SECTION 5-2

OPEN CHANNEL
FLOWS

1 Prof. Dr. Mehmet F. Sevimli

 5.5.3 Determination of Manning’s roughness coefficient

The biggest difficulty in applying the Manning equation is the

determination of the roughness coefficient, n. There is no exact method of
selecting the n value. Selecting a value of n actually means to estimate the
resistance to flow in a given channel. To experienced engineers, this means
the exercise of engineering experience; for a new engineer, it can be no
more than a guess and different individuals will obtain different results.

The value of n is highly variable and depends on a number of factors. The

factors that exert the greatest influence upon the roughness coefficient in
both artificial and natural channels are described below.

a) Surface Roughness: The surface roughness is represented by the size

and shape of the grains of the material forming the wetted perimeter.
This usually considered the only factor in selecting the roughness
coefficient, but it is just one of the several factors. Generally, fine
grains result in a relatively low value of n and coarse grains in a high
value of n.
2 Prof. Dr. Mehmet F. Sevimli
 b) Vegetation: Vegetation is regarded as a kind of surface roughness, and is
reduces the capacity of the channel. This effect depends mainly on height,
density, and type of vegetation.
c) Channel Irregularity: Channel irregularity includes irregularities in wetted
perimeter and variations in cross-section, size, and shape along the channel
length.
d) Channel Alignment: Smooth curvature with large radius will give a relatively
low value of n, whereas sharp curvature with severe meandering will
increase n.
e) Silting and Scouring: Generally, silting may change a very irregular channel
into a comparatively uniform one and decrease n, whereas scouring may do
the reverse and increase n.
f) Obstruction: The presence of blockages, bridge piers, and the like usually
increases n.
g) Size and Shape of the Channel: There is no definite evidence about the size
and shape of the channel as an important factor affecting the value of n.
h) Stage and Discharge: The n value in most flows decreases with increase in
stage and discharge.
i) Seasonal Change: Owing to the seasonal growth of aquatic plants, the value
of n may change from one season to another.

3 Prof. Dr. Mehmet F. Sevimli

 5.5.3.1 Cowan Method

Considering the factors affecting the roughness coefficient, Cowan (1956)

developed a method for estimating the value of n. The value of n may be
computed by,

𝑛 = 𝑛0 + 𝑛1 + 𝑛2 + 𝑛3 + 𝑛4 × 𝑚 (5.12)

n0: a basic value for straight, uniform, smooth channel in the natural
materials involved,
n1: a value for the effect of surface irregularities,
n2: a value for variations in shape and size of the channel cross-section,
n3: a value of obstructions,
n4: a value for vegetation and flow conditions,
M: correction factor for meandering of channel

These coefficients are given in Table 5.2 depending on the channel

characteristics.

4 Prof. Dr. Mehmet F. Sevimli

 Description Recomended
Variable
Channel Value
Earth 0.020
Rock 0.025
Basic, n0
Fine Gravel 0.024
Coarse Gravel 0.028
Smooth 0
Minor 0.005
Irregularity, n1
Moderate 0.010
Severe 0.020
Gradual 0
Table 5.2 Cross section, n2 Occasional 0.005
Alternating 0.010-0.015
Negligible 0
Minor 0.010-0.015
Obstructions, n3
Appreciable 0.020-0.030
Severe 0.040-0.060
Low 0.005-0.010
Medium 0.010-0.020
Vegetation, n4
High 0.025-0.050
Very High 0.050-0.100
Minor 1.00
Meandering, m Appreciable 1.15
Severe 1.30

5 Prof. Dr. Mehmet F. Sevimli

 5.5.3.2 Empirical Formulas for n

Many empirical formulas have been presented for estimating manning’s

coefficient n in natural streams. These relate n to the bed-particle size.
(Subramanya, 1997). The most popular one under this type is the Strickler
formula,
𝑑50 1 6
𝑛= (5.13)
21.1
Where d50 is in meters and represents the particle size in which 50 percent
of the bed material is finer. For mixtures of bed materials with considerable
coarse-grained sizes,
𝑑90 1 6
𝑛= (5.14)
26
Where d90, size in meters in which 90 percent of the particles are finer than
d90. This equation is reported to be useful in predicting n in mountain
streams paved with coarse gravel and cobbles.

6 Prof. Dr. Mehmet F. Sevimli

 5.5.3.3 Equivalent roughness
In some channels different parts of the channel perimeter may have different
roughnesses (Fig. 5.14). Canals in which only the sides are lined, laboratory
flumes with glass walls and rough beds, rivers with sand bed in deepwater
portion and flood plains covered with vegetation, are some typical examples.
For such channels it is necessary to determine an equivalent roughness
coefficient that can be applied to the entire cross-sectional perimeter in using
the Manning’s formula.
𝑛 2
𝑖=1 𝑛𝑖 𝑃𝑖
𝟏) 𝑛𝑒𝑞 = (5.15)
𝑃𝑖

𝑛 1.5 2
𝑖=1 𝑛𝑖 𝑃𝑖
3
𝟐) 𝑛𝑒𝑞 = (5.16)
𝑃𝑖

𝟏 𝟐 𝟑 𝟏 𝟐
𝑸= 𝑹 𝑺𝒆 𝑨 (𝟓. 𝟏𝟕)
𝒏𝒆𝒒

7 Figure 5.14 Multi-roughness type perimeter Prof. Dr. Mehmet F. Sevimli

 5.5.4 Wetted Perimeter and Hydraulic Radius in Circular Pipes
Wetted Perimeter (P): The wetted perimeter is the length of the line which is
the interface between the fluid and channel boundary (Fig. 5.15).
1
Wetted perimeter for half a circular conduit: 𝑃 = 𝜋 𝐷
2

Hydraulic Radius (R) : Hydraulic radius is defined as the area of the flow
section (A) divided by the wetted perimeter (P).

Hyrdaulic radius for half a circular conduit:

𝜋𝐷 2
𝜋𝐷 2 1 𝐴 8 𝑫
𝐴= 𝑃= 𝜋𝐷 𝑅= = =
8 2 𝑃 1
𝜋𝐷 𝟒
2

Figure 5.15

8 Prof. Dr. Mehmet F. Sevimli

 Partially full pipe flow parameters:

𝜃 𝜋𝐷2 𝐷 𝜃 𝐷 𝜃
𝐴= × − ( × 𝐶𝑜𝑠 × × 𝑆𝑖𝑛 )
2𝜋 4 2 2 2 2

𝜃 𝜋𝐷2 𝐷2 𝜃 𝜃
𝐴= × − ( × 𝐶𝑜𝑠 × 𝑆𝑖𝑛 ) Figure 5.16
2𝜋 4 4 2 2

𝜃𝐷2 𝐷2 𝐷2 𝜃 𝜃𝐷
𝐴= − 𝑠𝑖𝑛𝜃 𝐴= (𝜃 − 𝑠𝑖𝑛𝜃) 𝑃= 𝜋𝐷 =
8 8 8 2𝜋 2
𝐷2
𝐴 (𝜃 − 𝑆𝑖𝑛𝜃) 𝑫 𝜽 − 𝑺𝒊𝒏𝜽
𝑅= = 8 𝑹= (𝟓. 𝟏𝟖)
𝑃 𝜃𝐷 𝟒 𝜽
2
2 3
𝑄 𝐴 𝑉 𝑅
= × = (5.19)
𝑄𝑑 𝐴𝑑 𝑉𝑑 𝑅𝑑
9 Prof. Dr. Mehmet F. Sevimli
 Figure 5.17 Circular section proportional depth, velocity and discharge curve

10 Prof. Dr. Mehmet F. Sevimli

 Some hydraulic parameters to be used in trapezoidal and rectangular channel
problems for uniform flows are given below.

Table 5.3 Channel section geometric elements

11 Prof. Dr. Mehmet F. Sevimli

 5.5.5 Approach the uniform flow problems

Two situations are encountered in the calculation of uniform flow

problems:
a) The unknown is S0 or Q. Since b and h are known;
1
𝑄 =𝑉×𝐴 𝑎𝑛𝑑 𝑄= 𝑅2/3 𝑆01/2 𝐴 or
𝑛

1
𝐼𝑓 𝑏 ≥ 10𝑦 𝑡ℎ𝑒𝑛 𝑃 ≅ 𝐵, 𝑅 ≅ 𝑦 𝑎𝑛𝑑 𝑄= 𝑦 2/3 𝑆01/2 𝐴
𝑛

b) Unknown cross-section values ​(b and h). Since Q, n and S0 are known. In
this case, one of b and h is chosen from the fig. 5.18, the other is found
by trial and error (iteration) from the equations below.

1 2/3 1/2 𝑄𝑛
𝑄 = 𝑅 𝑆0 𝐴 = 𝐴𝑅2/3
𝑛 𝑆0

12 Prof. Dr. Mehmet F. Sevimli

 Figure 5.18
13 Prof. Dr. Mehmet F. Sevimli
 The curve given in Figure 5.19 is used in the application-oriented design of
trapezoidal open channels. Calculated cross-section values ​(b and h)
are ​intersected as in the Fig. 5.13. If the intersection point is on or within two
curves in the graph, the channel is designed for the application-oriented.

Figure 5.19
14 Prof. Dr. Mehmet F. Sevimli
 The fig. 5.20 below is used to calculate the lining freeboard and earth
freeboard in a trapezoidal channel.

Figure 5.20
15 Prof. Dr. Mehmet F. Sevimli
 5.5.6 Best Hydraulic (The Most Efficient) Cross-Section
Let K or n and the channel slope S0 be known in the Manning Stickler
equation. Is there a minimum cross-sectional area A for a flow rate Q
to pass through this channel? The minimum cross-sectional area and
the minimum lining area will reduce construction expenses

According to the continuity equation, the smallest A for the constant

Q means the maximum velocity (Vmax). For a constant K or n and S0 it
is necessary to maximize R to maximize V in the Manning-Strickler
equation.
𝑨
𝑹=
𝑷
According to the above equation, in order to get the maximum
hydraulic radius (R), wetted perimeter (P) must be the minimum.
The section with the smallest wetted perimeter (P) from the sections
with the same area (A) is called the most efficient cross-section.

16 Prof. Dr. Mehmet F. Sevimli

 Figure 5.21
Sections with the same area are shown in the fig. 5.21. Since the smallest
wetted perimeter is the section given in option c, this is the best hydraulic
cross-section.

The section that meets these conditions is actually a semicircle. The closest
section to this is the trapezoidal channel section. Except for sewer and
drainage channels, trapezoidal or rectangular channels are made in practice.

17 Prof. Dr. Mehmet F. Sevimli

5.5.6.1 Best Hydraulic Cross-Section for trapezoidal channel

𝑨
𝐴 = 𝑏 + 𝑚ℎ ℎ → 𝒃 = − 𝒎𝒉
𝒉
𝑃 = 𝑏 + 2ℎ 1 + 𝑚2

𝑃 = 𝐴ℎ−1 − 𝑚ℎ + 2ℎ 1 + 𝑚2 𝟓. 𝟐𝟎

𝑑𝑃
= −𝐴ℎ−2 − 𝑚 + 2 1 + 𝑚2 = 0 𝑨 = 𝒉𝟐 𝟐 𝟏 + 𝒎𝟐 − 𝒎 (𝟓. 𝟐𝟏)
𝑑ℎ

𝑃 = ℎ2 ℎ−1 2 1 + 𝑚2 − 𝑚 − 𝑚ℎ + 2ℎ 1 + 𝑚2

𝑃 = 2ℎ 2 1 + 𝑚2 − 𝑚 𝟓. 𝟐𝟐

18 Prof. Dr. Mehmet F. Sevimli

 Eq. (5.21)

𝐴 ℎ2 (2 1 + 𝑚2 − 𝑚 𝒉
𝑅= = → 𝑹=
𝑃 2ℎ(2 1 + 𝑚 − 𝑚
2 𝟐
Eq. (5.22)

𝐴 = ℎ2 (2 1 + 𝑚2 − 𝑚 𝟐 𝑎𝑛𝑑 𝐴 = 𝑏 + 𝑚ℎ ℎ

ℎ2 (2 1 + 𝑚2 − 𝑚 = 𝑏 + 𝑚ℎ ℎ

𝑏 + 𝑚ℎ = 2ℎ 1 + 𝑚2 − 𝑚ℎ 𝑏 = 2ℎ 1 + 𝑚2 − 2𝑚ℎ
𝒃
𝑏 = 2ℎ 1+ 𝑚2 −𝑚 =𝟐 𝟏 + 𝒎𝟐 − 𝒎 (𝟓. 𝟐𝟑)
𝒉

𝑏 𝑏
𝐹𝑜𝑟 𝑚 = 1 = 0.828 𝐹𝑜𝑟 𝑚 = 1.5 = 0.606
ℎ ℎ
𝑏
𝐹𝑜𝑟 𝑚 = 2 = 0.472
ℎ
19 Prof. Dr. Mehmet F. Sevimli
 The most suitable side slope for the trapezoidal section is the one that
minimizes P:
𝑃 = 𝐴ℎ−1 − 𝑚ℎ + 2ℎ 1 + 𝑚2

𝑑𝑃 2𝑚ℎ 2𝑚ℎ
= 0 → −ℎ + =0 ℎ= → ℎ 1 + 𝑚2 = 2𝑚ℎ
𝑑𝑚 1+𝑚 2 1 + 𝑚2

2 2 2
1
1+ 𝑚2 = 2𝑚 → 1 + 𝑚 = 4𝑚 3𝑚 = 1 𝑚=
3
1 1
𝑡𝑔𝜃 = = = 3→ 𝜽 = 𝟔𝟎°
𝑚 1
3
𝑏 1 𝒃
=2 1+ 𝑚2 −𝑚 𝑣𝑒 𝑚 = = 𝟏. 𝟏𝟓𝟓
ℎ 3 𝒉

The best hydraulic cross-section parameters for a trapezoidal channel

𝒉 𝟏 𝒃
𝑹= ; 𝜽 = 𝟔𝟎°; 𝒎 = ; = 𝟏. 𝟏𝟓𝟓
𝟐 𝟑 𝒉
20 Prof. Dr. Mehmet F. Sevimli
5.5.6.2 Best Hydraulic Cross-Section for rectangular channel

𝑨
𝐴 = 𝑏ℎ → 𝒃=
𝒉
𝐴
𝑃 = 𝑏 + 2ℎ = + 2ℎ
ℎ
𝑑𝑃
= −𝐴ℎ−2 + 2 = 0 𝐴 = 2ℎ2
𝑑ℎ
𝒃
𝑏ℎ = 2ℎ2 =𝟐
𝒉
𝐴 𝑏ℎ 2ℎ ℎ 2ℎ2 𝒉
𝑅= = = = 𝑹=
𝑃 𝑏 + 2ℎ 2ℎ + 2ℎ 4ℎ 𝟐

21 Prof. Dr. Mehmet F. Sevimli

 5.5.7 Compound Sections
Some channel sections may be formed as a combination of elementary sections.
Typically natural channels, such as rivers, have flood plains which are wide and
shallow compared to the main channel. Fig. 5.22 represents a simplified section
of a stream with flood banks.

Figure 5.22
Compound section

22 Prof. Dr. Mehmet F. Sevimli

 Consider the compound section to be divided into subsections by arbitrary
lines. These can be extensions of the deep channel boundaries as in Fig.
5.23. Assuming the longitudinal slope to be same for all subsections, it is
easy to see that the subsections will have different mean velocities
depending upon the depth and roughness of the boundaries. Generally,
overbanks have larger size roughness than the deeper main channel. If the
mean velocities Vi in divided subsections are known then the total
discharge is ΣViAi.
𝑛

𝐴= 𝐴𝑖 (5.24)
𝑖=1

Figure 5.23
𝑛
𝟏 𝟐 𝟑 𝟏 𝟐
𝑽𝒊 = 𝑹𝒊 𝑺𝟎 (𝟓. 𝟐𝟓) 𝑄= 𝑉𝑖 𝐴𝑖 (5.26)
𝒏𝒊
𝑖=1

23 Prof. Dr. Mehmet F. Sevimli