Gradually Varied Flows: Mild & Steep Surface

Water Profiles
By Mohamed S. Ghidaoui
2007

Introduction
1. Uniform flow theory is often used to size artificial channels (e.g., sewers).
2. Uniform flow serves as the standard reference for experimental and theoretical work
(e.g. studies aimed at understanding dissipation and turbulent flow behavior in
channels; stability to gravitational waves etc).
However, it needs to be emphasized that uniform flow (UF) in open channels is the
exception rather than the rule. Factors that cause the flow to depart from being uniform
include (i) irregularities in cross-section, alignment, roughness and slope of natural
channels; (ii) man- made obstructions such as dams, bridge piers; (iii) control devices
such as gates; and (iii) unsteadiness of flow caused by dynamic control structures and/or
by time and spatial varying inputs and outputs such as runoff and infiltration. Even in the
laboratory it is difficult, if not impossible, to produce a truly uniform flow because the
length of the flume is often not sufficient to establish this flow regime.
The departure from uniform flow means that the flow velocity and depth vary from
location to location along the channel. If flow non-uniformity occur over a relatively
short distance, then the flow velocity and depth vary rapidly. That is, the rate of change
of depth and velocity with distance is very large. This type of non- uniform flow is
referred to as rapidly varying flow (RVF). Examples of RVF include hydraulic jumps and
hydraulic bores. When the flow is a RVF, skin friction effects can frequently be
neglected, and in many instances, solutions may be obtained by simultaneously
considering the mass conservation law and the momentum equation. It must be stressed
that momentum and energy are not equivalent concepts across hydraulic jumps and bores.
Indeed, across hydraulic jumps and bores, momentum is conserved but energy is not.
On the other hand, if the flow varies gradually, skin friction resistance gives rise to
surface profiles of considerable extent (sometimes called ``Backwater Curves'') whose
quantitative description requires the integration of either the equations of continuity and
momentum or continuity and energy. The fact that we can use the momentum or the
energy implies equivalency of these two concepts for Gradually Varied Flow (GVF)
problems. To be sure, most cases of non-uniform flow in open channels represent a
combination of both RVF and GVF.

Types of Surface Water Profiles

Recalling that the differential form of the energy equation is:

dE
d
V2
d
Q2
y +
= S 0 S f y +
= S0 S f
dx
dx
2g
dx
2 gA2

= S 0 S f

Or
dy Q 2 d 1
dy Q 2 d 1 dy
dy
+
+
= S0 S f
=
2 = S0 S f
2
dx 2 g dx A
dx 2 g dy A dx
dx

Or
dy
=
dx

S0 S f
1+

Q2 d 1

2 g dy A2

S0 S f
S0 S f
S S
S S
S S
=
= 0 2 f = 0 2 f = 0 2f
2
Q d 1
dA
Q T
Q T
V

1+
1 2
1 2
1
2
2 2A
3
Q
dy
2 g dy A
gA
A gA
c
1+
4

2g A

Therefore, the differential equation that governs 1-D steady flow in a channel is:
dy S0 S f
=
dx 1 F 2
According to the above relation, the flow profile y(x) and its slope depends on the
channel slope; frictional slope and the Froude number. The possible types profiles are
now investigated.

Mild Profiles (Denoted by M):

The profile is said to be mild if the uniform flow depth is larger than the critical depth.
That is, yo > y c implying that the uniform flow is subcritical.
y
yo

F<1

yc
F>1
E

Example: Water flows uniformly in a long and very wide river of width 72m towards a
lake. The river's cross sectional geometry is approximately rectangular. The flow rate is
50m 3 /s, the slope of the river is 0.000019 and its Manning roughness is 0.03. Is this flow
mild?
Solution:
The flow rate per unit width is: q=Q/B=50/72=0.694 m 2 /s. Therefore, the critical depth
is:
q2
y c =
g

1/ 3

1/3

0.694 2
=

9.81

= 0.366m

The uniform flow depth can be obtained from the Manning equation, namely,
Q=

A 2 / 3 1/ 2
R S0
n

The hydraulic radius for a wide (i.e., B>>y) rectangular channel is:
R=

A
By
=
=y
P B + 2y

Therefore, the uniform depth is:

Q=

A 2 / 3 1/ 2 A 2 / 3 1/ 2
50 y 0 2 / 3
1/2
R S0 y 0 S0 50 =
y0 (0.000019) y0 = 2.556m
n
n
0.03

This is a mild channel since yo =2.556m > y c = 0.366

Also note that the Froude number associated with uniform flow conditions is:
Fr0 =

V0
q / y0
0.694 / 2.556
=
=
= 0.0173
gy 0
gy 0
9.81 2.556

That is, the uniform flow is subcritical. Note the very small value of the Froude number;
this is typical of most rivers.
Example: Calculate the uniform flowrate in an earth- lined trapezoidal canal having
bottom width of 3 m, sides sloping 1 vertical to 2 horizontal, laid in a slope of 0.0001and
having a depth of 1.8 m. Is this a mild channel?
Solution:

1.8
2

3m

A 2 / 3 1/ 2
R S0
n
A = 1.8 * 3 + 2(1.8)( 3.6) * 1 / 2 = 11.8m2

Q=

P = 3 + 2 1.8 + 3.6 = 11.05

n = 0.075( earth)
2

2/ 3

11.8 11.8
0 .5
3
Q=

(0.0001) = 4.93m / s
0.025 11.05
The critical area and depth are:
Q 2T
4.932 (3 + 2 2 yc )
1

Ac =
3 y c + 2 y c 2 y c =
2
9 .8
g

This is a mild channel since yo > y c .

1 /3

1 /3

M1
Sf <So ; F < 1
M2

Sf >So ; F < 1
yo

M3
Sf >So ; F > 1

e.g.

yc

yc =

Profile?

yo
yc

M1-Profile:
y > yo > yc

Sf < So and F < 1

dh/dx > 0

M1

yo
yc

Note that the water depth of M1 profile is larger than the original uniform depth and
increases with distance. It is this increase in depth that forces the displacement of people
to higher elevation at the upstream side of dams. This M1 profile is also called backwater
curve.

M2-Profile:

Profile?
yo
yc

yo > y > yc

dh/dx < 0

Sf > So and F < 1

M2
yo
yc

M3-Profile:

yo
Profile?

yc

yo > yc > y

Sf > So and F > 1

dh/dx > 0

yo
yc

M3

Steep Profiles (Denoted by S):

The profile is said to be steep if the uniform flow depth is smaller than the critical depth.
That is, yo < yc.
Example: The gutter of a roadway is confined by a curb 0.15 m high. The cross-slope of
the road is 30 to 1, as shown in the diagram below, and the roadway itself has a
longitudinal slope of 1 to 100. A heavy rain creates a runoff discharge of 0.4 m 3 /s.
Assuming that n=0.014, is this a mild or steep channel.

0.15m

1 road center line

30

2
Q
0 .4
1
15 y0
2
V= =
=
0
.
0267
y
=
0
A 15 y0 2
n y + y 2 + 30 2 y 2
0
0
0

0 .5

2/3

(0.01)0.5

0.724 y 0

2/ 3

3/8

0.0267
y0 =
= 0.147m
4 .4
The critical area and depth are:

Q 2T
Ac =
g

1/ 3

15 y c

0.4 2 30 y c
=
9 .8

1/ 3

1/3

yc

5/ 3

0.4 2 30
=

9 .8

This is a steep channel since yo < y c .

y
F<1

yc
yo

F>1
E

/ 15 y c = 0.17m

S1
S2
yc
S3
yo

S1-Profile:
1

Profile?

2
3

y > yc > yo

Sf < Sc < So and F < 1

dh/dx > 0
S1

yc

dh/dx > 0

S2-Profile:

Profile?
yo
yc

yc > y > yo

Sf < So and F > 1

dh/dx < 0

M2
yo
yc
S2
Control point
yo

S3-Profile:
1
2
yc
3

Profile?

yc > yo >y
S1

yo

Sf > So and F > 1

yc
S3

yo

dh/dx > 0

S1

S1

HJ
HJ

Sf2 < Sc

Sf1 > Sc

S f > Sc

B1

B2 > B1

S2
Sf1 < 0

S2
Sf2 > Sc
S f > Sc

gate
S3

S3
S f > Sc

Sf1 > Sf2 > Sc

Sf2 > Sc

There are 7 other profiles (please see Appendix).

Examples: Sketch the possible flow profiles in the channels shown in the following
figures.
(a)

Mild

(b)

Milder

Steep

Steep

Horizontal

Mild

(c)

Mild
(d)

Steep
Legend:
Critical-depth line
Normal-depth line

Computation of Flow Profiles:

The computation of the flow profile amounts to solving (integrating) the following
governing equation of 1-D steady flow in a channel is:
dy S0 S f
=
dx 1 F 2
Recall that the above is equivalent to:
dE
d
V2
= S0 S f
= S 0 S f y +
dx
dx
2 g

Integrating the above between station 1 and station 2 who are a distance x apart (see
Figure below) gives:

V2 2

d
y
+
1 2g = 1 (S 0 S f )dx
2

Or
y1 +

V12
V2
+ S f x = y 2 + 2 + S0 x
2g
2g
2

where S 0 is the channel slope and S f = S f dx = average friction slope between station 1
1

and station 2. The simplest way to obtain the average is:

2

S f = S f dx
1

1
(S + S f 2 )
2 f1

Solving for x one gets:

( y y 2 ) + (V1 V2 ) / 2 g ( y1 y 2 ) + (V1 V2 ) / 2 g
x = 1

1
S0 S f
S 0 (S f 1 + S f 2 )
2
2

The local friction slope can be obtained from the Chezy equation, the Darcy-Weisbach
equation or the Manning equation. The Manning equation is often used for this purpose.
As a result,
nV
S f = 2/3
R
Therefore:

( y y 2 ) + (V1 V2 ) / 2 g
x = 1

S0 S f
2

( y1 y 2 ) + (V1 V2 ) / 2 g
2
2
1 n1V1 n 2V2

S0
2 R12 / 3 R2 2 /3

Gradually varied flow (GVF) and velocity profile

The computation of the flow profile using Direct Step Method proceeds as follows.
1. Identify the control point (i.e., the mechanism responsible for establishing the flow
profile such as a dam, a gate etc). Mathematically, this is called boundary condition.
2. The computation proceed in the upstream direction of the control point if the flow is
subcritical. On the other hand, the computation proceed in the downstream direction
of the control point if the flow is supercrictical.
3. Let i be a counter that identifies a section whose coordinate is x i . Therefore,
y i , Vi , ni , Ri , Ai , S fi , Ei , Pi are the water depth, velocity, Manning coefficient,
hydraulic radius, area, frictional slope, specific energy and wetted perimeter
respectively.
4. The recursive formula to solve for the flow profile is:
2
2
( y i+1 y i ) + (Vi +1 Vi ) / 2 g
x = xi +1 xi
2
2
1 n iVi ni+1Vi+1

S0
2 Ri 2 / 3 Ri+12 / 3

5. Assuming that the Manning coefficient is constant and let i=1 at the control point.
The flow rate Q is known and y1 ,V1 , R1, A1, S f 1 , E1 , P1 at the control point are also
known. Now, assume a value for the water depth at i=2. The assumed value,
y 2 needs to be consistent with the flow profile. Using your assumed value of y1 and
the fact that you know the flowrate Q determine V2 , R2 , A2 , S f 2 , E2 , P2 . Now, the

distance between station 0 and station 1 can be computed as follows:

2
2
( y 2 y1 ) + (V2 V1 ) / 2 g
E2 E1
x 2 x2
=
2
2
2
2
1 nV1 nV2
1 nV1 nV2

S0
S 0 2 / 3 + 2 /3
2 R2 2 / 3 R2 2 / 3
2 R2 R2

6. Repeat step 5 for other values of i.

7. At the end of your calculation you have the following pairs: ( y1 , x1 ) ; ( y2 , x2 ) ,
( y3 , x 3 ) etc. Using these pairs, one can obtain the flow profile by potting y versus x.

The computation just described is best performed in a tabular form as follows:

i

yi Ai

Pi Ri 2 / 3

Vi

Vi 2 / 2 g S fi

( S fi + S fi+1 ) / 2 Ei +1 E i X i+1 X i X i

1
Control
Point
2

Example: Water flows uniformly in a long and very wide river of width 72m towards a
lake as shown in the figure below. In order to raise the water level of the lake by 0.6m, a
dam will be constructed. The location of the dam is as indicated in the figure. The flow
rate in the river is 50 m3 /s. This river has a slope of 0.000019 and a roughness, n, of 0.03.
Estimate the increase in depth at a station that is approximately 30 km upstream of the
lake.
Solution:
Final level with dam
30km

river

50m3 /s

0.6m
Initial level with no dam
Lake

So=0.000019
N=0.03

Dam

The flow rate per unit width is: q=Q/B=50/72=0.694 m 2 /s. Therefore, the critical depth
is:
q2
y c =
g

1/ 3

1/3

0.694 2
=

9.81

= 0.366m

The uniform flow depth can be obtained from the Manning equation as follows.
As a result,
A
1/ 2
Q = R 2 / 3 S0
n
The hydraulic radius for a wide (i.e., B>>y) rectangular channel is:
A
By
=
=y
P B + 2y
Therefore, the uniform depth is:
R=

Q=

A 2 / 3 1/ 2 A 2 / 3 1/ 2
50 y 0 2 / 3
1/2
R S0 y 0 S0 50 =
y0 (0.000019) y0 = 2.556m
n
n
0.03

This is a mild channel since yo =2.556m > y c = 0.366. The flow profile is an M1 type (i.e.,
backwater curve).
The control point is station 1 (see figure below). This control point (CP) is established
because of the presence of the dam.

0.6m
2.556m

1
The water depth at the control point is the original uniform depth plus the increment in
depth induced by the dam. That is:
y1 =2.556+0.6=3.156m

3.156m

M1 profile

0.6m
yo =2.556m
ycr=0.366m

river

x (k
m)

V=q/y V /2g E

SE

SE

1
CP

3.156

227.232

78.312

2.0344

0.2199

0.00246

3.15846

0.0000105

-0.0559

0.000010
8

-6.82

-6.82

5.1

223.2

78.2

2.0121

0.2239

0.00256

3.10256

0.0000111
-0.0983

0.000011
75

-13.77

-20.59

-0.04991

0.000012
75

-7.99

-28.58

-0.00998

0.000013
15

-1.71

30.29

3.0

2.95

2.94

216

212.4

211.68

78

77.9

77.88

2/3

1.9721

1.9517

1.9476

0.2313

0.2353

0.2361

0.00273

0.00282

0.00284

3.00273

2.95282

2.94284

0.0000124

0.0000131

0.0000132

At distance x 30.29 Km upstream of the control point, the water depth is 2.94m.
The negative sign means that station 2, 3 etc are upstream of the CP.
Assume that the depth is changing linearly from region i=4 to i=5, the depth at 30 Km
upstream of the CP is:
2.95-0.01*(30-28.58)/(30.29-28.58)=2.942m.

(km)