2/10/21

CIVL 4720: Hydraulics

FLOW CONTROL AND MEASUREMENT

FLOW CONTROL AND MEASUREMENT

• Note:
• Material related to this section is partly covered in the chapter: Rapidly Varied Flow, pages
212 to 215 of the textbook.
• Related problems are also in the same section of the book

1
 2/10/21

Flow Control
• In an open-channel flow system, we can recognize that there may be cases in which the
flowrate Q needs to be maintained at a specific level
• This can be required due to a number of concerns, including channel stability and
ecological sensitivities
• In pressurized pipe flow, we understand that Q can be controlled using valves
• However, OCF is characterized by its free surface, which is open to the atmosphere;
pressurized flow control is not possible in such a case

Q: How can Q be controlled in an open-channel system?

A: By making use of channel obstructions, such as gates and weirs

• The above-mentioned so-called channel obstructions are also called channel control
structures

Flow Measurement
Q: Why are we interested in flow measurement?

• So far, we have mainly discussed cases and problems in which the majority of flow
characteristics are given
• For example, in order to generate a specific energy diagram and glean information about
the critical depth, minimum specific energy and so forth, we need to be provided with the
flowrate Q, or the flowrate per unit width q, where q = Q/b and b = channel width in the
transverse or cross-stream direction for a rectangular channel

Q: How can Q in OCF be measured?

• Channel control structures are a very reliable method by which the flowrate through a
given channel can be quantified

2
 2/10/21

Underflow Gates: Sluices, Drums, etc.

• An underflow gate is an obstruction with an adjustable opening at the bottom which allows
fluid flow underneath
• Underflow gates are located at the bottom of a wall, dam or an open channel

Gate

Tonegawa sluice gate, Ryoso Rock Creek Dam with drum gate
Aqueduct (Wikimedia Commons (Davis Hydro 2008)
2018)

Underflow Gates: Outflow

• Consider the following cases in OCF

• When an underflow gate is opened, the fluid upstream of the gate accelerates and
reaches Vc at the gate
• A supercritical jet issues from beneath the gate in the downstream direction
• If the jet is open to the atmosphere, this is called a free outflow
• If the discharged fluid becomes a roller in which the jet is submerged, it is called a
drowned or submerged outflow; this is the outflow condition for a hydraulic jump

3
 2/10/21

Underflow Gates: Specific Energy Diagram

• For idealized, ‘frictionless’ gates, E remains constant (e.g. from 1 to 2a)

• However, E decreases for actual gates (e.g. from 1 to 2b)
• For a free outflow, the flow downstream of the gate is supercritical (2b)
• However, for a drowned outflow with a hydraulic jump, the flow downstream of the gate is
subcritical (2c)

Discharge Coefficient, Cd

• If we assume that the frictional effects are negligible

and the velocity upstream of the gate is low, the
Bernoulli equation gives

𝑉= 2𝑔𝑦!

• However, since we know that frictional effects do exist,

it is preferred to account for them by using the
discharge coefficient, Cd
• Note: this is not to be mistaken for the drag coefficient
for flow around bluff bodies, CD
• The discharge velocity at the gate are then given by

𝑉 = 𝐶" 2𝑔𝑦!

4
 2/10/21

Discharge Coefficient, Cd

• For idealized flow (i.e. wherein

frictional effects are taken as
negligible), Cd = 1
• For practical flows, Cd < 1
• Cd values have been determined
experimentally and are shown in the
figure as a function of the
contraction coefficient, y2/a and the
depth ratio y1/a
• We can see that Cd drops sharply
for drowned outflow
• For a given value of y1/a, Cd
decreases with increasing y2/a

Data from Henderson, Open Channel Flow, 1st Edition, © 1966. Pearson
Education, Inc., Upper Saddle River, NJ

Overflow Gates: Weirs

• A weir is a flow control device in which the
water flows over the obstruction

A broad-crested weir (a) © Design Pics RF/The Irish Image

Collection/Getty Images RF; (b) Photo courtesy of Bryan
Lewis.

A sharp-crested rectangular weir in a natural flow system

Hlom & Balvin (2019)

A sharp-crested triangular weir

International Commission on Irrigation
and Drainage (2012)

10

5
 2/10/21

Overflow Gates: Specific Energy Diagram

• In OCF with negligible frictional effects, we recall
# &!
that hL = 0 and 𝐻 = 𝑧 + + is constant
$% '%
• The one-dimensional energy equation for OCF
between any upstream section 1 and a
downstream section 2 is
𝑉!' 𝑉''
𝑧(! + 𝑦! + = 𝑧(' + 𝑦' +
2𝑔 2𝑔

𝐸' = 𝐸! − Δ𝑧(

• In the above expression, Δzb is the elevation of

the bottom of the channel at section 2 relative to
that at section 1
• We can see that the specific energy of a fluid
stream increases by Δzb during downhill flow,
decreases by Δzb during uphill flow and remains
constant during horizontal flow

11

Overflow Gates: Specific Energy Diagram

• For a channel of constant width b, Q = AcV =
byV and the specific energy can be expressed
as

𝑄'
𝐸 =𝑦+
2𝑔𝑏 ' 𝑦 '

• This diagram is extremely valuable as it shows

the allowable states during flow
• Once the upstream conditions at section 1 are
specified, the state of the liquid at any section 2
on a specific energy diagram must fall on a point
on the specific energy curve that passes through
point 1

12

6
 2/10/21

Flow Over a Bump with Negligible Friction

• Consider flow over a bump with negligible friction
• Here, y2 is taken as the water depth over the bump
• From 𝐸' = 𝐸! − Δ𝑧(

𝑉'' 𝑉!' 𝑦!'

𝐸' = 𝑦' + → 𝐸! − Δ𝑧( = 𝑦' +
2𝑔 2𝑔 𝑦''

• From the above expression

𝑉!' '
𝑦') − 𝐸! − Δ𝑧( 𝑦'' + 𝑦 =0
2𝑔 !

• This gives us two possible values for y2 (since the

third root of the equation is negative and can be
discarded)

13

Flow Over a Bump with Negligible Friction

• How then do we know which value of y2 is correct?

Q: Does the free surface rise or fall over the bump?

• If flow is subcritical upstream of the bump, then y1

> y2; if y1 - y2 > Δzb, flow is supressed (because y1 >
y2 + Δzb)

• (Note: If y1 = y2 + Δzb, flow depth does not change

and y1 – y2 = Δzb)

• If flow is supercritical upstream of the bump, y1 <

y2 and the flow depth rises over the bump

14

7
 2/10/21

Flow Over a Bump with Negligible Friction

• Let us return to the specific energy diagram

• As Δzb increases, E2 (point 2a for subcritical flow and

point 2b for supercritical flow) continues to shift left
until reaching Emin

• This indicates that flow over the bump is at the critical

state when, E1 – Δzb(c) = E2 or Δzb(c) = E1 – Emin

• E2 cannot reduce any further, even if Δzb is

increased, because E = Emin

• Therefore, flow over the bump must remain critical,

and is known as choked

15

Broad-Crested Weirs
• From the preceding discussion, we should now
recognize that flow over a sufficiently high
obstruction (in a sufficiently wide channel) is
always critical
• If we then view a broad-crested weir as a bump
in the channel bed, we should recognize that the
depth of flow over a broad-crested weir is yc and
the velocity is Vc
• A broad-crested weir is defined as a rectangular
block of height Pw and length Lw that has a
horizontal crest over which critical flow occurs
• Q over the weir can then be expressed as

)/'
𝑄 = 𝐴* 𝑉 = 𝑦* 𝑏 𝑔𝑦* = 𝑏𝑔!/' 𝑦*

• H is the upstream depth above the top surface of

the weir

16

8
 2/10/21

Broad-Crested Weirs
• The energy equation between a section upstream
of the weir and a section over the weir for a flow
with negligible friction can be written as

𝑉!' 𝑉*'
𝐻 + 𝑃, + = 𝑦* + 𝑃, +
2𝑔 2𝑔

• Substituting 𝑉* = 𝑔𝑦* gives

2 𝑉!'
𝑦* = 𝐻+
3 2𝑔

• From which Q for an ideal weir (e.g., neglecting

friction) can be determined from

)/' )/'
2 𝑉!'
𝑄=𝑏 𝑔 𝐻+
3 2𝑔

17

Broad-Crested Weirs

• In actual practice, frictional effects cannot be

neglected in weir design and are accounted for
using the weir discharge coefficient for broad-
crested weirs, Cwd,broad
• Cwd,broad has been determined experimentally for
a range of weir characteristics and can be
calculated from

0.65
𝐶wd,broad = For H/(H + Pw) ≤ 0.35 Alternate expression:
1 + 𝐻/𝑃, /.!
𝐻 . 𝐻 !)
( ) + 1500 , )
𝐿, 𝐿,
• The velocity head term is very small and can be
taken as negligible 𝐶," = 0.5 + 0.1
𝐻 )
• Our expression for Q then becomes 1 + 1000 ,𝐿 )
,

)/'
2 )/' Empirical equation for finite-crest weirs, Swamee (1998)
𝑄 = 𝐶wd, broad 𝑏 𝑔 𝐻
3
Develop a functional form for Q using Dimensional Analysis

18

9
 2/10/21

Broad-Crested Weirs
• We have already established that broad-crested
weirs should be sufficiently high and the channel
in which they are placed should be sufficiently
wide in order for the assumptions made in the
preceding analysis to apply

• Similarly, the length of the weir, Lw, influences the

flow over the weir surface

• If Lw > 12H, wall shear dominates the flow, the

velocity decreases and the flow becomes
subcritical; the weir is too long

• Furthermore, if Lw < 2H, the fluid will not be

allowed to accelerate to critical velocity before
the overfall is reached; the weir is then too short

19

Broad-crested Weir

Most designers will adapt 0.08 for the lower end of H/L

20

10
 2/10/21

Other methods for designing of Broad-Crested Weirs

Denoting V1 as the approach flow velocity, He as the energy head on the weir crest and neglecting energy losses:

𝑉!'
𝐻1 = 𝐻 +
2𝑔
𝑄' 3 𝑄⁄𝐵 ' !⁄)
𝐻1 = 𝑦* + ' =
2𝑔𝐴* 2 𝑔
If energy losses are absorbed in the discharge coefficient, Cd, we can solve for Q in terms of the measurable
quantity, H:
' ' !⁄'
𝑄 = 𝐶3 𝐶" 𝑔 𝐵𝐻 )⁄' B is the width of the channel
) )

Here, Cv = (He/H)3/2. During the design, we can assume Cv = 1 and then the approach velocity head can be
calculated to update the value of Cv.
'⁄) !⁄'
Some designers use: 𝐶" 𝐴∗ 𝐶3 − 1
= 𝑤ℎ𝑒𝑟𝑒 𝐴∗ = 𝐵𝐻 𝑎𝑛𝑑 𝐴! = 𝐵 𝐻 + 𝑃,
𝐴! 0.385𝐶3

21

Design a broad-crested weir for a laboratory flume with a width of 15 in. The discharge range is 0.1 to 1.0 cfs.
The maximum approach flow depth is 18 in. Determine the height of the weir and the weir length in the flow
direction. Plot the expected head-discharge relationship.
For: H/(H + P) ≤ 0.35 and 0.08 ≤ H/Lw ≤ 0.33 for broad-crested behavior.
!⁄'
2 2
𝑄 = 𝐶3 𝐶" 𝑔 𝐵𝐻 )⁄'
3 3
The head-discharge relationship is solved for Hmax when Q = Qmax = 1.0 cfs, taking the coefficient
of discharge has been taken to be 0.54, and the crest width is 1.25 ft.
!⁄'
2 2 Substitute Q = 1 cfs, Cv = 1, Cd = 0.54, B = 1.25 ft
𝑄 = 𝐶3 𝐶" 𝑔 𝐵𝐻 )⁄'
3 3
This yields, Hmax = 0.61 ft
Similarly, the minimum head is determined to be 0.13 ft for Q = 0.1 cfs.
Set Hmax/Lw = 0.33, and calculate Lw = 0.61/0.33 = 1.85 ft, which is the length of the weir in the flow direction.
For the minimum head this gives Hmin/Lw = 0.071, which is only slightly less than the allowable value.
In addition, set H/(H + P) = 0.35 for H = Hmax and solve for the weir height,
P = (Hmax/0.35) – Hmax = (0.0.61/0.35) – 0.61 = 1.13 ft.
So the weir should have a length in the flow direction, Lw = 1.85 ft, and a height, P = 1.13 ft.

To plot the head-discharge relationship, use the discharge equation with Cd = 0.54; B = 1.25 ft; and Cv
from the last equation in the previous slide, which has to be solved for a calculated value of CdA*/A1 for each head
to plot H vs. Q.

22

11
 2/10/21

Sharp-Crested Weirs
• Sharp-crested weirs are thin plates mounted perpendicular to the flow direction
• The top of the plate is beveled in order to encourage formation of a nappe
• The plate has an opening which can vary in cross-sectional shape (rectangular, triangular,
circular, etc.)
• The forthcoming analysis of flow over a weir and subsequent design methodology is
based on the assumption that the pressure is atmospheric above and below the overfall,
which highlights the importance of a distinct nappe which flows clear of the weir

beveled edge
nappe

Aerated

23

Flow Over a Sharp-Crested Weir

• As with a broad-crested weir, the upstream flow is
subcritical and accelerates, reaching criticality as it
approaches the weir
• The rapid fall of the free surface as the approach flow
passes over the weir causes the flow to accelerate as
the elevation head of the flow is converted to velocity
head
• The resulting overflow (comprising the nappe) takes the
form of a supercritical flow resembling a free jet
• If we once again assume negligible friction in the flow,
the Bernoulli equation between a point in the upstream
flow (point 1) and a point over the weir (point 2) at some
level h below the upstream free surface is

𝑉!' 𝑢''
𝐻 + 𝑃, + = 𝐻 + 𝑃, − ℎ +
2𝑔 2𝑔

24

12
 2/10/21

Flow Over a Sharp-Crested Weir

• Solving for u2 over the weir gives

𝑢' = 2𝑔ℎ + 𝑉!'

• In order to determine the flowrate over the weir, we

integrate the product of velocity and a differential area of
flow (wdh) over the entire area of flow, where w is the
width of flow at distance h from the upstream free
surface location

8
𝑄 = A 𝑢' 𝑑𝐴*' = A 2𝑔ℎ + 𝑉!' 𝑤 𝑑ℎ
5! 67/

• We can now derive similar expressions for specific weir

opening shapes

25

Rectangular Sharp-Crested Weirs

• For a rectangular weir, w = b and Q for flow over a
rectangular weir for idealized flow with negligible friction
is given by

)/' )/'
2 𝑉!' 𝑉!'
𝑄 = 𝑏 2𝑔 𝐻+ −
3 2𝑔 2𝑔

• However, it should be noted that in practical flows, the

free surface falls slightly as flow accelerates and passes
over the weir (known as the drawdown effect)
• Furthermore, there is a separation of flow at the top
edge of the weir, causing the nappe to narrow (the
contraction effect)
• Therefore, the use of a weir coefficient is required in
order to account for both frictional effects as well as
drawdown and contraction effects

26

13
 2/10/21

Rectangular Sharp-Crested Weirs

• Our previous expression is then superseded by

2
𝑄 = 𝐶wd, rec 𝑏 2𝑔𝐻)/'
3
• In the above expression, the rectangular weir coefficient
is determined from
𝐻 𝐻
𝐶wd, rec = 0.598 + 0.0897 for ≤2
𝑃, 𝑃,

Develop a functional form for Q using Dimensional Analysis

27

Example: Measuring Flow Rate Using a Sharp-Crested Weir

The flow rate of water in a 5-m-wide horizontal open channel is being measured with a 0.60-
m-high sharp-crested rectangular weir of equal width. If the water depth upstream is 1.5 m,
determined the flow rate of water.
• Recall
H
' )/'
• 𝑄 = 𝐶wd, rec 𝑏 2𝑔𝐻
) y1
Pw
8 8
• 𝐶wd, rec = 0.598 + 0.0897
#"
for #"
≤2

Develop a functional form for Q using Dimensional Analysis

28

14
 2/10/21

Example 2. Measuring Flow Rate Using a Sharp-Crested Weir

The flow rate of water in a 5-m-wide horizontal open channel is being measured with a 0.60-
m-high sharp-crested rectangular weir of equal width. If the water depth upstream is 1.5 m,
determined the flow rate of water.

8
• H = y1 – Pw = 1.5 m – 0.60 m = 0.90 m; ≤2 H
#"

/.9/ : y1
• 𝐶wd, rec = 0.598 + 0.0897 = 0.733
/.;/ <
Pw
$
' 9.=! :
• 𝑄 = 0.733 5m 2 0.90 m # = 9.24 m) /s
) >#

29

Rectangular Sharp-Crested Weir

𝐿 𝐻 𝐻
𝐶" = 𝑓𝑛 , , , 𝑅𝑒, 𝑊𝑒
𝑏 𝐿 𝑃
Neglecting surface tension and viscous effects:
𝐻
𝐶" = 0.611 + 0.08
𝑃
Corrections for Re and We effects can be included by making
corrections to H and crest length L by defining an effective
discharge coefficient. 2
𝑄 = 𝐶"1 2𝑔 !⁄'𝐿1 𝐻1 )⁄'
3
𝐿1 = 𝐿 + 𝑘? 𝐻1 = 𝐻 + 𝑘8
kH was found to be a constant (0.001 m or 0.003 ft)
Discharge coefficient for the Kindsvater-Carter formula

From Kindsvater and Carter, 1957, Georgia Tech

30

15
 2/10/21

Example
Find the height of a sharp-crested free flow rectangular weir to pass a discharge of 0.4 m3/s with an upstream
depth of 1 m. The width of the weir is 2 m and is equal to the width of the channel.

Given H + P = 1 𝐻
From the figure: L/b = 1; kL = 0, kh = 0.001 m 𝐶" = 0.611 + 0.08
𝑃
From Table: Cde = 0.602 + 0.075(H/P) If you use Cd, use in Q equation with L

Substitute in the discharge equation Q = 0.4 m3/s

Cde = 0.602 + 0.075(H/P) = 0.602 + 0.075(1 – P)/P 2
𝑄 = 𝐶"1 2𝑔 !⁄'𝐿1 𝐻1 )⁄'
3
0.4 = (0.602 + 0.075(1 – P)/P) (2/3)(2 x 9.81)1/2(2)(1- P + 0.001)3/2 𝐿1 = 𝐿 + 𝑘? 𝐻1 = 𝐻 + 𝑘8
P = 0. 77 m

31

Triangular Sharp-Crested Weirs

• The flowrate for triangular weirs (also called v-notch
weirs) can be similarly quantified
• By taking the notch width as
𝜃
𝑤 = 2 𝐻 − ℎ tan
2
• We arrive at the following expression for triangular
weirs in frictionless flow

8 𝜃
𝑄= tan 2𝑔𝐻./'
15 2

• Again, a weir coefficient is required in order to account

for friction and nappe effects

8 𝜃
𝑄 = 𝐶wd,tri tan 2𝑔𝐻./'
15 2

32

16
 2/10/21

Triangular Sharp-Crested Weirs

• Unlike broad-crested weirs and rectangular-crested sharp weirs, the weir coefficient for
triangular weirs is not derived from a single expression
• Rather, as with the discharge coefficient for underflow gates, Cwd,tri is determined from the
figure below based on experimentally derived curves for various notch sizes

33

Example 3.
The flow rate of water flowing in a 3-m-wide channel is to be measured with a sharp-crested
triangular weir 0.50 m above the channel bottom with a notch angle of 60°. If the flow depth
upstream from the weir is 1.5 m, determine the flow rate of water through the channel. Take
the weir discharge coefficient to be 0.60.
• Recall
H
= @
𝑄 = 𝐶wd,tri tan 2𝑔𝐻./'
!. '
y1

Ans. 0.818 m3/s Pw

θ
H

Pw

34

17
 2/10/21

8 𝜃
𝑄 = 𝐶de tan 2𝑔𝐻𝑒 ./'
15 2

Where He = H + kh

35

Plot and compare the head-discharge relationships for a rectangular sharp-crested weir having a
crest length of 1.0 ft in a 5-ft wide channel with that for a 90o V-notch, sharp-crested weir if both
weir crests are 1 ft above the channel bottom. Consider a head range of 0–0.5 ft.
For the rectangular, sharp-crested weir, P = 1.0 ft; L = 1.0 ft; and b = 5 ft. Then L/b = 0.2 and from Table (slide 30),
Cde = 0.589 – 0.0018 H/P.
In addition, kL = 0.0082 ft (0.0025 m) from Figure and kH = 0.003 ft. The head-discharge relationship is given by:

The triangular weir has θ = 90o and P = 1.0 ft. From the figure in the previous slide, Cde = 0.58 and kh = 0.0033 ft (1 mm).
The head-discharge relationship is given by:

Triangle

36

18
 2/10/21

Circular Sharp-Crested Weirs

• Circular sharp-crested weirs (in which a circular
hole with beveled edges is cut out of a weir
plate) are also used for flow control and
measurement

• Circular weirs can be a preferred option in that

the weir crest does not have to be levelled, and
hence the point of zero flow is easy to
determine

• Furthermore, they are easy to construct

• If H is sufficiently high, the weir behaves as an

orifice

Balachandar et al. 1991

37

Circular Sharp-Crested Weirs

• The flowrate through a sharp-crested circular
weir is calculated from

2
𝑄= 2𝑔𝐶,",*AB 𝐴𝐻!/'
3

• As with the preceding weir types, a weir

coefficient is required
• The following expression has been developed
to determine the value of Cwd,cir

𝐶,",*AB
𝐻
= 0.517 + 0.066 − 0.105(𝐻/𝐷)'
𝐷
+ 0.123(𝐻/𝐷))

𝐻 𝐻 𝐷
for 0 ≤ ≤ 1.0, 0 ≤ ≤ 1.0 and 0 ≤ ≤ 0.6
𝐷 𝑃 𝐵

38

19
 2/10/21

Practice problems

A control structure is built in a 10-m wide rectangular channel. Find the discharge, if the control structure is:
1. A sharp-crested weir with depth measured upstream above the weir crest is 1.2 m (Ans. Q ~ 24 m3/s)
2. A broad-crested weir with depth measured in the middle of the weir in the zone of parallel flow as shown in
figure (b) in slide 20 is 1.2 m (Ans. Q ~ 41.2 m3/s)
Make suitable (logical) assumptions after identifying if any data is missing

39

Example: Subcritical Flow Over a Bump

Water flowing in a wide horizontal open channel encounters a 15-cm-high bump at the
bottom of the channel. If the flow depth is 0.80 m and the velocity is 1.2 m/s before the
bump, determine if the water surface is depressed over the bump. If so, what is the depth of
flow over the bump?

• Recall y2
& y1
• 𝐹𝑟 =
%6
Δzb
• 𝐸' = 𝐸! − Δ𝑧(

40

20
 2/10/21

Example: Subcritical Flow Over a Bump

Water flowing in a wide horizontal open channel encounters a 15-cm-high bump at the
bottom of the channel. If the flow depth is 0.80 m and the velocity is 1.2 m/s before the
bump, determine if the water surface is depressed over the bump. If so, what is the depth of
flow over the bump?
• The Froude number in the upstream flow is
!.' :/>
𝐹𝑟 = = 0.428
!.#$%
( &' )(/.= :) y2
y1
• The critical depth as determined by setting Fr = 1 Δzb
!/) !/) !/)
𝑄' (𝑏𝑦!𝑉!)' (𝑦!𝑉!)'
𝑦* = = = = 0.455 m
𝑔𝑏' 𝑔𝑏' 𝑔
• The upstream flow is subcritical since Fr < 1, and therefore the water surface is depressed over the bump
𝐸' = 𝐸! − Δ𝑧(

𝑉'' 𝑉!' 𝑦!'

𝐸' = 𝑦' + → 𝐸! − Δ𝑧( = 𝑦' +
2𝑔 2𝑔 𝑦''

𝑉!' '
𝑦') − 𝐸! − Δ𝑧( 𝑦'' + 𝑦 =0
2𝑔 !

41

Example: Subcritical Flow Over a Bump

Water flowing in a wide horizontal open channel encounters a 15-cm-high bump at the
bottom of the channel. If the flow depth is 0.80 m and the velocity is 1.2 m/s before the
bump, determine if the water surface is depressed over the bump. If so, what is the depth of
flow over the bump?
m '
1.2
𝑦') − 0.80 m s − 0.15 m 𝑦'' +
9.81 m y2
2 y1
s'
&
(!.' )# Δzb
'
(.*+ & (0.80 m)' = 0
'( # )
'

• The roots of the above equation are determined to be 0.59 m, 0.36 m and -0.22 m
• We can discard the negative solution as well as y = 0.36 m, because this is less than yc
• Therefore, the water surface over the bump is 0.59 m

42

21