OPEN CHANNEL FLOW

Any fluid flow that is bounded by the sides and the bottom but free on the surface and subjected to a
constant (usually atmospheric) pressure is referred to as open channel
Unlike in pipe flow, where the pressure difference between two points within the pipe is the driving
force of flow, in open channel flow, gravity force is the main driving force of flow and therefore
elevation changes (drop) in the direction of flow is very essential.

As in highway drains
As in urban storm drains As in hydrotechnical pipes

As in river or stream flows

As in sanitary sewers
Examples of open channel flows

Since open channel flow is characterized by a constant pressure on the free surface, then the
pressure distribution with depth follows the hydrostatic law.
Depending upon the shape, open channels can be either prismatic or non-prismatic.
A prismatic channel is one in which the cross-sectional area along the direction of flow remains
constant. This implies that along the direction of flow, the area is only dependent on the depth of
flow. In non-prismatic channel, the cross-sectional area changes with distance along the direction of
flow. In this case the area is a function of depth and distance along the flow.
Depending upon the form, an open channel is either exponential or non-exponential. A channel is
exponential when its area of cross-section may be expressed in the form:

A = khm --------------------(OC-1)

Where A = area of cross-section, k = constant; d = depth of flow and m = exponent.

Rectangular, parabolic and triangular channels are exponential, whereas trapezoidal and circular
channels are non-exponential.
Classification of flow in open channel

Below is a classification attributed to Ven Te Chow taken from (Fluid Mechanics – A. K. Jain,
2003).

1 Dr. F. O. K. Anyemedu
 Open channel flow

Steady flow Unsteady flow

Uniform flow Non uniform or Uniform flow Non-uniform flow

varied flow or varied flow

Gradually Rapidly
Gradually varied flow Rapidly varied varied flow varied flow
flow

HYDRAULIC (GEOMETRIC) ELEMENTS OF CROSS-SECTION

Depth of flow (h): it is the depth normal to the direction of flow

Cross-sectional area (A): it is the area normal to the direction of flow
Wetted Perimeter (P): It is the length of the cross-section wetted by the flowing fluid.
Hydraulic Radius (R): by definition, it is the ratio of the cross-sectional area to the wetted
perimeter.
Top Width (B): It is the width of the channel at the free surface
Hydraulic Mean Depth (D): By definition is the ratio of the cross sectional area to the top width

Hydraulic elements of some common open channels

2 Dr. F. O. K. Anyemedu
 B B
B

h 1 h m m h

m
b b

B = b + 2mh B=b B = 2mh

A = (b+mh)h A = bxh A = mh2

P=b+2h(1+m2)1/2 P = b +2h P = 2h(1 +m2)1/2

UNIFORM FLOW IN OPEN CHANNELS

Consider a uniform flow in a prismatic channel shown below. The channel has a bed slope of So
Since the flow is uniform, it implies that the depth in section (1) and section (2) are the same and
therefore the slope of the water surface S must be equal to the slope of the bed. Also since flow is
uniform, then the velocity v1 must be equal to the velocity v2 and hence the velocity heads at section
(1) and section (2) must also be the same. This means the slope of the water surface is equal to the
slope of the energy line. Therefore by the condition of uniformity of flow in the channel, we have:

So = S= SE ----------------------------(OC-2)

Now let us consider the flow of the mass of fluid within the sections 1 and 2 with cross –sectional
area of A. Since the flow is uniform, then according to Newton’s second law the sum of forces on
the fluid mass in the direction of flow must be equal to zero.

3 Dr. F. O. K. Anyemedu
 (2)
(1)

Slope SE Energy line

V12/2g
Slope S Liquid surface
V22/2g

h1
Slope So = sin θ h2

Z1 θ L Z2
Channel bottom
O O

Arbitrary datum O –O

Let us find all the forces acting on the fluid mass which include i) the weight of fluid, ii) the
frictional force acting on the wetted perimeter and iii) the pressure forces
However since the fluid is subjected to a constant atmospheric pressure and h1 =h2 then the pressure
forces resolve in the direction of flow to be zero.
Now, resolving all forces in the direction of flow, we shall obtain:

γALsin θ - τoPL = 0 ------------------------(OC-3)

where A – cross-sectional area; τo = average shear stress on the wetted perimeter;

P = wetted perimeter and L = length of section.
The average shear stress is then given by:

τo = γ (A/P) L sin θ = γRS ----------------(OC -4)

At this stage, we make the assumption in respect to the flow resistance, according to which the
force resisting the flow (τoPL) per unit wetted area is proportional to the square of the velocity head.
i.e τo = K (v2/2g) where K = const.
The above assumption can be put in this form:

τo/γ = (f/4)x(v2/2g) ------------------------(OC-5)

where K = γ.f/4; f is the frictional resistance factor.

Substituting eq. (OC-5) into eq. (OC-4) and making velocity the subject, we shall obtain:

-----------------(OC-6)
where C = (8g/f)1/2 referred to as the Chezy’s coefficient

If we represent K =AC(R)1/2 -------conveyance of channel

4 Dr. F. O. K. Anyemedu
Then,

Equation (OC-6) represents the Chezy’s formula; a formula developed by Antoine Chezy, a French
engineer in 1770 for uniform flow.
The Chezy’s formula is simple and easy to use; however the main difficulty with it is the correct
determination of the constant C in the formula, which is generally obtained by empirical means.
Fortunately, a lot of work has been done on the constant C and there are recommendations for
determining the value of C. Below are a few of the available recommended formulae.

1) Ganguillet and Kutter’s formula

In 1819, two \Swiss engineers, based on extensive experimental studies on European rivers
proposed the following empirical formula to determine the Chezy’s constant.

-------------------(OC-7)

2. Manning’s Formula
Robert Manning, an Irish engineer in 1889 presented a formula for velocity of flow in open channel
from which the Chezy’s constant could be defined as:

C = 1/nR1/6 ---------------------------(OC -8)

3. Bazin’s formula

The French Hydraulic engineer proposed the following formula for the Chezy’s constant in 1897.

-----------------------(OC -9)

where k = m/1.81. The values of m is given in a table below

n = Manning’s roughness coefficient given for various surfaces in a table beloew.

Table X. Manning’s Roughness Coefficient

No. Description of channel surface Manning’s Roughness

coefficient, n
1 Well planed timber, glass or brass 0.009
2 Wood-shaved flumes, finished concrete 0.010- 0.013
3 Glazed tiles, vitrified sewers, concrete pipes 0.013-0.016
4 Bricks in cement mortar, ashlars 0.012- 0.20
5 Rubble masonry in cement 0.020 – 0.030
6 Straight unlined canals in earth 0.020
7 Unlined canals in gravel, unlined canals in earth 0.225

5 Dr. F. O. K. Anyemedu
 with some curves
8 Corrugated metal flumes, unlined canal winding 0.025
9 Canals with rough stony bed or weeds 0.030
10 Winding natural streams in good conditions 0.035
11 Rivers of irregular cross –section winding 0.04 to 0.10
Cast iron 0.013 – 0.017
Riverted steel 0.017 – 0.020

Table XX. Bazin’s Coefficient

No. Description of Channel Surface M

1 Very smooth cement or planned wood 0.11
2 Unplanned wood, concrete or bricks 0.21
3 Planks, ashlars 0.29
4 Rubble masonry 0.83
5 Earth channels in perfect condition 1.54
6 Earth channel in average condition 2.36
7 Earth channel in rough condition 3.17

HYDRAULIC MOST EFFICIENT CROSS-SECTION

Let us assume that we are to design a channel (i.e. to determine the dimensions) with the following
initial parameters given:
i) form of the cross-section or channel= trapezoidal
ii) the longitudinal slope S = So
iii) Side slope of channel m = mo
iv) The roughnest coefficient n = no and
v) Flow rate Q = Qo
The above problem has so many solutions; in fact an infinite number of solutions

6 Dr. F. O. K. Anyemedu
 h1 m1
β1=b1/h1
b1

m1=m2=m3= --=mo
h2 m2 β2=b2/h2
i1=i2=i3----=io
b2
n1=n2=n3=---no

h3 β3=b3/h3
Q1=Q2=Q3=-----Qo m3

b3

In the diagram above, only three of such solutions are given and it is clear that the first solution is
characterized by wide channel and shallow depth while the last is characterized by small base width
and deep depth.
We see that the first and third solutions are characterized by relatively high friction, defined by the
magnitude of the wetted perimeter and hence their velocities will consequently be small.
β1 ≠ β2 ≠ β3 and
P1 ≠ P2 ≠ P3
Where β = relative width
From the above discussion, it can be seen that, among the infinite number of solutions available,
there must be one which has the least wetted perimeter and cross-sectional area and hence the
maximum velocity of flow and least resistance to flow.

Hydraulic most efficient cross-section: is that cross-section which for the given parameters Qo,,
mo, no, io will characterize the maximum possible velocity of flow and hence the minimum wetted
perimeter and minimum cross-sectional area or a channel that gives the maximum discharge for a
given shape, slope area and roughness. Such a channel offers the least resistance to flow.

7 Dr. F. O. K. Anyemedu
 P, Q, V
P, A, V A = f(β)
Q=f(β)

P=f(β)

V=f(β) V=f(β)

P= f(β)

βb.s βb.s

Best Section: Q, m, i, n, = const Best Section: A, m, i, n, = const

To obtain the best section, the mathematical concept of minimum and maximum is used.
For a given discharge, the best section is obtained when the area and wetted perimeter are minimum
or the velocity is maximum.

For a trapezoidal channel, the area

A = (b + mh)h = h2β + h2m and
P = b +2h(1 + m2)1/2= hβ + 2h(1+m2)1/2
Therefore δA = (2hβ δh +2hm δ h) +h2 δβ and
δP = β δh + 2(1+m2)1/2 δh +h δβ

By solving the two systems of equations, we shall obtain the best section for trapezoidal channel as
--------------(OC-10)

Uniform flow in close conduits

Examples of conduits designed on uniform flow concepts could be sewage pipes, drainage pipes,
hydrotechnical conduits etc. They come in different forms with relatively complex cross-sectional
areas, and wetted perimeter.
Closed conduits working under uniform flow conditions have the following specific characteristics:
1) The maximum velocity of flow in the conduit occurs at a depth:
2)
h = (0.80 – 0.85)D ----------------------(OC-11)

2) The maximum discharge Qmax occurs at a depth

8 Dr. F. O. K. Anyemedu
 h = (0.93 – 0.95)D ---------------------(OC-12)

It is quite difficult to understand why maximum velocity and discharge do not occur at the top when
conduit is full but one could explain the observation as follows.
At higher depths, (h.>0.8D), any small increase in depth brings about increases in cross-sectional
area and wetted perimeter of flow. However, the changes in the wetted perimeter far exceed the
increases in cross-sectional area. And since the wetted perimeter defines the frictional losses of
flow, the velocity consequently reduces and hence the flow rate.
Also, between the free surface and the suffit of the pipe is always a layer of a gaseous mixture of
gas and liquid, which is in motion, the motion getting faster as the depth in the conduit increases.
The presence of this gaseous mixture creates resistance at the free surface, the magnitude getting
bigger as depth of flow increases.
Equation (OC-11 & OC -12) are quite complex to prove for most sections, however, for circular
pipes, they can be done analytically.

Conditions for maximum discharge and velocity in circular conduits under uniform flow
conditions
Consider a pipe of diameter D with changing depth of flow, h. In circular pipes, the cross-sectional
area and the wetted perimeter are a function of the angle β, where 2β is the angle subtended at by
the free surface at the centre
Now, using the Chezy’s formula to define the discharge, we have

---------------(1*)

For maximum discharge, δQ/δ β=0

-------------(2*)

From geometry,
P = D β and -----------------------------------(3*)

9 Dr. F. O. K. Anyemedu
 β β
D O
h

-------------------(4*)

Substituting the wetted perimeter and area from (3*) and (4*) into (2*) we shall ontain:

The solution of this equation is obtained by trial and error as:

β = 2.68 radians or β = 154o

Therefore the depth of flow for maximum discharge is

h = D/2 + (D/2)cos (180- β)

h = (D/2) (1-cosβ)

And substituting β = 154o,

10 Dr. F. O. K. Anyemedu
 h = (D/2) (1-cos 154o)
or h =(D/2) (1+0.90) = 0.95D β

Therefore the maximum discharge in a circular pipe occurs when the depth of flow is 0.95 D.
If the Manning’s formula were used, the result would have been slightly different i.e
Hmax. Q = 0.94D
Generally the maximum flow occurs at h = (0.93 -0.95)D.

By similar analysis as above, the maximum velocity could be obtained as:

hmax. V = 0.81 D

Generally the maximum velocity occurs at h = (0.80 – 0.85)D

Although the above conclusion has been obtained for circular pipes, it hold true for all closed
conduits.

Problems encountered in the computation of channels in uniform flow conditions

As an example, we shall consider trapezoidal channel. Trapezoidal channel is characterized by six
parameters; b, h, m, (for cross-sectional parameters), n, S, Q or v. Some of these parameters would
be given and so the hydraulic computation of the channel involves finding the parameters that are
not known. Below are some examples of some of the problems that could be encountered in
practice.

a) Cross-sectional parameters to be found. Given: m, b, n, I, Q; and required to find h

This problem can be solved either analytically or graphically.
i) Analytical solution.
The discharge , Q = vA = AC(RS)1/2 = A(1/n)R2/3 S1/2

A = (b+mh)h; P = b +2h(1=m2)1/2
From these, the hydraulic radius is obtained as:

R = A/P = [(b+mh)h]/[b+2h(1+m2)1/2]
Therefore Q = (b+mh)h x (1/n) x {[(b+mh)h]/[b+2h(1+m2)1/2]}2/3 x S1/2

In the above equation of Q, the only unknown is h which can be solved by the method of successive
approximation.

ii) Graphical solution.

1. For the graphical we use the formulae Q = AC(RS)1/2 and K = ACR1/2
2. We determine the necessary conveyance to carry the flow Q
Knec = Q/(S1/2)
3. By giving different values of h we try to find the value of h that will give us the necessary
conveyance. This is better done in a tabular form

No. Quantitties Different values of given depths

h h1 h2 h3 h4 h5
2 A = (b +mh)h
3 P = b + 2h(1+m2)1/2
4 R= A/P
5 C = (1/n))xR1/6

11 Dr. F. O. K. Anyemedu
 6 K = AC(R)1/2

4. For each depth, h, we obtain the conveyance, K and therefore we can plot K =f(h)

h
K =f(h)

hnec

h
m

O
b Knec K
Fig. Graphical Solution

5. From the plotted graph and knowing the necessary conveyance, we determine the depth hnec

b) Given: b, h, m, S, n. Required to find Q

Solution: 1. Calculate A and P
2. Calculate R = A/P
3. Calculate C = (1/n)R1/6
4. Calculate v = C(RS)1/2
5. Calculate Q = vxA

c). Given: b, h, m, n, Q Required to find S

From v = C(RS)1/2 then S= v2/(C2R.)

1. Calculate A and P
2. Calculate R
3. Calculate C
4. Calculate v = Q/A
5. Calculate S = v2/(C2R)

d) Given: m, n, Q, S, v. Required to find b, and h

From v = C(RS)1/2 Let CR1/2 = W then v = WS1/2
1. Calculate A = Q/v
2. Calculate W = v/S1/2
3. Write an expression for A= (b +mh)h = f(h, b)
4. Write an expression for W = CR1/2 = f(h,b)

Worked Examples
1. Water flows uniformly at a depth of 1.2 m in a rectangular canal 3m wide laid on a slope of 1 in
1000. What is the mean shear stress on the sides and bottom of the canal?.

12 Dr. F. O. K. Anyemedu
Based on the Reynolds and Froude numbers, determine the type of flow in the canal. Take Chezy’s
C = 70 and ν = 10-6m2/s. (Taken from A. K. Jain)

Solution
The mean shear stress on sides and bottom of canal is given by

τ0= γRS

The hydraulic radius R = A/P = (b x h)/(b + 2h) = (2 x 1.2)/(3 + 2 x1.2) = 0.667m

The mean shear stress τ0 = 9810 x 0.667 x(1/1000) = 6.54 N/m2

Average velocity, v = C(RS)1/2 = 70x(0.667 x/1000) = 1.806 m/s

Reynolds number of flow Re = vR/ν = (1.806 x 0.667)/10-6 = 1.203 x104

Since Re is greater than 500, then flow is not laminar but turbulent since it is far greater than 2000.
The Froude Number is given by Fr = v2/gD; D = A/B = b xh /b = h where D = hydraulic mean
depth..
Fr = 1.806)2/(9.81 x1.2) = 0.527
Since Fr, 1, then flow in channel is sub-critical
Therefore flow is turbulent regime and sub-critical state.

2. A wide channel carrying silt free water has a depth of 2.0 m. The maximum tractive stress
permissible on the bed to prevent scour is 1.96 N/m2. What is the maximum slope that can be given
to the channel?

Solution
For a wide channel the hydraulic radius R ≈ the depth of flow.
The critical tractive stress τc = γRcS
from which S = τc/γRc = 1.96/(9810 x 2 = 1/104
Therefore the maximum slope that can be given to the channel is 1 in 10,000.

3. Show that the friction factor of the Darcy-Weisbach formula, and the Manning’s roughness
factor are inter-related.

Solution
We showed that the Chezy’s coefficient C = (8g/f)1/2
Also the Manning’s formula is given by C = (1/n) xR1/6
Therefore equating the two, we have
(8g/f)1/2 = (1/n) R1/6
Squaring both sides and making f the subject, we shall obtain
f = (8g/R1/3) x n2+
f = 8 x 9.81/R1/3) xn2 = 78.5n2/R1/3

4. The cross-section of an open channel is a trapezium with a bottom width of 4 m and side slope of
1V to 2H. Calculate the discharge if the depth of water is 1.5 m and S = 1/1000 Use a)Chezy’s
formula, C.= 50, b) Bazin’s formula, m =2.30

Solution:
a) i) Use the continuity equation Q = vA

13 Dr. F. O. K. Anyemedu
 ii) Express the velocity by Chezy’ formula v = C√RS; Q = AC√RS
iii) Determine the area of flow A = (b+mh)h = (4+2x1.5)1.5 = 10.5m2
iv) Determine the wetted perimeter P = (b+2h √(1+ m2) = (4+2x1.5√5)=10.71 m
v) Determine the hydraulic radius R = A/P = 10.5/10.71 = 0.98 m
vi) Determine Q = AC√ RS = 10.5 x 50√0.98 x (1/1000) = 13.02 m3/s

b) Step i) to v) is the same as in a) above.

vi) Determine C using Bazin formula C = (87)/[1+(k/√R)]
but k = m/1.87; therefore C = (1.87 x87)/[1.87+2.30/√0.98 = 38.13
vii) Determine Q = AC√RS = 10.5 x 38.13√0.98x1/1000 = 9.90 m3/s

5. A rectangular channel has a width of 2.5 m and a slope of 1:400. Find the depth of flow if the
dischargeis 10 cumecs. Use Chezy’s formula, C = 50

i) Q = vA = AC√RS
ii) Determine A = bxh =
iii) Determine P = b + 2h
iv) Determine R = A/P = (bxh)/(b+2h)
v) Determine Q = 10 = AC√RS = (2.5 x h) x 50 x √(2.5 x h)/2.5+2h)x1/400
vi) Solving for h, we shall obtain h = 1.85 m

Example 6.Ariver used for navigation has the cross-section shown in the figure below. If the
channel cross-section if of earth (Manning’s n = 0.025) and has a slope of 1 in 1000, what rate of
flow should it carry at the given depth?

J I

A H
3

B G
C F
7

D E
6 6
30 15 125 15 30
6

Solution:
The channel is a complex one consisting of the main channel and the flood plains so we are going to
divide it into three parts: ABCJ, JCDEFI and FGHI and determine the flow passing through each.

a) Portion JCDEFI
i) The flow is determined by Q = vA = AC√RS = (A/n)R22/3S1/2
ii) Determine A =area CDEF +area JCFI = (b+mh)h +(b+2mh)h*

14 Dr. F. O. K. Anyemedu
 A = [125+(15/7)x7] 7 + [125 + 2(15/7)x7] x 3= 980+465 = 1445m2
iii) Determine the wetted perimeter P = b +2h√(1+m2) = 125+2x7√(1+4.5)= 158.1m
iv) Determine the hydraulic radius, R = A/P = 1445/158.1 = 9.15m
v) Determine the flow Q = (A/n)R22/3S1/2 =(1445/0.025) x 9.152/3 x (1/1000)1/2=
Q2= 8320m3/s

b) Portion ABCJ
i) Determine the area A = (BC+AJ)/2 x CJ = [30+(30+6)]/2 x 3 = 99m2
ii) Determine the wetted perimeter, P = AB + BC = (h*√1+m*2)+30 = 3√5+30=36.71m
iii) Determine the hydraulic radius, R = A/P = 99/36.71 = 2.7m
v) Determine the flow rate Q1 = (A/n)R22/3S1/2 = (99/0.025) x 2.72/3(1/1000)1/2 = 243 m3/s

c) Portion FGHI; Q3 = 243 since it is similar to portion ABCJ

Therefore total flow rate Q = 8320 + 2 x 243 = 8806 m3/s

Example 7.
A trapezoidal channel with side slope 1 :1 has to be designed to convey 10 m3/s at a velocity of 2
m/s, so that the amount of concrete lining for the bed and sides is the minimum.
Calculate i) the area of lining required for one metre length of the canal
ii) If the rugosity coefficient n = 0.015, calculate the bed slope of the canal for
uniform flow.

Solution.
i) For minimum lining, means we must use the best section. From eq. (OC-10), best section of a
trapezoidal channel is given by :
βbs = b/h = 2(√1 + m2) –m = 2(√2)= 2x1.42 -1= 1.84
b = 1.84h
ii) Determine the cross-sectional area, A = Q/v = 10/2 = 5 m2 = (b + mh)h =
= (1.84h + 1 x h)h=2.84h2 therefore, h = √(5/2.84) = 1.32
b = 1.84 x1.32 = 2.43m
iii) Determine the wetted perimeter, P = (b + 2h(√1 + m2)= (b+2h√ 2) = =(1.84h+2.84h)=4.68h =
4.6 x 1.32 = 6.07
iv) Determine the hydraulic radius , R = A/P = 2.8h2/4.68h = 0.598h = 0.82
v) Determine the flow rate Q = (A/n)R2/3S1/2 = (5/0.015)x0.822/3S1/2 = 10
By making S the subject, we find S = [(10)/(333.3 x 0.879)]2=(0.03413)2 =0.001165
S = 1.165 x 10-3 = 1 in 858

Example 8.

A hydraulically efficient trapezoidal channel has side slopes of 1 : 1. It is required to discharge 14

m3/s with a gradient (channel slope)of 1 in 1000. If unlined, the value of Chezy’s C is 45. If lined
with concrete, the value is 70. If the cost per m3 of excavation is three times the cost per m2 of
lining, will the lined or the unlined channel be cheaper?

Solution:
To understand and appreciate the problem, one must know that the lined channel will be able to
pass the required flow with much smaller cross-sectional area due to the larger velocities it can
pass.

15 Dr. F. O. K. Anyemedu
Let us find the areas of both unlined and lined channels to pass the required flow
For a trapezoidal section, A hydraulic efficient section is given by: =
2(1.42)-1= 1.84 Therefore b = 1.84h.
a) for the unlined channel; m = 1, S = 1/1000, Q = 14; C = 45
i) Determine the area A= (b+mh)h= (1.84h+h)h= 2.84h2
ii) Determine the wetted perimeter P = b+2h(1+m2)1/2 = 1.84h+2h x 1.42= 4.68h
iii) Determine the hydraulic radius R = A/P = (2.84h2)/4.68h= 0.607h
iv) Determine the flow rate Q = AC√RS = 2.84h2 x 45 (0.607h x 1/100)1/2 =14
v) By method of successive approximation, we obtain h = 1.82m
Hence b = 1.84 x1.82 =3.35m
Area of section = 2.84 h2 = 9.407m2
If E is the cost per m2 of lining of the channel, then 3E will be the cost of excavation per m3 .
Considering a unit length of channel, cost of excavation of unlined channel will be:
Costunlined= Area x1 x 3E = 9.407 x1 x 3E = 28.22E

b) i) We do same for the lined and since slope and flows are the same, everything will be the same
except the flow equation where instead of C = 45 we put in C = 70
ii) Q = AC√RS = 2.84h2 x 70 (0.607h x 1/100)1/2 =14
form whence h = 1.53; b= 1.84 x 1.53 = 2.815m
iii) Area of section A2 = 2.84 x 1.532 = 6.648m2
iv) Cost of excavation = Area x 1 x 3E = 6.648 x 1 x 3E = 19.944E
v) Cost of lining = Area x E = 6.648 x E = 6.648E
vi). Total cost of lined channel = cost of excavation + cost of lining =
= 19.944E + 6.648E= 26.692E

Therefore the lined channel is cheaper.

Example 9.
Water flows in a channel of the shape of an isosceles triangle of bed width a and side making an
angle of 45o with the bed. Determine the relation between depth of flow d and the bed width for a
maximum velocity and maximum discharge conditions. Use the Manning’s formula and note that d
is less than 0.5a

Solution

16 Dr. F. O. K. Anyemedu
 D F

G H

a/2
d

45o 45o
B
E C
a

Manning’s formula for velocity

i) Determine the area of flow A = ½(GH + BC) x d = ½{(a-2d) +a}x d = (a - d)d

ii) Determine the wetted perimeter, P = a + 2BG = a +2√2 d

iii) Determine R = A/P =

For maximum velocity, dv/dd = 0

which upon simplification, results in for maximum velocity

a2 – 2ad - 2√2d2 =0 from which
d/a = 0.3382

For maximum discharge, dQ/dd = 0

which for maximum discharge simplifies to

5a2 – 1.5147 ad – 22.6274 d2 =0

17 Dr. F. O. K. Anyemedu
from which d/a = 0.4378

Velocity Distribution in Open Channel

The point velocity in an open channel keeps changing across the cross-section, just like in pipes
because of friction near the boundaries. In an open channel, an additional factor which influences
the velocity profile is the presence of a free surface. The surface tension of the liquid-air interface
produces a resistance to flow and causes the maximum velocity not to occur at the surface, where
the shear stress is negligible but to occur at some distance below the free surface. Another factor
that can affect and modify the velocity profile is the presence of wind. A strong wind blowing
upstream or downstream can modify the velocity distribution considerably. In a vertical, the
minimum velocity occurs at the bottom, whiles the maximum value occurs at a distance from the
free surface which varies from 0.05 to 0.25 times the depth of flow. The higher values are
applicable to sections closer to the boundary

y
Z Y X

at Z

at Y
y

at X

v
b
Position of maximum velocities
Channel cross-section with distances from boundary

The sketches above show the pattern of velocity distribution in the horizontal direction of a
rectangular channel

18 Dr. F. O. K. Anyemedu
The velocity distribution in channels depends upon the shape of channel section, roughness and its
alignment. However, in wide channels lateral boundaries have negligible effects on the velocity
distribution in the central portions. A wide channel, according to Franzini and Finnemore is one
with the width of channel being more than 10 times the depth. The velocity distribution in wide
channels according to (Vanoni, 1941) can be approximated by;

where V = depth average velocity, k = von Karman constant (≈ 0.4), d = depth of flow, So = slope
of channel and y = the distance from the bottom of the channel

The average velocity can be estimated by

V = v(0.4d) or

V = {v(0.2d) + v(0.8d)}/2

Vanoni in general showed that the Plandtl-Karman universal logarithmic velocity distribution law
for pipes is also applicable to two-dimensional open channel flow. Using Nikuradze’s constants for
pipe he obtained the following velocity distribution in open channel flow as:

where v = velocity at a distance y from channel bed, k = height of surface roughness

The theoretically average velocity analytically obtained by Keulegan,using Plandtl-Karman
universal velocity distribution law and Nikuradze and Bazin’s data is given below as:

V= mean velocity of flow.

From measurements of velocities in natural streams and velocity distribution curves plotted from
such observations, one can make the following inferences:

1) The mean velocity in a vertical is at 0.55 to 0.65 of the flow depth. The velocity at 0.6 depth
is usually within 5%of mean velocity.
2) 2The mean velocities at 0.2 depth and 0.8 depth usually gives the meanb velocity in the
vertical within 2% The maximum velocity in the vertical is at 0.80 to 0.95.

GRADUALLY VARIED, NON-UNIFORM STEADY FLOW IN CHANNELS

19 Dr. F. O. K. Anyemedu
We shall be considering only turbulent flow i.e. flow in the quadratic region of resistance. If a solid
body (object) begins to fall through the atmosphere, at the beginning the motion of body will be
accelerated (ie the body will have acceleration).
As the velocity of fall increases, the resistance to motion by the atmosphere also increases and so
the accelerated motion of the object gradually changes into a uniform motion; ie a uniform motion
such that the force of gravity of the body is balanced by the resistance of the air. Similarly, if we
should discharge water into a prismatic canal, then at the beginning the water entering the canal will
have a non-uniform motion. However as velocities change, so also changes the resistance (frictional
forces), until at a very far distance from the beginning the frictional forces balance the body
(gravity) forces of the fluid, and we shall have a uniform flow. Every stream seeks to have a
uniform flow where the work done by the body (gravity) forces will be equal to the work of the
frictional forces.
Non-uniform flows in channels arise when we in one way or the other disturb (disrupt) the regime
of uniform flow.

Examples of non-uniform flow: which are characterised by these general conditions:

i) hconst in the direction of flow
ii) const in the direction of flow

1. PRISMATIC (CONSTANT GEOMETRY) CHANNELS WITH (LONGITUDINAL)

SLOPE OF BED i 0

In this case the uniform flow will be disrupted in the following cases.

a) When a dam is constructed across the channel.

The water begins to flow over the dam. As can be seen as a result of constructing the dam, we fix
(artificially) in the channel the point A on the free surface of the stream and also the depth h f which
differs from the normal depth which characterises a uniform flow NN.

B
N A
H huniform
Q N
h
ib0
dam

In this case, we have hf  huniform

Therefore in a certain segment of channel AB we shall have non-uniform flow.

b) When a drop structure is constructed across the channel.

20 Dr. F. O. K. Anyemedu
At the crest of the structure, we shall have a depth h f, which is less than the normal depth, which
characterises uniform flow. In this case also we fix (artificially) a depth in the channel, which
differs from the normal depth.
hf  huniform

N
B
Q huniform N
Ib>0 A
hf

Drop structure

c) When a sluice gate is constructed across the channel

Q
huniform flow B N hf

Here too we have

hf  huniform
Inference:
Non-uniform flow in prismatic channels with bed slope ib 0 arises only when we in some way
artificially fix the depth hf which differs from the normal depth which characterises uniform flow.

2. PRISMATIC CHANNEL WITH BED SLOPE ib =0 or ib < 0

Q
i=0 Q i<0

By analysing Chezy’s formula

 = CRi
If i = 0 then  = 0

21 Dr. F. O. K. Anyemedu
Therefore we may conclude that if i=0 and more especially if i < 0 then uniform flow cannot
generally exist. In this case there can only be non-uniform flow.

3. NON-PRISMATIC CHANNELS
(EITHER EXPANDING OR CONTRACTING CHANNELS)

Q
Q

Expanding Channel Contracting Channel

In this case there can only be a non-uniform flow. In conclusion, there is the need to know and
remember the following. Non-uniform flow occurs:

I) In the case of prismatic channels

a) When ib<0 and when we artificially fix a depth in the channel which differs from the normal
depth of uniform flow.
b) When bed slope ib=0 or ib<0

II) In the case of a non-prismatic channel.

Therefore uniform flow can only exist in a prismatic channel with positive bed slope (i b=0) on
condition that the channel is long enough and that we do not in any way (ie by putting across the
channel structures.) disturb the uniform regime.
In the study of non-uniform flow in channels, the basic problem that usually arises is the
determination of the free surface line (i.e. the line of intersection of a vertical longitudinal plane
with the free surface line),which is very important in hydraulic engineering and has so many
practical uses.

For example:
1 2

h1 hconst
h2 h1 h2
1 2

a) If we draw the free surface curve for a canal, we can find the depth h at any cross section.
Knowing the depth of water, we can then solve the problem of how deep the cut or fill should
be.

22 Dr. F. O. K. Anyemedu
b) By drawing the free surface curve for a river on which a dam is to be built, we can know the
area to be unadated (flooded) and take the necessary precautions to reduce the damage of
flooding.

BASIC DIFFERENTIAL EQUATIONS OF GRADUALLY VARIED NON-UNIFORM

FLOW
E Energy Line

dhw
P Piezometric line E

h dH
O H h1 P
h2
L z

z1 dL z2 S

O O

We shall at first consider the general case of a non-prismatic channel. To remind you once again

1. The shape (form) and width of cross section of a prismatic channel is the same all along the
direction of flow. Therefore for such channels the cross sectional area A is a function of only the
depth ie A = f(h)…………………………(1)
2. The shape (form) of cross section of a non prismatic channel changes in the direction of flow
and so the cross sectional area A is a function of both depth and distance L from the origin ie
A = f (h,l)……………………….(2)
Let us consider a longitudinal section of a stream h - depth of flow; z – height of bed above an
arbitrary chosen datum. H = h + z
Let us further consider cross section 1-1 at a distance S from the reference section W-W and 2-2
situated at an infinitesimal distance dS from 1-1.
Now let us write the Bernoulli's equation for the sections 1-1 and 2-2.

3. ……………….(3)

P1 = P2 = Pa. Lets represent z1 +h = H

After differentiation we shall obtain
4. …………………………………………………...(4)

Now let us represent

dhw = -(dh + dhv)

23 Dr. F. O. K. Anyemedu
 or

5. dH = dhw + dhv ……………………………………………..(5)

Eqn 5 is the basic differential equation of non-uniform flow.

Here dH is the change in free surface level within the distance dL (dH) may be positive or negative.
From equation 5 we can see that the fall (or increase) in the specific potential energy of stream
equals the increase (or decrease) in specific kinetic energy plus the losses due to friction. Let us
divide equation 5 by dL.
…………………………………………………..…(6)

- free surface slope

Since in free surface flow the piezometric line coincides with the free surface of the fluid dh/dS is
also the piezometric slope.
- Hydraulic gradient (slope of the energy line)
Therefore equation (6) becomes
……………………………………………………(7)

Further on we make the following assumptions:

The head loss due to friction hw in a non uniform flow may be expressed by the same relationship as
in a uniform flow.
With this assumption we can now use Chezy's formula to find the value of if.
 = CRif

Where , C, R, K are the flow and channel characteristics at cross section 1-1.

………………………………………..…A

Equation A is the first form of the differential equation of a non-uniform flow.

SECOND FORM OF THE DIFFERENTIAL EQUATION OF A NON UNIFORM

GRADUALLY VARIED FLOW
Let us transform equation (6). Let us begin with the left hand side.
Since H = z + h, then dH = dz + dh

But ; - bed slope

……………………………………………………………….(8)
Let us transform the right hand side of equation (6)

24 Dr. F. O. K. Anyemedu
 …………………………(9)

Since in the general case

A = f(h,s)

Then the total differential of the function A will be given by

Therefore

But equals the width of the channel at the surface of water.

Equation 9 can therefore be written as

The value

Therefore equation 6 becomes after substituting all values

………………………………………….(10)

Solving equation 10 relative to we shall obtain

………………………………………………….B

Which is the second form of the differential equation for a non-uniform flow.

I. For a prismatic channel,

A = f(h) and so the partial differential and the basic equation B takes the form

……………………………………………B'

25 Dr. F. O. K. Anyemedu
Which is the 2nd form of differential equation for a non-uniform, gradually varied flow for prismatic
channels.

II. If channel has a zero slope. (i.e. ib =0)

……………………………………………….B''

III. For a channel with reverse slope (ib 0)

Let's adopt the symbol ib = -ib' (where i' 0)

………………………………………………B'''

FREE SURFACE NON-UNIFORM FLOW

4 AUXILIARY CONCEPTS

1) SPECIFIC ENERGY OF A CROSS SECTION

Energy line
Piezometric line
P free surface
 M M
h
z Obed
Obed z' Obed 26 Obed Dr. F. O. K. Anyemedu
O O
O O Datum
The total energy (Ht) per unit weight of a flowing fluid is given by the sum of the potential head
above an arbitrary datum, the pressure head and the kinetic head. For a particle at the point M, the
total energy Ht is given by

By definition, specific energy  of a cross section is the total energy when the bed level is chosen as
the datum i.e. when the datum is chosen to coincide with the bed of the channel.
The specific energy for the particle at the point M is given by
………………………………....(10)

But

………………………………….….(11)

…………………………………..(+)
Equation (+) has three roots of which two are positive and real and h
the other is negative and unreal. For a constant value of specific
energy  is constant, there are two and only two alternate depths for
a given discharge. The larger of the two values corresponds to
conditions of deep slow flow (tranquil, or straining flow). The hcr
smaller value is that for shallow fast flow (shooting flow). As the
discharge per unit width increases for tranquil flow, the depth qmax
decreases while for shooting flow the depth increases.

There is a depth at which the two depths coincide, when the two discharges for a given specific
energy is a maximum and the energy required for a given discharge is a minimum.

………………………………………….(11a)

For a rectangular channel A = bh 

q – unit discharge – discharge per unit length of channel

………………………………………….(11b)
Let us analyse the equation 11 or 11a.

27 Dr. F. O. K. Anyemedu
In analysing it we shall assume the flow is given and that the given flow through a given cross
section may pass at different depths depending on the bed slope, roughness etc. For different depths
we shall obtain different specific energies. We can therefore say that  = f(h)

From the equations 11, 11a, 11b we can easily see that:
a) As h 0;   better seen in equation 11b, since the second term on the right hand side
tends to infinity.
b) As h 0;  

If a continuous function tends to plus infinity at the extreme of the dependant variables then it
means in between extreme values there must be a minimum value of the function.

The graph shows the function .

The shaded portion gives the changes in  M
the velocity head v2
2g I h
In the case of a uniform flow v2
(where h=constant in the direction of hcr 2g
flow) the total energy Ht decreases h O
downstream (due to frictional losses) 450 II
but the specific energy E remains N
constant since the datum is at the bed
min 
level.

A plot of specific energy of a cross-section

2) CRITICAL DEPTH hcr

By definition critical depth is the depth corresponding to the minimum value of the specific energy
of a section. The velocity corresponding to this depth is called the critical velocity. To find the
value hcr we find the differential of the function and equate it to zero.

……………………………..(12) h A3 =f(h)
B
Then …....(12a)
hcr
Since we have

A3
or …………………………………(12b) Q2 B g
Finding the critical depth

The solution of the equation 12b gives you the value of the critical depth. Equation 12b can easily
be solved by plotting a graph of A3 = f(h)
B
From 12b we may write

28 Dr. F. O. K. Anyemedu
 ………………………………(12c)
12b is the general expression for the critical depth in any open channel. For a special case of a
rectangular channel (A=Bh)

………………………………….(12d)

where q=Q/B

Eqn. (12d) can be rewritten as

v2=1hk
2g 2
E E
P P
h=hk
Therefore in rectangular channels, when h =hcr
the velocity head equals ½ the depth in the channel.

The specific energy for a rectangular section is

At critical conditions
hkp

hkp

Thus the critical depth is two-thirds of the specific energy. E

Note: Critical depth does not depend on the slope of the bed.
N
3) NORMAL DEPTH ho:
K
The diagram shows a non-uniform
N hcr ho
flow with a free surface curve AB.
The flow is considered given Q.
B K A
For a given discharge, normal depth
is that depth which will be established
Q

29 Dr. F. O. K. Anyemedu
in the channel if the flow were to be
uniform.
from here ho can be obtained where o, co, Ro, are parameters of normal
depth.

4) CRITICAL SLOPE
Let us assume that we are given the flow Q, the cross section of a channel and the roughness. Then
for a given bed slope, we shall have one normal depth ho which can be obtained from Chezy's
formula. Fig A below is a channel with adjustable bed slope. By giving the bed different slopes we
can find the corresponding normal depth ho against bed slope using Chezy's formula.
ho
ho = f(ib)

ho'' ho'
hcp = f2(ib)
A
i'
i''
hcr ho=hcr

fig A icp ib
fig B

The critical depth, unlike the normal depth which (for a given cross-sectional dimensions) depends
on the bed slope, depends only on the discharge. The critical depth does not depend on the bed
slope. As can be seen from the right diagram. There exist a slope of channel bed for which the
normal depth ho will be equal to the critical depth hcr ie ho=hcr.
That slope for which ho=hcr is called the CRITICAL SLOPE icr.

DEFINITION
CRITICAL SLOPE
It is the slope, which must be given a cylindrical prismatic channel so that the normal depth ho
equals the critical depth (ho=hcr)
It is evident that;
a. If icr  ib , then hcr  ho ib - is called mild slope
b. If icr  ib, then hcr  ho ib - is called steep slope
c. If icr = ib, then hcr = ho ib - is called critical slope
Let us find an expression for critical slope
Q = AC(Ri)
Qcr = AcrCcr(Rcricr)
Earlier on hcr was obtained

………………………………………..(*)

By putting Ccr into (*) we shall obtain

30 Dr. F. O. K. Anyemedu
Or

Hence

CRITICAL, SUPERCRITICAL AND SUBCRITICAL FLOW

CRITICAL FLOW:
When channel depth equals the critical depth ie h = hcr, the flow is called critical and its
corresponding velocity called the critical velocity. Fr=1

SUPERCRITICAL (TORRENTIAL) FLOW

When h hcr, the velocity is greater than the critical velocity and the flow is said to be supercritical.
Fr1

SUBCRITICAL FLOW (TRANQUIL) FLOW

When h hcr, the velocity is lower than the critical velocity and the flow is said to be subcritical.
Fr1
Subcritical flow lies in the region I of the specific energy curve i.e. E = f(h) where we can see that
subcritical flow is characterised by the condition. dE  0
dh
Supercritical flow lies in the region II of the specific energy curve i.e. E = f(h) and is characterised
by the condition. dE  0
dh
Note:
If flow in a channel is subcritical, then changes in the flow pattern in the lower parts (reaches) of
the channel affect the flow pattern in the upper parts (reaches) of the channel; however changes in
the upper reaches do not affect the flow pattern in the lower reaches.
Therefore in subcritical flow, the depth in a given cross-section is determined by the conditions of
the lower sector of the channel.
If flow is supercritical we have an opposite situation. i.e. the depth in a given cross-section is
determined by the conditions in the upper reaches of the channel.
Let us look at a situation in which we have two flows; a subcritical and supercritical flow. Let us
examine how the transition is going to take place. The equation (2 nd differential) of a non-uniform

flow is given by …………………………..(#)

It was shown above that when h = hcr

31 Dr. F. O. K. Anyemedu
If this is the case then the denominator of the 2 nd differential equation tends to zero and therefore
with the condition h hcr the value dh 
ds
This means that the curve h = f(s) when h = kcr have a vertical tangent at the point where h = hcr
i.e.
W - W. It must be noted that the above inference was derived on the basis of formal analysis of the
equation (2nd differential). In practice at the point h = hcr we obtain a very (sharp) abrupt in the flow
pattern.

Experiments and practice show the following.

1. Changes from supercritical flow to C
subcritical flow, as a rule, are
accomplished only through hydraulic Cr
jumps.
Hydraulic jump is the abrupt increase in
stream depth from a value h less than hcr hydraulic jump
to a value h greater than hcr.

2. Changes from subcritical flow to

supercritical flow is accomplished
through a drop which must pass
through the critical depth.

INVESTIGATION OF FLOW PROFILE IN A STEADY GRADUALLY VARIED

(NON-UNIFORM) FLOW – CYLINDRICAL, PRISMATIC CHANNEL
FLOW PROFILE:
It is the outline of the water surface in a longitudinal section of a channel.
Let us write the 2nd form of differential equation for a non-uniform gradually varied flow.

Let us transform this equation so that it becomes convenient for analysis.

Let us first consider the numerator of the right-hand side.
It is known that Q=koi where ko= conveyance of channel (moduli of discharge) for the normal
depth ho.
Therefore Q2 = Ko2ib
K = ACR
K2 = A2C2R

32 Dr. F. O. K. Anyemedu
Therefore the numerator becomes

Now let us consider the denominator

For prismatic channel A=Bh

but - Froude Number
The denominator becomes 1-Fr

Now let us consider a longitudinal section of a channel.

N Zone A
C Zone A
N

C ho Zone B N Zone B C

hcr C N
Zone C ho Zone C
i > icr
i  icr
Three possible zones can be defined
Zone A – the zone of space above the critical and normal depths.
Zone B - the zone of space between the normal depth and critical depth
Zone C – the lowest zone of space above the channel bed and below both the critical and normal
depth lines.
As three types of slopes exist (mild M, steep S and critical C) and there are three zones for C slope,
eight different curves can be drawn.
Now let us consider mild curves i.e. when ib icr

M1A Curve-backwater flow

N N

MB curve-drawdown curve

C M2
C
M3
MC curve flow through sluice
gate
i < icr

33 Dr. F. O. K. Anyemedu
First let consider zone A. Here the depth is greater than the normal depth. Therefore the modulus of

discharge KKo. The ratio . Therefore the numerator is positive. Since h  hcr then it means

Fr1 and therefore the denominator is also positive.

 This means the depth must increase in the downstream direction and the flow
profile is of the type MA or backwater curve.
Now let us examine the conditions at the limits. Downstream, the depth gets larger and in the limit
tends to infinity (i.e h ). In this case K  and therefore the numerator tends to 1. As h 
The value of Froude Fr 0 and the denominator also tends to 1.
Therefore and dh = ids …………………….(*)
The equality (*) shows that the water surface tends to the horizontal, as an increase in depth dh
equals the drop in the channel bed ids. Therefore the curve Ma downstream tends to be asymptotic
to the horizontal line.
Upstream, the depth decreases and remains in zone A and tends to h o. But in this case K K o and

the ratio and therefore the numerator tends to zero. The denominator remains positive

since h is still greater than hcr. Therefore . If decreases and tends to zero in the
limit, then it means that upstream the MA curve is asymptotic to the normal depth.

Zone B:
Here the depth h ho, but h hcr and therefore K  Ko and Fr 1.

Therefore and the numerator becomes negative. The denominator is positive

This implies that the depth decreases downstream and the profile thus obtained is the M2 type
(drawdown curve).

The Limit Conditions

In the zone B the depth h is greater than hcr but downstream tends to hcr. Then KKcrKo u Ko/K1.
As the depth tends to hcr, Fr 1

i.e. the tangent to the water surface tends to the vertical. Upstream the depth increases and h ho,
then K Ko and Fr becomes less than unity.

Therefore the MB curve upstream is asymptotic to the normal.

Zone C:
Here h hcr, therefore K Kcr Ko and Fr 1

34 Dr. F. O. K. Anyemedu
Therefore
Therefore the depth increases downstream and the profile is the MC type
In the downstream limit h hcr and therefore

Therefore the water surface sharply rises and tends to be tangent to the vertical.
Upstream h can tend to zero, which does not have any physical sense (i.e. if h=0 Q=0)

STEEP CURVES (i  icr)

SA curve (ponding curve)

C
zone A

C
N zone B

N
zone C

Zone A
THE SA CURVE

The depth increases downstream. The condition at the extremities.

Downstream h  K   Ko2 0
K2
Fr 0 dh i or dh = ids
ds
The water surface tends to the horizontal.
Upstream h decreases and tends to hcr  ho  Ko2  0
K2
As h hcr Fr 1

Therefore the water surface falls and tends to be tangent to the vertical.

Zone B
S2 THE SB CURVE

The depth decreases downstream

The conditions at the limits.

Downstream h decreases and tends to ho. K Ko. At the extreme, . Fr > 1

35 Dr. F. O. K. Anyemedu

Downstream the curve is asymptotic to the normal depth. Upstream h increases and tends to hcr> ho.

 . At the limit h = hcr  Fr 1

Upstream the water surface rises and tends to the vertical.

Zone C
S3 THE SC CURVE

The depth increases downstream.

The conditions at the limits.

Downstream h decreases and tends to ho : K < Ko. K Ko. At the extreme, ; Fr > 1


Downstream, the curve is asymptotic to the normal depth. Upstream h decreases and tends to zero
which is meaningless since h=0 means Q=0.

S1 CRITICAL, C-CURVES
S3

C-N Zone A

i = icr
N-C
Zone C

If i = icr, the lines N-N and C-C coincides, so zone B does not exist and we are left with zones A and
C.
Zone A
The depth increases downstream
Extreme conditions

Downstream h increases and h  K   Fr 0 

Therefore curve tends to the horizontal.

Upstream h decreases and h hcr (ho)

K Ko  Fe (1) not defined.

36 Dr. F. O. K. Anyemedu
Zone C

Downstream h increases.

Extreme conditions
Downstream h increases and h hcr (ho), K Ko from (-)
not defined

If the channel is wide enough and the Chezy’s coefficient do not change with depth (C=const) then
these curves are straight horizontal lines.

CHANNEL WITH HORIZONTAL BED (i=0)

If i=0 then ho = , therefore zone A does Zone
not exist since line N-N is located at infinity B
from the bed. Therefore we shall be left with Zone C
zones B and C.

CHANNEL WITH ADVERSE SLOPE (i0)

Just like the (i=0) no normal depth can occur Zone
[see ho = f(I)] so the normal depth disappears B
and we have zones B and C. Zone C

INTEGRATING THE GRADUALLY VARIED FLOW DIFFERENTIAL EQUATION

(We shall consider i0 only). By B. A. Bachmetot’s method.

We must note that for non prismatic channels solution of the differential equation B does not exist
since in general the function A = f(L) is unknown and therefore the differential is also not
known. Therefore, solution of the differential equation is only applicable to prismatic channel.

We have the equation ………………..(*)

K = wCR C = 1/nRy
In the equation above the parameters K, B, A are all functions of depth h and the remaining
parameters are constants. Therefore the right side of the above equation can be expressed as a

37 Dr. F. O. K. Anyemedu
certain function of depth ie F(L). This function is quite complex and more especially when we are
to look for the integral for the above equation. To make the solution of the problem easier B. A.
Bachmetot proposed (for the purpose of integrating (*) a more simpler relationship between K and

h. This relationship is of the form

Where h’’ and h’ are two arbitrary depths taken for a given cross section K’’ and K’ modulus of
discharge corresponding to these depths.
X – HYDRAULIC INDEX OF A CHANNEL
X – is a constant for a given channel and does not depend on the value of the depths h’ and h’’.

The numerator may be written as

Let us transform the denominator

But A/X = R ; A2C2R = K2

Therefore

Let
Dividing both numerator and denominator by (ho/h)x we shall obtain

…………………… ( )

Let us now represent h/ho by a new variable 

Integrating

38 Dr. F. O. K. Anyemedu
Introducing javerage and representing ………we shall finally obtain;

……………()

In this equation,
S2-S1 = l - length of segment
2 and 1 corresponds to h2/h0 and h1/h0
for the given sector
(2) and (1) corresponds to 2 = h2/h0 and 1 = h1/h0
The value of the function is obtained from a table for different values of the
hydraulic index of channel, X.

C
h1

h0 S1
h2
L=S2 – S1
S2

Equation () makes it possible to solve 2 main problems.

Problem 1.
Finding the distance l=s2-s1 between two cross sectional areas of a given channel, when the depths
h1 and h2 are known. It is supposed that all parameters of channel Q, ho, i, n and channel form are
also known. In this case

With X calculated we find the values of the function(2) and (1) from a table.

Problem 2.
Finding the depth h1 in a given cross section at a known distance from another cross section whose
depth h2 is known. All other parameters of the stream and channel form are known. The solution of
this problem is more complex since the depth h1 would be found from the condition h1 = h01.
Therefore the value 1 from equation ( ) can be found by method of approximation (by giving
arbitrary values h1, h1’, h1’’, …. then finding l1, l1',…. ) or by constructing a graph of l = f (h1).

Note :
Remark:

39 Dr. F. O. K. Anyemedu
Since the first problem is less difficult solving, then in practical calculations (for example in the
construction of flow profiles) it is necessary to use the first.
For example if we are requested to draw the flow profile of a river, which has been dammed,
knowing h2, we give arbitrary values of h and find the lengths l1.

Before demonstrating the use of the different curves in predicting the channel profiles that occur in
various circumstances, it is necessary to examine a phenomenon that is of frequent occurrence in
open channel, namely the formation of hydraulic jumps.

THE HYDRAULIC JUMP

Hydraulic jump: is an abrupt (rapid) increase in depth of a stream from a value less than the
critical depth to a value greater than the critical depth.

GEOMETRIC ELEMENTS OF THE HYDRAULIC JUMP

Height of jump aj = h'' –h'
Length of jump lj = horizontal projection of the roller.

40 Dr. F. O. K. Anyemedu
The depths h' and h'' measured just before and after the jump are called conjugate depths.
Note that whenever there is an increase in free surface depth to pass through the critical depth
hydraulic jump occurs.
Let us examine the flow in the region of the jump. In the stream between the sections 1-1 and 2-2
we observe a separation plane ABC; below this plane, we have a stream (transitional stream) which
rapidly expands from a depth h' to a depth h'' above ABC we have surface roller, backward flow. In
this zone, strong turbulence is developed and this results in large energy losses.
Just after the jump we have the post-jump section. In section 2-2 just after the jump the velocity
profile (average) is shown, the velocity at C is zero and the velocities just near the bed are relatively
high. The jump facilitates the rapid increase in the fluctuation of the actual velocities and pressure.
As a result, the post-jump region is characterised by a very strong turbulence.
The presence of large velocities at the bottom and high pulsation of actual velocity give rise to a
high erosion ability of the stream. Between section 2-2 and 3-3 the velocity profile gradually
changes until it attains the normal profile corresponding to a uniform flow.
The length lpj is given by lp.j = (1030) h''
Let us examine hydraulic jump from the energy viewpoint.
Let us consider a hydraulic jump caused as a liquid flows through a sluice gate as shown in the
diagram. The flow profile is also shown in the diagram.

Sluice gate II
Fr <1

c c'
C C
Let us consider also the curve ofespecificd energy of this channel discharging
d' the given flow. The
a on the
points a’b’c’d’ b specific energy curve corressponds to points a,b,c,d
b' on theFr flow
>1 profile. As
we move in the direction of flow from point a to point d on the flow profile, we move a'on specific
a
energy curve E=f (h) following the route a’b’c’d’. At the point b’ (corresponding to point b) there is
a jump (transition) from the lower part I of the specific energy curve E=f (h) to part II and the point
falls to point C’. Lastly the stream obtains its minimum energy E at the point d’, and therefore here
critical depth is established.
If however we were to suppose that in nature jumps do not exist, then the curve ab at a certain point
e would have come to the critical depth C-C and the stream having minimum energy could not have
moved on further since there would not be energy to overcome frictional losses.

BASIC EQUATION OF HYDRAULIC JUMP

Let us consider a longitudinal cross section of a jump as shown on the diagram. Our task is to find
the analytical relationship between the conjugate depths h’ and h’’.
1 2

C y2 C
C.G B
h'' y1 C.G
G P2
h' C.P(centre of pressure) Q P1
C.P
A T0 D S

1 V 2
41 Dr. F. O. K. Anyemedu
We are going to use the momentum equation and we are going to apply it to the section of fluid
ABCD.
Fdt = MV  MV/dt = F
0Q(v2 – v1) = T0 + Gs + Rs + Ps ……………………..(1)
Where
v1, v2 – average velocities in sections AB & CD
Tos – projection of frictional force on section ABCD in the s-dxn.
Tos – will be neglected since it is small compared to other forces
(1-2 is considered very small); Tos = 0
Gs – projection of weight of liquid on the s-axis; Gs – 0
Rs – projection of reaction from the bed on the liquid ABCD; Rs = 0
Ps – P1 – P2 = A1(y11) – A2(y22)

Eqn. (3) is the general equation of hydraulic jump

FUNCTION OF THE HYDRAULIC JUMP (HYDRAULIC JUMP FUNCTION)

Q = const h
E (h)
E2

y (h) E
h

E1 y(h') = y(h'')
For a given cross section and discharge, and since A=f (h) and y = f (h) and the rest of the
parameters being constant we can write the following.
……………………..(4)
where h – depth in the section
A- Area of section

42 Dr. F. O. K. Anyemedu
 y – position of centre of gravity of section
The function (h) is called hydraulic jump function.
We may therefore write the general equation of hydraulic jump as
 (h’) =  (h’’) ………………………..(5)
where
 (h’) – hydraulic jump function when h = h’
 (h’’) – hydraulic jump function when h = h’’

From equation (5) it is evident that for conjugate depths the hydraulic jump function has the same
value or hydraulic jump functions for conjugate depths are equal to each other.

The hydraulic jump function has the following properties

i) The hydraulic jump function has a minimum and the minimum coincides with the minimum
of the specific energy curve E(h).
ii) When h 0;  (h) 
iii) When h  ;  (h) 

Using the hydraulic jump function, we can find the conjugate depth h' when given h'' and vice
versa. Approximate empirical formula for conjugate depth.

Rachmanov's formula

HYDRAULIC JUMP EQUATION FOR A RECTANGULAR CHANNEL

For a rectangular channel
A = bh ; y = h/2 ; Q/b = q;  Q = qb
Where q – discharge per unit length of channel.
The hydraulic jump function becomes

……………………………………….….(6)

- Unit function of hydraulic jump

………………………………..…….….(7)

Instead of equation (5) we may write y (h') = y (h'') ……………(8)

But

Substituting into eqn (8)

43 Dr. F. O. K. Anyemedu
Or
Solving it relative to h' or h'' we shall obtain

………………………(9)

Let us consider But

Therefore,

…………………………..(10)

ENERGY LOSS IN A HYDRAULIC JUMP

Energy loss in a hydraulic jump for a rectangular channel can be obtained using the Bernouilli's
equation and hydraulic jump equation.
Thus applying Bernouilli's equation to sections 1-1 and 2-2 we shall obtain

--------------(11)

By finding hcr3 from equation (9) and putting into equation (11) we shall obtain

Using the hydraulic jump function, we can find the conjugate depth h' when given h'' and vice
versa.
Approximate empirical formula for conjugate depth by A .N Rachmanov:

44 Dr. F. O. K. Anyemedu
where ξ’, ξ’’ are the relative depths; ξ’ = h’/hcr, ξ’’=h’’/hcr

LENGTH OF JUMP
Length of jump are found by empirical formuli
1. Pavloski's formula (1937)
Lj = 2.5(1.9h''-h')
2. Safranecs' formula
Lj = 4.5h''
3. Baulmelov's formula
Lj = 5(h'' – h')
4. Aivazian's formula
L = k(h'' – h')3
4h'h''
Where
k = 8.10 + Fr
Fr

TYPES OF HYDRAULIC JUMPS

Hydraulic jumps appear in various shapes. The value of the Froude number of the approach flow
decides the nature of the jump. The U.S. Bureau of Reclamation has classified S patterns of the
hydraulic jump.

45 Dr. F. O. K. Anyemedu
46 Dr. F. O. K. Anyemedu