a) is modified as a
parabola, face B’C’H’ in the z-plane (as shown in Figure 1b). This is in anticipation with the shape of the obtained
flownets in the z-plane (refer Figure 2), and correction should be made to account for this discrepancy [2]. The focal
distance x0 can be estimated by putting y = 0 at equation (6), i.e.,
q
x0 = ; ⇒ q = 2kx0 (7)
2k

1
 H td tdm
G H'
G
C Earth Dam C' Earth Dam
D ψ
Phreatic Line D' Phreatic Line C' ψ = q E
(ψ = q) (ψ = q)
hd

hd
hw

φ=0
hw

φ = -khw
z-Plane z-Plane
φ=0 φ=0 w-Plane
-y Filter -y Filter φ
φ = -khw φ = -khw
A x A B' ψ = 0 A
B x
E F B' E F
ψ=0 ψ=0
f f
x0 x0 (c)
bd bdm

(a) (b)

Figure 1: (a) Realistic geometric shape of an earth dam in z-plane, (b) modified geometric shape of an earth dam in
z-plane, (c) boundary conditions of flownets in w-plane.

Therefore, Kozeny’s basic parabola can be expressed as:

y2 y2
 
1
x= 2x0 − = x0 − (8)
2 2x0 4x0
Here, the value of x0 can be estimated by considering the coordinate of the point C(xC , yC ) from Figure 1a as
 q 
2 2
x0 = 0.5 xC + xC + yC ; xC = hw ; yC = −bd + f + 0.5hw (bd − td )/hd (9)

Finally, the complex transformation (of equation (3)) can be expressed by using equation (5) and (7) as

z = f (w) = −0.25w2 /k 2 x0 . (10)

The orientation of the flownets for an earth dam is given in Figure 2.

z = f(w) = -0.25w2/(k2x0) Dam outline

15

-07
=

e Filter
3.

6.7
0e

20 = Water level
-0

1
7

Flow lines
-07
.7e Equipotential lines
=5
= 2.6e-07 3
13 -07
15 .8e
=4
= 2.2e-07 5
11 -07
.8e
=3
Height (m)

= 1.7e-07 7
9 -07
10 .9e
=2
= 1.3e-07
-07
9
7
.9e
=1
= 8.7e-08 11
5 5 -08
.5e
=9
3
= 4.3e-08 13

=0
0 1
=0 15

-25 -20 -15 -10 -5 0 5 10

Width (m)

Figure 2: Flownets for a homogeneous isotropic earth dam

2
1.2 Flat-bottomed concrete gravity dam
In this section the realistic boundary conditions for flownets underneath a flat-bottomed concrete gravity dam (see
Figure 3a) are identified in the physical domain or z-plane and then accordingly considered in the w-plane as shown
in Figure 3b. Thereafter, the analytical expressions for the flownets will be solved in the w-plane and then a
transformation scheme will be used to map the solutions onto the physical plane. For this problem, the complex
velocity W [3], can be expressed as
dw ∂φ ∂ψ
W = = +i = vx − ivy (11)
dz ∂x ∂x
W was expressed in terms of z to obtain the transformation function
M
W = vx − ivy = p (12)
b21 − z2

where M is the real constant and 2b1 is the width of the dam. The main reason behind choosing the function is that
it follows the velocity conditions at the boundaries in the physical domain. It should be noted that W is real along
line BC (where −b1 < x < b1 ) and imaginary along lines A∞ B and CD∞ (where |x| > b1 ) according to equation
(12). This is consistent with the flow velocities being horizontal along BC (i.e. vx ̸= 0, vy = 0) and vertical along
A∞ B and CD∞ (i.e. vx = 0, vy ̸= 0). Now, from equations (11) and (12) we can write that
Z
M
dw = W dz ⇒ w = p dz ∴ w = φ + iψ = M sin−1 (z/b1 ) + N (13)
b21 − z 2

where N is constant of integration. The following boundary conditions are used to find the constants M and N .
Mπ
At point B: ψ = 0, φ = −k(h1 − h2 ) = −kh, w = −kh, z = −b1 ; ⇒− + N = −kh (14)
2
Mπ
At point C: ψ = 0, φ = 0, w = 0, z = b1 ; ⇒ +N =0 . (15)
2
Therefore, M = kh/π and N = −kh/2, and we can write that
      
kh z kh kh π z kh kh z
w= sin−1 − = − cos−1 − =− cos−1 ⇒ z = f (w) = b1 cos(πw/kh) (16)
π b1 2 π 2 b1 2 π b1

The orientation of the flownets for a flat-bottomed concrete gravity dam is given in Figure 4.

b2 = 1.5 m
-ψ
-y z-Plane
B ψ=0 C
a1 = 20 m
a2 = 18 m

Concrete Dam φ
h1 = 15 m

a3 = 4.5 m
φ = -kh

kh
φ=0

h2 = 2 m
w-Plane
x A D
A B C D
8

ψ=0
8

2b1 = 18 m (b)
φ = -kh φ=0
(a)

Figure 3: (a) Geometry of a flat-bottomed concrete gravity dam in z-plane, (b) boundary conditions of flownets in
w-plane.

3
 30
Dam outline
z = f(w) = b1cos( w/kh) Water level
Flow lines
20
Equipotential lines

= 3.0e-07

= 2.2e-07

= 1.5e-07
= 7.4e-08
10

17

13
=0

9
5
1 =0
0 17
Height (m)
1
= 4.6e-07

-10
15
= 5.7e-08
3
= 4.0e-07

-20
13
= 1.1e-07
5
= 3.4e-07
-30 = 1.7e-07
= 2.8e-07 11
7
9
= 2.3e-07
-40

-50 -40 -30 -20 -10 0 10 20 30 40 50

Width (m)

Figure 4: Flownets underneath a flat-bottomed concrete gravity dam

1.3 Flat-bottomed concrete gravity dam with sheet pile

Most often a hydraulic structure such as concrete gravity dams founded on permeable soil layer(s) are provided with
cut-off, like sheet piles to decrease the seepage forces and uplift forces resulting from the water flowing underneath
the structure. In these cases, the complicated flownets can not be transformed from w-plane onto the physical plane
directly by using some simple mathematical expressions of parabolas, ellipses or hyperbolas as shown in the previous
sections. To solve such problems, Pavlovsky considered an additional auxiliary complex variable space t-plane, upper
or lower half of which could be used to be mapped onto the z and w-planes by the following transformations

z = f1 (t) and w = f2 (t) (17)

∴ z= f1 (f2−1 (w)) = f (w) . (18)

Schwarz-Christoffel transformation technique is used to find the conformal mapping (i.e. the functions f1 (t) and
f2 (t)) of the dam profile in the z-plane and the complex seepage potential in the w plane from the lower or upper
half space of t-plane and finally the solved analytical flownets in the w-plane can be transferred onto the physical
plane by the function z = f (w).

1.3.1 Inclined sheet pile

Abbas Z. Ijam [4] provided the solution technique to find the flownets underneath a flat-bottomed concrete gravity
dam provided with an inclined sheetpile as shown in Figure 5a. The inclination of the pile with the vertical axis is
1 πθ
θ, so γ = − .
2 180
1. Transformation from t-plane onto z-plane: Points C and E are placed at -1 and +1, whereas the points B
and D are placed at tB and a, respectively (see Figures 5b) in the t-plane. The following Schwarz-Christoffel
transformation provides the required mapping
Z
z = M (1 + t)−γ (t − a)(1 − t)γ−1 dt + N (19)
dz
⇒ = M (1 + t)−γ (t − a)(1 − t)γ−1 (20)
dt
where M and N are constants. Since the domain in the z-plane is comprised of radial slit, the mapping of the

4
 b2
t-Plane
z-Plane
A B C D E F

8 8

8 8
Concrete Dam
- tB -1 a +1 +
a1

a3
a2

h2
h1
(b)
-ψ

x B C D E
A B ψ=0 C E F ψ=0 φ
8

φ = -kh
s
γπ
φ = -kh b1 φ=0 kh

φ=0
θ
w-Plane
y D Sheet Pile A F

8
(a) (c)

Figure 5: (a) Geometry of a flat-bottomed concrete gravity dam provided with an inclined sheetpile in z-plane, (b)
an auxiliary complex variable space t-plane, (c) boundary conditions of flownets in w-plane.

flow domain, according to Verigin’s transformation [5], is given by

z = c(1 + t)1−γ (1 − t)γ (21)

dz
⇒ = −c(1 + t)1−γ γ(1 − t)γ−1 + c(1 − γ)(1 + t)−γ (1 − t)γ
dt
dz
⇒ = −c(1 + t)−γ (1 − t)γ−1 [γ(1 + t) − (1 − γ)(1 − t)]
dt
dz
⇒ = −c(1 + t)−γ (1 − t)γ−1 (t + 2γ − 1) . (22)
dt
From equations (20) and (22), we can write that M = −c and a = 1 − 2γ. Now, putting t = a and z = se−iγπ
in equation (21) for point D we obtain

se−iγπ
c= (23)
(1 + a)1−γ (1 − a)γ
 1−γ  γ
−iγπ 1+t 1−t
∴ z = f1 (t) = se . (24)
1+a 1−a

Now, putting t = tB and z = −b1 in equation (24) for point B we obtain

1−γ 
 γ
1 + tB 1 − tB
−b1 = se−iγπ
1+a 1−a
 1−γ  γ
b1 tB + 1 tB − 1
⇒ − = [∵ e−iγπ = (−1)γ , a = 1 − 2γ] . (25)
s 2(1 − γ) 2γ

tB can be estimated from the above equation iteratively.

2. Transformation from t-plane onto w-plane: The required Schwarz-Christoffel transformation for mapping be-
tween the w-plane and the lower half of the t-plane is given by
Z
w = M1 (t − tB )−1/2 · (t − 1)−1/2 dt + N1 (26)
 
2t − tB − 1
⇒ w = iM1 sin−1 + N1 (27)
1 − tB

where M and N are constants of integration. The following boundary conditions are used to find the constants

5
 M and N .
iM1 π
At point E : t = 1 and w = 0; ⇒ + N1 = 0
2
iM1 π
At point B : t = tB and w = −kh; ⇒− + N1 = −kh
2
kh kh
∴ iM1 = and N1 = − . (28)
π 2
Therefore, we can write using equation (27):
 
kh 2t − tB − 1 kh
w= sin−1 −
π 1 − tB 2
1 n  πw  o
⇒ t = f2−1 (w) = (1 − tB ) cos + (1 + tB ) . (29)
2 kh

Finally, flownets in the z-plane can be obtained using equations (24) and (29), where the parameter tB is estimated
iteratively from equation (25).

1.3.2 Vertical sheet pile

The seepage analysis of the concrete gravity dam with vertical sheet pile (see Figure 6a) can be carried out as a
special case for concrete dam with inclined sheet pile, where the angle created by the sheet pile with the vertical
axis is zero. Hence, θ = 0, γ = 1/2 and a = 0. Therefore, the transformation between the physical z-plane and the
complex w-plane can be obtained using the following equations.
p
z = f1 (t) = s t2 − 1 (30)
1 n  πw  o
t = f2−1 (w) = (1 − tB ) cos + (1 + tB ) (31)
p 2 kh
b21 + s2
tB = − . (32)
s

b2 = 1.5 m
z-Plane t-Plane
A B C D E F
a2 = 18 m

8 8

8 8
a1 = 20 m

Concrete Dam
h1 = 15 m

- tB -1 a +1 +
a3 = 4.5 m (b)
h2 = 2 m -ψ
B C D E
x φ
A B C E F ψ=0
8

b1 = 18 m
s = 12 m

φ = -kh

φ = -kh kh
φ=0

Sheet Pile w-Plane

D A F
y
8

(c)
(a)

Figure 6: (a) Geometry of a flat-bottomed concrete gravity dam provided with a vertical sheetpile in z-plane, (b) an
auxiliary complex variable space t-plane, (c) boundary conditions of flownets in w-plane.

The orientation of the flownets for a flat-bottomed concrete dam with vertical sheet pile is given in Figure 7.

6
 Dam outline
Sheet pile
Water level
20 Flow lines
Equipotential lines

= 1.1375e-07

= 1.5925e-07
= 1.365e-07

= 2.275e-08

= 6.825e-08
= 1.82e-07

= 4.55e-08
= 9.1e-08
10

17

13

11

15
=0

7
1
0
Height (m)
1
= 4.55e-07 17
=0

-10
3
= 3.9813e-07 = 5.6875e-08
15

-20
5
= 3.4125e-07 = 1.1375e-07
13

7
= 2.8438e-07 = 1.7063e-07
-30 11
9
= 2.275e-07

z=s (t2-1); t = 0.5(1-t )cos( w/kh)+0.5(1+t ); t = - ((b /s)2+1)

B B B 1

-40 -30 -20 -10 0 10 20 30

Width (m)

Figure 7: Flownets underneath a flat-bottomed concrete gravity dam with vertical sheet pile

References
[1] J. Kozeny, “Grundwasserbewegung bei freiem spiegel, fluss-und kanalver-sicherung,” Wasserkraft und Wasser-
wirtschaft, vol. 26, no. 3, pp. 28–31, 1931.

[2] A. Casagrande, “Seepage through dams–in contributions to soil mechanics 1925–1940,” Boston Society of Civil
Engineers, Boston, Massachusetts, pp. 295–336, 1940.
[3] P. Y. Polubarinova and N. Kochina, Theory of ground water movement. Princeton University Press, Princeton,
NJ, 1962.

[4] A. Ijam, “Conformal analysis of seepage below a hydraulic structure with an inclined cutoff,” International
Journal for Numerical and Analytical Methods in Geomechanics, vol. 18, pp. 345–353, 1994.
[5] N. Verigin, “Seepage through the foundation of dams with oblique screens and cutoffs,” Gidrotek. Stroitel, no. 2,
1940.