Missouri University of Science and Technology

Scholars' Mine

International Conference on Case Histories in (2004) - Fifth International Conference on Case

Geotechnical Engineering Histories in Geotechnical Engineering

15 Apr 2004, 1:00pm - 2:45pm

Estimation of Phreatic Line Using Dimensional Analysis

D. R. Phatak
Government College of Engineering, Pune, Maharashtra State, India

S. R. Pathak
Government College of Engineering, Pune, Maharashtra State, India

K. C. Birid
Geotechnical Engineer, Mumbai, Maharashtra State, India

Follow this and additional works at: https://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons

Recommended Citation
Phatak, D. R.; Pathak, S. R.; and Birid, K. C., "Estimation of Phreatic Line Using Dimensional Analysis"
(2004). International Conference on Case Histories in Geotechnical Engineering. 43.
https://scholarsmine.mst.edu/icchge/5icchge/session02/43

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been
accepted for inclusion in International Conference on Case Histories in Geotechnical Engineering by an authorized
administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including
reproduction for redistribution requires the permission of the copyright holder. For more information, please
contact [email protected].
 ESTIMATION OF PHREATIC LINE USING DIMENSIONAL ANALYSIS
D. R. Phatak (Mrs.) S. R. Pathak K. C. Birid
Faculty of Civil Engrg. Dept. (Retd.) Faculty of Civil Engrg. Dept. Geotechnical Engineer
Govt. Engrg. College Govt. Engrg. College Mumbai, Maharashtra State, India
Pune, Maharashtra State, India Pune, Maharashtra State, India

ABSTRACT

A new technique using dimensional analysis (D.A.) is presented here to draw profile of phreatic line through an earthen dam. A
universal equation is formulated using D.A. to get numerous points on a steady-state phreatic surface. The prediction of phreatic line
using D.A. is then compared with those of conventional methods by Kozeny, A. Casagrande, Stello and also with centrifuge model test
results and large scale prototype field results to evaluate applicability of this equation. It has been observed that D.A. can predict the
results reasonably well. Thus Dimensional Analysis method can prove to be an easy and sufficiently accurate method to predict
solutions to complex and multiparameter problems.

INTRODUCTION

Butterfield (1999) has explained the procedure for application In case of homogeneous and non-homogeneous earthen dam,
of D.A. to civil engineering problems. Taking a clue from this the variables used in the analysis are explained below with
work authors applied D.A. method to locate phreatic surface respect to figure (1).
through homogeneous as well as zoned earthen dam. Various
methods, graphical and analytical, developed by Kozeny, A. L
Casagrande, L. Casagrande, Van-Iterson, Dupuit, Pavlovsky, 0.3 L
Stello etc. are available in literature to estimate the phreatic
surface through an earthen dam. A trial and error procedure is
H
also used to draw flow nets. This method is simple and useful Y
a
in many types of application. However, it is cumbersome for θ φ
problems involving unconfined flows in inhomogeneous soil X
(Chang 1988). The D.A. equation to draw phreatic line B
D
eliminates such trial and error method and graphical method.
The phreatic line as determined by D.A. equation is then
compared with the existing methods using various geometrical
combinations, laboratory model test results and field
observations. α
H
Y
θ γ
BASIC FORMULATION OF D.A.
X
B
The aim of D.A. is to reduce to a minimum the dimension
space in which the behavior of a specific system might be
studied by combining and arranging systematically the Fig. 1 Dam sections (Homogeneous and Non-Homogeneous)
assumed governing variables (v) = (v1, v2, v3,……,vn)
encompassing total of m independent primary dimensions Here H= height of water level (in meter)
(D)= (D1, D2, D3,….., DN) into N= (n-m) D.G.s, that are (π1, B= bottom width of dam section in case of homogeneous dam
π2, π3,……, πN), N being less than v. Primary dimensions are and bottom width of core section in case of non-homogeneous
expressed in mass (M), length (L), time (T), etc. dam section (in meter)

Paper No. 2.76 1

θ= Angle of upstream sloping portion in case of homogeneous dam section having H=90m, B=220m, θ=450, K=5, X=330m,
dam section and angle of upstream face of core in case of non- and Y=16m. The values of X and Y are selected by taking
homogeneous dam section (in degrees) guidance from the Stello’s method.
X = horizontal distance (in meter) Substituting these values in equation (4) we get
Y = corresponding vertical distance of point on phreatic line.
0.0727 = [ β1 (0.4090 β2 ) (45 β3) (1.5 β4) ] π5 (5)
Here we have v = { H, B, θ, X, Y} hence n = 5
Writing the dimensions of the variables In logarithmic form
0 0 0 0 0 0 0 0
{v} = [M LT , M LT , 0, M LT , M LT ], m = 1, i.e. no. of -2.6214 = π5 [logβ1 - 0.8940β2 + 3.8066β3 + 0.4054β4] (6)
repeating variables forming a set Q, which can not be reduced β5 β5
where π5 = K =5
further.
Thus no. of dimensionless Pi – group is Solving equation (5) and (6) by trial and error procedure we
N = (n-m) = (5-1) = 4 get
now R = set of variables in v which have dimensions totally β1 = 6.0538, β2 = 1.4, β3 = -0.55, β4 = -0.655,and β5 = 0.2301
distinct from each other.
Substituting above values in eq. (4), we get
Here in v, the dimensions of all the variables are same hence;
any one variable can be selected as R. Y = B [6.0538 H1.4 θ –0.55 B –0.745 X –0.655] π5 (7)
Therefore R = {B} 0.2301
where π5 = K
Q is to be selected from R, therefore Q = {B}
Isolated variables are {H, θ, X, Y} Using the equation (7) a reasonably accurate top seepage line
Therefore the dimensionless groups are can be plotted for homogeneous as well as non-homogeneous
dam sections. The steps to be followed while drawing the
π1 = {B, H} π2 = {B, θ}
seepage line using D.A. equation are explained as follows.
π3 = {B, X} π4 = {B, Y}
Therefore, represent
PHREATIC LINE FOR HOMOGENEOUS DAM SECTION
π1 = (Ba, Hb) (1)
In case of homogeneous dam section, following steps shall be
Therefore in dimensional form
followed while drawing the phreatic line using D.A. equation.
[ M0L0T0] = [MLT] a x [ MLT]b Step 1
Start drawing the line using D.A. equation (7) from point of
Hence comparing the indices of LHS and RHS, for M, L, & T,
intersection of U/S slope and U/S water level (say A),
Now b = 1, then a = -1
considering K=1.
Substituting the values in equation (1)
Step 2
π1 = (B-1 x H) = H/B Draw the line until it meets the D/S slope of the section.
Step 3
Similarly π2 = θ
Apply the correction at exit point as suggested by Van-Iterson,
π3 = X/B i.e.
π4 = Y/B
D D2 H2
a=
Now, let π4 = φ (π1, π2, π3), where φ is an unknown function. cos φ cos 2 φ sin2 φ

Therefore, we can write

π4 = β1 (π1) β2 (π2) β3 (π3) β 4 (2) where a, D, H and φ are as shown in fig. 1
Step 4
Here a separate π term i.e. π5 has been introduced and raised
At point A, phreatic line is improved & oriented to meet the
to the power of above equation to take into consideration the
physical requirements such that it meets the point of
difference in the permeability of core and shell section in case
intersection of water level with U/S slope of dam section; i.e.
of non-homogeneous dam section. Thus
the starting point of the phreatic line.
π5 = K β5
where K = Ks/Kc = (permeability of shell material) / PHREATIC LINE FOR ZONED SECTION
(permeability of core material)
Therefore, equation (2) becomes, Figure 2 shows non-homogeneous dam section with H=90m,
β2 β3 β4 π5 B=220m, K=20, U/S slope of core (θ)= 1:1 i.e. 450, D/S slope
π4 = [ β1 (π1) (π2) (π3) ] (3)
of core = 1:1, D/S slope of shell = 2.5:1.
β2 β3 β4 π5
Y/B = [ β1 (H/B) (θ) (X/B) ] (4) Step 1
In order to draw the phreatic line for this section start from
To find the values of constants β1, β2, β3, β4 and β5 a guiding
point A (fig. 2). Calculate various values of Y for different
point has been selected from a fictitious Non-homogeneous

Paper No. 2.76 2

values of X and taking K =1, draw the line until it meets the Table 1. Homogeneous dam sections
point B; which is the intersection point of seepage line with
D/S slope of core. Dam Water
Fig. U/S D/S Base
Step 2 Ht. level H/HD
To draw the further profile in shell portion take K=20 and start No. slope slope width
(HD) (H)
drawing seepage line from the D/S part of shell portion from
any point say F on the phreatic line, which is found out using 3a 2:1 2:1 50 30 216 0.60
D.A. equation for any value of X within the shell portion. 3b 2:1 1.5:1 60 25 229 0.41
Draw the line for various values of X towards the boundary of 3c 2.5:1 2.5:1 70 55 372 0.78
shell and core CC’ until it meets the boundary CC’ at point E. 3d 3:1 3:1 70 60 442 0.85
3e 3.5:1 3.5:1 80 75 585 0.93
4 2:1 2:1 23 17.25 95 0.75
100 C 5 2:1 3.3:1 20.33 11.67 126 0.57
6 3:1 2.5:1 58 50 332 0.86
VERTICAL DISTANCE(Y) ,m

80 A
60 B Table 2. Non-Homogeneous dam sections
40

20 U/S D/S D/S

F Fig. Dam Water Base
E D slope slope slope K
0 No. Ht. level width
0 50 100 150 200 C' 250 300 350 400 (core) (core) (shell)
HORIZONTALDISTANCE(X),m

7a 1:1 1:1 2.5:1 100 90 220 20

7b 1:1 1:1 2.5:1 100 90 220 5
Fig. 2 Phreatic line using D.A. for Non-Homogeneous section
7c 1:1 1:1 2.5:1 100 50 220 5
7d 0.7:1 0.5:1 3.0:1 30 20 41 5
Step 3
7e 0.7:1 0.5:1 3.0:1 30 20 41 15
There are some rules for entrance and exit condition of
8 0.6:1 0.6:1 2.5:1 47.35 39.93 60.62 CD
seepage line along the boundary CC’. Measure the angle of
* All dimensions are in meter
intersection of line in the core section with boundary CC’ at CD = Chimney Drain
point B which is α (fig. 1). Use equation tan α / tan γ = K.
Knowing all the parameters except γ, determine γ which is the
angle of intersection of line in the shell portion with the 60

boundary CC’ or the exit angle as shown in fig. 1. 50

Kozeny
VERTICAL DISTANCE (Y),m

A.Casagrande
Step 4
Stello
Draw the straight line from the point B at an angle equal to γ
40

D .A .
with the boundary CC’ and join this line by a smooth curve to 30

the already drawn seepage line EF. 20

Step 5
Completely draw the seepage line near D/S part of shell, 10

starting from point F until it intersects D/S slope of core at 0

point D.
0 50 100 150 200
HO R IZO NTAL DISTANC E (X),m

Step 6
At point A, phreatic line is improved & oriented to meet the a)
physical requirements such that it meets the point of
intersection of water level with U/S slope of core; i.e. the 70
starting point of the phreatic line. 60
Kozeny
VERTICAL DISTANCE (Y), m

A .Casagrande
50
Stello

COMPARISON 40 D .A .

30

The phreatic line drawn by D.A. for homogeneous and non- 20

homogeneous dam sections is compared with conventional 10

methods (Kozeny, A. Casagrande, & Stello) and also by 0

Centrifuge model test results (Sutherland et al. 1984) and case 0 50 100 150 200 250
study. In table 1and 2, the geometrical parameters of various HORIZONTAL DISTANCE (X), m

dam sections used for comparison are shown. In fig. 3a-e, the
seepage line for homogeneous section drawn using D.A. b)
equation is compared with other conventional methods.

Paper No. 2.76 3

 80

Kozeny
Jumbo Dam
Seepage Line
VERTICAL DISTANCE(Y),m

70
Colorado
A .Casagrande
60 Reconstructed
Stello section
50
D .A .
40

30

20

10

0
0 50 100 150 200 250 300 350
a)
HORIZONTAL DISTANCE(X),m

30

VERTICAL DISTANCE (Y),m

c) 25
Measured
80
20
Predicted (DA)
Kozeny
VERTICAL DISTANCE(Y),,m

70
A .Casagrande 15
60
Stello
10
50
D .A .
40 5

30
0
0 20 40 60 80 100 120
20
HORIZONTAL DISTANCE (X), m
10

0
0 50 100 150 200 250 300 350 400
b)
HORIZONTAL DISTANCE(X),m

Fig. 5 Comparison with Jumbo dam, Colorado

d)
90

80 Kozeny 80
Saluda Dam
VERTICAL DISTANCE(Y), m

70
A .Casagrande 70
South Carolina
60
Stello
60
50

50
D .A .
40 Semi-Hydraulic Fill
40 30

20
30
Seepage Line
20
10
T ailwater
0

10

0
0 100 200 300
HORIZONTAL DISTANCE(X),m
400 500
a)

e) 80
VERTICAL DISTANCE (Y), m

Fig. 3 Homogeneous dam sections 70 Measured

60
Predicted (DA)
Figure 4, 5 and 6 shows the comparison of D.A. with 50

centrifuge model test results and actual phreatic line for Jumbo 40

dam and Saluda dam respectively. The base of the dam section 30
20
in fig. 5b and 6b is joined by a straight line for the
10
convenience. 0
0 50 100 150 200 250 300
HORIZONTALDISTANCE(X),m

30

b)
VERTICAL DISTANCE (Y), m

25 Centrifuge Model

D. A.
20 Fig. 6 Comparison with Saluda dam, South Carolina
15

10
Figure 7a-e shows the comparison of seepage lines for non-
5 homogeneous dam sections. For zoned section, in order to
account for the effect of different soil permeability in core
0
0 10 20 30 40 50 60 70 80 90 (Kc) & shell portion (Ks), the phreatic line has been drawn for
HORIZONTAL DISTANCE (X),m
different values of K ranging from 5 to 20. Since for K = 20,
Fig. 4 Comparison of D.A. with Centrifuge model test results the top seepage line by all methods closely follows the base of

Paper No. 2.76 4

 the dam, the authors feel that there is no need to consider K > 35

20. Similarly for K < 5 there is not much difference in the Kozeny
Kozeny

VERTICAL DISTANCE(Y),m
30

profile & the dam section is as good as homogeneous section

A .AC.aCasagrande
sagrande
25
S te llo
(Bharatsingh & Sharma,1976) 20
Stello
D . A.
D .A .
15

10

100 Kozeny 5
VERTICAL DISTANCE(Y),m

A.Casagrande 0
80 0 20 40 60 80 100 120
Stello HORIZONTAL DISTANCE(x),m
60
D .A .

40

d)
20

0
0 50 100 150 200 250 300 350 400
35
HO RIZO NTA L DIS TA NCE (X),m
Kozeny

VERTICAL DISTANCE(Y), m
30

A .Casagrande
a) 25

Stello
20

D .A .
15
100
VERTICAL DISTANCE(Y),m,

Kozeny 10

80 A.Casagrande
5

Stello
60
0
0 20 40 60 80 100 120
D .A . HORIZONTAL DISTANCE(X),m
40

20 e)
0
0 50 100 150 200 250 300 350 400
Fig. 7 Non-Homogeneous dam sections
HORIZONTAL DISTANCE (X),m

b) 60
VERTICAL DISTANCE (Y), m

50 Measured

100 40
Predicted (DA)
VERTICAL DISTANCE(Y), m

Kozeny
30
80 A.Casagrande

Stello 20

60
D .A . 10

40
0
-20 0 20 40 60 80 100 120 140

20 HORIZONTAL DISTANCE (X), m

0
0 50 100 150 200 250 300 350 400
HORIZONTAL DISTANCE(X),m
Fig. 8 Comparison with Camanche dam
c)
verified. The dam sections and other parameters used for the
comparison of zoned section have been summarized in table 2.
A different conjectured dam section is selected as shown in
fig. 7d with H=20m, B=41m, K=5, U/S slope of core (θ) =
0.7:1 i.e. 550, D/S slope of core = 0.5:1, D/S slope of shell = CONCLUSION
3.5:1. The height of dam section here is only 30m compare to
100m of previous sections. The comparison of profile drawn The comparison (Fig. 3a-e) reveals that the D.A. equation can
by D.A. with the other methods indicates applicability of D.A. be applied effectively to draw phreatic line for geometrically
for geometrically different sections also. different homogeneous dam section. The phreatic line drawn
using empirical correction (Van-Iterson correction) closely
Thus with such variation in H, K, B, & θ, the location of follows with the Centrifuge model test results (with
phreatic line for the zoned section using D.A. has been configuration A and 90 g scaling) and actual field results of
Jumbo dam and Saluda dam (Creager et al. 1944)

Paper No. 2.76 5

The comparison for non-homogeneous sections shows that the
same D.A. equation can be effectively used for dam section
with zones of different permeabilities. The line thus drawn
usind D.A. equation is sensitive to variation in the soil
permeability of different sections.

The time required to draw the phreatic line using above D.A.
equation is very less and is less tedious compared to other
conventional graphical methods.

The incorporation of various influencing parameters in the

D.A. analysis can be done without much effort.

It is thus possible to formulate a D.A. equation for other

problems with a prerequisite that the various parameters
involved in the problem should be known either
experimentally or assumed using conventional analysis, along
with their inter relationship with each other.

REFERENCES

Balasubramaniam, A.S., Tomiolo, A., Yudhbir, and Younger,

J.S. [1980] “Geotechnical problems and practice of dam
engineering”, Proceedings of the international symposium
held at the Asian Institute of Technology, Bangkok, A.A.
Balkema publishers, Rotterdam-1982, pp. 57

Bharatsingh, and Sharma, H.D. [1976] “Earth and Rockfill

Dams” Sarita prakashan, Meerut, pp. 16-65.

Butterfield, R. [1999] “Dimensional analysis for geotechnical

engineers”, Geotechnique, vol.49, No. 3, pp. 357-366.

Cedergren, H.R. [1967] “Seepage, drainage, and flownets”,

John Wiley and Sons, NY, pp. 489.

Chang, C.S. [1988] “Boundary-element analysis for

unconfined seepage problems”, Journal of Geotechnical
Engineering, ASCE, Vol.114, No.5, pp. 556-572.

Creager, W.P.(late), Justin, J.D.(late) and Hinds, J. [1944]

“Engineering for Dams, Vol. III-Earth, Rock-fill, Steel and
Timber Dams”, John Wiley and Sons, NY, Chapman & Hall
Ltd., London, pp. 678, 680.

Stello, M.W. [1987] “Seepage charts for homogeneous and

zoned embankments”, Journal of Geotechnical Engineering,
ASCE, Vol.113, No.9, pp. 996-1012.

Sutherland, H.J., and Rechard, R.P. [1984] “Centrifuge

simulations of stable tailings dam”, Journal of Geotechnical
Engineering, ASCE, Vol.110, No.3, pp. 390-402

Paper No. 2.76 6