Unit 1
Free Surface Flows  Introduction
Historical Development of Hydraulics
History of Hydraulics in India
Classification of Flow
Unit 2
Channels and their Geometric Properties
Examples
Pipe Flow and Free Surface Flow
Basic Equations
Unit 3
Continuity Equation
Energy in Free Surface Flow
Basic Momentum Equation
Velocity Distribution
Unit 4
Velocity Measurement and Distribution
Discharge Measurement by Velocityarea Method
Radioactive tracer technique for Measurement of River
Discharges
Measurement of Flow of Water and the Limitations of Velocity
area Method
Errors in Depth Measurement in High Velocity Flows
Unit 5
Secondary Current and Spiral Flow
Unit 6
Energy and Momentum CoefficientsDerivation
Energy and Momentum Coefficients for Different Velocity
Distributions
Comparison between Momentum and Energy Equation
Unit 7
Pressure Distribution
Specific Energy
Unit 8
Specific Energy Equations for Rectangular Channels
Application of Specific Energy
Problems
Unit 9
Specific Force
TransitionProblems
Application of Specific Force and Specific Energy
Transition in Field
Critical Flow
Unit 10
Characteristics of Critical Flow
Occurrence of Critical Flow
Unit 11
Critical Depth in Trapezoidal & Circular Channels
Hydraulic Exponent for Critical Flow
Problem
Unit 12
Critical Flow Depth Computations
Problems
Flow Measurement
Unit 13
Measuring Flumes
Critical Depth Flumes
Unit 14
WerisIntroduction
Types of Control Structures
Proportional weirs
Flow Over weirs
Polygonal weirs
Special types of weirs
Broad Crested weirs
Different types of Broad Crested weirs
Bear Trap weir
Unit 15
Flow below a Sluice Gate
Brink Depth
Modern Measurements of Flow Measurements
Outlets & Modules
Errors in Measurements
International Standards for Flow Measurement in Open Channel
Uniform Flow
Unit 16
Concept of Uniform Flow
Derivation of Uniform Flow Equations
Resistance in Open Channel Hydraulics
History of Uniform Flow Velocity and Resistance Factor
Unit 17
Friction
Ganguillet and Kutter Formula
Conveyance
Section Factor for Uniform Flow Computation
Unit 18
Hydraulic Exponent for Uniform Flow Computation
Maximum Discharge
Classification of bed Slope
Computations
Unit 19
ProblemsMaximum Discharge
ProblemIrregular Channel
Solution of algebraic or Transcendental Equation by Bisection
Method
Solution of Manning Equation by Newton Raphson Method
Unit 20
Slopearea Method
Normal & Critical Slopes
Design of Canals
Unit 21
Design of Canals
Typical Canal Cross Sections
Unit 22
Lining the Canals
Seepage Prevention with Impermeable membranes
Failure of Canal Lining
Most Efficient Hydraulic Section
Design of Unlined Channels
Examples & Problems
Gradually Varied Flow
Unit 23
Introduction
Dynamic Equation for Steady Gradually Varied Flow
Classification of Gradually Varied Flow Profiles
Unit 24
Real Life Cases of Water Surface Profiles
Sketching of Composite Water Surface Profiles
Examples
Unit 25
Computation of Gradually Varied Flow
Example
Unit 26
Standard Step Method
Example
Unit 27
Integration of Differential Equation
Improved Euler Method
Fourthorder RungaKutta Method
HEC2
Hydraulic Jump
Unit 28
Normal Hydraulic Jumps
Classification of Jumps
Momentum Equation
General Hydraulic Jump Equation
Unit 29
Energy loss in the Jump
Turbulent Characteristics of the Jump
Pressure Distribution in the Jump
Velocity Distribution in Hydraulic Jump
Length of the Jump
Unit 30
Air Entrainment Characteristics of the Jump
Pre Entrained Hydraulic Jump
Air Concentration Distribution along the Jump
Decay of Turbulence Downstream from a Stilling Basin
Unit 31
Hydraulic Jumps in Sloping Channels
Unit 32
Sequent Depth Tail Water Relationship Stilling Basin
Baffle Stilling Basin
Bhavani Type Stilling Basin
Stilling Basin in Sudden Expansion
Slotted Bucket Stilling Basin
Spillways
Unit 33
Spillways  Introduction
Unit 34
Siphon Spillway
Unit 35
Chute Spillway
Stepped Spillway
Flow in Bends
Unit 36
Introduction
Classification of River Bends
Transverse Water Surface Slope in Bends
Superelevation
Velocity Distribution in Bends
Unsteady Flow
Unit 37
Introduction
Basic Terminology
Classification of waves
Ocean Waves
Tides
Nature of waves
Unit 38
Surge Computation
Example1
Example2
Unit 39
Gradually Varied unsteady Flow
Celerity
Unit 40
Method of Characteristics
Method of Specified Intervals
Unit 41
Dam break ProblemIntroduction
History of Dam Failures
Causes of Dam Failures
Routing
Case StudyDam Break Analysis for Kali River
Self Aerated Flows
Unit 42
Self Aerated FlowDefinition of Terms and Instrumentation
Characteristics of Self Aerated Flows
Unit 43
Measurement in Self Aerated Flows
Experimental Investigation
Bhakra Dam SpillwayA Case Study
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
HYDRAULICS  FREE SURFACE FLOWS
1.1 Introduction
A fluid is any substance that deforms continuously when subjected to shear stress, no
matter how small the shear stress is.
Shear force is the force component tangent to the surface. Average shear stress is the
shear force per unit area.
Fluids can be classified as ideal fluids and real fluids.
Ideal fluids are those which are incompressible with zero viscosity and, shear stress is
always zero. Ideal fluid is hypothetical.
Fluids with viscosity are known as real fluids.
Example: Water, Milk, and Honey etc.,Then real fluids are classified as Newtonian and
nonNewtonian. Box 1.1.
Examples of nonNewtonian fluids are
Thixotrophic substance (thixotrophic jelly paints), ideal plastic, Bingham plastic (sewage
sludge), pseudo plastic (clay, milk, cement), dilatant substance(quick sand) etc. Fig 1.1.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Yield stress Shear stress
Figure 1.1 Rheological diagram
Box 1.1 Newtonian fluids follow the law of viscosity
du
=
dy
1.1
in which , is the Shear stress, is the viscosity coefficient and
du
dy
is the
velocity gradient in y direction.
Viscosity is a fluid property and is known as dynamic viscosity. The equation
1.1 is known as Newton's law of viscosity.
The kinematic viscosity is given by the ratio of dynamic viscosity to mass
density of fluid .
= 1.2
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Dimensions and units
Coefficient of Dynamic viscosity
2 1 1
1 1
M L T Ns m kg m s Poise
1 1
10 poise = 1 kg m s , Pa s (Pascal seconds)
3 1 1
Example: Water: 1.14 x 10 kg m s ;
5 1 1
Air : 1.78 x 10 kg m s
2 1 2 1 4
= L T m s , 10 Stokes =1
or or
=
=
2 1
m s
6 2 1 5 2 1
Example: water 1.14 x 10 m s at 15 C , air 1.46 x 10 m s
However viscosity depends on temperature.
Physical properties of water at atmospheric pressure and S.I units are given
Mass Density of water : Mass per unit Volume.
3 3 3
=[ML ]; kg m , =1000kg m
3
Mass density of air = 1.23 kg m
5 2
at atmospheric pressure of 1.013 x 10 N m and temperature 288.15 K.
weight per unit volume
is known as specific weight
3
N m
3 3
9.81 x 10 N m
of water
3
= 12.07 N m
of air
g
=
=
In free surface flows water is the dominating fluid. Water is a basic element and
supports the life system.
Proper control and management of water is required for sustaining the life on earth.
Hydraulics forms a part of water resources engineering. The free surface flows deals
with the movement of surface water in rivers, stream, canals etc. In order to understand
the mechanism of free surface flows, the different classification of them is to be
understood.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1.2 HISTORICAL DEVELOPMENT OF HYDRAULICS
Hydraulic Engineering has served the mankind althrough the ages by providing drinking
water as well as protective measures against floods and storms. In the course of history
it has made the water resource available for human uses of many kinds. Management
of the world's water is a complex task and both its scope and its importance continue to
grow.
In course of time mankind has not only diverted and used the waters of the world for its
purposes, but by engaging nature into its service has turned deserts into fertile land
(e.g. Rajasthan Indira Gandhi Canal Project). Natural habitat is threatened in more and
more parts of the world by an evergrowing human population. Time has come for
formulation of the new value system. Thus long term needs are not only food, water and
shelter but also for an aesthetically pleasing, healthy, nurturing environment.
Sustainable development is "mantra" of the future.
Method of teaching Hydraulic Engineering has undergone several changes considering
the availability of computers, GPS, GIS, Remote sensing data, and web based tools.
1.2.1 The stages of Development
1950s Experimental hydraulics  empirical Hydraulics  Development of Engineering
hydraulics.
1960s Fundamental Research in unsteady flows, Open channel and ground water.
1970s Gathering of large data  hydrologic engineering  Flood control.
1980s
Initial awareness on the Environmental aspects. Large scale water Resources
planning, stochastic hydrology, System Analysis, distributed rainfall runoff
modeling.
Early
1990s
Modeling, urban hydrology, disaster management including floods,
computational engineering, CAD in hydraulics, Environmental hydraulics,
water quality  quantity integration, GIS based distributed modeling in
hydrology, Decision support systems.
Late
1990s
Integrating of hydraulics with water resources engineering for sustainable
development using GIS, GPS, Remote sensing  Hydro 2004 informatics,
Enviro informatics, Physical hydrology , space and Time scales, Climate
change and its impact on river basin, planning and management. Soft
computing (ANN, GA etc.,) IT impact on Water Data base and knowledge,
Integrated River basin Development. Reliability and Risk tools. WEB  Water
Earth Biota. Alternate sources of energy.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1.2.2 Future
Broad scope for specialisation in aggregation of many integrated aspects of the water
system.
To design integrated systems and integration of numerical modeling into information
systems.
Globalisation of water research and exchange through Internet and its impact on
sustainable development.
Integrating sociology, economics, biology, environment  Hydro bio modeling.
Global water markets, participatory approach.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1.3 HISTORY OF HYDRAULICS IN INDIA
"ONE WHO SOLVES THE PROBLEM OF WATER IS WORTH OF TWO NOBEL
PRIZES,ONE FOR PEACE AND ONE FOR SCIENCE"  J OHN.F. KENNEDY.
1.3.1 Growth of Hydraulics and Irrigation Research In India
Introduction (CBI&P 1979)
During the nineteenth and early part of twentieth century, hydraulic and irrigation
problems were being tackled mostly by engineering judgement based on experience.
However, many engineers, with intuitive insight and initiative gave deep thought to
various problems and arrived at valuable conclusions. They were the pioneers of of
individual research exploring virgin ground in advance of the era of organised research
with the aid of models and other experimental facilities and techniques. Roorkee
professional papers on Indian Engineering (18631886) contain many original and
useful ideas on the theory of flow in artificial earthen channels, measures for efficient
distribution of irrigation waters and the design of hydraulic structures justifying high
tribute to these pioneer researchers. In 1864, fundamental ideas on the causes of silting
and scouring were initiated. At about the same time, tables of mean velocities and
depths were evolved for North Indian Canals. The Ogee type fall was originated on the
Ganga Canal (by 1870). Between 187479, Cunningham made a valuable contribution
in the techniques of the measurements of discharges and determination of velocities. By
about 1880, training of rivers with embankments combined with a system of groynes
was experimented in the field. During 1881  82, Kennedy made important estimations
of the losses by evaporation and absorption in the Bari Doab Canal. Cotton in the south
and Cauteley in the north produced some of the most imaginative river conservation
schemes over a hundred years ahead of the time they were realized to be essential and
taken up for implementation.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Cottonreddypalem, Andhra Pradesh, several other villages rooted in the Cotton name
and several generations of males with variations of Cotton name, all celebrated Sir
Arthur Cotton's bicentenary in the Godavari District, David Abbott of the British Deputy
High Commission, was present at the Rajahmundry celebrations. Cottons contribution
to making the Krishna Godavari area the granary of South India.
"Father of Irrigation", "Sculptor of Deltas". It is to be noted that the 3.685 km long
Dowleswaram Barrage across the Godavari, built at a cost of 120,000 over five years,
turned a flood and drought prone area into million acres of flourishing paddy and
sugarcane, where the rent of an acre of paddy land today is Rs.1 lakh. "When the
farmer tills his land (here) or receives the money for his produce, he thinks one man
Sir Arthur Cotton".
A Sir Arthur Cotton Museum is to be set up at the dam site at a cost of Rs. 1 crore and,
more significantly, a Sir Arthur Cotton Memorial Agricultural Service Centre is being set
up over 15 acres, at Bobbarlanka, 20 km from Rajahmundry and near Dowleswaram, at
a cost of Rs. 1 1/4 crore.
He was the beloved of the Ryots (farmers).
General Sir Arthur Cotton: His life and work, is described as "a classic on India's
development". "India had taken hold of him. Not the India of Romance, but the India of
need". The 500 page book was reprinted by the Institution of Engineers (India, in 1964).
Cotton had spent two years in Vishakhapatnam before moving on to Rajahmundry and
his greatest work. While at Vizag, he had built the St. J ohn's Church in Waltair, and
groynes to protect the beach. He also predicted that Vizag would one day be a great
port. Truly was he a farsighted engineer.
The reports of the select committee admitted the success of all the irrigation works in
the Madras delta with which Sir Arthur Cotton's name is so honourably associated,
namely the Cauvery, Kistna and Godavari, and indicated that if there was any financial
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
failure in other case in the past, the main cause appears to be the want of ability or
energy on the part of the officers of the Public Works Department and their reliance on
hasty generalisation.
Cotton use to use to tell his daughter, "Do something, my girl, do something. Never be
idle for a single moment. Remember Time is short, Eternity is near."
He was 96 and had not suffered any major illness. On the night of J uly 14th, 1899 he
became feverish and restless and began slowly sinking. The end when it came was
'perfect peace'.
"His life, judged by any test was one of the true greatness, such as is only given to vary
few to attain in the world. He has left behind him a fame and a name which must
endure to all times". Sir Richard Sankey, R.E., K.C.B., wrote in a letter to Lady Arthur
on hearing of her husband's death.
Reference
Madras Musings, October 1  15, 1999.
During 18th and 19th centuries, the irrigation works in India were neglected by East
India Company so much so that Arthur Cotton, Royal Engineer working with Madras
Presidency complained bitterly in 1821 against the policy of apathy of the government.
In the history of India, 18th and 19th centuries saw some of the worst famines in the
north as well as south. As a result, efforts were made for saving agriculture. In the field
of irrigation, these included reopening of Western and Eastern Yamuna canals,
renovating Hissar branch canal and repairing Grand Anicut on Kaveri during 1810 
1836 period.
Col. Proby T. Cautley of the Royal Artillery (1802  1871), was the superintendent of the
canals in the NorthWestern Province and director of the proposed Ganga Canal. In
1838, Cautley submitted to the government the first proposal to take a canal from
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Ganga at Haridwar.
Governor General Lord Hardinge visited the site personally and authorized the
construction of canal in 1842.
J ames Thomason (1804  1853) who was then Lt. Governor of Northern Province fully
supported the proposal of Ganga canal.
The excavation of the canal was started in 1842 and water entered the canal in 1854. It
is interesting to note that when the canal was designed, the only hydraulic principles
known were continuity equation and resistance law. And yet the unlined canal designed
to carry discharge of approximately 300 cumecs as well as the cross drainage works
such as Solani aqueduct, siphons and level crossings which are still intact and
functioning well and have stood the test of time.
It is worth mentioning that Cautley became involved in public controversy over the
design of Ganga canal against Arthur Cotton in 1863  65 and was publicly censured in
the columns of the Times. However, he was officially exonerated by the Governor
General in 1865.
LOOKING BACK
If we have done our duty at least to this part of India, and have founded a system which
will be a source of strength and wealth and credit to us as a nation, it is due to ONE
MASTER MIND Which, with admirable industry and perseverance, inspite of every
discouragement, has worked out this great result. Other able and devoted officers have
caught Colonel Cottons spirit and have rendered invaluable aid under his advice and
direction, but for this first creation of genius we are indebted to him alone.
Colonel Cottons name will be venerated by millions yet unborn, when many, who now
occupy a much larger place in the public view, will be forgotten; but, although it
concerns not him, it would be, for our own sake, a matter of regret if Colonel Cotton
were not to receive due acknowledgement during his lifetime.  Minute by the
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Government of Madras. Sir Charles Trevelyan, Governor, in his review of the Public
Works Department on May 15th, 1858.
General Sir Arthur Cotton, R.E., K.C.S.I., was born in Cheshire, England on May 15,
1803, the tenth son of Henry Calveley Cotton. Lt. Arthur Cotton arrived in Madras in
September 1821 and was attached to the office of the Chief Engineer for the
presidency. In May 1822, he was posted as an Assistant to the Superintending
Engineer of the Tank department, Southern Division.
Survey of the Pamban Pass to propose an enlargement of the pass for the passage of
oceangoing steamers from the West Coast to the East Coast ports. This was the
beginning of the Sethusamudram Project we have been talking of for a century ! .
In 1829, he was promoted as Captain and given separate charge of the Cauvery
irrigation. He soon saw the need for saving the district from the ruin that was staring it
with barely any flow in the cauvery due to heavy silting at the Grand Anicut. He soon
evolved the scheme for erecting a control structure on the Coleroon at the Upper Anicut
and the opening up of scour vents in the old Grand Anicut. On J anuary 1, 1830 the
great work of seven sluices was started. In 1832, got the project reports both for Upper
Anicut and the Lower Coleroon Anicut on the Coleroon ready. They were sanctioned by
the Government in time to get the preliminary work started before the freshes arrived in
J une.The first bold step taken by Cotton was the construction of the Upper Coleroom
Dam at Mukkombu.
Mr. W.N. Kindersley, the Collector of the district, wrote there was not one individual in
the province who did not consider the Upper Anicut the greatest blessings that had ever
been conferred upon it. The name of the projector would, in Tanjore, survive those of
all the Europeans who had ever been connected with it.
At this distant date we fail to realize the great truth in these statements made and the
valuable contributions of this pioneer, Sir, Arthur Cotton. He always insisted on saying
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
that the value of irrigation works was not to be measured simply by the additional
revenue yielded to the Government treasury, but that a much truer criterion would be
found in the enhancement of the income of the people and in the consequent saleable
value of the land itself. Irrigation brings with it prosperity to the region, some perceptible
and much more imperceptible and intangible.
The work that made a magical change in the hinterland of the delta of the River
Godavari, the masterpiece of the great thinker, the planner, the designer and the maker,
Major Arthur Cotton, was to come soon after.
Cotton, after a careful study of the sufferings of the people in the delta, while huge
volumes of floodwaters were being carried out to the sea day in and day out by the
mighty Godavari, reported to the Board of Revenue in May 1844 that the only way to
turn the Godavari district from being the poorest to nearly the richest in the presidency
was bringing in irrigationcumnavigation facilities in the Delta by building an anicut
across the wide river.
Reference
Madras MusingsSeptember 1630, 1999.
Outstanding contributions to subsurface and surface flow research came from Col.
Clibborn and Kennedy during 1890's. Col. Clibborn carried out the historic experiments
(189597) with Khanki sand to investigate the laws of flow of water through sand in
relation to weir design. Col.Clibborn's other contribution was on investigations on the
replenishment and velocity of flow of ground water in the Gangetic plains. In 1895, after
field experiments on the Upper Bari Doab Canal, Kennedy propounded his classical
relations between the critical velocity and channel depth as influencing channel design.
The early twentieth century has been notable for the rapid extension of irrigation in the
country and with it for the rigorous efforts on the investigations on the economic and
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
reliable design of hydraulic structures, design of stable channels, efficient distribution
devices, weed control, antiwater logging measures and land reclamation.
Kennedy's classical equations for the design of channels were followed by Lindley's
relations in 1919 indeed the very concept of the regime theory itself. Between 192939,
Lacey's sustained and pioneering work led to the development of comprehensive
formulae for designing stable channels in alluvium. The thread was picked up by various
workers principally, Inglis, Bose, Malhotra, Blench, et al. and this subject has continued
to be a subject of sustained interest in India.
Investigations for the control of sand entering channels attracted the attention of many
engineers also, Inglis, the father of hydraulic model research in India, demonstrated that
curvature of flow or nature's way was the dominant factor affecting surface and bed
flow and, therefore, the most effective way of controlling sand. In 1922, Eldsen initiated
the idea of the tunnel type of excluders, and in 1934 Nicholson built the first excluder at
the head of the Lower Chenab Canal at Khanki. King's investigations for exclusion of
heavy silt from canal by vaned pitching (1918) and with silt vanes (1920) were earlier
notable investigations in the same field.
India's contribution of the development of subsoil flow hydraulics in relation to the
design of weirs has indeed been unsurpassed. After Col. Clibborn's historic experiments
(189597) with Khanki sand, Khosla propounded (192936) the very valuable theory of
subsoil flow in relation to the design of weirs on permeable foundations. The first full
size experiments in the world was conducted during 192936 on the Panjnad Weir. This
was followed by laboratory research on models of Rasul Weir (193034) and Panjnad
Weir (193435) by Taylor and Uppal, and on electrical analogy models by Vaidyanathan
(1936) and others.
Efficient distribution of water from canals was another subject which attracted the
attention of engineers from early times. Up to the end of the nineteenth century,
ordinary canal outlets in the form of open cuts, pipe or barrel outlets were in vogue. In
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1882, Beresford introduced a general type of outlet with a 15.2 cm pipe with flap and
face walls. Since the beginning of this century, a number of investigators have studied
the various aspects of canal outlets and several types have been developed. The
earliest semimodular type was in 1902 by Kennedythe sill outlet. Kennedy's gauge
outlet was introduced in 1906which was further improved in 1915. By 1922 Kirkpatrick
on the J amras (Sind) and Crump in Punjab developed semimodules of the open flume
and the orifice types. Among the modules with moving parts, Visvesvarya's self acting
module (1904), Kennedy's outlet module (1906), Wilkins type (1913), J oshi's module
(1919) and Kenti's 'O' type module (1923) were the important developments. A module
without any moving parts had been developed by Gibb as far back as 1906 and it was
greatly improved later by experiments in Poona. Many silt extracting outlets were also
developed, the outstanding one being the Haigh's type in 1937. Valuable experiments
conducted on broadcrested weirs were utilised by Burkitt in developing the 'Headless
meter.
Bharat Rathna Dr. Sir. Dr. Mokshagundam Visvesvaraya (1861  1962)
September 15 is a memorable day in the annals of the engineering community in
particular in this country. On this day 135 years ago, one of the greatest sons of India,
Dr. Sir. Mokshagundam Visvesvaraya, the towering personality in the history of Indian
engineering  was born at Muddenhalli in the Kolar district of Karnataka. Graduated from
the college of science, Poona in 1883, Visvesvaraya joined the Bombay PWD and rose
to the position of Chief Engineer. He worked ceaselessly throughout his life to bring
fruits of advanced science and technology to the doorsteps of the common man. On
retirement, his services were requisitioned by the Maharaja of the erstwhile Mysore
State, who appointed him as Dewan. The following years witnessed an era of planned
development and allround growth. A visionary who could think ahead of his time,
Visvesvaraya realised that there could be no salvation for the people of the country
except judicious use of the results of technological innovations. In recognition of his
services to national development and for the cause of engineering, he was honoured by
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
presentation of the country's highest award  Bharat Ratna  in 1955.
To perpetuate the memory of this great engineerstatesman, the Council of Institution of
Engineers India decided to observe September 15 each year as Engineer's Day and
evolved guidelines for celebrating the Day.
The State of Mysore has been well known for its engineers. Modern research as such in
engineering started about 1870's. The first claimant for leadership in engineering
research was Sri Adil Shah Dabe who constructed in the first decade of the 20 th
Century the Mari Kanave Dam with masonry in Surki mortar. It was easily the highest
dam at that time in the world constructed with a matrix other than cement.
The second decade of the 20th Century started with the advent of the world famous
Engineer Bharat Ratna Dr. Sir. M. Visvesvaraya at the helm of affairs in Engineering
and Administration. His pioneering works in the block system of Irrigation, Invention of
the automatic gates are well known. Under his leadership considerable progress in
research in the use of surki mortar for construction of hydraulic structures, gauging of
rivers, evaporation and seepage losses, etc,.
Ganesh Iyer during 1930's initiated research and experimentation on Volute siphons.
In the development of canal falls, the Ogee type was in use as early as 1870. The
trapezoidal notch fall was developed by 1894. With the mechanism of the energy of
flowing water and the formation of the standing wave becoming known better, the
standing wave flume type of fall was developed by Inglis by 1930.
Numerous investigators worked on the theory of the hydraulic jump which has helped
immensely in tackling various hydraulic problems. Important investigators on this
problem were Inglis and J oglekar (1924  1940), Coyler (1926), Lindley (1927), Montagu
(1929) and Crump (1930). Energy dissipation works below river and canal structures by
means of a cistern with baffles, deflectors and blocks were evolved with the help of
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
model experiments by Bhandari and Uppal (1938) in the Punjab and by Inglis in Poona
(1935).
The control of rivers flowing through bridges and other structures by a system of guide
banks, first introduced by Bell in 1888, has subsequently been investigated extensively,
both on the model and in the field, and the system is now widely in use.
Losses by evaporation and percolation in canals were investigated by Kennedy on the
Bari Doab Canal as early as 1882 and further work was carried out by various
engineers.
The special Irrigation Research Division, created in the Bombay P.W.D. in 1916,
through efforts of Inglis, contributed a great deal in the field of organised irrigation
research. During 19161928, valuable investigations were made on the problems of
land drainage and reclamation, canal losses, canal lining, weed growth and improved
irrigation methods. In the field of hydrodynamic research with the aid of hydraulic
models, experiments on standing wave flumes, energy dissipation devices below falls,
cutwater and easewater experiments for the best design of Sukkur Barrage piers are
few examples of early organised research.
With the realisation of the importance of model investigations, research centres at
Poona and Lahore were developed and new Research station started in United
Provinces (1938) and some other states. The attainment of Independence and
formulation of plans for a number of River valley Projects posed a multiplicity of
problems and it became necessary to expand the facilities at the existing research
centres and to open new centres of research, today, laboratories equipped for dealing
with the problems connected with River Valley Projects, including reservoir surveys,
testing of soils, concrete and other construction materials have been set up in most of
the states.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1.3.2 COORDINATION OF RESEARCH
The creation of the Central Board of Irrigation in 1927 was a sequel to the realisation of
the need for coordinating research activities at various centres. After Independence,
with growing realisation of the need for development of power the Board was
redesignated as the Central Board of Irrigation and Power. In addition, it coordinates
the national activities and functions as Indian National Committee for the International
Commission on Large Dams (ICOLD), International Commission on Irrigation and
Drainage (ICID), International Association for Hydraulic Research (IAHR), International
Water Resources Association (IWRA) and International Conference on Large High
Voltage Electric System (CIGRE). The board also actively collaborates with the Bureau
of Indian Standards, the Central Road Research Institute, the Council of Scientific and
Industrial Research, the Indian Council of Agricultural Research, the Department of
Science and Technology, the Seven Indian Institutes of Technology, the council of
Technology Education, Indian Institute of Science.
On the recommendations of an expert committee appointed by the board in 1958, a
scheme of research on fundamental and basic problems, relating to river valley projects
and flood control works was sanctioned. To start with 12 main topics were included for
study under the scheme. Till 1980's, the work under the scheme has increased to the
extent that there are 44 main topics presently under study at 16 State and Central
Research Stations and 12 technical institutions under the supervisory control of the
Board. The Board publishes every year the Annual Review Summaries of the work done
on these problems. A quarterly journal 'Irrigation and Power' brought out by the Board
contains papers on both basic and applied research in water and power engineering.
The papers contributed and discussed at Annual Research Sessions are brought out as
proceedings of these sessions.
Besides the journal and proceedings, publication of important researches relating to
specific subjects carried out by individuals or institutions are compiled as Board's
publications and these form useful authentic reference manuals with the irrigation and
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
power engineers of the country. As part of the Research Studies the research stations
have prepared Reviews with Bibliographies as well as status reports on a number of
topics. These are also issued as publications of the Board. In late 70's a new periodical
'Irrigation and Power Research Digest' has been started to furnish the latest research
work done at various research stations to the research community.
1.3.3 RESEARCH ACTIVITY IN INDIA TODAY
There were sixteen major research stations in India (in 1980's) which were undertaking
research studies on various aspects of river valley developments and which usually
participated in the Research Scheme applied to River Valley Projects. A number of
technical institutions are also associated with this programme and they are mostly
tackling the problems with a great academic bias. The background and the special
features of some of the State and Central Government research stations are given
below.
(1) Andhra Pradesh Engineering Research Laboratory, Hyderabad
The Engineering Research Department, established by then Hyderabad State
Government in the year 1945 became the Research Laboratories of Andhra Pradesh
when the new state was formed in November 1956.
(2) Central Soil and Materials Research Station, New Delhi
To meet the need for research wing, for soils and material testing on the pattern of the
central water and Power Research Station, Pune, (Described subsequently) the Central
Soil and Materials Research Station came into existence at New Delhi during the year
195354. The research station undertakes field and laboratory investigations for river
valley and other projects in the disciplines of soil mechanics, rock mechanics, concrete
technology, sediment investigation, preirrigation soil surveys and chemical analysis of
construction materials. The station has extended its service of consultancy to a number
of foreign countries including Bhutan, Nepal and Afghanistan. Highly sophisticated
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
testing facilities such as 1,000 tonne testing machine, have been installed and it is one
of the best equipped laboratory of the country in its field.
(3) Central Water and Power Research Station, Pune
As a sequel to the need for organised research, a special Irrigation Research Division
was created under the auspices of Bombay P.W.D. in 1916, by the efforts of Sir C.C.
Inglis, who did pioneering work on various aspects of the irrigation problems and laid
the foundation of organised research in the country. Problems concerning laid drainage
and reclamation, canal losses, canal lining and improved irrigation works were taken for
investigation. Soon the Research Division expanded its activities in new branches and
this centre was subsequently taken over by Government of India in 1937. Irrigation and
river training research were added to its scope and was renamed as 'Indian Waterways
Experiment Station'. In 194647, the expansion and reorganisation of the station was
sanctioned with seven new branches for dealing with navigation, soils, materials of
construction, statistics, physics, mathematics, hydraulic machinery research problems.
The station was redesignated the 'Central Water and Power Research Station' and
brought under the administrative control of Central Water Commission. The quality of
research work turned out by the Research Station won it acclaim not only within the
country but abroad as well. In recognition of the tremendous progress made, it has been
chosen as Regional Laboratory for the United Nations Economic Commission for Asia
and Far east. CWPRS has extreme built up expertise in many fields during its life span
of more than 85 years. Some of the notables are: hydraulic structures, earth sciences,
ship model testing, coastal engineering and the application of methods from the
different disciplines of physics, chemistry, mathematics, statistics, botany, geology,
instrumentation and computer science.
The station extends its activities to prototype testing, digital data acquisition, field
investigations, testing of turbine and pump models in cavitation tanks and developing
techniques for the use of radioactive and fluorescent tracers in tidal as well as fluvial
flow conditions for various purposes.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The station has been offering technical assistance and consultancy services to other
countries also, which include Burma, Afghanistan, Tanzania, Iraq, Philippines,
Singapore, Libya, Nepal, Sri Lanka, Egypt and Zambia. Notable engineers from these
stations are Sir C.C. Inglis and Dr. D.V. J oglekar.
(4) Gujarat Engineering Research Institute, Vadodara
On the bifurication of the Bombay State, the development and Research Division at
Vadodara, which was a branch of the Central Research Institute, Nasik was transferred
to the Gujarat State in 1960 and was renamed as Gujarat Engineering Research
Institute, with headquarters at Vadodara. The institute's major contribution related to
the study of ground water flow and its recharge, river training, sediment studies in canal
and reservoirs, canal lining, soil mechanics and materials testing specially pozzolana.
(5) Hirakud Research Station, Hirakud, Orissa
During the planning of the Hirakud Dam Project in 1947, this research station was
started at the dam site for observations of data on the silt load of the Mahanadi and for
testing construction materials for the project. Subsequently, this station was expanded
to take up the quality control work during the construction and for the fixing and
observations of the instruments provided both in the earth dam and the masonry and
concrete dams. With the transfer of this station, along with the Hirakud Dam Project to
the Government of Orissa in April 1960, the activity of the Research Station has been
extended to cover the whole of the Orissa State.
A Masonry Testing Unit for testing large size masonry and concrete blocks, has been
set up about 11.3 Km away and it is one of the few such units in the country.
The Station also undertakes the sedimentation survey of the Hirakud Reservoir by
echosounding.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(6) Institute of Hydraulics and Hydrology, Poondi (Tamil Nadu)
Abundant water and land becoming available with completion of the Poondi Reservoir
Irrigation Research Station came into being at Poondi, 60 Km from Chennai, in April
1944.
This Research Station deals with all hydraulic problems of the river valley and flood
control projects. Tshaped blocks have been evolved for effective and economic
dissipation of energy below spillways. Implemented in Bhavani Sagar project. Similarly,
lined canal chutes have been developed and considerable savings have been effected
in the cost of the Lower Bhavani Project Canal System by work at this Station. A special
mention may be made of the studies conducted for the improvement of the coefficient of
discharge of tank weirs, which has enabled the irrigation of additional areas from the
remodeling of a large number of tanks in the Tamil Nadu State.
The Irrigation Research Station was functioning as a part of the State Public Works
Department and as such it concentrated on applied research having relevance to the
immediate functional needs of the department. Observing the switchover from hydraulic
to hydrologic research all over the world urgent need was felt to bring about a change in
the outlook of this statement also.
The station was upgraded into a full fledged Institute of Hydraulics and Hydrology in the
year 1973 making it possible to deal with problems in ground water and coastal
hydrology and surface water management using computer simulation methods, system
analysis and the like.
The need for instrumentation, especially on the electronics side had also been realised
fully. As a result an electronic laboratory has been established.
The activities of the Institute are spread over area of Ground Water Hydrology,
Hydrology of River Basins including Flood Prediction, Hydrological Modeling,
Instrumentation and Water shed Management Schemes.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(7) Irrigation Research Directorate, Bhopal
The research station has been started in 1964. It is mainly dealing with hydraulics, soils,
and concrete and model prototype conformity problems.
(8) Irrigation Research, Jaipur
With the advent of irrigation projects in the State of Rajasthan and use of local materials
for the constructional purposes, the Irrigation Research has been conceived.
(9) Irrigation Research Institute, Khagaul, Patna
The research station was opened in 1956 at Khagaul, 10 km from Patna. The Institute
has done considerable work on soil, use of micaceous sand in mortar and concrete, and
other construction material problems. It has recently taken up studies regarding
sedimentation survey of reservoirs and ground water problems including optimum
spacing of tubewells in various regions of Bihar State.
(10) Kerala Engineering Research Institute, Peechi (Kerala)
On the formation of the Kerala State on 1 November 1956, the systematic and intensive
development of the water resources of the state assumed great importance.
The State Government sanctioned a Research Institute in Kerala which started
functioning on J une 1960.
The main Research Institute is located at the foot of the Peechi Dam, about 22.5 km
from Trichur.
Being a coastal State the Institute has mainly concentrated on the problem of coastal
erosion and has evolved cheaper designs of sea walls which have been constructed to
protect the land against sea erosion successfully. Other studies being carried out are
use of laterite as pozzolana, water requirement for rice, etc.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(11) Karnataka Engineering Research Station, Krishnarajasagar
Though it started as a small section attached to Gauging SubDivision dealing with
hydraulic investigations only, with the increase in demand for the testing of soils and
various engineering materials, the Soil Mechanics Branch and the Material testing
Branch were added during 1940.
The Hydraulic Research Station was later strengthened in 1945 and made a seperate
wing of Public Works Department under the direct administrative control of the chief
Engineer and redesignated 'Mysore Engineering Research Station'. During 1974 due to
the redesignation of Mysore State to Karnataka State, the station was also redesignated
'Karnataka Engineering Research Station'.
The outdoor hydraulic laboratory and the indoor laboratories (material testing, soil
mechanics, chemical, road research, etc.) are all located at Krishnarajasagar, just below
the Krishanarajasagar Dam overlooking the famous Brindavan Gardens.
One of the important contributions from this Research Station has been the
development of the volute siphons, initially designed and promoted by Ganesh Iyer, an
eminent engineer of the Mysore State. One of the important studies carried out by this
Research Station in collaboration, with other research stations was to determine the
prototype behaviour of the siphons when running full under likely cavitation conditions
under excessive head.
Other notable studies carried out by this Research Station are the twin surge tanks, the
approach channel to the Vodenbyle twin tunnel, and the surplussing arrangements of
the Linganamakki Talakalale, Kali Complex and other projects of the state. Experiments
for restriction of evaporation, cheaper canal lining, model prototype conformity,
sedimentation survey of reservoirs, problems of soil mechanics, materials testing and
rock mechanics are some other important achievements of the station.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
During 1971, an Engineering Staff Training College has been started under aegis of
Karnataka Engineering Research Station, to impart training to in service engineers of
P.W.D. by running shortterm and longterm refresher courses.
(12) Land Reclamation, Irrigation and Power Research Institute,
Punjab, Amritsar
Around the year 1925, the Government of Punjab constituted a Water logging Enquiry
Committee to study and report on the extent and causes of water logging in irrigated
areas and the preventive measures which should be adopted. A small farm at
Chakanwali for field experiments regarding the reclamation of waterlogged areas and a
laboratory at Lahore for the analysis of soil and water sampleslater designated as the
'Scientifc Research Laboratory' was set up in this connection.
In 1931, the Hydraulic Section was started and, by 1932, under the redesignated name
'Irrigation Research Institute, Lahore' there were six independent Sections: Hydraulics,
Physics, Chemical, Statistical, Mathematical and Land Reclamation. During the next 15
years, the Institute was able to carry out a great deal of work which gained recognition
in the scientific and engineering circles.
The Hydraulic Section initiated (1932) smallscale model experiments for tracing subsoil
flow under structures on permeable foundations, by treating the sand in the model with
a chemical and allowing another chemical to flow from one side of the work to the other
through the sand. Arrangements were made to measure the pressures under the work
at different points. The comparisons of the results with theoretical expectations pointed
to the need for a mathematical technique to give more exact results and standard cases
were successfully tackled from 1936 to 1940 to obtain the effects of various
components of a structure on the pressure distribution under it. The physics section
developed, at the same time, the electric analogy model for a rapid determination of the
pressure distribution comparable with those given by theory and the hydraulic model.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In 1936, Khosla put forward his 'method of independent variables' for determining the
pressure distribution based on the concept that each component had an individual effect
and the superposing of these individual effects have the overall effect. The theoretical
results and the laboratory experiments were used to verify and, where necessary,
modify Khosla's method, which ultimately became the standard method, which
ultimately became the standard method for the design of works on permeable
foundations. This was indeed a signal contribution by a cooperative group of Indian
workers to a difficult engineering problem.
Dr. A.N. Khosla made a name in the the field of Research through his work on seepage
theory and design of weirs on permeable foundations. He was appointed the first
chairmen of the newly constituted Central Waterways, Irrigation and Navigation
Commission in 1945 and developed it into a front rank organisation. When Bhakra
control of board was set up in 1950, Dr Khosla was appointed its Vice Chairman and
Chairman of the board of Consultants. He remained associated with the project till its
commissioning in 1963.He served as the Vice Chancellor of the Roorkee University
from 1954 to 1959 and virtually transformed it from a small though reputed college to a
leading technical university. In 1962 he was appointed as Governor of Orissa, the first
and so far the only professional engineer to have been given such a responsibility.
Another name worth noting is that of Dr Kanwar Sain. He was responsible for planning
of the gigantic planning of the gigantic Rajasthan Canal project still under completion.
For nine years he worked on the planning of the complex Mekong River project under
the auspices of the United Nations.
Another important contribution of those years was in regard to the design of stable
channels in alluvium. The Institute developed, for the first time, appropriate scientific
instruments capable of collecting and analysing samples of silt from irrigation channels.
The results of analysis were processed to obtain the mean size of the silt and to
correlate it with the other hydraulic elements of the channel.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Another field of study related to the engineering works connected with the control and
training of rivers. This required comparatively largescale methods and a field research
station was opened at Malakpur in Gurdaspur District where the requisite facilities were
available. This station, which was started around 1934, subsequently grew into one of
the most advanced station in India and handled the model work for most of the
important projects in the Punjab.
Yet another development was the largescale work on land reclamation undertaken by
Punjab Government in 1940. This ultimately led to a seperate department of Land
Reclamation being formed under a 'Director, Land Reclamation'.
Immediately after partition in 1947, East Punjab set up a new Institute at Amritsar and
work at the Malakpur Station was continued. Since then, the institute has grown
considerably and has now been made a zonal institute for the North Zone, consisting of
Himachal Pradesh, J ammu and Kashmir, Punjab and Rajasthan.
In the field of hydraulics, a substantial contribution was made in regard to the design of
spillway and outlets for Bhakra and Nangal Dams and of the flood control, drainage and
reclamation problems of Kashmir Valley.
The Hydraulic Research Station, Malakpur has been recognised to help and solve many
complicated problems in connection with Beas Dam at Pong, Beas Sutlej LinkPart II,
Sirhind, Ferozpur and Rajasthan Feeders and recently for Shah Nahar Project,
Anandpur Hydel Project, Mukerian Hydel Project and the prestigious Thein Dam and its
appurtenant works. The station specializes in developing sediment excluding devices
from rivers and channels.
A Field Lining Research Station has been set up at Doburji (Near Amritsar) for
Investigations relating to the economical specifications of lining material for reducing
seepage from the earthen channels and water courses. Research for development of
pressure release values behind canal lining is also being undertaken at this station.
Excellent work regarding vortex suppressors in the intake has been carried out.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(13) Maharashtra Engineering Research Institute, Nasik
Considering the importance of achieving efficiency, economy and progress of large
development works undertaken in the Bombay State, the State Government approved
the creation of a Central Engineering Research Institute, and it was set up with
headquarters at Nasik in 1959. On the creation of Maharashtra State and bifurcation of
research station it has been redesignated 'Maharashtra Engineering Research Institute'.
The Institute carries out investigations on soil mechanics, materials testing, hydro
dynamic problems and public health and rural engineering. The Institute specializes in
Environmental Engineering with special reference to water quality and its measurement
throughout Maharashtra State. Recently field studies have been conducted on
breaching and dismantling of Old Waghad Dam.
The Soil Survey Division at Poona does systematic soil surveys of the areas under the
command of various irrigation projects in the state.
(14) River Research Institute, West Bengal, Kolkata
Due principally to the abandonment of the BhagirathiHoogly course by the Ganga,
many of the rivers of West Bengal have decayed and the drainage of West Bengal
during the flood Season has been seriously affected. A Research Station to study the
various river problems and to evolve measures for controlling the destructive causes of
the dying rivers was set up in the State in the year 1943.
Investigations for foundations of hydraulic structures for borrow materials for
construction of dams and soil surveys for irrigation projects have also been taken up.
Facilities are also available for conducting aggregate and concrete tests. With the
passage of time the institute has acquired specialization in a number of fields such as
River training for the purpose of conservancy of the river, prevention of erosion and
flooding, Navigation and irrigation, Design of channels, Meandering of streams and
conservation of tidal rivers, Tidal computation, closure of estuaries, tidal channel and
reclamation and Engineering properties of soils.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(15) Soil Mechanics and Research Division, Chennai
The Research Station was initially formed as Physics and Soil Mechanics Office in
1946. The Concrete Laboratory was established in 1947. In 1953 the two were merged
to function as "Soil Mechanics and Research Division" of the Tamil Nadu Public Works
Department. The Research Station had the benefit of guidance of K.L. Rao, the noted
engineer statesman in the early stages.
The laboratory has successfully evolved Ennore sand as the Indian standard sand. This
sand is now supplied to engineering research institutions and cement factories all over
India and has resulted in considerable saving of foreign exchange.
The laboratory, in its thirty years of useful service has made significant contributions in
the various fields of engineering research. Intensive soil investigation work has been
carried out for all the irrigation projects executed in the state, regular quality control
work has been organised. For building works, regular foundation analysis by load tests
has been carried out for almost all major buildings. The station has done notable work
on Design of Weirs on permeable Foundations of Finite Depth.
(16) Uttaranchal Irrigation Research Institute, Roorkee
A small Hydraulic station was established at Lucknow in 1938 to study the problems of
scour and erosion below falls and bridges on irrigation channels. To meet the needs of
an increasing number of problems, an Irrigation Research Station at Bahadrabad, about
20 km from Roorkee, started functioning in 1947. This Station was further expanded in
1955. Earlier it was known as Uttar Pradesh Irrigation Research Institute, Roorkee.
The activities of this Institute cover both basic and applied problems in hydraulics, soil
mechanics, ground water, mathematics, physics, instrumentation, hydrology and
measurement of discharges of rivers and canals. Specific problems concerning the
development projects, such as river training and protection works, soils and construction
material problems, etc., constitute its main activities, but the station has been also doing
remarkable basic research work in a number of fields.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Few of the important contributions of the Institute relates to the design of the 1.8 m high
dentated sill for dissipating energy below Sarda Barrage sluices, which had collapsed
during the floods of 1956. This was the first kind successfully tested and adopted in
India under boulder river conditions.
Hydraulic design of Surge tanks for all major projects constructed / under construction in
Himalayan region and its computer simulation, design of gravel pack and prepacked
filter for tube wells, design of stilling basin for low Froude Number, design of stilling
basin for low Froude Number, design of guide bunds at bridges and barrages, intake
structures, stilling basins, design of bifurcations and trifurcations for tunnels, assortment
of river training problems, prototype load test, design of channels and evolving formula
for design of channels, design of structures founded on stratified soils, design of
barrages and canal regulators on threedimensional flow consideration, etc., are a few
of the fields of the specialization of the Institute. The Institute offers technical assistance
not only to State Irrigation Department but to other States and departments. The
Institute also takes up the foundation investigations for dams, power houses and other
hydraulic structures, Instrumentation in dams, in situ testing of rocks and model
prototype conformity studies. Recently due to reorganisation of states, this is now in
Uttraranchal.
Reference
Water Resources Research in India, Publication No. 78 (Revised) CBI&P, New Delhi,
1979.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1.4 CLASSIFICATION OF FLOW
Uniform flow Nonuniform flow
Subcritical
Critical
Supercritical
Steady Unsteady
Gradually varied flow
Rapidly varied flow
Spatially varied flow
Fluid flow
Froude number
Reynolds number
Spatial
Temporal
Compressible / incompressible
Pressure Flow
Free Surface Flow
Single phase
Two phase
Multi phase
Reciprocating upstream flow
Unidirectional upstream flow
Highly irregular
Highly variable upstream flow
Fluid flow
One dimensional
Two dimensional
Three dimensional
Classification of flow is done based on different criteria. A brief description of the
classification is given in the following paragraphs.
CLICK ON THE TITLE FOR FURTHER DETAILS
(a) Based on Ideal and Real fluid flows
(b) Pressure flow and Gravity flow
(c) Based on ratio of Inertial and Gravitational forces
(d) Based on Inertial and Viscous force ratio
(e) Compressible and Incompressible flow
(f) Based on Spatial variations
(g) Based on dimensions
(h) Based on Time
(i) Based on Rotational and Irrotational flows
(j) Based on Mono phase and Multi phase flows
(k) Based on Stratification
Examples of some combination of flows
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2. CHANNELS AND THEIR GEOMETRIC PROPERTIES
2.1 INTRODUCTION
An open channel is a physical system in which water flows with a free surface at the
atmospheric pressure. In other words the pressure is impressed on free surface. A
channel can be classified as either natural or artificial channel according to its origin.
Natural channels include all watercourses of varying sizes from tiny hillside rivulets,
streams, small and large rivers to tidal estuaries that exist naturally on the earth.
Subsurface streams carrying water with a free surface are also treated as natural open
channels.
The cross sections of natural channel are irregular and hence hydraulic properties may
vary from section to section, and reach to reach. A comprehensive study of the behavior
of flow in natural channels (the mobile boundaries) requires knowledge of other fields,
such as hydrology, geomorphology and sediment transportation. Generally, these
aspects are dealt in detail in river mechanics (fluvial hydraulics).
Artificial channels are those constructed or developed by human effort such as gutters,
drainage, ditches, floodways, tunnels, log chutes, navigation channels, power canals
and trough, spillways including model channels that are built in the laboratory for
experimental investigation studies. Long distance canals have been constructed to
achieve the interbasin transfer of water at National and International levels.
The artificial channel is known by different names, such as " canal "," chute", "culvert",
"drop", "flumes" and "open  flow tunnel", Aqueduct.
However, these names, are used rather loosely and can be defined only in very general
manner.
The canal is usually a long and mildsloped channel built in the ground, which may be
lined or unlined with stone masonry, concrete, cement, wood or bituminous materials
etc.
Eg: Ganga Canal, Indira Gandhi Canal, Narmada Canal.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The chutes are a channel having steep slopes. The culvert, flowing partly full, is a
covered channel of comparatively short length provided for draining water across
roadways and through railway embankments.
The drop is similar to chute, but the change in elevation is effected with in a short
distance.
The flume is a channel of wood, metal, fiber reinforced plastic, concrete, or masonry,
usually supported on or above the surface of the ground to carry water across a
depression.
The open flow tunnel, fall, is a comparatively long covered channel used for carry water
through a hill or any obstruction on the ground. Normally these artificial canals are with
rigid boundaries.
The channels can be classified as prismatic and nonprismatic. A channel built with
constant cross section and constant bottom slope and fixed alignment is named as
prismatic channel. Otherwise, the channel is nonprismatic.
Example: spillway having variable width and canals curved alignment. (Meandering).
The term channel section refers to the cross section of channel taken normal to the
direction of the flow.
A vertical channel section, however, is the vertical section passing through the lowest or
bottom point of the channel section. For horizontal channels, therefore, the channel
section is always a vertical channel section.
Natural sections are in general very irregular, usually varying from an approximate
parabola to an approximate trapezoid shapes and for streams subject to frequent
floods, the channel may consist of a main channel section carrying normal discharges
and one or more side channel sections for accommodating overflows. These are called
compound channel.
Artificial channels are usually designed with sections of regular geometrical shapes.
Table gives the geometric properties for the cases of rectangular, trapezoidal,
triangular, circular, parabolic channels. In addition the details of Round bottomed
triangular and round bottom rectangular are also given.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2.1.2 Geometrical Properties
Unlined trapezoidal section is the most common channel section used in the field for it
provides side slopes for stability. The rectangular channel with an angle 90 and
triangular channel with a bed width equal to zero are special cases of the trapezoidal
channel. Since the rectangular channel has vertical sides, it is commonly used for
channels built of materials, such as lined masonry, rocks, metal, or timber. Precast
concrete sections are also used for small size canals. The triangular section is used
only for small ditches, roadside gutters, and for experimental investigations in the
laboratory. The circular shape is the popular section for sewers and culverts of small
and medium sizes. The parabola is used as an approximation for section of small and
medium size natural channels. Practical sections are also used as shown in figure (as
recommended by Central Board of Irrigation and Power).
b
1
m
1
m
Lined channel section for Q >55 m
3
/s
y
1 1
y
y y
1 1
A = by + y
2
( 1+ Cot1)
P = b + 2y( 1+ Cot1)
R = by + y
2
( 1+ Cot1)
b + 2y ( 1+ Cot1)
__________________
1
m
m
1
0
Lined channel section
for Q <55 m
3
/ s
y
y
1
1
2
1 A = 2(1+y
2
Cot1) + y
2
21
1
2
__
= y
2
(1+Cot1)
P=2yCot
1
+2y
1
= 2y(
1
+Cot
1
)
R=
A
P
__
y
2
(1+Cot1)
2y(1+Cot1)
____________
=
y
2
__
=
y
Closed geometric sections other than circular section are frequently used in sewer
system, particularly for sewers large enough for a person to enter. These sections are
given various names according to their form, they may be eggshaped, ovoid,
Semielliptical, Ushaped, catenary, horseshoe, baskethandle, etc. The complete
rectangular and square are also common for large sewers.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
A special geometric section known as hydrostatic catenary or lintearia is the shape of
the cross section of trough, formed of flexible sheets of negligible weight, filled with
water upto the top of the section, and firmly supported at the upper edges of the sides
but with no effects of fixation. The hydrostatic catenary has been used for the design of
the section of some elevated irrigation flumes in UK (United Kingdom). These flumes
are constructed of metal plates so thin that their weight is negligible, and are firmly fixed
to beams at the upper edges.
Hydrostatic Catenary
Cartesian equation: y = a cosh(x/a)
Click here for Geometric elements of channel sections
Geometric elements are properties of a channel section that may be defined entirely by
the geometry of the section and the depth of flow. These elements are extensively used
in computations of flows.
The geometric elements for simple regular channel sections can be expressed
mathematically in terms of the depth of flow and other dimensions of the section. For
complicated sections and sections of natural streams, however, no simple formula can
be written to express these elements, but graphs representing the relation between
these elements and the depth of flow can be prepared for use in hydraulic
computations.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2.1.3 Definitions of several geometric elements of basic importance
are given below
Depth of flow
The depth of flow y is the vertical distance from the lowest point of channel cross
section to the free surface. This term is often used interchangeably with the depth of
flow section. Strictly speaking, the depth of flow section is the depth of flow normal to
the direction of flow, or the height of the channel section containing the water. For a
channel with a longitudinal slope angle , it can be seen that the depth of flow is equal
to the depth of flow section divided by. In the case of steep channels, therefore, the two
terms should be used discriminately.
x
y
90
0
horizontal
Normal and vertical depths
Box
10 , cos =0.9848,thus there would be an error of 1.51%.
y =d cos
=
If x is measured along the horizontal direction instead of the sloping bed, then the
2% error occurs at about 11 S 0.20 or
= = , which is an extremely steep slope in open channels.
However, there is exception in cases such as spill ways, falls, chutes.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1
1
m =2
m =2
T
P
A Channel cross section
b
A =Area of flow
T =free surface width (m)
m =side slope defined in horizontal to 1 vertical; m:1
m =cot
l
m
0
P =Wetted perimeter is the boundary which is in
contact with the flow (m)
b =bed width in (m)
y =depth of flow
y
Stage
Datum
0
Water surface
Bed
Definition of stage
H (M.S.L)
(Above Mean Sea Level)
Datum
EL 210.00 m
EL 205.00 m
EL 200.00 m
The stage H is the elevation or vertical distance of the free surface above the datum. If
the lowest point of the section is chosen as the datum, the stage is identical with the
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
depth of flow. Free surface width T is the width of channel section at the free surface.
dA
T
dy
The water flow area A is the crosssectional area of the normal to the direction of flow.
The wetted perimeter P is the length of the line of intersection of the channel wetted
surface with a cross sectional plane normal to the direction of flow.
The hydraulic mean radius R is the ratio of the water flow area to its wetted perimeter,
A
R =
P
When a shallow channel of b is used and
b
y thenR
2
.
b
b
__
2
y
R
Hydraulic mean radius
Wide Rectangular
y
y
R
dR
dy
__
Trapezoidal
y then R
b
__
2
b R
y R
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The hydraulic mean depth D is the ratio of the water area to the free surface
width,
A
D=
T
. The section factor for critical flow computation m is the product of the
water area and the square root of the hydraulic depth,
A
Z = A D = A
T
. The section
factor S.F for uniformflow computation in case of Manning formula is the product of the
water area and the twothirds power of the hydraulic radius
2
3
S.F = AR other wise for
chezy's formula it is i.e.,
2
3
AR . The details of circular channel are given in OPEN 
CHANNEL HYDRAULICS by VEN TE CHOW  pp 632  639(1959).
Earlier the nomographs for trapezoidal and parabolic sections were used for specific
side slopes see reference. The geometrical characteristic of the irregular cross section
can be obtained using a set of co  ordinates describing the cross section, with the help
of interpolation between any inter mediate depth. The typical programme is given in the
appendix. The computations can be done either by from top or from the bottom most
point.
Actual area up to depth y =Total area A  dA
Area up to (y +dy) =Area up to y +dA
y
distance
dy
River bed elevation as a fuction of distance from
the river bank
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2.1.4 Circular channel
Normalised geometric characteristics are shown in figure. When the flow is full the
hydraulic mean radius is (i.e)
2
d
d 1 A
4
d = =
4 P d 4
for
maximum value of
2
3
AR when the discharge is maximum.
do
y
T ___
do
Z
do
2.5
___
D ___
do
A
___
Ao
P ___
Po
R
Ro
___
P
o
=
d
o
A
0
=
d
o
2
___
4
R
0
=
d
o ___
4
Normalized geometric elements for a circular section
Subscript zero indicates full flow condtion
0 0.1 0.2 0.3 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 0.4
0.1
0.2
0.3
0.5
0.6
0.7
0.8
0.9
1.0
0.4
y
__
d
0
Problem: Write a computer program to obtain the geometrical elements of a circular
shape channel and obtain the
2/3
2/3
0
0 0
y AR
Vs
d
A R
Compute the geometric elements, area, hydraulic mean radius, hydraulic mean depth
for the following cases:
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Rectangular channel: Bed width is 10 m, Depth of embankment is 15.15 m, Depth of
flow is 8.870 m.
Trapezoidal channel: Bed width is 10 m, Depth of embankment is 15.15 m, Depth of
flow is 7.77 m, side slope m:1 =2:1.
Triangular channel: Depth of embankment is 15.15 m, Depth of flow is 9.75 m, side
slope m:1 =2:1.
Circular channel: Diameter is 15.15 m, Depth of flow is 6.47 m.
2.1.5 Natural channel
The depth of flow 7.567 m.
The program could be developed using spread sheet.
The INPUT for the natural channel is as follows
Distance of the embankments form reference Stage of flow (m)
Left embankment Right embankment (m)
2.000 10.000 10.000
3.000 9.000 11.000
4.000 8.500 12.500
5.000 8.000 13.600
6.000 7.000 15.000
7.000 6.300 16.900
8.000 5.400 18.000
9.000 5.000 19.500
10.000 4.300 21.000
11.000 3.900 22.000
12.000 3.000 23.300
13.000 2.700 25.000
14.000 2.200 26.300
15.000 1.900 27.000
16.000 1.300 28.200
17.000 1.000 29.000
18.150 0.700 30.000
The depth of flow =7.567 m
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Solution:
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Distance from reference (m)
Natural channel
0.0 12.0 18.0
24.0 30.0 6.0
4
8
12
16
20
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Variation of area with depth of flow
Depth of flow (m)
0.0
3.0
6.0
9.0
12.0
15.0
100.0
200.0
300.0
400.0
500.0
600.0
Natural
Triangular
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
* *
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
Variation of hydraulic mean depth with depth of flow
Depth of flow (m)
0.0
3.0 6.0
9.0
12.0 15.0
10.0
20.0
30.0
40.0
50.0
60.0
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3.0 6.0 9.0
12.0
15.0
2.0
4.0
6.0
8.0
*
*
*
*
*
*
*
*
*
*
*
*
*
Depth of flow (m)
Variation of Hydraulic radius with the depth of flow
Table showing the geometrical elements for the above channels (metric units)
Section
y A P T R D
Z A D =
Trapezoidal 7.77 198.800 44.748 41.080 4.434 4.830 437.665
Rectangular 8.870 88.700 27.740 10.000 3.196 8.870 264.316
Triangular 9.750 190.500 43.603 39.000 4.360 4.875 421.324
Circular 6.470 73.488 21.575 14.954 3.397 4.910 163.428
Natural 7.567 58.895 39.0007 15.747 1.504 3.724 114.067
Problem: Compute the geometric elements for the horse shoe tunnel shown in figure
below. Plot the normalised graphs representing the geometrical elements.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Horse shoe tunnel
d
0
If d
0
is 10 m and the depth of flow 7.5 m, what would be the area of flow, wetted
perimeter, hydraulic mean radius, section factor for uniform flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2.2 Verify the geometrical elements for Circular channel
2
2 2
4
=
=
2
2
2
2
2
0
2 2
2
Top width of water T
d T
y  = r
2 2
d T
r y 
4 2
d d
T = 2 y 
2 2
d d d
T = 2 y +  2y
4 2
or
2
d
= 2
4
2
2
d
y
4
( ) ( )
180
2 2
=
=
2yd
2
T = 2 y y +d T = 2 y d y
T
sin ( 180  ) r
2 2
or T = d sin d sin
Area of flow = Area of circle  Area above the chord
or
180
2
180 180
2 2 2
42
=
= = =
= =
d T
Area of triangle = x y 
2 2
d d
sin y 
2 2
d
y 
d d d
2
cos( ) or y  cos( ) cos
r 2 2 2
Area of full circle
Area for = x
2
d d
.
( )
8
8 8
8
2 2
2
d d
sin
d
Area of flow = sin
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
( )
2
2
2
2
1
8
4
2
1
2
2
2
1
8
= =
= = =
d d 1
P x
2 2A
sin
sin
d sin
R =
4
A
Z =A D A
T
1
sin d
A
8
D = =
T
d sin
d A sin
D =
T 8
sin
d sin
D =
8
sin
d
d A
R
d P
Z ( )
( )
15
05
2
2
32
2
.
5/2
0 .
d sin
sin d
8
sin
sin
d
sin
Z
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2.3 Difference between Pipe Flow and Free Surface Flow
OPEN CHANNEL FLOW PIPE FLOW
Defines as a passage in which liquid
flows with its upper surface exposed to
atmosphere.
The flow is due to gravity
Flow conditions are greatly influenced
by slope of the channel.
Hydraulic grade line coincides with the
water surface
The maximum velocity occurs at a little
distance below the water surface.
The shape of the velocity profile is
dependent on the channel roughness.
A pipe is a closed conduit which is
used for carrying fluids under pressure.
The flow in a pipe is termed as pipe
flow only when the fluid completely fills
the cross section & there is no free
surface of fluid.
Hydraulic grade line does not coincides
with the water surface.
The maximum velocity occurring at the
pipe centre.
Velocity Distribution is symmetrical
about the pipe axis.
Horizontal
TEL
HGL
Datum
PIPE AXIS
Velocity
head
Piezometric
head
(2)
Piezometer
Z
1
Z
2
hf V
1
2
____
2g
V
2
2
____
2g
P
1
____
P2
____
V
(1)
2.3.1 Hydraulic Grade Line (HGL)
Definition: A curve drawn above the datum which has ordinates equal to the piezometric
head at every point is called HGL or Hydraulic gradient.
The vertical intercept between the datum and pipe axis is the elevation head.
the datum and pressure gradient (HGL) is the peizometric head.
the pipe axis and the HGL is the pressure head.
HGL and TEL is the velocity head. Datum and TEL is the total head.
The TEL always falls on the direction of flow because of loss of head. The HGL may
rise or falls depending on the pressure variation in the pipe.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In a pipe of uniform section the velocity head remains the same, if the rate of flow is
constant. hence TEL and HGL are parallel if the pipe axis is horizontal.
HGL is always below the TEL. At point where pressure is equal to the atmospheric
pressure, HGL meets the pipe axis.
Shear stress distribution in pipe flow
Velocity distribution Shear stress distribution
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3.1 Continuity equation
Continuity equation represents the law of conservation of mass.
In general for unsteady flow the continuity equation is
(Mass flow rate into the system)  (Mass flow rate out of the system) =Rate of change
of storage.
For steady state condition
(Mass flow rate into the system)  (Mass flow rate out of the system) =0.
Example: Inflow: The flow that is coming into a system or an elemental volume such as
rainfall in y direction, flow entering into the river or a channel.
Outflow: The flow escaping from the system such as evaporation, seepage, water
released from a system.
Elemental volume
Inflow
Inflow
Outflow
Outflow
x
y
Generally, the mass balance is written in all the three directions namely x, y and z.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
u v w
0
x x x
inwhich
u,v andwarethevelocity components in x, y, z directions respectively,
is themass density of thefluid. If themass density is constant theabove
equationcan berewritte
+ + =
n as
u v w
0
x x x
If for one dimensional flowit reduces to
u
0
x
Mass
Massdensity
Volume
u
*elemental area=constant
x
Integrating one gets
UA =constant
Volumeratecouldbe express
v=0, w=0 i.e.,
+ + =
( )
3
3
edas m /s. This is generally known as
flow rateor discharge and expressed as cubic meter/second andis
abbreviatedas cumec (m /s).
Q Area*Velocity AV
Q
= =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3.2 ENERGY IN FREE SURFACE FLOW
It is known in basic fluid mechanics that the total energy in ( Newtonmeter per Newton )
of water along any streamline passing through a channel section may be expressed as
the total head in meter of water, which is equal to the sum of the elevation (above a
datum), the pressure head, and the velocity head. For example, with respect to the
datum plane, the total head H at a section containing point X on a streamline of flow in a
channel of large slope may be written as
2
V
H cos z
2g
x
x
d
x x
= + +
o
y = d cos
z
90
0
v
2g
Datum
Energy in gradually varied open channel flow
y
x
z
_
Total Energy Line
1
v
2g
__
1
1
y1 = d1 cos
v
__
1
z1
z2
y2 = d2 cos
v
2g
__
1
2
hf
__
Streamlines
Y
Y
Section YY
H
H = z + y +
v
2g
__2
2
2
2
in which z is the elevation of point Y above the datum plane, d is the depth of flow
normal to the bed, y is the vertical depth below the water surface measured at the
channel section, is the angle of the channel bottom with horizontal and
2
V
2g
is the
mean velocity head of the flow in the streamline passing through point X. In view of the
variation in velocity over the depth, the velocity head would be differing. The mean
velocity obtained by integrating the velocity distribution is considered for the entire
section
A
V = v dA
0
. In order to account for the variation of the velocity due to non uniform
pattern of velocity distribution, an energy correction factor is used.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Theoretical uniform flow
velocity distribution
(Ideal)
Linear
velocity
distribution
depth
of flow
y
y
Logarithmic
velocity
distribution
Power
Law
Typical velocity distribution
Inner wall Outer wall C
L
STATION A
STATION B
STATION D STATION C
ISOVELS in open channel bend [Normalised with V ]
Q = 83.5 lps, F = 0.41, R = 103460
e
max
1.30
Inner wall
Outer wall
C
L
Inner wall
Outer wall
C
L
Inner wall Outer wall
C
L
1.25
1.15
1.10
0.77
1.25
1.30
1.15
1.10
1.00
0.93 0.78
1.20
1.15
1.10
1.08
1.00
0.83
1.15
1.05
1.00
0.95
0.80
NonDimensionalised isovels (v/vmax)
0.740
0.775
0.813
0.85
0.888
0.905
0.998
0.665
0.628
0.687
0.722
0.765
0.791
0.825
0.860
0.895
0.963
Section A
Q = 33.61 l/s, F = 0.2457
Section 4
Re = 179574, n = 0.009834
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In general, every streamline passing through a channel section will have a different
velocity head, because of the non uniform velocity distribution in actual flow. Only in an
ideal parallel flow of uniform velocity distribution, can the velocity head be truly identical
for all points on the cross section. In the case of gradually varied flow, however, it may
be assumed, for practical purposes, that the velocity head for all points in a channel
section are equal, and the energy coefficient ( ) may be used for correcting for the
overall effect of the nonuniform velocity distribution. Thus, the total energy at the
channel section may be written as
2
V
H cos
2
d z
g
= + + .
y
90
0
d
Normal and Vertical depths
for channel of small slope, 0 thus, the total energy at the channel section is
2
V
H= y + z
2g
+ .
The slope of the energy line is denoted by S
f
, the slope of water surface is denoted by
S
w
and, the slope of the channel bottom by S = sin
= = = difference is 0.0027
cos = cos 0.9848 10
=
Thus there would be an error of 1.51 % when y d . If distance x is measured
along the horizontal instead of the sloping bed, then an error of order of 2%
occurs. If 11 S 0.20 or
1
2
Datum
y
1
1
2
Figure  Momentum equation
_
_
( )
2 1
2 1 1 2 f
Q
V  V =P P +WsinP
g
in which subscripts refer to sections 1 and 2; P
1
and P
2
are the resultants of pressure
forces acting on the two sections; W is the weight of water bounded between the
sections; and P
f
is the total external force due to friction and resistance acting along the
surface of contact between the water and the channel. The above equation is known as
the momentum equation and was first suggested by Belangar.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
For gradually varied flow, the values of P
1
and P
2
in the momentum equation may be
computed by assuming a hydrostatic pressure distribution. For a curvilinear or rapidly
varied flow, however, the pressure distribution is no longer hydrostatic; hence the
values of P
1
and P
2
cannot be so computed but are to be corrected. For simplicity, P
1
and P
2
may be replaced, respectively, by
1 1 2 2 1 2
P and P in which and are correction
coefficients at the two sections. These coefficients are called pressure distribution
coefficients. Since P
1
and P
2
are forces, the coefficients may be specifically called force
coefficients. It can be shown that the force coefficient may be expressed as
0 0
1 1
1
A A
h dA c dA
Az Az
= = +
in which z is the depth of the centroid of the water area A below the free surface, h is
the pressure head on the elementary area dA, and c is the pressure  head correction
factor. It can be shown that it is >1.0 for concave flow, <1.0 for convex flow, and equal
to 1.0 for parallel flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.1 Velocity Measurement and distribution
One of the basic components in Hydraulics is the understanding of velocity in the flow
field. Generally the average velocity or the mean velocity is computed using the
continuity equation namely
Q
v
A
= . The velocity varies locally and spatially depending on
the type of channel (straight, steep, bends, meandering, etc.,) and the flow (uniform,
nonuniform, laminar, turbulent etc.,). Therefore it is essential to measure the velocity
vectors in the flow field. There are different approaches for measurement of velocities.
a. Velocity measurements using Hydrogen bubble technique: This technique is used
basically for flow visualisation purposes in the laboratories.
b. Velocity measurement using Laser Doppler Velocimeter:
This is yet another technique to measure the flow field very precisely in the laboratory
using Laser Doppler Velocimeter. This can also give us the turbulence level. The
fundamental requirement for this is the transparent sides of the channel.
c. Velocity measurement in free surface flows in laboratories:
In general, in the laboratories and to an extent in the field, velocities can be measured
using different devices such as Pitot tube (One dimensional), Pitot cylinder (Two
dimensional) and Pitot Sphere (Three dimensional). However, these devices have their
limitations and are restricted to low velocity fields.
d. Stream Gauging:
In case of flow measurements in channels and in rivers different approaches are
adopted. Current meter is used in measuring the flow in canals and in rivers. While
using current meter it is necessary to calibrate. For this purpose the towing tank is used.
The details are given in different links. One of the very popular methods is the velocity
area method. Also float rods are used for estimating the surface velocity. In order to
understand the accuracy of measurements the error analysis is to be carried out.
Some typical Velocity distributions in a river are shown below:
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1.524 m
3.05 m
Godavari Western Main Canal  Mile 5/2
Q = 78.96 m
3
/s
Q = 70.68 m
3
/s
Q = 49.21 m
3
/s
0.0
1.0
2.0
0
47427 mm
45897.1 mm
0.0
0.5
1.0
1.5
0.0FT
1.524 m
3.05 m
Q = 145.96 m
3
/s
Q = 135.67 m
3
/s
Q = 129.73 m
3
/s
Godavari Western Main Canal  Mile 5/2
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0.883
0.898
0.213
0.93
0.96
0.990
0.998
0.805
0.775
0.742
0.62
0.948
0.960
0.970
0.98
0.99
SECTION 7
Q = 1.187 CFS, F = 0.2457, Re = 179574
SET III SECTION 0
O
NONDIMENSIONALISED ISOVELS (v/vmax)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Inner wall
Inner wall
Inner wall
Inner wall
Outer wall Outer wall
Outer wall
Outer wall
C
L
C
L
C
L
C
L
0.85
1.30
1.25
1.15
1.10
1.30
1.25
1.15
1.10
1.0
0.77
1.30
1.27
0.92
1.00
1.08
1.00
0.95
0.75
1.30
1.25
1.23
1.220
1.00
0.95
1.30
0.80
Station A Station B
Station C
Station D
Isovels [Normalised with Vmax] Q = 71.9 lps, F = 0.44, Re = 95420
Further the maximum velocity does not occur always at the free surface. It occurs below
the free surface due to presence of differential shear distribution on the boundary.
Hence secondary currents play an important role. The isovels reveal the presence of
secondary currents when there are more than one location of the maximum velocities.
Secondary
currents
Isovels
(a) Open channel
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The moving boat method, ultra violet measuring technique are the modern
measurements of measuring the flow. In order to access the water resources and to
have proper management it is essential to measure the discharges at various gauging
stations in rivers. This aspect is dealt in detail under river flow measurements.
Moving boat technique
Acoustic Doppler Current Meter
River Flow Measurements
Calibration characteristics of Current Meter: Rating of current meter is to be completed
before it is used in the field.
Towing Tank: Towing tank is used for calibrating (rating) the current meter.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.2 Discharge measurement by Velocity Area Method
(Chitale, 1974)
This method comprises measuring the mean velocity V and the flow area 'A' and
computing the discharge Q from the continuity equation. The site which satisfies the
requirements such as straightness, stability, uniformity of crosssection is chosen for
discharge measurement. The requirements of the site are dealt with in detail in
standards of the ISI 1192, (1959). The discharge measurement site is then marked by
aligning the observation crosssection normal to the flow direction.
The crosssection is demarcated by means of masonry or concrete pillars on both the
banks, two on each side 30 m apart.
ISI 1192, (1959), "Velocity area methods for measurement of flow of water in open
channels, Bureau of Indian Standards".
4.2.1 Segmentation
The interval at which the depth of water is measured along the crosssection for
channels with different widths is given in Table.
Description of Channel
(m)
Number of Observation
verticals
Maximum width of
segments (m)
Width less than 15 15 1.50
Width between 15 and 90 15 6.0
Width between 90 and
150
15 15.0
Width greater than 150 25 
The intervals specified are also such that not more than 10 percent and preferably not
more than 4 percent variations in the discharge between two adjacent segments occur.
The discharge through any segment is also not allowed to be more than 10 percent of
the total discharge.
For measurement of velocity, the maximum spacing between adjacent verticals is so
maintained that the mean velocity does not differ by more than 20 percent with respect
to the lower value of the two velocity measurements. In no case less than five velocity
verticals are permitted.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In case of canals allowing the variation of 2 percent in discharge and adopting as 15
verticals as a standard of comparison a lesser number of verticals 15 is adopted. The
verticals for depth and velocity measurements are kept the same according to Table
shown below.
Widths of segments for measurements of depths and velocities in canals
Channel capacity m
3
/s
Approximate surface
width (m)
Number of verticals for
depth and velocity
(a) Above 85 Above 35 11
(b) Between 1585 Between 15  35 9
(c) Between 0  15 Between 0  15 5
Method of marking segments varies according to the method of discharge observation.
Pivot point method is common, the details of which are available in the ISI : 11921959.
Angular, Stadia method and method of linear measurement are also used for locating
depth and velocity verticals under special circumstances.
4.2.2 Measurement of Depth
When velocities and depths are smaller and width up to 0.9 m, observations can be
made using wading or suspension rods. However, when wading observations are found
difficult, sounding rods of wood and bamboo are used. When depths are in excess of
about 4.6 m or current is too swift to permit the use of sounding rod, hand line is used
for depth measurement. But when the depths are large and velocities are high even the
handlines cannot be used. Under such circumstances a cable line is lowered by means
of a crane. Echo sounders of indicator as well as recorder type are being used for depth
measurements.
4.2.3 Measurement of Velocity
For the measurement of velocity the current meters are most commonly used. IS: 3910
 1966 gives specifications for cup type current meter and IS: 3918  1966 gives the
code of practice for use of this type of current meter.
To obtain a mean velocity in a vertical, velocity distribution observations can be made at
a number of points along the vertical. This is done when results are required to be
accurate, or for purpose of calibration. In twopoint method the velocity observations are
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
made at 0.2 and 0.8 depth below the surface while in one point method observations is
made at 0.6 depth below the surface. Both the twopoint and one point methods are in
common use in India, though subsurface method comprising making velocity
observations just below the surface is also used during floods when other methods are
not feasible.
In high floods at times, even surface measurement of velocity by current meter may not
be possible, float measurements are then used using surface floats, double floats or
special types of floats (IS 3911  1966). Velocity rods (IS 4858  1968) are also used
generally for velocity observations in canals. Details of the method are given in IS: 1192
(1959).
In adopting the float method or the surface velocity method in which current meter is
used, a reduction coefficient is used to change surface velocity into mean velocity in
each vertical. Measurements on Indus River in Sind at Mithankot, Sukur and Kotri
during 19111920 (Indus River commission records, "discharge, silt, velocity and
miscellaneous observations", parts I to IV, 1911  1920 printed at Commission press
1922, part II, pages 1 to108) showed that reduction coefficient varied between 0.74 and
0.92 when the depth variation was from 2.44 to 13.72 m and surface velocity from 0.19
to 5.09 m/s.
The studies in canal were similarly made by Mysore Engineering Research Station at 32
sites. The mean velocity V of the crosssection was obtained by the current meter
whereas the surface velocity was measured using floats.
The following relationship was obtained
V (m/s) =0.8529 V
s
+0.0085
A relationship between the surface velocity Vs and the mean velocity V in terms of
Chezy C has been developed and is given by:
s
V
=1 +2.5 g / C
V
The usual assumption made in practice is that
s
V
=0.85
V
which corresponds to 'C' value
of 52.4 m
0.5
s
1
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Theoretical considerations based on the logarithmic velocity distribution law indicate
that the reduction coefficient would be applicable only to a particular stream for which it
has been determined, since it would depend on the relative roughness of the channel,
depth, slope, etc., and hence it would be different for different streams, and for
fluctuating flood stages even in a given stream. It is therefore, recommended by the BIS
that the reduction coefficients should be found out from actual field observations made
by a current meter and only if such determination of the coefficient is not possible during
high flood stages then the reduction coefficient should be extrapolated to the stage from
data collected at lower stages.
4.2.4 SlopeArea Method
In the event of infeasibility of velocity area method due to either rapid rise and fall of
stage or lack of equipment, the slope area method is adopted for rough estimation of
the discharge.
The requirements of the site are mostly similar to those for area velocity method. The
crosssectional area is measured adopting the procedure as in case of area velocity
method. The velocity formula used is that of Manning, the energy slope for nonuniform
flow . The roughness coefficient value to be used is related to bed material size and
condition of the channel. These recommendations are given in Indian Standards
Institutions IS : 2912 (1964).
4.2.5 StageDischarge Relationships
Regular recording of discharges over a period of time is essential for correct estimation
of water resources of river basins and subsequent planning and utilization. Daily
discharge observations over a long period are sometimes not feasible. The estimation
of the discharge is then achieved by using proper stage discharge relation. The method
adopted for the preparation of the stage discharge relationship for the different river
basins as well as the the exhaustive instruction for adopting the method of estimation of
discharge by establishing stage discharge relationship are contained in the Indian
Standard Recommendations IS: 2914(1964).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.2.6 Details of Existing Indian Standards
A  Stream Gauging:
1. Printed Standards / Under Print:
(a) Measurement/ Estimation, Analysis and Recording:
IS: 1191 Glossary of terms and symbols
IS: 1192 Velocity area methods
IS: 1193 Notches, wiers and flumes
IS: 1194 Forms for recording measurement
IS: 2912 Slope area method
IS: 2913 Flow in tidal channels
IS: 2914 Stage discharge relation
IS: 2915
Instructions for collection of data for
analysis of errors
IS: 3918 Use of current meter
IS: 6059 Weirs of finite crest width
IS: 6062 Standing wave flumefalls
IS: 6063 Standing wave flume
IS: 6330 End depth method for rectangular channels
(b) Instruments
IS: 3910 Current meters
IS: 3911 Surface floats
IS: 3912 Sounding rods
IS: 4073 Sounding weights
IS: 4080 Vertical staff gauge
IS: 4858 Velocity rods
IS: 6064 Sounding and suspension equipment.
Reference:
Chitale S.V., Discharge Measurement  Technology and Data Analysis, Hydraulics of
Alluvial Streams, Central Board of Irrigation and Power, a Status Report Number 3, New
Delhi, J une 1974. Page 13 to 24.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.3 Radioactive Tracer Technique for Measurement of River
Discharges
The conventional method of measurement of discharge in open channels by adopting
area velocity method necessitates measurement of river crosssection at a site and also
observing velocities on several verticals across the measuring section. But this
procedure may not be feasible in all the cases. The radioactive tracer method (total
count) dispenses with the measurement of cross section and velocities and, where, it is
applicable, is much simpler, cheaper and quicker. These methods have been tested for
measurement of discharge up to 227 m
3
s
1
, and accuracy as high as 98 percent is
attained. Central Water Power Research Station, Pune in collaboration with Bhabha
Atomic Research Centre, Bombay conducted experiments using (i) radioactive tracer
technique on River Mutha, in the recirculation system of the CWPRS and in River Tapi
and
(ii) Chemical Salt dilution method downstream of tailrace tunnel of Koyna Power House
and in Vaitarni River. The measurement of discharge by these methods require pre
knowledge of mixing length. The mixing length is defined as the minimum distance at
which the mass transfer and the concentration are equal, i.e.,
c m
d d
c m
=
The mixing length depends upon many factors such as: (i) degree of turbulence, (ii)
geometry of the crosssection, (iii) number and position of tracer injection, (iv) properties
of tracer used, and (v) velocity distribution.
CWPRS, Pune by using the radio isotope method found that in case of Tapi River the
mixing length is 40.23 km for a river discharge of 756 m
3
s
1
. Further it was also
observed that mixing length is higher in case of side injection compared to the central
injection of the tracer.
Uttar Pradesh Irrigation Research Institute, Roorkee conducted experiments in
mountainous rivers of Himalayan origin. The data obtained from these experiments
showed that the mixing length (
mix
l ) in mountainous rivers varies linearly with the
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
average river width
o
B in the experimental reach and is governed by the relation
mix o
l KB C = +
in which, K and C are constants and found to be 77 and 120 respectively in a set of
experiments given in Table. The flow rate obtained by dilution method compared well
with that obtained by area velocity method.
Name of
the River
Discharge m
3
s
1
River
slope
(m/km)
Average
top water
surface
depth (m)
Observed
mixing
length
(km)
Remarks
Ganga
as per
area
velocity
method
(m
3
s
1
)
as per
dilution
method
(m
3
s
1
)
Ganga 296.00 319.00 1.21 70.00 4.30
Ganga 136.00  1.21 57.00 
Power
mixing
was not
achieved
Ganga 150.50 147.10 1.21 55.00 4.00
Song 148.50 154.20 4.48 50.00 3.40
Tons 14.00 13.70 5.59 20.00 1.11
Ganga 425.00 453.25 2.00 52.00 4.53
Ganga 771.70 763.00 3.16 136.00 10.10
Song 629.60 640.60 3.16 104.00 9.00
6.20 6.90 7.30 23.20 1.60
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.4 Measurement of flow of water and the limitations of velocity area
method (CBIP, 1978)
Systematic observations of river gauges and discharges have been practiced in some
parts of the country for several decades. In Punjab, Sind, Tamil Nadu, Karnataka,
Andhra Pradesh, Uttar Pradesh, West Bengal and some other states, rivers have been
regularly gauged. Ganga Basin Water Resources Organization under the administrative
control of Central Water Commission is observing gauges, discharges, sediment
charges in entire of Ganga Basin which is one of the largest basins in Asia. The work
has also been carried out in other river basins such as Mahanadi, Krishna, Cauvery,
Brahmaputra.
4.4.1 Relation between Surface Velocity and Mean Velocity
When surface velocities are measured by surface floats or current meters, a coefficient
is applied to obtain the mean velocity on the vertical. Earlier experiments on different
streams have indicated that this coefficient is not constant and lies between 0.79 to 0.9.
In the Punjab and Sind a large number of observations gave a value of 0.89. This value
was in use in other parts of India. The studies conducted under Research Scheme
sponsored by Government of India, on 24 sites of different canal reaches in Karnataka,
the analysis of a set of 46 observations indicated that the ratio of mean velocity to
surface velocity works out to.
4.4.2 Point of Mean Velocity
Experiments carried out on Sind canals showed that in 79 percent of the cases the
mean velocity occurred between 0.51 and 0.75 of depth on each vertical. Data collected
on 43 sites on the Sukkur Barrage canals was similarly examined. Five verticals were
selected out of a crosssection for purposes of study. Analysis showed that the average
position of mean velocity on the two and verticals was obtained at 0.67 depth, for the
intermediate two verticals at 0.63 depth and for the central vertical at 0.61 depth.
Data of velocity distribution on 951 verticals on the River Indus during the years 1916 to
1932 yielded the following statistical relationship
( )
0.6
V (m/s) =0.3048 1.004 +0.041
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
in which
0.6
is the velocity at 0.6 depth in fps (foot per second).
Velocity observations made on an equally large number of verticals during the years
193638 on the various Sukkur Barrage canals gave the following relationship between
the velocity observed at 0.6 depth and the mean velocity over the vertical
( )
0.6
V (m/s) =0.3048 1.010  0.059
in which
0.6
is the velocity at 0.6 depth in fps (foot per second).
4.4.3 Velocity by Float Rods
Cunningham in his Roorkee experiments showed that velocity of the float rod ( )
r
V
would be equal to the mean velocity of a vertical
( )
V when the submerged length of the
rod was 0.95 to 0.927 depth of water, the exact value depending on the position of the
maximum velocity on the vertical. The Indian practice has been to use rods having
submergence of 0.94 depth to account for variations of stream depth along the float
track, the following relation between
( )
V and ( )
r
V has been used.
(m/s) 03048 1012 0116
in which is in FPS.
=
y l
V . V . .
r
y
V
r
where, y is the depth of water in feet and L is the submerged length of the rod in feet.
Lacey proposed the use of a special tabular rod which was named after him. He
suggested that the following formula could be used:
( ) 0.3048 2
080 04
in which and in FPS at 0.8 and 0.4 depth of flow respectievely.
080 04
=
V m/ s
. y . y
. y . y
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0.930
0.940
0.950
0.960
0.970
0.980
0.990
V
__
____
V
r
0.6 1.2 1.8 2.4 3.0
Depth of Flow in meter
X is the ratio of Length of the measuring rod to Depth of water
Typical Correction factors for velocity rods for log velocity
distribution (This depends on Manning, n value)
Manning, n = 0.020
X = 0.90
X = 0.85
X = 0.80
X = 0.75
X = 0.70
X = 0.65
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.4.4 Errors in Discharge Measurements in Large Rivers by the
Velocity Area Method
Accuracy of discharge measurement in natural and artificial channels depends upon a
particular method being selected. Out of many practical methods, the one which has
found wide applicability and use in field and project studies is the "Area Velocity
Method". ISO/BIS have brought out standards for adopting this method. Detailed
investigations have been carried out in the past as well as recently for evaluating, the
errors in discharge measurement by the area velocity method.
Measurements of flow in open channels by the area velocity method are subjected to
systematic errors in the measuring instruments and random errors caused by their lack
of sensitivity in the range in which they are used. Random errors can also be due to
pulsations and personal errors during observation. The total error in discharge
measurement by area velocity method comprises components of errors due to width
and depth and velocity measurements which are subject to random and systematic
error of observations, besides the one due to using finite number of verticals along the
entire crosssection.
4.4.5 Error in Width
In the Pivotpoint method, which is the standard practice in India for positioning of the
boat at various observation points in a wide river, the position of the station at which the
depth or velocity is to be observed is located by a geometrical layout of points on the
bank or banks of the river. The distance from the bank is not measured, but the boat is
brought to the desired position by aligning it against the crosssection line pegs and
prefixed pivotpoint flags on the bank. The error in positioning has been determined by
comparing the distances with those determined by the angular method with the help of a
precise theodolite. The latter method is presumed to yield true distances. Observations
taken on 10 days for a total of 154 verticals have been statistically analyzed and the
mean standard deviation determined. The results are summarized below
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Range of width
observed (m)
Absolute error (m) Mean standard
deviation (%)
Remarks
300 to 600 2.34 0.386 From one bank
600 to 1200 6.72 0.564
From each of two
banks
It would appear that with increase in width, the percentage error increases in magnitude
with the same equipment.
Errors due to measurement in width could be minimized by taking segments at equal
distances and the total surface width could be measured with more sophisticated
instruments available now.
4.4.6 Error in Depth
Depth is usually measured by a rigid sounding rod up to a depth of 6 m and by a log line
beyond this depth.
To work out the error in depth, two readings are taken with the same sounding rod at
the same place. The average of these two readings are used for comparison with the
individual readings to work out the standard deviation.
Observations for 10 days for a total of 80 verticals have been statistically analyzed and
the following results are obtained.
Range of width observed
(m)
Absolute error (m) Mean standard deviation
(%)
0.41 to 6 0.039 0.65
6 to 14 0.049 0.35
The percentage standard error would appear to decrease with depth, though the
absolute error increases.
4.4.7 Error in Mean Velocity at Verticals
The normal Indian practice is to measure the velocity at 0.6 depth and take it as the
mean velocity, unless the point of mean velocity is observed by preliminary
observations to be at different depth. This is compared with the mean velocity obtained
by the sixpoint method, i.e., observing velocities at 0.2, 0.4, 0.6 and 0.8 of the depth
below the surface and as near as possible to the free surface and at the bottom. The
mean velocity was worked out from the following equation
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
1
2 2 2 2
02 04 06 08
10
V V V V V V V
s . . . . bottom
= + + + + +
Observations for 21 days at different sites for a total of 390 verticals yielded the
following results
Range of velocity Mean standard deviation (%)
0.087 to 1.3 m / sec 4.75%
Error due to Limited Number of verticals:
Investigations by the Rijkswaterstaat show that the standard error reduced
progressively with the increase in the number of verticals, as given in the Table
Table: Progressive reduction in the standard error with the increase in the number of
verticals
Number of verticals Standard error in percent of discharge
8 2.35
10 1.35
12 0.90
15 0.60
20 0.38
25 0.30
4.4.8 Components of Error in ISO/ISI Specifications
When equidistant verticals spaced at 'b' unit apart in a water surface width 'T' are used.
The systematic part of the error in discharge measurement dependent on the number of
verticals was found to be
( )
50 50
T 1
m
b
X b or
m
=
+
in which, ( )
m
X b is systematic error in discharge due to 'm' number of verticals. Random
errors
v
2 2
y
X and X due to velocity and depth after analyzing 43 sets of observations
having approximately 100200 number of velocity observations on the Maharashtra
canals, Gole et al. (1973) have suggested the following two equations:
3
2
v 2
10
X =
m
and
2
y 4 3
28
X
/
m
=
The average coefficient of variation for horizontal distribution of velocity was obtained as
32 and for mean depth as 5.3 percent.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1 10 100 200
0.5
1.0
5.0
10.0
50.0
100.0
0.1
x
__
2
v
__ m = [
T
__
b
1
1]
m
x
__
2
v
__ 2
x
2
v
__ ( )
+
_
10
3
m
2
x
2
v
__
__
=
__
x
2
v
__
__
Variation of
with
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
x
_
2
y
_
1
m
=
_
2
y
_
X
1 5 10 50 100
0.5
1.0
2.0
5.0
10.0
0.1
m is the Number of verticals)
m>70
x
_
2
y
_
Variation of with the number of verticals
m<70
The total error could thus be worked out as
( )
05
1
2 2 2
.
X X b X X
Q m
v y
m
= + +
4.4.9 Intrinsic Error in ISO/ISI Specification for 15/50 Verticals
ISI have recommended 15/50 equidistant verticals for computation of discharge in
channels. It is found that intrinsic errors in discharge due to number of verticals being 15
and 50 are +3.12 and +1.0 percent respectively. The total root mean errors due to
velocity and depth measurements works out to be 3.8 and 1.4 percent respectively.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0 10 20 30 40 50 60
7.0
8.0
0
1.0
2.0
3.0
4.0
5.0
6.0
Showing theoretical curve for X
m
and experimental
data as a function of number of verticals
Number of verticals m =
w
b

1
__
ISO data
Carter and Anderson
Delft data
CWPRS data
4.4.10 Effect of Deployment on Random Errors
2 2
X and X
v y
When the 'm' number of verticals are deployed according to a particular scheme for
precisely locating the depth (crosssection) profile, it is expected that it would have
some effect on the estimate of
2 2
X and X
v y
. Since deployment help in getting the
nearest estimate of the representative profile and hence coefficient of variation in
velocity and depth over a crosssection, it will affect the contribution to the random error
in so far as the estimate of the coefficient of variation in error are compared to the true
value. But errors are inversely proportional to m and hence the difference is not
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
significant unless 'm' is very small. To verify the above hypothesis
2 2
X and X
v y
were
estimated for different deployment of vertical and the total contribution against the mean
function for
( )
2 2
v y
1
X and X
m
(See Figure). It is found that the difference in the
2
Q
X for
different deployments from the mean curve for equidistant verticals are not significant
and for all practical purposes the theoretical curve may be utilized to get the contribution
of random errors due to sources velocity and depth.
X
2
q
__
__
0 5 10 50 100
2
5
10
20
50
100
0.1
0.2
0.5
1.0
3.0
X
2
q
__
__
=
10
3
__
m
2 +
28
m
4 3
(
__
/
)
m>30
1
__
m
=
1
__
m
x
2
y
+
x
2
v
( )
X
2
q
__ __
__
m<30
m number of verticals
m
Variation of with
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Table Shows computation of
m
X from experimental data (CBI&P)
Sl.No % error with
reduced number
of verticals, E
m
=8
Number of
verticals for
standard
discharge
m
X Intrinsic error
in standard
discharge due to
verticals (%)
1 3.74 11 4.17
2 8.17 11 7.17
3 4.76 10 4.55
4 5.68 16 2.94
5 0.99 16 2.94
6 5.21 16 2.94
7 1.22 20 2.38
8 0.82 21 2.27
9 0.89 41 2.19
10 18 3.06
( )
( ) ( )
( )
( )
2
2 2
2
2
Computation done by CBI & P (Central Board of Irrigation and Power)
E 330
293
1
115 293
2
113125
336 306
336 306 642
m m
m m
m
m
Average .
Set Error .
E . .
or X X .
X X . but X .
X . . . percent
=
=
+ = +
=
= =
= + =
in which,
( )
m m
X X is the estimate of the error due to reduced number of verticals,
relative to the standard discharge.
4.4.11 Total Error
Q
X
Total error could be worked out using Equations (9) and (10) for
2 2
X and X
v y
and the
theoretical value
m
X of for respective cases. Experimental data on total error obtained
independently by Delft, Carter and Anderson and worked out at CWPRS, Pune using
data of Maharashtra with equidistant verticals, are compared. The experimental data
conformed closely to the semitheoretical equation on total error obtained at CWPRS.
Area velocity method results in biased estimation of the discharge which is in
agreement with the findings of Delft and Dickinson. Theoretical mean bias, i.e.,
systematic error could be estimated as a function of the number of verticals or mean
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
width of the equidistant segments. Theoretical mean bias has been found to be close to
independently observed data of Delft, Carter and Anderson and CWPRS.
0 5 10 15 20 25 30 35 40 45 50 55
2
4
6
8
10
12
14
16
X
Q
Deift data
Carter and Anderson data
ISO specification (CWPRS) data
MERS data
m (Number of verticals)
Variation of total error with number of verticals
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Deift data average
CWPRS data
MERS data
APERL data
(assuming Xm as error
in standard discharge)
0
20 40 60 80 100 120 140
2.0
4.0
6.0
8.0
10.0
12.0
Theoretical mean error (percent)
100 32 19 15 11 9 7 5 4 3
Number of vertical or
(
w __
b
1
)
Figure showing the variation of mean observed error against theoretical error
The above method is applicable subject to following limitations:
(a) Above method is applicable subject to the data are free from systematic errors in
observations. If data are expected to contain errors which are systematic in nature,
while estimating the total error, the amount of systematic error entering into
observations as per standard formula of root mean error is to be accounted.
(b) For computation of random error due to sources, velocity and depth, attempt should
be made to compute them by using ISO/BIS formulae. When repetitive observations are
not available then only the error may be worked out for actual deployment.
(c) When no observational data are available the empirical formulae (9) and (10) may
be used for determining the approximate total error in discharge measured with finite
number of verticals spaced nearly equidistant. This assumes that channel is straight
and has got characteristics similar to canals whose data have been used in evolving the
above formulae.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(d) The value of
m
X could be obtained from figure for a given number of verticals with
equal spacing. The formula
50b
W
could be used only when end segment spacings are
nearly equal. For completely unsymmetrical deployment this formula cannot be used.
4.4.12 Error in Area Measurements
As often contemplated, the error in estimation of area contributes the major source of
error in this discharge estimation by area velocity method.
The total error in area can be obtained using
2
2
50 1
y
A
b
X X
W m
= +
Since the systematic error (50b/W) is dominant one in
A
X appreciable error reduction is
possible if number of sounding verticals are more than the velocity verticals. But the
gain in accuracy is not possible if area velocity method is used, since this method uses
the information on the same number of depth verticals which are having velocity
measurements to obtain
i
' q s. Moreover, in view of systematic error due to discrete
number of verticals being estimated precisely, there is no need for increasing the sound
verticals, since with the same number of 'm' verticals the correction in discharge could
be made to gain the accuracy equivalent to very large number of sounding verticals,
made use for minimizing
A
X .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.4.13 Optimum Deployment
The Bureau of Indian Standard has laid down that the segmentation for measurement of
depth in channels having different widths would be as given in Table.
Sl.No Description No. of verticals in
a crosssection
Maximum space
of verticals in m
1
Width of channel not
exceeding 15 m where the
river bed changes abruptly
15 1.5
2
Width of channel from 15 to
90 m
15 6.0
3
Width of channel from 90 to
180 m
15 15
4
Width of channels greater
than 180 m
25 
The spacing of verticals required a variation in discharge between adjacent segments
not to exceed by 4 to 10 percent, preferably the smaller. It further stipulated discharge
through any segment not to exceed 10 percent of the total discharge. For making
velocity observations the maximum spacing of verticals has been so specified that the
mean velocities on the adjacent vertical would not differ by 20 percent with respect to
higher values of the two. In no case there could be less than five verticals. These
specifications are for natural streams like rivers, drainage channels, etc.
Closer the interval of verticals, the more accurate will be the calculated discharge.
Moreover, allowing a variation of 2 percent in discharge and adopting 15 verticals as
standard of comparison a lesser number of verticals than 15 was recommended for
adopting so as to enable a discharge observation to be completed in a working day of 6
to 8 hours. The criterion recommended for the number of depth and velocity verticals of
Channels of various capacities has been indicated in Table.
Sl.No Channel capacity m
3
s
1
Approximate
surface width (m)
No. of verticals for
depth and velocity
1 Above 85 Above 35 11
2 Between 15  85 Between 15  35 9
3 Between 0  15 Between 0  15 5
It was also recommended to increase the number of depth verticals to have a better
appraisal of the cross sections of the channel in regard to its trend of silting or scouring.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
If it is desired to determine the errors in discharge with 15 verticals, and also to
determine the minimum number of verticals for measurement of canal discharge in
order to restrict this error to 2 percent, more number of depth and velocity verticals
should be observed.
Total errors in CBI&P deployment for different ranges of discharge varies from 3.9
percent for 11 verticals to 8.10 percent for 5 verticals.
If discharge with 50 verticals is taken as standard the error in discharge for CBI&P
deployment would be of the order of 2.75 percent, 3.5 percent and 6.7 percent for the
number of verticals 11, 9 and 5 respectively. Since intrinsic error in discharge due to
number of verticals, measured with 15 vertical works out to be of the order of +3
percent, the CBI&P deployments for different ranges of discharge with less than 2
percent error, investigated earlier in fact leads to total intrinsic error, systematic in
nature of the order of +5 percent. With the knowledge of the systematic error in CBI&P
deployment, the necessary correction could be made to get unbiased estimate of true
discharge.
The data obtained from the Godavari Canals and the K.C. Canal were analyzed and it
was found that for the range of discharges between 14m
3
/s to 85 m
3
/s , 5 or 7 verticals
gave discharge values within a range of 5% error, as compared to 15 verticals. For 85
m
3
/s to 225 m
3
/s range, with 9 verticals, the deployment being four verticals at either
ends upto quarter length and a central vertical, discharges which were within and error
of 4% were obtained.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
45897.1 mm
0.0
0.5
1.0
1.5
0.0FT
1.524 m
3.05 m
Q = 145.96 m
3
/s
Q = 135.67 m
3
/s
Q = 129.73 m
3
/s
Godavari Western Main Canal  Mile 5/2
1.524 m
3.05 m
Godavari Western Main Canal  Mile 5/2
Q = 78.96 m
3
/s
Q = 70.68 m
3
/s
Q = 49.21 m
3
/s
0.0
1.0
2.0
0
47427 mm
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0
0.5
1.0
1.5
1.524
3.05
Q = 28.49 m
3
/s
Q = 52.78 m
3
/s
Q = 73.60 m
3
/s
Nizamsagar canal M 2/0
Distance
4.0233.6 mm
Q = 82.5 m
3
/s
4.4.14 A note on the Optimum number of verticals to ensure required
accuracy in current meter gauging
Instructions regarding the choice of number of verticals for current meter gauging have
been drawn up by i) the I.S.I. ii) the C.B.I.P and iii) the C.W.P.R.S. The objective of
these instructions is to ensure that the order of accuracy attainable by following these
instructions is 2%.
The first point that attracts attention as for as the recommendations of the various
authorities is the relative importance of a vertical and its spacing according as it is used
for measurement of velocity or depth, the latter being intended to enable the
measurement of area.
While the C.B.I.P. appears to imply that the number of verticals and their deployment
recommended applies to the measurement of both depth and velocity, the C.W.P.R.S.
is categoric that depth measurement is essential on verticals spaced at 60 cm intervals
irrespective of the requirement of the verticals for velocity observations. The I.S.I. on the
other hand bases its recommendation for the number of verticals for depth
measurement on the criterion of variation of area from segment to segment and for
velocity measurement on the criterion of variation of mean velocity on a vertical to the
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
mean velocity on an adjacent vertical. In all these cases the requirement implied
appears to be a determination of such spacing wherein the variation is gradual and
unidirectional. In an artificial channel where essentially uniform and stable conditions of
flow may be expected, the elaborate requirement of the I.S.I. which included primarily
stream gauging in its scope may be neither be necessary nor desirable.
Figures indicate the velocity measurement details for Nizamsagar canal, Godavari
central main canal and Godavari western main canal for a wide range of discharge
(28.49 m
3
/s to 145.96 m
3
/s). The velocity in plan is normalized with respect to mean
value obtained at 0.6 times at the depth to the mean value at the vertical.
Reference:
Central Board of Irrigation and Power Problem No. 4 APERI design of channels 1978.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4.5 Errors in Depth Measurement in high velocity flows in
laboratory using Point Gauges: (Jayaraman and Sethuraman,
1973)
Water depth measurements in highvelocity flow in open channels are subject to
inaccuracies due to the presence of surface pulsations that are a characteristic feature
of super critical flows. Since the depths of water in supercritical flow in laboratory
channels are usually small, the importance of accurate depth measurement cannot be
overemphasized.
The most common device for measuring the depth of a supercritical flow in a laboratory
flume still continues to be the goodold pointgauge. Although more sophisticated
electrical probes are available for depth measurement, these invariably give rise to flow
disturbance when immersed in a high velocity channelflow. In order to improve the
accuracy of pointgauge measurement, Brock suggested the provision of pressure taps
on the bed of the flume and a visual check of the tip of the point gauge at a depth
setting corresponding to the mean hydro static pressure indicated by the pressure cell.
This method evidently cannot be applied for plotting surface profiles, as in transitions,
where a large number of depth measurements scattered all over the flume may be
necessary. Moreover, the assumption that the mean pressure indicated by the pressure
cell corresponds to the hydrostatic pressure for the mean depth needs experimental
verification in view of the unknown dynamic effects of the water surface pulsations.
A simple and inexpensive instrument, the Gauge ContactTime Indicator that can be
used with any pointgauge to improve the accuracy of depth measurement.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
7.56 (Mean)
32%
0 20 40 60 80 100
50%
7.0
7.2
7.4
7.6
7.8
7.49
Percentage contact time
Water level fluctuations at Froude Number equal to 3.5
(after Jayaraman and Sethuraman)
Figure shows an experimental plot of the gauge reading (with the channel bed as
datum) versus percentage contact time at a Froude number of 3.5.
Repeated tests made by more than one observer showed that for a specified contact
time of 50%, the gauge readings could be repeated to an accuracy of 0.1 mm, even
though the water surface had pulsations of about 8 mm amplitude.
The following conclusions are drawn regarding the use of the Contact Time Indicator for
pointgauge measurement in highvelocity flows:
1. Pointgauge measurement in highvelocity flows by visual observation of the tip of the
gauge inevitably involves a positive error in the measured depth of flow. This error can
be serious when the flowdepth is very small as is often the case in the study of
supercritical flow through channel expansions.
2. By specifying a particular contact time  say 50%  the technique of depth
measurement using the point gauge can be refined and standardized. Errors due to
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
persistence of vision, that are present in a visual observation of the tip of the gauge, are
eliminated.
3. The use of the Contact Time Indicator eliminates the "personal error" inherent in
pointgauge measurement in high velocityflows. With this instrument, all observers can
get readings within 0.01 cm.
4. Where a large number of depth observations are to be made in a single testrun,
such as in the plotting of surface profiles in supercritical transitions, the instrument
reduces considerably visual fatigue of the observer. The observer need not even look at
the water surface while making depth observations.
Reference:
R. J AYARAMAN and V. SETHURAMAN "IMPROVING THE ACCURACY OF POINT
GAUGE MEASUREMENT IN HIGHVELOCITY FLOWS", J ournal of Hydraulic
Research, Volume 11, Number 4, 1973, Page Number 317 to 323.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
5. Secondary Current and Spiral Flow
The curve of constant velocity for rectangular and triangular crosssection obtained by
Nikuradse are shown in Figures 1 and 2. In all cases the velocities at the corners are
comparatively very large with stems from the fact that in all straight pipes of noncircular
crosssection there exist secondary flows. These are such that the fluid flows towards
the corner along the bisectrix of the angle and then outwards in both directions. The
secondary flows continuously transport momentum from the centre to the corners and
generate high velocities there. Schematic diagrams of secondary flows in triangular and
rectangular pipes are shown in Fig. 3. It is seen that the secondary flow in the
rectangular crosssection which proceeds from the wall inwards in the neighborhood of
the ends of the larger sides and of the middle of the shorter sides creates zones of low
velocity. They appear very clearly in the picture of curves of constant velocity in Fig1.
Such secondary flows come into play also in open channels, as evidenced by the
pattern of curves of constant velocity in Fig. 4. The maximum velocity does not occur
near the free surface but at about one fifth of the depth down of the free surface.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Fig. 1. Curves of constant velocity for pipe of rectangular crosssection,
after Nikuradse
Fig. 2. Curves of constant velocity for a pipe of equilateral triangular
crosssection after Nikuradse
a b
Fig. 3. Schematic representation of Secondary flows in pipes
of triangular and rectangular (open channel) crosssection
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
water level
Fig. 4. Curves of constant velocity for a rectangular
open channel after Nikuradse
Secondary circulation is that flow wherein the velocity can be resolved into two
components, one in the longitudinal direction of the channel and the other in transverse
to the direction of the channel. The transverse component of the velocity gives rise to
the secondary circulation. It can occur in both straight and curved channels and for
different reasons. Secondary circulation is affected by temperature gradients, sediment,
turbulence, nonuniformity of boundary shear, and the curvature of streamlines.
Secondary circulation has been associated with turbulent flow in prismatic channels
wherein the shear at the boundary is not constant. In straight circular pipes as shear at
the boundary is constant for both laminar and turbulent flow the secondary circulation
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
has not been observed. When secondary circulation does occur, it seems to take place
in an even number of cells as depicted in Figure 5. The nonuniformity of sediment
across a channel has been associated with secondary circulation.
Fig. 5. Secondary circulation in straight channel
Secondary current is the flow taking place in transverse direction of the main flow. The
secondary currents are of four types viz.
1. The 'weak' secondary currents in straightnoncircular channel sections and in pipes
due to boundary resistance (figure 5).
2. Secondary flow developed due to nonuniform bed configuration as in case of alluvial
channels.
3. The ' strong ' currents caused in bends due to centrifugal force.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
SPIRAL FLOW
O
y
OUTSIDE WALL
INSIDE
SECTION ON AA
ILLUSTRATION OF SECONDARY FLOW AND SPIRAL CURRENTS
IN A 90 BEND
4. Secondary currents due to the unsteadiness of the oscillating boundary layer.
The occurrence of the maximum velocity filament in a straight channel just below the
free surface (see figure below) to the findings of secondary current.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Secondary
currents
Isovels
(a) Open channel
(b) Equivalent closed conduit
Comparison of Open Channel Flow with ClosedConduit Flow
The lens shaped figure is drawn such that it is orthogonal to each isovel. It may be noted
that the maximum velocity occurs slightly below the free surface. On the lens shaped line
no velocity gradient exist. The shear on the free surface is negligible and their is no shear
resistance to balance the component of the weight of the prism along the main flow
direction. The equivalent closed conduit is symmetrical about the central line and the
shear stress is distributed along the boundary line.
0.750ySo
0.750ySo
0.970ySo
y
4y
Side Slope, m: 1 = 1.5 : 1
Tractive force distribution obtained using membrane analogy
This distribution varies depending on the cross section and material
Gibson, explained the origination of the secondary current. Darcy, Cunningham, Sterns,
Moseley, Francis and Wood (Thandaveswara, 1969) recognized the presence of this
secondary current and superposition of the main flow leads to spiral flow. If there is any
slight disturbance in approach flow conditions instead of double spiral, then single spiral
exists. Kennedy and Fulton established that the secondary current has a definite effect
on the frictional resistance of the channel.
The second type of secondary currents were observed by Schlichting, J acob, Schultz
Grunov. The projection of spheres from the surface is just similar to the spherical sand
particles fixed uniformly over the surface, then this type of secondary current can be
expected when the sand roughness is used.
The flow pattern which exists behind an obstacle placed in the boundary layer near a
wall differs markedly from that behind an obstacle placed in the free stream. This
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
circumstance emerges clearly from an experiment performed by Schlichting and shown
in figure. The experiment consisted in the measurement of the velocity field behind a
row of spheres placed on a smooth flat surface. The pattern of curves of constant
velocity clearly shows a kind of negative wake effect. The smallest velocities have been
measured in the free gaps in which no spheres are present over the whole length of the
plate; on the other hand, the largest velocities have been measured behind the rows of
spheres where precisely the smaller velocities.
d
5d
1
2
3
3
2
1
10d
10d
10d
5d
6.00
5.75
5.50
5.25
5.00
4.75
4.50
4.25
4.00
V
[m/s)
measuring
station
Isovels behind a row of spheres as measured by Schlichting. Secondary flow
in the boundary layer is marked behind (1), as calculated by K. SchultzGrunow.
In the neighbourhood of the wall, the velocity behind the spheres is larger than
that in the gaps. The spheres produce a "negative wake effect" which is explained
by the existence of secondary flow. Diameter of spheres d= 4mm
When the spacing of roughness is close, the wavy water surface will not exist as the
formation of vortices will be confined to roughness elements and forms a pseudowall
and does not affect the main flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
s
y
k
Isolated  roughness flow (k/s)  Form drag dominates
s
The wake and the vortex are dissipated before the next element
is reached. The ratio of (k/s) is a significant parameter for
this type of flow
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
s s
Wake interference flow (y/s)
j
j
j
Quasi smooth flow  k/s or j/s becomes significant acts as Pseudo wall
s
y
k
y
k
s s
s
j
k is surface roughness height
s is the spacing of the elements
j is the groove width
y is the depth of flow
Concept of three basic types of rough surface flow
When the roughness elements are placed closer, the wake and the vortex
at each element will interfere with those developed by the following
element and results in complex vorticity and turbulent mixing. The height
of the roughness is not important, but the spacing becomes an important
parameter. The depth 'y' controls the vertical extent of the surface region of
high level turbulence. (y/s) is an important correlating parameter.
Quasi smooth flow is also known as skimming flow. The roughness elements
are so closed placed. The fluid that fills in the groove acts as a pseudo wall
and hence flow essentially skims the surface of roughness elements. In such
a flow (k/s) or (j/s) play a significant role.
In the following paragraphs 3rd type of secondary current has been discussed briefly.
The third type of secondary currents will come into picture while the fluid flows in a
curved channel. The fluid in a curved channel will be subjected to centrifugal force. Due
to this centrifugal force, a pressure gradient normal to the direction of the main flow is
created. Then the particles near the inside wall are thrown outside and they reach the
outside boundary moving in transverse direction. Thus a sort of centripetal lift will be
created causing the heaving up of the fluid. If the flow is irrotational and the fluid enters
with uniform velocity into bend, then it is analogous to the potential vortex.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
VELOCITY DISTRIBUTION IN POTENTIAL FLOW
IN A CURVED CHANNEL
r
Vr =CONSTANT
O
r
i
r
c
B
v
r
0
But in actual case due to the presence of shear stress at the boundary, the velocity of
main flow decreases abruptly at the boundary setting a velocity gradient in the boundary
layer. It may be observed that the energy in the boundary regions is less than in the
potential zone. It follows that at the outside of the bend the pressure intensity falls away
abruptly towards the wall, unless a secondary flow takes place in the direction of outer
wall. Continuity equation requires an inward flow along the side walls to compensate
since the pressure gradient normal to the wall is exactly opposite to that of potential
motion.
The spiral flow motion induced by the centrifugal force is very pronounced and irregular
in the bend. The complicated pattern of flow is caused by the superposition of
secondary current in the bend over the spiral flow of the approach channel. The spiral
flow of bend begins as a lateral boundary current near the point where the stream line
curvature begins and at the bottom inside corner of the bend.
This type of spiral motion also called helicoidal flow and was recognized by Thomson in
1876 and was demonstrated by him in the laboratory in an 180 circular bend with
rectangular channel section in 1879. This was supported further by Engles, Beyerhams
and others. During 1883 to 1990 several researchers while investigating the flow
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
characteristics in meanderings observed the action of scouring and deposition in the
river bends.
Several investigators (refer to Thandaveswara's Thesis, 1969) mostly conducted the
experiments in channel whose aspect ratios were of the same order of magnitude. Thus
the mean flow occurring was essentially three dimensional in character.
But Betz, Wilcken, Maccol and Wattendrof conducted experiments in two dimensional
channel (rectangular conduit). Watterdrof showed the potential character of the spiral
flow and drew the following conclusions.
(i). There is only slight increase in channel resistance due to the presence of bends as
indicated in pipe bends.
(ii). The velocity distribution follows free vortex law.
(iii). Rayleigh's stability criterion based on the calculation of mixing length and exchange
factor showed the instability and increased mixing at the outer walls of the curved
channels and decreasing mixing and stability at the inner wall.
(iv). If the depth to breadth ratio is large enough so that the lateral currents occupy only
a relatively small part of the area of the crosssection near the bottom and if form losses
are ignored near the bend, then the bend loss scarcely exists.
5.1 Strength of spiral
The term "Strength of Spiral" is defined as the percentage ratio of the mean kinetic
energy of the lateral motion to the kinetic energy of flow and is denoted by
xy
S .
( )
2
xy
2
xy
m m
xy
2 2
V
2g V
S = * 100 = * 100
V V
2g
The strength of secondary current can be qualitatively estimated to be proportional to
the extent of distortion of isovels. The concentration of velocity near boundary means
the secondary flow concentration near boundary. This bears the hypothesis that the
mechanism of secondary motion arises out of the boundary shear turbulence.
It may be noted that the approach flow plays an important role and has a direct effect on
the number of spirals, strength of spiral and other characteristics of spiral flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Following equations relate the deflection angle
1
along the centre line of bed, geometry
of the channel and the hydraulic properties of flow, in channel bends.
(i) For a smooth rectangular bend
(ii)
1
1
c
1 e
0.25
e
P
r
tan =17.4 for 2000 R 45000
R
(iii) For a smooth triangular channel
1
1
c
1 e
0.25
e
P
r
tan =13.4 for 2000 R 15000
R
In general,
1
c
1 3
0.25
e
P
r
tan =K
R
If the channel is wide then
1
0 5
c
1 4
0.25
e
y
r
tan =K
R
.
But Russian authors found that for a rectangular wide channel
1
c
y
tan =11
r
In general for a wide rectangular channel,
( )
1
1 0 e
c
b
tan =K R
r
for smooth flow
1 0
c s
b y
tan =K
r K
for rough flow
a
1 0
c
b
tan =K f
r
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
where f =friction coefficient and "a" is an exponent >1. The last equation can be
expressed in Chezy terms of coefficient
8g
C=
f
in the form
a
1 0
2
c
8g b
tan K
r
C
=
The value of
1
tan can be assumed to indicate the strength spiral to some scale.
Reference:
Thandaveswara B.S., "Characteristics of flow around a 90 open channel bend",
M.Sc (Engineering), Department of Civil and Hydraulic Engineering, Indian Institute
of Science, Bangalore, 1969.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6.1 Energy and Momentum Coefficients
Generally, in the energy and momentum equations the velocity is assumed to be steady
uniform and nonvarying vertically.
This assumption does not introduce any appreciable error in case of steady (or nearly
uniform) flows. However, the boundary resistance modifies the velocity distribution. The
velocity at the boundaries is less than the velocity at a distance from the boundaries.
Further, in cases where the velocity distribution is distorted such as in flow through
sudden expansions/contractions or through natural channels or varying cross sections,
error is introduced.
When the velocity varies across the section, the true mean velocity head across the
section,
( )
2
2g
m
, (the subscript m indicating the mean value) need not necessarily be
equal to 2g
2
V . Hence, a correction factor is required to be used for both in energy
and momentum equations (See Box). The mean velocity is usually calculated using
continuity equation.
Keulegan presented a complete theoretical derivation of energy coefficient and
proved that the selection of and (Momentum coefficient) depends solely on the
concept of the coefficient of friction which is adopted. If the equation of motion is derived
by the energy method, the concept underlying the friction coefficient in that equation is
that of energy dissipation in the fluid per unit length of channel and is the proper
factor to use. To understand proper use of factors and and the energy principle or
momentum principle is used appropriately.
Box:
The weight of flow through an element of area dA is equal to dA g ; the kinetic energy
per unit weight of this flow is 2g
2
V ; The rate of transfer of kinetic energy through this
element is equal to
dA . dA (1)
g
g
=
2 3
2g 2g
Hence, the kinetic energy transfer rate of the entire flow is equal to
A
dA (2)
2g
0
g
3
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
and the total weight rate of flow is equal to g dA
Q = VA g g
3
mass
mass density =
volume
kg
mass = * volume =
m
3
x m
3
=
= =
=
= =
2 2
2
kg
Force N mass * acceleration
kg m m
= kg *
s s
kg m N
specific weight g *
m s m
3
=
Velocity distribution in a Trapezoidal Section
Real ideal
Velocity v Velocity v
A
A
Velocity distribution along section AA
The mean velocity is by definition equal to Q / A. Hence, the mean velocity head, or
kinetic energy per unit weight of fluid, is equal to
A
dA
2
2
V 2g
0
(3)
2 2 VA g g
m
= =
3
in which is a correction coefficient to be applied to the velocity head as calculated
from the mean velocity. It is also known as the Coriolis coefficient. Hence
N
3
i
i =1
3 3
A
dA
dA
0
i = 1.....N (4)
V A V A
=
3
Similar approach can be applied for computing the momentum term VQ . The rate of
transfer of momentum through an element of area dA is equal to
2
dA V ; Following
similar logic as above the momentum correction coefficient can be obtained as
A
N
2 2
i
i =1
0
2
dA dA
i=1,2....N (5)
2
V A
V A
=
is also known as Boussinesq coefficient.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In general, the coefficients are assumed to be unity for channels of regular geometrical
cross sections and fairly straight uniform alignment, as the effect of non uniform velocity
distribution on the computation of velocity head and momentum is small when
compared to other uncertainties involved in the computations. Table shows the values
of and and for selected situations.
Table: Values of and for selected situations (after Chow, 1958)
Channel
Minimum Maximum Average Minimum Maximum Average
Regular
channels,
flumes,
spillways
1.10 1.20 1.15 1.03 1.07 1.05
Natural
streams
and
torrents
1.15 1.50 1.30 1.05 1.17 1.10
River
under ice
cover
1.20 2.00 1.50 1.07 1.33 1.17
River
valley,
over
flooded
1.50 2.00 1.75 1.17 1.33 1.25
The kinetic energy correction factor and momentum correction factor can be
expressed as (see box).
N
3
i
i =1
3 3
A
dA
dA
0
i = 1.....N (4)
V A V A
=
3
A
N
2 2
i
i =1
0
2
dA dA
i=1,2....N (5)
2
V A
V A
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6.1.1 Determination of and
Many investigators have done extensive investigations on the computation of and .
Chow (1958) has summarised different equations for determination of and for
various velocity distributions.
2
2
Rehbock assumed a linear velocity distribution and obtained
1
1
3
= +
= +
and for logarithmic velocity distribution.
2 3
2
max
max
1 3 2
1
V
in which 1 , V is the maximum velocity and V is the mean velocity
V
= +
= +
=
If the velocity distribution is along a vertical is logarithmic, then the relation between
and , as shown by Bakhmateff, is that exceeds unity by about onethird of the
amount by which exceeds unity. If 1+ n and 1 3 + n then
2
3
+
=
approximately. Generally, the coefficients and are greater than one. They are
both equal to unity when the flow is uniform across the section, and the farther, the flow
departs from uniform, the greater the coefficients become. The form of Equations (4)
and (5) makes it clear that is more sensitive to velocity variation than , so that for a
given channel section, > . Values of and can easily be calculated for
idealized twodimensional velocity distributions.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Velocity Distribution
n
0 0
y
y
=
0
v
n 1
=
+
( )
3
n 1
3n 1
+
=
+
( )
2
n 1
2n 1
+
=
+
( )( )
( )
n 3 2n 1 1
1 3n 1
+ +
=
+
1
If n
7
=1.043 =1.015
=
,
The high value of appropriate to laminar flow is of limited interest, since laminar flow
is rare in free surface flow problems. For turbulent flow in regular channels seldom
exceeds 1.15. In view of the limited experimental data on values of , the question
always arises whether the accuracy attainable with channel computations warrants its
inclusion!.
A practical method of arriving at the values of and for other than and idealised
velocity distribution is a semi graphical and arithmetical solution based on planimetered
areas of isovels plotted from data measurable at the cross section. Measured velocities
are plotted to draw the Isovels. The Isovels are constructed for each cross section and
cross sectional areas, A , of each stream tube are calculated with planimeter and
computations performed.
6.1.2 The Methods of computation of and may be classified as
1. Theoretical Methods
Based on experimental studies Strauss in 1967, has given empirical formulae for
computing and for general channel section based on the velocity distribution
given by the following equation.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1 n
V=ay
in which v is the velocity at a point located at a height y from the bed a is a constant
and n is an exponent such that 1 n .
and can easily be computed using following equations , if the velocity distribution
is known.
3
A
dA
0
V A
=
3
A
2
0
dA
2
V A
=
Strauss states that the general velocity distribution of the type given by above equation
covers all possible distributions by suitably choosing the value of n. In the limiting case
when n the velocity distribution tends to become rectangular. At the other
extreme when n=1, the velocity distribution is linear for which case =2 and =1.33.
Strauss showed that
( )
( )
1 1 1
1 1 1
=
=
f n, ,B ,
f n, ,B ,
in which n is the exponent of the velocity distribution, and,
1
is normalized depths, B
1
is
the normalized width of free surface to bed width,
1
is normalized bed width of berm
(including) to channel bed. The velocity distribution plays a dominant role in influencing
and and in trapezoidal channel in addition to
1
T
B =
b
. For rectangular channel the
exponent n of velocity distribution has a dominating effect. But Strausss method has
limited practical utility. It is not always true that the same velocity distribution prevails
along all the verticals of the crosssection, especially in nonrectangular channels. Also
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
this method is not applicable when there is a negative velocity zone over the cross
section as in the case of a diverging channel, a bend or a natural channel.
Theoretical uniform flow
velocity distribution
(Ideal)
Linear
velocity
distribution
depth
of flow
y
y
Logarithmic
velocity
distribution
Power
Law
Typical velocity distribution
2. Graphical Method
In Velocity area method, the flow area is divided into number of grid cells and local
velocities are measured using one of the measuring devices and finally integrating one
will get the average velocity. The velocities are measured at the intersecting grid lines
(nodes). Example: a1, b1, c1 etc......a5, b5.......e5.
The average velocity over the elemental area is v
cell
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Grid for measuring Velocity
a b c d e
1
2
3
4
5
dx
dy
i
i+1
j
j+1
Co ordinates of the nodes are (i, j), (i+1, j), (i+1, j+1), (i, j+1)
v(i, j) + v(i+1, j) + v(i+1, j+1) + v(i, j+1)
_______________________________________
Corresponding velocities are v (i, j), v(i+1, j), v(i+1, j+1), v(i, j+1)
Average velocity of the cell vcell =
_
4
Average velocity of the flow
( )
y b
cell
0 0
1 v dA
v v dy*db
A A by
in which dAis the elemental area of the cell
=
=
The other alternative is to draw the isovels (isovel is a line having the same value of
velocity sometimes it is also known as isopleths) assuming the linear variation between
two values and interpolating the value in between two nodes. It may be noted that the
velocity would be zero on the solid boundary. Hence the gradients are sharper very
close to the boundary. Typical isovels are shown in Figure. In this method, velocities are
measured at several points of crosssection and the lines of equal velocities called
isovels (also called isotachs) are drawn as shown in Figure.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0.3505
0.2987
0.2499
0.3639
Q = 17.95 l/s
y = 0.332 m
= 1.041
= 1.01
Graphical Method
While drawing isovels it is assumed that the velocity varies linearly between two points.
Next the area within each isovel is plain metered. Assuming that the velocity through the
area bounded by, two isovels is equal to the average of their values and and are
calculated using the following expressions.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
a
1 a
2
a
3
a
4
A
v
v
2
v
3
v Vs elemental area
Graphical Method of determining and ( , , ) av av
2
av
3
3
3
A A
3 3
dA
dA
(4)
AV AV
=
and
2
2
A A
2 2
dA
dA
= (5)
AV AV
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Rehbock used a graphical method and reduced the computational work in the above
procedure. After planimetering the areas within each isovel, he plotted the curves of
v,v
2
, and v
3
against the corresponding planimetered areas as shown in Figure. It is
evident that the areas under v
2
, and v
3
curves are equal to
3
dA
and
2
dA
respectively. V, and are computed as shown in the box.
0 4 8
12 16 20 24 28
Graphical method of computing V, and
0
4 8 2 6 10
0
2 1
3
0
4
8
12
0
4
8
12
0
4
8
12
v
3
, m
3
/s
3
v
2
, m
2
/s
2
v, m/s
__
Shaded area = A
0
Shaded area = A
1
Shaded area = A
2
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0 1 2
y
0
1
1
2
2
3
Shaded areas A , A , A are planimetered.
The average velocity
shaded area A
V v dy=
y
shaded area A
Similarly,
V y
shaded area A
and =
V y
=
=
6.1.3 Grid Method
In this method, the flow area is divided into suitably chosen grids an velocities at the
centers of gravity of these grids are measured as shown in Figure 3. Assuming that the
effective velocity through each grid is equal to that at the center of gravity of the grid,
the quantities da
,
2
da
,
3
da
=
Assuming a logarithmic velocity distribution law proposed the following expressions.
2 3 2
1 3 2 ; 1 = + = +
In which
max
is the maximum velocity and V is the mean velocity.
It should be noted that the above approximate formulae are applicable only when the
flow is free from any reverse flow occurring over any part of the crosssection of flow.
6.1.5 Computation of and for Reverse Flow
In case of the reverse flow one of the four methods presented above is directly
applicable. If the reverse flow is occurring over any part of the crosssection of the flow,
and can be calculated using either the graphical or the grid method. While using
these methods it should be noted that the velocity in the reverse flow region should be
assigned a negative sign and all the computations should be done taking the sign also
into consideration.
6.1.6 Values of and in Several Practical Cases
Actual values and in many practical cases (which are frequently met with in
Hydraulic Engineering) are presented in Table I. Some of these values are listed by
OBrien and Hickox OBrien and J ohnson and King. They are reproduced here along
with several other cases for the sake of a comprehensive table of and values.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Sl.
No.
Channel Dimensions Hydraulic
elements
Coefficients Remarks
width
(m)
Max.de
pth (m)
Hydraulic
Radius
(m)
Area
(m
2
)
Critical
depth
(m)
Mean
velocity
(m/s)
Graphi
cal
Rehb
ock
Grap
hical
1 0.60 0.862 0.222 0.519 0.198 0.320 1.20 1.10 1.07
2 1.00 0.862 0.3250 0.895 0.216 0.53 1.22 1.20 1.08
Rectangular
channel 0.9144
m above weir
and
obstructions
upstream
3 1.00 0.874 0.3249 0.893 0.219 0.365 1.41 1.37 1.12
4 1.01 0.429 0.2316 0.431 0.496 2.56 1.07 1.04 1.03
Simson Tunnel
 centre of
straight reach
49.98 m long
5 10.54 3.23 1.86 23.27 1.42 1.01 1.10 1.07 1.05
6 1.987 1.50 0.6309 2.898 0.76 1.48 1.07 1.03 1.034
Horse shoe
conduit straight
reach
7 159.4 3.81 2.438 4.055 1.91 1.024 1.35 1.43 1.121
Rhine 365.76
m below bridge
on a long curve
8 2.59 1.38 0.6949 3.429 0.685 0.886 1.06 1.02 1.01
9 2.67 1.22 0.6492 3.009 0.658 0.874 1.04 1.04 1.014
10 2.74 0.914 0.548 2.19 0.600 0.792 1.04 1.03 1.014
11 2.71 0.618 0.411 1.415 0.53 0.658 1.04 1.02 1.010
12 2.65 0.460 0.326 1.014 0.499 0.569 1.04 1.03 1.012
Sudbury
Aqueduct with
a bottom slope
0.000189
13 0.264
0.053 0.35 2.31 1.161
Computed with
Bazin series 10
14 0.244
0.0366 0.14 0.205 1.138
Computed from
Nikurade's data
15 1.286 0.762
1.07
Series (E)
schoder and
Turner
16 1.286 1.524
1.08
Series (I)
Schoder and
Turner  Run
54 to 58.
17 1.286 1.524 1.60 Series I ibid
18 1.286 1.524 2.08 Series I ibid
19 1.286 3.07
1.80
Series D
Schoder and
Turner Runs
101 to 105
20 1.286 2.743
2.00
Series D, L, M.
Schoder and
Turner
21
1.528 1.105
Triangular
channel
22
1.665 1.225
Trapezoidal
channel
23 1.365 1.085 Pipe
24 1.460 1.164 Shallow ditch
25 1.422 1.136 Natural channel
26 0.45 0.0911
1.222
Experiment
number 2C 
Rajaratnam
Muralidhar
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
27 0.45 0.350
3.72 2.14
Diverging
channel
28
1.76 1.41
Rectangular
open channel
bend
29 0.61
15.40 5.00
Maximum
and in a
hydraulic jump
with an inflow
Froude number
of 7.4.
30
3.87
At the outlet
section of a
draft tube
31
7.40
Spiral flow
under a model
turbine wheel
(Serial No. 1 to 20 are from O'Brien and J ohnson, Enr, Vol. 1113, page 214  216, 1934
August 16 th after J agannadhar Rao and others).
From the table it may be seen that values are larger in nonrectangular channels
compared to rectangular channels and also that the values for natural channels are as
high as 1.422. When there is a reverse flow in the crosssection, the values of are
still larger. The value in the case of a diverging channel is 3.72. For spiral flows a value
of as high as 7.4 has been quoted . All these examples show that there are several
practical cases in which the neglect of and in hydraulic flow computations for a
proper assessment of energy and momentum at any flow section may lead to large
errors.
6.1.7 Variation of and along the Hydraulic jump
The variation of and along the length of hydraulic jump is given in figure below.
J agannadha Rao (1970) conducted the experiments in a flume of 0.6 m width at Indian
Institute of Technology, Kharagpur. The data given is for the case of a hydraulic jump
with an approach flow Froude number of 7.4.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
x
______
y2y1
0
15
0
 0.25
0.5
1.0
5
10
0.75 0.25
0
 0.25
0.5
1.0
0.75 0.25
2
4
6
8
10
12
14
16
Jump Profile Roller Zone
x
______
y2y1
+ +
=
+ +
+ +
=
+ +
+ +
=
+ +
1
3
n = 0.035
1
2
4 m
n = 0.015
1
2
2.5 m
n = 0.035
1
3
S0 = 0.001
5 m 10 m
COMPOUND CANAL CROSSSECTION
10 m
This is particularly true in time of flood, when the river overflows on to its flood plains, or
"berms,". These are known as Compound channel. In this case there are in effect three
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
separate channels. The mean velocity over the berms will be less than that in the main
channel (MC), because of higher resistance to flow (basically due to, smaller depths
over the berms , and due to the higher roughness in the berms. This variation in mean
velocity among the different flow zones (Main channel and berms) is mainly responsible
for values of much higher than those produced by gradual variation within a given
section, so much higher as virtually to nullify any contribution to the value of
produced by gradual velocity variation. However, it is usually accurate enough to
compute by assuming the velocity to be constant within each subsection (zone) of the
waterway; then the following may be written.
( )
( )
( ) ( )
( )
3 3 3
1 2 3
3 3 3
1 2 3
1 2 3
2
3 3 3
1 2 3
3
1 2 3
N
3
i = 1
1 2 3
3
V
1 2 3
1 2 3
1 2 3
1 2 3
1 2 3
1 2 3 1 2 3
1 2 3
=
i i i
A A A
A A A
A A A
A A A
A A A
A A A
A A A A A A
A A A
A A
+ +
=
+ +
+ +
=
+ +
+ +
+ +
+ + + +
=
+ +
( )
N
2
i = 1
N
3
i = 1
Similarly expression for can be obtained.
i
A
i
( )
2
2 2 2
1 1 2 2 3 3
1 2 3
+ +
V + +
A A A
A A A
=
in which
1 1 2 2 3 3
V
1 2 3
A A A
A A A
+ +
=
+ +
.
When flow resistance formula (Manning, Chezy, other formulae) is combined with the
above equations numerical values of , may exceed much higher than 2 under certain
situations. Generally, the value is taken as 1.0 when the information is lacking.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
References:
1. Chow Van Te Open Channel Hydraulics, McGraw Hill Publications, 1958.
2. Henderson F.M. Open Channel Flow, MacMillan Publishing Company, 1966.
3. J aganadha Rao, M.V., Lakshmana Rao, N.S., and Seetharamiah, K. "On the use of
Energy and Momentum coefficients in Hydraulic flow computations"  J ournal  Irrigation
Power CBIP , Volume 27, part 3, pp 315  326, 1970.
4. Strauss.V. The Kinetic Energy Correction Factor and the Momentum Correction
Factor in Open Channels. Proceedings of Twelfth Congress of I.A.H.R., Vol.1, Sept
1967,pp.314323.
5. O' Brien, M.P. : "Discussion on stream flow in general terms" by Casler", Trans.
A.S.C.E. Vol. 94, 1930, pp. 42  47.
6. O' Brien, M.P. and J ohnson, J . W. : "Velocity Head Connections for Hydraulic Flows".
Engineering News Record. Vol. 113, No. 7, pp. 214  216, Aug. 16, 1934.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6.2 Energy, Momentum coefficients for different velocity
distributions
Rehbock obtained
1) For Linear Velocity Distribution
{ }
( )
= +
= + =
= +
=
+ +
= =
max
2
2
V
in which
V
Given:
=>
Substitution for " the expression for " ",
 1
=1 +
3
+2
= (Linear relation)
3
2
2
2
1
1 , 1
3
1
1
"in
3 1 2
3 3
Linear velocity distribution
V
max
y
o
y
v
v __
Vmax
=
y
yo
__
1 1.6 2.2 2.8
1 1.2 1.4 1.6
The plot is shown below
Kinetic energy correction factor
1.1
1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
Scale
Xaxis 1 cm = 0.1
Yaxis 1 cm = 0.1
(2) He obtained for Logarithmic Velocity Distribution the following equations.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
( )
= +
=> =
=> =
2
2
V
Kinetic Energy correction factor,
Momentum correction factor, = 1 +
2.5
in which =
Given : = 1 +
Substituting for " " in the expression for " ",
=1 + 3 1
2 3
2
*
1 3 2
( 1)
1
v
( )
=> =
3
3 / 2
2 1 = 1 + 3 (  1 )  2 (  1 )
2
3/2
3 2( 1) 2
1.0 1
1.1 1.237
1.2 1.421
1.3 1.571
1.4 1.694
1.5 1.793
1.6 1.8705
1.7 1.929
1.8 1.969
1.9 1.992
2.0 2
2.1 1.993
2.2 1.971
2.3 1.9355
2.4 1.887
2.5 1.826
2.6 1.752
2.7 1.667
2.8 1.57
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The plot is shown below
Relationship between and
=1+3
2
2
3
=1+
2
1.0 1.2 1.4 1.6 1.8 2.0
1.4
1.6
1.8
2.0
1.2
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6.2.1 Derivation of relationships
=
=
n
Assuming a wide channel with the two  dimensional velocity distribution given by
y
sin and
2y
y
determine " " and " " ( as a function of exponent n in second case). Hence show
y
0
0
v
V
v
V
( )
=
=
=
 1
that (a) For laminar case and
1
n+3 (2n+1)  1
(b) for turbulent case .
1 (3n+1)
Solution:
y
Case ( a ) : sin
2y
where is the velocity at a depth of " y " from boundary
2.76
0
v
V
v
y y
0
0
, y is the total depth of flow in wide channel.
Let B the width of wide channel.
y
= sin
2y
1
Mean velocity = V = dA
A
y y 1
V = sin B dy = sin dy
By 2y y 2y
2y
=
y
0 0
cos
0
0
0
V V
v
V
V
V
{ }
{ }
y
0
y
y
2y
y 2 2
=
2y 2
 2
=
2
V =
0
cos cos cos(0)
0 1
V V
V
V
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6.2.2 Kinetic Energy Correction Factor
:
{ }
=
=
=
y
3 3 3
3
0
y
3
0
3 2
2 2
2
3
y 1
= dA = sin B dy
2 y
V A
By
y
= sin dy
8 y 2 y
y y y
sin sin sin
2 y 2 y 2y
y y
sin cos 2A = 1 2 sin A
y 2y
y y
sin
2y y
y
sin
2 y
3
3
1
2
cos 1 2
1 1
cos
2 2
0
v V
V
= =
+ +
=
2
y y y y
sin sin sin
2 y 2 y 2 y y
y y y 1 1
= sin sin
2 2 y 2 2 y 2 y
sin ( A+B ) +sin (A B )
sinA cosB =
2
y y
sin s
2 y y y y
sin
2 y 2 y
1 1
cos
2 2
cos
cos
=
3
y y
in
2 y y
2
y y 1 1
= sin sin
2 2 y 2 2 y
y y 1 1
= sin sin sin ( A) = sinA
2 2 y 2 2 y
y
sin s
2 y
3
3
1
2
0
y
3
y y y
in sin
2y 2y 2 y
y y y
= sin sin sin
2y 2y 2 y
y
= sin dy
8y 2y
3
0
1
cos
2
3 1 1 1
2 4 4
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( ) ( ) ( )
{ }
+
+
0 0 0
y y y
3 3 3
y y y
y y y 1 1
= sin dy  sin dy+ sin dy
8y 2y 4 2y 4 2y
y y y y y y
=
8y 2y 2y 2y
y y y
=
8y 6 2
y
=
8y
3
0 0 0
3
3
0 0 0
3
3
3 1
2
2 2 2 1 1 1
cos cos cos
2 4 3 4
0 1 0 1 0 1
{ } { }
+ =
=
o
2
y y y y 3y
6 8y 6
y
=
8y 6
=
6
3
3 2
6
2
8
6
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6.2.3 Momentum correction factor
{ }
=
=
o
o
o
y
2 2
2
0
2
y
2
0
2
2
0
y
2
0
2
y 1 1
= dA= sin B dy
2 y
2
V A
B y
y
= sin dy
4y 2 y
A = 12 sin A
y y
sin
2 y y
y
sin dy
4 y 2 y
=
2
2
2
cos 2
1 1
cos
2 2
0
0
V V
V
[ ]
{ }
= =
o o
0
0
y y
0 0
0
y
2
y
0
0
2
2
2
y
dy dy
4y y
y y
= y sin
4y y
y
= y  0 )
4y 2
y
=
4y y
=
0
2
2
1 1
cos
2 2
1 1
2 2
1
( (0 0)
2
8
1
1
6
1
1
8
2
8
2
6
2.76
6 8
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
=
n
n
y
case (b ):
y
where velocity at a depth " y " from boundary, y is the total depth of wide channel.
Let B the width of wide channel
y
=
y
Mean velocity =
v
V
v is the
v V
( )
( )
( )
( )
( )
y y
n
n
n n+1
0
0
y
n+1 n+1
n+1 n+1
1
V = dA
A
y 1
V = B dy = y dy
B y y y
y y
=
n+1 n+1
y y
=> V =
n+1
Kinetic energy correction factor :
0 0
0
.
0
v
V
V
V V
V
{ }
( )
( ) ( )
( )
0
n
o
y
3n
3
3 3n
o
0
y
3n
3n
0
y
3n+1 3n+1
3n+1 3n+1
0 0
y 1 1
= dA = B dy
y
V A
B y
n+1
n+1
= y dy
y y
n+1 n+1 y y
=>
3n+1 3n+1 y y
n+1
=> =
3n+1
3
0
3
0
3 3
0
3
0
3
0
v V
V
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
{ }
( )
( )
0
y
2n
2 2
2 2n
0
y
n
2n
0
y
2n+1
2n+1
Momentum correction factor
y 1
= dA B dy
y
V A
B y
n+1
n+1
= y dy
y y
n+1 y
=
2n+1 y
2
0
2
2
0
1
0
V V
V
( )
( )
( )
( )
( )( )
+ +
= =
+
2n+1
0
2n+1
3 2
2
n+1 y
=
2n+1 y
n+1
=> =
2n+1
n n n+13n1
n+1
(3n+1)
3n+1
n n+12n1 n+1
(2n+1)
2n+1
n+ 3 2n+1
n+1)
If n =
2
2
3
2
0
3 3
1
1
1 2
1
1
1 (3
( )
{ }
( )
{ }
+
= =
+
=> = =
+
= =
+
=> = =
1
7
n+1
1
n+1
*
7
n+1
1
n+1
7
3
3
2
2
1
1
7
3
3 1
1.4927
1.0449
1.4285
1
1
7
2
2 * 1
1.3061
1.0158
1.2857
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Example:
Obtain and for the velocity distribution given below
( )
( )
( )
( )
( )
1
0
1
2
0
h 1
3
3
3 3
0 0
1
3 2
0
y
u = 0.4 + 0.6 , h=1.0,
h
Solution:
1 1
u = udy 0 4 0 6
h 1
y
0.4 y + 0.6
2
= 0.7 m/s
1 1
= u dy = 0.4+0.6y dy
u h 0.7 *1
1
= 0.064+0.216y 0 432 0 288 dy
0.343
y
. .
h
. y . y
= +
=
+ +
1
4 3 2
0
1
0.064 +0.216y 0 432 0 288
0.343
= 1.18
y . y . y
= + +
Problems:
1. The velocity distribution ( in m/s ) in an open channel 2m deep can be
represented by the equation,
0
y
v(y) = 0.6 +1.4 (
y
1/2
)
Calculate the energy correction factor. Here in y is the height above bed and
y
o
=2m.
2. In a channel of trapezoidal cross section the velocities were measured at mid
depth at various sub areas. Compute the average values of and for a given
cross sections.
105 m
2 : 1
2 : 1
15 m
2.8 m/s 2.9 m/s 3.0 m/s 3.1 m/s 3.1 m/s 3.0 m/s 2.9 m/s
2.8 m/s
y = 10 m
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3. For an assumed velocity distribution
*
30y
V 5.75V log
K
=
Prove that
2 3 2
1 3 2 1 and = + = + in which
max
max
V
1, V is the maximum velocity, Vis the mean velocity
V
. =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6.3 Comparison Between Momentum and Energy Equation
Theoretically when the flow is gradually varied, energy and momentum equation should
yield same results. Consider a gradually varied flow. The pressure distribution in the
sections is taken as hydrostatic, the channel bed slope as small. For a rectangular
channel of small slope and width b, in a short reach the expression for pressure forces
can be written as
2
1 1
2
2 2
1
2
1
and
2
Force due to friction can be written as
f f
P by
P by
If P
=
=
h by =
in which is the friction head and
'
f
h y is the average depth, or ( y
1
+y
2
) / 2. The
discharge through the reach is equal to
( ) 1 2
1 2
1
V V
2
Also, the weight of the body of water is
sin
Then the momentum equation, after substituting these expressions simp
Q by
W byL
z z
and
L
= +
=
=
2 2
'
1 2
1 1 1 2 2 2
lifies (see box) as
V V
2 2
f
z y z y h
g g
+ + = + + +
2 2
V V
1 2
z z h
1 1 1 2 2 2 f
12 2g 2g
y y + + = + + +
This equation appears to be practically the same as the energy equation (Bernoulli
equation). However, the energy loss given by momentum equation is due to external
forces whereas the loss given by energy equation is due to internal forces. One is a
vector quantity and other is scalar quantity. However, if the flow is uniform, then h
f
=h'
f
if
the difference between and is ignored. Similarity ends here. There are cases where
either momentum equation or energy equation can be used with the continuity equation.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Momentum Application
1 2 1 2
1 2
1 2
1 2 2 2 '
2 2 1 1 1 2 f
2 2
2 1
2 1 2 2 1 1 1 2
y +y V +V V +V
Q = b =by
2 2 2
W = byL= Specific weight * (Volume)
z z
sin =
L
V +V 1 1
by V  V = by  by + by Lsin  h by
g 2 2 2
by 1
V V V V V V
2g
+ =
2 2 ' 1 2
1 2 f
2 2
2 2 ' 1 2 1 2 1 2
2 1 1 2
2 1 2 2 1 1 1 2 1 2 f
2 2
2 2 2 1 2 2 1 2 1 1
2 1 2 1 1 2
z z 1
by  by + by L  h by
2 2 L
divided by b
y +y y +y y +y y 1 1
V V V V V V y  y + z  z h
2g 2 2 2 2 2
y +y V V V V V V 1 1
y  y +
2 2g 2g 2g 2g 2 2
+ =
+ =
( )
2 2
'
1 2 1 2 1 2
f
1 1 2 2
2
2
'
2 2 2
1 2 1 1 1
f
2 1 1 2 1 2
2 2
2 1 1 1
1 2
'
1 2
f
1 1 1 2 2 2
y y y y y +y
z + z z z h
2 2 2 2 2
Simplifying
V V V V V V
y y z z h
2g 2g 2g
V V V V
If we can neglect 0
2g
V V
z z h
2g 2g
y y
+
+ =
+ + = + + +
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
7.1 Pressure Distribution
The atmospheric pressure is impressed on the free surface boundary. Hence, the
reference pressure on the free surface is taken as zero pressure. The pressure
distribution in free surface flows is governed by the acceleration including gravity. Thus
Euler's equation in s and n directions can be written as
( )
( )
s
n
p Z a
s
p Z a
s
+ =
+ =
m direction
binormal
rectifying
plane
tangent s direction
normal
plane
stream line
osculating
plane.
n direction
(Principal normal)
The direction of the normal to s direction is towards the plane Centre of curvature is
considered as positive.
Thus the acceleration a
n
is given by
2
n
v
a
r
=
in which v is the velocity of flow along the streamline, r is the radius of curvature of the
streamline.
(i) If a
n
is zero then (a) v =0, no flow and (b)r , the streamlines are straight
lines.
(a) v =0, then
( ) p z 0
s
p
z constant.
+ =
+ =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
h =h
s
h
Hydrostatic pressure distribution in parallel flows
At free surface
p
0
= , hence constant =z
1
Therefore, at any point x below the free surface, the pressure p
x
x
x
p
the distance from the free surface say 'h'
p h
=
=
h
Ho
X
h
P=
h
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(ii) In general, when the flow is in the channel with small slope bed , then the
streamlines are nearly parallel to the bed. The vertical depth and the depth
normal to boundary are nearly same. Hence, one can assume the hydrostatic
pressure distribution to be valid.
(iii) In case of large channel slope, expression for pressure can be written as
Pressure at a point
x
can support the weight of the fluid.
x
x
P x y xcos
or P ycos
=
=
y
A
A
A'
A
A'
c B B'
h = y cos
2
y
Pressure distribution
on A'C
Pressure distribution in parallel flow in channels of large slope
If h is the total depth normal to the boundary, then the vertical depth d can be related
to h =dcos
2
p
hcos dcos
= =
Thus the hydraulic grade line does not match with the water surface.
(iv) Pressure distribution over curved boundaries.
In field situation when the flow has to pass over a spillway, smooth curves are
provided near the crest. Similarly for energy dissipation the buckets are provided.
The streamline have a large curvature. Hence, pressure distribution requires to be
converted. The curves could be either convex or concave. Theoretically this flow is
known as curvilinear flow. The curvature introduce appreciable acceleration
components or centrifugal force normal to the direction of flow. Thus the connection
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
for the hydrostatic pressure distribution is to be introduced and thus it can be written
as
s s
h h c h h c = + = for convex.
r
concave and convex profile on spillway
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
hs
c
h
A
c
h
B
B'
h = hs  c
Convex surface: Centrifugal force
opposing Gravity force
Example: Spillway Crest
Non Hydrostatic Pressure distribution
hs
c
h
c
h
A
B B'
Concave surface: Centrifugal force
in the same direction of
Gravity force
Example: Flip Bucket
Non Hydrostatic Pressure distribution
h = hs + c
n n
n
n
s
2
n
a a p p
z r c, for Concave 1 y for Convex section
g g
For a Concave vertical section
a p
y 1
a
thus h=h c in which c =
g
v
In a curvilinear flow a
r
+ = + =
= +
=
2
v
c =
gr
If the variation of v w.r.t to r is known, then acceleration could be evaluated.
The following three situations arise in the field
(i) v =constant and equal to mean velocity.
c
(ii) v = (free v
r
2 2
0.5
0.5
ortex)
(iii) v =rc (forced vortex)
v v
(iv) , R is the radius of curvature at the mid depth.
(r d 2) R
=
+
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Problems:
Show that for a circular spillway bucket having a radius of
curvature R the effective pressure distribution is
(a) If the velocity is constant over thedepth y it can be shown that the
pressure at any
( )
2
2
c
c
c
c
cp 2
c
c
point r and is
p v r
r R y cos ln
g R y
(b) Effective piezometric head.
y 1
v ln
y
R
1
R
h Z ycos
h
gR cos
R
= + +
= + +
1
V
__
y
R
c
Flow in a bucket
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Example:
Compute the overturning moment due to pressure on a retaining wall
soln: (i) Assume to be small
P
y
3
__
y
y
Force acting on the retaining wall, P =Area of pressure triangle.
=
2
1 y
y y
2 2
=
Overturning moment =P * distance from the base at which P is acting
2 3
1 1 1
y * y y
2 2 6
= =
(iii) If is not negligible, y =ycos
2
( )
2
2
2
4
2 3
4 4
ycos
y
P cos
2 2
y y y
over turning moment cos * cos
6 3 6
= =
= =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
7.2 Pressure correction coefficients
( )
( )
A
0
A
0
s
A
s
0
A A
s
0 0
A
0
Az hdA
1
hdA
Az
but h h C
1
h C dA
Az
1 1
= h dA C dA
Az Az
1
1 C dA
Az
=
=
=
=
+
= +
A A
0 0
2
s
Show that
1 1
hvdA 1 cvdA
Qy Qy
in which is the pressure distribution coefficient.
dv
c= , d is the depth of flow in the section.
g r
Solution: h=h +C
Head recorded in a curve =static pressure cor
= = +
( )
A
0
A
0
A A
s
0 0
A A
s
0 0
rection factor.
pressure =Mass of water * depth
= g v dA h
Also pressure is yQ g
Thus yQ g = g v dA
1 1
v dA h v dA h C
Qy Qy
1 1
h v dA + v dA C
Qy Qy
= = +
=
A
0
s
1
1 v dA C
Qy
for uniform flow h vdA Qy.
= +
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
8.1 Specific energy equations for rectangular channels
Specific energy in Open channel is defined as the energy per
N m
N
of water at any
section of a channel measured with respect to the channel bottom. Thus it is
the total energy with z being zero.
A
E
Total Energy E
r
Z
A
d
d
b
section AA
Specific energy equation
explanation of symbols.
See also Notation Table
Datum
V
__
Notations
r
E =Total energy above datum =
2
V
z + d cos +
2g
E =Specific energy =
2
2
2
V q
d cos +
2
2
d cos
g
gd
+ =
Q =Discharge, b =channel width, d =flow depth,
q =Discharge per unit width =Q/b,
tan =Bed slope, =Velocity coefficient, g =Acceleration due to gravity
Thus specific energy can be written as
2
V
E = d cos +
2g
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The concept of specific energy as it applies to open channels with small slopes is given
below.
2
2
2
Total energy equation is
p v
z cons tan t
2g
In other words it can be rewritten as
v
y z cons tan t
2g
If z 0 then
v
E y
2g
+ + =
+ + =
=
= +
which indicates that the specific energy is the sum of the depth of water and the velocity
head.
8.1.1 Specific energy diagram
Solution of the specific energy equation for rectangular channels
Consider a specific energy equation for the case of a rectangular channel.
2
v
E y
2g
= +
2
2
2
2 2
Disch arge Q V A
Q
Therefore V
A
Q Q
V
A
b y
in which b is the width of the channel and y is the depth of flow.
=
=
=
=
Substituting this in the specific energy equation it can be written as
( )
( )
( )
2
2 2
2 2
2
2
2
Q
E y
2gy b
Q
Defining q
b
Q q
Then E y y a constant
2g
2gb
E y y constant
=
=
= =
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The above algebraic equation is a cubic equation and there are three routes for y for
given specific energy E. Out of which two are positive roots and one is negative root. i.e.
y<0 which is physically impossible. Therefore it is only an imaginary solution. The two
positive depths are called alternative depths. Normally indicated as y
1
and y
2
for
supercritical and sub critical condtions and are known as low stage and high stage
values of depths.
This is graphically shown in Figure, where the specific energy is plotted against the
depth, for a given discharge per unit width, rendering the familiar representation.
d
depth
E<E
c
E=E
c
E>E
c
two imaginary roots
one negative real root
two positive real roots
one negative real root
Specific energy where E=dcos + (q /2gd )
2
2
When
c
E E > (minimum energy for a given q) three real unequal roots are obtained:
two positive ones (sub critical and supercritical depths) and a third one negative (no
physical meaning). When
c
E E = the two positive roots become equal and this depth is
the critical depth. When
c
E E < the two positive roots become imaginary and the third
one remains negative.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Figure below shows the variation of the specific energy as a function of depth when the
discharge per unit width changes. when q increases the corresponding critical depths
increase and the positive and negative limbs of the function move away from the origin.
The opposite applies when q decreases. When q=0 the critical depth is equal to zero,
the sub critical depth equals E / cos and the supercritical depth (and the negative root)
are equal to zero.
B
A
C
45
o
q1
q2
q3
q3 > q2 > q1
SubCritical flow
SuperCritical flow
y
1
y
2
y
c
Specific energy E
Imaginary
Specific energy diagram
The Specific energy curve is confined between two asymptotes namely y =E and y =0.
The first asymptote is at 45 with respect to abscissa. However, if the effect of the bed
slope of the channel is considered the angel will be different from 45.
For a given Q, specific energy curve has two limbs BA and AC.
Line BA represents Sub critical flow
Line AC represents Super critical flow
C represents Critical flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
For a given Specific energy E there are three possible depths: Two positive values and
one negative value. Two positive values are y
1
and y
2
respectively representing Super
critical and Sub critical depths. The minimum value of specific energy for the given
discharge represents the critical flow condition.
The locus of this represented by
c
2
y E
3
=
For different values of discharges namely Q
1
, Q
2
, Q
3
different specific energy curves
would be there.
The minimum specific energy represents the critical condition.
2
2
3
2
3
2
2
2
2
V
2g
dE dV
2V
dd 2g dd
2 Q dA
0
dd
2gA
2 Q
T
2gA
Q V
2g
gA
V
1
gDcos
Making
E= d cos +
cos +
=cos 
cos =
D cos =
=
=
=
=
V
F
cos
This is for non rectangular channel.
If 1, is verysmall then it can be written as
V
F
gD
In which D is the hydraulic mean depth
Thus the specific energy is minimu
F can be defined as
=
gD
and
.
=
=
2
2 2
m when the flow is critical.
V
1
gD
V V D
D
g 2g 2
or
=
= =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
8.2 Application of Specific energy
Transition Problem:
y
1
V1
__
C
1
y
2
Sluice gate
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Example 1:
1 2
2
1
1
Flowbelow asluicegate
Problem: If y 2.5m, y 0.60m, b 3.5m, determine the discharge Q.
Solution:
Apply Bernoulli equation beteween sections 1and 2, assuming l osses are negligible.
V
y z
2g
= = =
+ +
( ) ( )
( )
2
2
1 2 2
1 2
1 2
1 2
1 2
1 2
1 2
1 2
2 1
2 1
2 2 2
2 2 1 1
V
y z
2g
z z andwidth bis constant
Q A V A V
by V by V
3.5 2.5 V 3.5 0.6 V
8.75V 2.10 V
8.75
V V
2.1
V 4.16V
V V V
4.16 17.36
2g 2g 2g
Substitutingthevalue
= + +
=
= =
= =
= =
=
=
=
= =
( )
( )
2 2
2 1 1
2 2
1 1
2
1
1
1 1
3
2
s intospecific energy equation
V V
2.5 0.6 4.16
2g 2g
V V
2.5 0.6 17.36 ,
2g 2g
V
16.36 2.5 0.6
2g
2.5 0.6 *2*9.81
V = V =1.5095 ms
16.36
Q =13.2081m/s
V =6.2795 m/s
Froudenumber in thedownst
+ = +
+ = +
=
2
2
2
V
reamF
gy
6.28
2.59
9.81*0.6
=
=
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Example 2:
Consider a transition with a vertical step of height z in bed, in a rectangular channel of
constant width b. upward step z is considered as positive. What is the depth over the
step?
y1
V
1
y2
V
2
y
Z
y1
y
2
V2
y
Z
__ __
V
1
__
__
Positive step of z height
Negative step of z height
2 2
1 2
1 2
2 2
1 2
2 2
1 2
2 2
2 1
2 2
2 1
2 1
1 2 2
V V
y y z
2g 2g
Q q q
q, y y z
b
2gy 2gy
q q
y y z
2gy 2gy
E E z
E and z areknown E is tobesolvedfor y by trial anderror or
using solution of cubic equation.
Note:
Subcritical f
.
+ = + +
= + = + +
+ = +
=
+ + =
=
+
+
2
2
Consider uniform
Q
E y constant (approximately)
2gA
Assume
E y 0
2
2
cos
anglebetween slope of straight line and horizontal axis as
cos is one asymptote
=
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
y
E
E y
y
1
E
1
1
The
2
2
2
2
1
2
tan
cos
cos from figure
1=tan cos
tan
cos
The expression for the slope of straight line to which upper limb of specific energy curve is
tan
cos
angle depends u
=
=
=
=
=
pon the bed slope of the channel.
y
d
y
d
y
E
Problem:
Plot the specific energy vs depth curves for Q =400, 600 and 800 m
3
/s in a trapezoidal
channel having bottom width of 20 m and the side slopes of 2(H) : 1(V). Assume the
bottom slope as small. From these curves, determine the critical depth for each
discharge. Write a computer program to obtain the above.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
8.3 Problems
1. A rectangular channel, 9.15m wide carries 7.65 m
3
/s when flowing 9l5 mm deep. (a)
What is the specific energy? (b) Is the flow sub critical or supercritical?
2. A trapezoidal channel has a bottom width of 6.0 m and side slopes of 2 horizontal to
1 vertical. When the depth of water is 1.07m, the flow is 10.50 m
3
/s. (a) what is the
specific energy?
(b) Is the flow sub critical or supercritical?
3. The discharge through a rectangular channel (n =0.012) 4.60m wide is 11.30m
3
/s.
When the slope is 1m in 1 00m, is the flow sub critical or supercritical?
4. A rectangular channel 3m wide, carries 11.3 m
3
/s.
(a) Tabulate (as a preliminary for preparing a diagram) depth of flow against specific
energy for depths from 0.30 m to 2.4m.
(b) Determine the minimum specific energy.
(c)What type of flow exists when the depth is 0.6m and when it is 2.4m?
(d) For C =55, what slopes are necessary to maintain the depth in (c)?
5. Water flows at a Velocity of 1m/s and a depth of 0.25m in a rectangular channel. Find
the critical depth. Find the alternate depths assuming no change in specific energy. 6.
As shown in Figure, the depths at a short distance u/s and d/s of sluice gate in the
horizontal channel are 3.0m and 0.60m respectively. The channel is rectangular in
section and 3m wide. Find the discharge under the gate.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3.0 m
V / 2g
Sluice Gate
0.6 m
V
2
1
2
__
__
6. The depth of flow and flow velocity upstream of a 0.2 m sudden step rise in the
bottom of 5 m wide rectangular channel are 5 m/s and 4 m/s respectively. Assuming
there are no losses at the transition, determine,
(i) The flow depth at downstream of the step and change in water level.
(ii) The flow depth of water level downstream of the step if the channel bottom has drop
of 0.2 m instead of rise.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
9.1 Specific Force
A short horizontal reach of a prismatic channel is considered. Further, the external
frictional force and the effect of weight component of water can be considered as
negligible. Then
( )
2 1
2 1 1 2 f
Q
V  V =P P +WsinP
g
( )
f 1 2
2 1
1 2
1 2
If = 0, and P 0 and also if 1, then the momentum equation simplifies can be written as
Q
V V
g
The hydrostatic pressure forces P and P are respectively
P P
= = =
=
1 1 1 2 2 2
1 2
1 2
in which z and z are the distances to the centroids below the surface of flow
of the respective water flow areas (A and A ).
P z A and P z A = =
z
_
centroid from free surface
1 2
1 2
2 2
1 1 2 2
1 2
Q Q
Also,V = and V = .
A A
Then, the momentum equation reduces to
Q Q
+ z A = + z A
gA gA
The two sides of the above equation are analogous and, hence, may be generally
2
expressed
for any channel geometry as
Q
M = + zA
gA
The first term is the rate of change of momentum of the flow passing through the
channel section per unit weight of water, and the second term is the force per unit
weight of water. Since both terms are essentially force per unit weight of water, their
sum is known as the specific force indicated as M. Accordingly, it may be expressed as
M
1
=M
2
. This means that the specific forces of sections 1 and 2 are equal, provided that
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
the external forces and the weight effect of water in the reach between the two sections
can be ignored.
9.1.1 The momentum Function  Rectangular channels
The general situation is shown in Figure in which there may or may not be an energy
loss between sections 1 and 2, and there may or may not be some obstacle on which
there is a drag force P
f
. In Figure the direction of P
f
is that of the force exerted by the
obstacle on the flow. It is this force (not the drag on the obstacle) which is to be
considered in the momentum equation.
Flow
Pf
P1
Definition Sketch  Momentum Equation
P2
If there are any bluff body offering resistance force (P
f
) to flow then
f
1 2
P
M  M =
The force P
f
should include the frictional resistance due to boundary surface, and weight
of the bluff body.
The following are some of the particular cases that occur in the field
1.Energy loss E = 0, P 0
f
(the sluice gate)
2. E 0, P = 0
f
(the simple hydraulic jump)
3. E 0, P 0
f
(the hydraulic jump with its formation assisted by some obstructions
in the flow such as dentated sill (Forced hydraulic jump)
Sequent depths of Normal Hydraulic jump
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
If P
f
=0 then the specific force equations can be simplified as
`
( )
( )
( )
2
1 1 1
2 2
1 2
2
1 2
2
1
. .,
2 1
2
1 2
The substitution q = v y leads to
1 1
2
1
1 2
2 1
2
1
2
1
2 1 2 2
F 1
1
2
1 1 1
=
= +
= +
= = +
q
y y
g y y
q
i e y y
gy y
v y
y y
g y
v y y
or
gy y y
which is the well known equation of the normal hyraulic jump (NHJ ). The Froude
number F plays a key role. The above equation is quadratic in y
2
/y
1
, whose solution is
given by
= +
y
1
2 2
1 8F 1
1
y 2
1
and
= +
y
1
2 1
1 8F 1
2
y 2
2
In general, there are three independent quantities, and knowing two of them initially
third one can be calculated. The downstream control can create appropriate conditions
to form the jump. The corresponding depths y
1
and y
2
are known as conjugate or
sequent depths.
9.1.2 Specific Force Diagram
The diagram shows the variation of the depth against the specific force for a given
channel section and discharge, is called specific  force diagram. This curve has two
limbs AC and BC. The limb BC approaches the horizontal axis asymptotically toward
the right. The limb CA rises upward and extends indefinitely to the right. For a given
value of the specific force, the curve has two possible depths y
1
and y
2
. These two
depths constitute the initial and sequent depths of a hydraulic jump (see box). At point C
the specific force is minimum at the critical depth (see box).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1
3
C
2
M
M
Specificforce
diagram
Specificforce
A
B
C is the point of minimum specific force for a given discharge This corresponds to
critical depth, AC is the sub critical limb, BC is the super critical limb. For a given
specific energy there are two depths (Points 2, and 3 respectively) known as sequent
depths. The difference between points 1 and 3 represent M =specific force at point 1
minus the specific force at point 3.
The phenomenon of the hydraulic jump occurs when flow changes from supercritical to
sub critical flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Minimum value of specific force:
The specific force to be of a minimum value then the first derivation of M with
respect to y should be zero, i.e.
( )
2
2
d zA
dM Q dA
=  + = 0
dy gA dy dy
For a change die in the depth, the corresponding change ( ) d zA click in the static
moment of the water area becomes ( ) d zA Ady. Then the above equation
simplifies as
2
2
2
dM Q dA
=  +A = 0
dy gA dy
Since, dA / dy = T, Q /A = V, and A / T = D. the above equation reduces to
V D
=
2g 2
This is the criterion for the critical flow condition (Froude number =1). Therefore,
the depth at the minimum value of the specific force is the critical depth. In other
words the specific force is minimum for the given discharge at the critical state of
flow.
9.1.3 Comparison between specific force and specific energy
For a given specific energy E
1
, the specific  energy curve indicates two possible
depths, namely, a low stage y
1
in the supercritical flow region and a high stage y
2
in the
sub critical flow region. For a given value of M
1
, the specificforce curve also indicates
two possible depths, namely, an initial depth y
1
in the supercritical region and a sequent
depth y
2
in the sub critical flow region. If the low stage and the initial depth are both
equal to y
1
. Then the sequent depth y
2
is always less than the high stage y'
2
.
Furthermore, the energy content E
2
for the depth y
2
is less than the energy content E
1
for the depth y
2
. Hence, in order to maintain a constant value of M
1
, the depth of flow
may be changed from y
1
to y
2
which results in loss of specific energy is
1 2
E E E = .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
E
0
2
c
1
M y
1
1
2
y
2
0
2
y y
2
0
c
M
y
c
Sluice
gate
Hydraulic
jump
Specificforce
diagram
Specificenergy
diagram
Hydraulic jump at sluice gate outlet
1
E
E
1
y
0
y2
'
Specificforce curve supplemented with specificenergy curve.
(a) Specificenergy curve; (b) channel section; (c) specificforce curve
y
E P
P'
2
P"
2
C'
P'
1
E
E1 E2
45 for a channel
of zero or small
slope
0
(a) (b) (c)
B
P
A
M1
P2
C
P1
y
y
y2
yc y1
Centroid
y'
2
yc
y1
T
dy
dA
z
_
0
M
M
E
Note:
Specfic energy diagram Specific force diagram
1. Given E
1
as initial depth y
1
(point
'
1
P ).
Initial depth y
1
is super critical depth.
2. Corresponding to E
1
the alternate depth
'
2
y on sub critical limb
'
2
P
3. The
sequent depth due to hydraulic
jump is y
2
and the corresponding specific
1. Corresponding to initial depth y
1
specific
force is M
1
(Point P1). Initial depth y
1
is
super critical depth.
2. Corresponding to alternate depth
'
2
y the
specific force is point P.
3. The corresponding specific force for the
sequent depth is M
1
. In other words for
Normal Hydraulic J ump, the upstream and
downstream specific forces are the same.
4. The specific force corresponding to
sequent depth is indicated by the point P
2
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In Hydraulic jump energy loss takes place. The depth corresponding to given E
1
at high
stage is known as alternate depth to y
1
and vice versa. Whereas the depths due to jump
are known as sequent depths.
y
1
, y
2
are sequent depths.
y
1
,
'
2
y are alternate depths.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
9.2 Transition: A Field Example
1. A horizontal channel converges from width b1 to b2 over a distance L. Approach flow
is sub critical. No hydraulic drop is permitted. Given Q, y
1
. Determine the water surface
profile.
Solution
b1
b
2
Critical depth line
b
x
x
L
0
yx
Specific energy line
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1
1 1
1
1 1
2
1
1 1
2
2
2
1
2 2 2
2 2
2 2
Steps
Given Qanddepth y
Q
1. Q A V V
b y
V
E y ,
2g
2. It is assumethat no energy loss takes placealongthetransition
Q V
3. E y y
2g
2gy b
4.
,
= =
= +
= + = +
( )
1 2 2
x
2
1 x
2 2
x x
1 2
x 1
E E , obtain subcritical depth y by trial anderror or by direct solution
5. Let Subcritical depth at any section x beis y
Q
E y
2gy b
b b x
b b
L
6.
=
= +
=
x
1
Solvefor y for various x.
a. Plot theprofile:
In this caseas thetransition is astraight wall transition,
water surfacecan bejoinedbetween y
2
andy .
2. In the above problem if a hydraulic drop is permitted at a distance x, determine the
water surface profile what would be the constriction width?
Solution
Hydraulic drop means flow changes from sub critical to super critical via y
c
2
3
c
2
x
2
1 1
2 2
1 1
c x
1
Step1:
Q
y
gb
Step2:
Q
E y
2gb y
2
y E b canbe determined.
3
=
= +
=
Thus maximum constriction at x is known. After determining the b
x
obtain super critical depths
(low stage depths) in the downstream of this constriction. If b
x1
is the width at a distance x
1
from the constriction then,
( )
2 x
x1 x 1
2
x1 x1
2 2
x1 x1
b b
b b x
L
Q
E y
2gb y
2
E
= +
= = +
Solve for y
x1
for super critical flow conditions. Figure shows the typical water surface profile.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
b1
b2
Critical depth line
bx
x
L
0
yc
Specific energy line
High stage
Low stage
Plan
Water surface profile with hydraulic drop
Sub critical flow
Super critical flow
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3. Super critical flow occurs in the approach channel of a transition in which b
1
changes
to b
2
as shown in figure. Locate the hydraulic jump if it occurs.
Solution
bx b1
b2
x1
Specific force line for high stage
Specific force line for low stage
Specific energy line
for upstream
Specific energy line
for downstream
Low stage
High stage
E
a) When jump occurs energy loss takes place
2
2
2
1
1 1
2 2
1 1
E E
V Q
y y
2g
2gy b
1
1
E
E
=
= + = +
b)
1
1 1
1
V
y is given Q is given F
gy
=
c) Given at the downstream section, should be sub critical depth if the jump
has to occur.
2 2
Q, b ,y
2
y
d) J ump occurs but (i) can occur in the transition reach, (ii) not in the transition reach.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1
2
Step1: ComputeE andplot theline.
ComputeE andplot theline.
2
1 1
1
2
2
2
Q
Step2: Computespecific forceM zA
gA
Q
Similarly compute M z
gA
= +
= +
Step 3:
When the specific force M
1
=M
2
, the hydraulic jump forms. It may be noted that jump will have
certain length. In this calculation it is assumed that it occurs in a section.
x bx
low stage
depth for
specific
energy E1
Specific
force for
low stage
Specific
force for
high stage
Remarks
0 b
1
y
1
M
1
M
2
2
5


x b
x
From the above computation plot a force lines and the intersection gives the location of
the jump.
The location of the jump is at the section where the specific forces are equal. Therefore
solving these two algebraic equations for specific forces simultaneously the location of
the jump x can be determined.
4. Elimination of the jump by a hump
In the above problem modify the transition to eliminate the jump by providing a hump.
Obtain the hump profile.
Solution
1. Assume a smooth water surface profile between approach flow depth and the
downstream depth. Thus an elevation H
x
of the water surface profile is known.
2. Assume that E
1
to E
2
loss is distributed linearly over transition.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
2
x x x
x x x
x 1 2
1 x
3. It can be written that
H y z ,
z H y
At any point x the specific energy is given by
E E
V
E  x H
L 2g
I nthe above equation only unknown is velocity. Hence calculate the
= +
=
= +
( )
2 2
2 1
x
x 1
2 2 2
x x x
x x x
velocity
b b
Q Q
But V . Calculate y knowing b x
L
A b y 2g
Obtain z fromtheknown H y
and plot the z as a function of x to obtain the hum
x
.
b
by subtracting the depth
= = = +
p profile.
Alternative solution for eliminating the jump is to increase the roughness
Inother words increasing the friction. Also by changing the width.
bx b1
b2
x1
Specific force line for high stage
Specific force line for low stage
Specific energy line
for upstream
Specific energy line
for downstream
zx
yx
V
__2
x ____
2g
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
5. The flow is taking part a section shown in Figure. The step height is 4.57 cm. The
upstream depth 45.7 cm. The water surface drops by 7.63 cm from its original level
on the step. Determine the discharge.
45.7
7.63 cm
Flow over a step
Z = 4.57 cm
Solution
2 2
2 2
2
1 2
1 1 2 2
1
2
1 2
1 2
2
1
1 2 2 1
1 2
1 2
y 45.7 4.57 7.63 45.7 12.2 33.5
V V
E y E y
2g 2g
y V 33.5
by V by V 0.733
y 45.7 V
So V 0.733V or V 1.364 V
V V
z y y
2g 2g
4.57 45.
cm
= = =
= + = +
= = =
= =
= + +
=
=
2 2
2
2
1 2
1
2
1
1
V V
7 33.5 1
2g
V
V 1
1 4.57 45.7 33.5
2g 0.733
V 131.866cm/s
Q 6.026l /s/cm
+
= +
=
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6. Water flows in a rectangular channel 3 m wide at a velocity of 3 m/s at a depth of
3 m. There is an upward step of 0.61 m. What expansion in width must take place
simultaneously for this critical flow to be possible?
Solution
3 m
3 m
3 m
3 m/s
0.61 m
3 m
3 m/s
b = ?
3
3
1/3 1/3
2 2
c
c
2 2
1
1 1
2
' 1
1
c
Q 3*3*3 27m /s
Q 27
q 9.0 m /s/ m
b 3
q 9.0
y 2.021 m
g 9.81
V 3 N
E y 3 3.46875
2g 19.2 N
3
3
E
19.2
E 1.72
y 2.021
= =
= = =
= = =
= + = + =
+
= = =
m
' '
2 1
c c
' 2
2 2
c2
Z 3.46875 0.61
Downstreamspecific energy E E 1.4145
y y 2.021
If theflowhas to becritical
E
E 3.46875 0.61 2.85875 E 1.5
y
= = =
= = = =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
c2
2 c2 c2 2 c2 c2
2
E 2.85875
y 1.9058
1.5 1.5
Q b y V b y gy
27
b 3.2765m
1.9058 9.81*1.9058
For critical flow to occur downstreamwidth must be3.2765m.
Minimumexpansion permittedis 0.2765min width
=
= =
= =
= =
7. A rectangular channel of width 23 cm expands to 481 cms over a horizontal
distance of 400 cm. The approach channel has a bed slope of 0.0016. At the
junction bed drops by 25 cm over a length of 400 cm. The discharge is 11 l/s.
The approach flow is uniform flow. In the downstream a minimum depth of 15 cm
is sustained. A maximum water level of 40 cm is expected. Study the flow profiles
for different downstream depths (between 15 cm to 40 cm). Locate the jump if it
occurs.
23 cm
481 cm
25 cm
400 cm
15 cm
S0 = 0.0016
y
n
Plan
70 cm
Q = 11 l/s
Longitudinal sectional Elevation
8. A transition is as shown in figure. Obtain the water surface profile if the width of
the approaching channel is 50 cm. A discharge of 150 l/s is allowed into the
channel at a depth of 35 cm. Downstream depth is controlled and a depth of 15
cm is maintained. Examine the possibility of a hydraulic jump after the
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
downstream after the transition and if the jump has to occur downstream of the
transition, what necessary modifications are required.
B
B
1.5 B
B
__
2
B
50
75 50
Longitudinal Section along a Transition
(All dimensions are in cm)
Plan
Section along "YY"
B/2
5
Y
Y
9. A rectangular channel of 3.0 m width is narrowed down to 2.5 m by a contraction
in a length of 20 m, built of straight walls and a horizontal bed. If the discharge is 3.5
m
3
/s and the depth of flow is 1.50 m on the upstream side of the transition,
determine the flow surface profile in the contraction (i) allowing no gradual hydraulic
drop (ii) allowing a gradual hydraulic drop having its point of inflexion at the mid
section of the contraction. Neglect frictional losses.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
9.3 Application of Specific Force and Specific Energy
1. Determine the energy Loss in a NHJ
Solution:
( )
( )
( )
( )
( )
( )
2 1 1 2
2
1 2 2 2
1 2
2
1 2
2
1 2 1 2
2
2
2
2
2
1 2 1 2
2 2
1 2
1 2
2 2
1 2
2 2 2 2
1 2
1 2
Applying Momentum equation
Q
V V P P
g
y y
Q
y y
y y
gb
y y y y
Q
4
2gb
Q
q
b
y y y y
q
(1)
2g 4
Specific energy equation
V V
y + y E
2g 2g
Q Q
E = y + y
2gy b 2gy b
q
E = y y
=
+
=
=
+
=
= + +
+
( ) ( )( )
( )
( ) ( )( )
( )
( )
( )
2
2 2
1 2
2
1 2 2 1
2 2
1 2
2 1 1 2 2 1
1 2
2 2
1 2
2
1 2
1 2 1 2
1 2
1 2
1 2
1 2
1 1
2g
y y
q 1
= y y 1 y y 1
2g
y y
Substituting from momentum equation
y y y y y y 1
= y y 1
4
y y
y y
= 4y y y y
4y y
y y
= 4y y
4y y
+ +
+ +
+
( )
2 2
1 2 1 2
3
2 1
1 2
y y 2y y
y  y
E =
4y y
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2. Problem of specific force
Determine the sequent depth in a trapezoidal channel of 3 m width. The initial depth is 0.5 m.
The side slope is 1:1. Initial flow velocity is 4 m/s.
Solution:
1
1
0.5
3
A= (b+my)y
(b+my)y
b+2 m
2
+1
y
______________
=R
3
2
1
Q = A. =
= (3+0.5) 0.5*4
3.5
= *4 = 7 m /s
2
Q
M = + z
gA
( )
2 2 3
3
1 2
2 2
3 2
1 2
2 2
2
A
7 3*0.5 1*0.5
= + +
3.5
2 3
9.81*
2
= 2.86 + 1.5*0.25 + 0.041666 = 3.2767m
M =M
3*y 7 2
M = + + y
9.81* 3+y y 2 3
Solve by trial and error y 1.05m.
Alternative approach is:
z
=
1
3 3
2 2 2
2
5 5
2
M 1*3.2767
= = 0.12135
b 3
m Q 1*7
= = 2.0510
gb 9.81*3
From graph y = 1.05m.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
9.3.1 Hydraulic J ump
1. For the case of hydraulic jump in a rectangular channel, complete the following table.
y
1
(
m )
1
V (m /s) q (m3 /s) y
2
(m)
2
V (m / s)
Head
1oss ( m )
0.20 1.204
2.50 1.00
1.91 26.18 50
2. A hydraulic jump occurs in a rectangular channel and the depths of flow before and after
the jump are 0.50 m and 2.0 m respectively. Calculate the critical depth and the energy loss
in the jump.
3. Two rows of baffle piers are to be installed in a stilling basin as shown in the figure in
order to assist the formation of the hydraulic jump with in the basin. It is found that such an
arrangement of blocks has an effective drag coefficient 0.3, based on the upstream velocity
and on the combined frontal area of the blocks. If the discharge is 50 m
3
/s and the upstream
depth is 0.6 m, find the downstream depth required to form a jump,
(a) If the baffle blocks are installed and (b) if they are not (c) In each case find the head loss
in the jump.
0.5 m
0.6 m
8 m
Plan
Section
Stilling basin
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
4. A rectangular channel 6m wide carries 11.5 m
3
/s and discharges onto a 6m wide apron
with no slope with the mean velocity of 6.0 m/s. what is the height of the hydraulic jump?
What energy is absorbed (lost) in the jump?
5. A rectangular channel 5m wide carries a flow of 6 m
3
/s. The depth of water on the
downstream side of the hydraulic jump is 1.30 m. (a) What is the depth at upstream? (b)What
is the loss of head?
6. After flowing over concrete spillway of a dam, 254.7 m
3
/s then passes over a level
concrete apron (n = 0.013). The velocity of the water at the bottom of the spillway is 12.8 m /
s and width of the apron is 54.86 m. Conditions will produce a hydraulic jump, the depth in
the channel below the apron being 3.05 m. In order that the jump be contained on the apron,
(a)How long the apron should be built? (b)How much energy lost from the toe of the
spillway to the downstream side of the jump?
7. Starting from first principles, show that the following equation holds true for a hydraulic
jump in a trapezoidal channel.
2 2 2
b y my Q
g + + =Constant
2 3 ( b+my )y
Draw the forcemomentum diagram for the following conditions and determine the initial depth
if the sequent depth is 0.2 m.
Q = 50 l / s; b = 0.46 m; m = 1.
8. A flow of 2.8 m
3
/s occurs in a circular channel of 1.8 m in diameter. If the upstream depth of
flow is 0.60 m, determine the downstream depth which will cause a hydraulic jump.
9. A flow of 100 m
3
/s occurs in a trapezoidal channel with side slopes of 2:1 and a base width of
5m. If the upstream depth of flow is 1.0 m, determine the downstream depth of flow which will
cause hydraulic jump.
10. A hydraulic jump occurs downstream from a 15 m wide sluice gate. The depth is 1.5 m, and
the velocity is 20 m /s. Determine
(a) The Froude number and the Froude number corresponding to the conjugate depth, (b) the
depth and velocity after the jump, and (c) the power dissipated by the jump.
11. A 10 m wide rectangular channel is carrying a discharge of 200 m
3
/s at a flow depth of 5 m.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(i) If the channel bottom has sudden rise of 0.3 m determine the depth of flow at a downstream
cross section. Does the water surface rise or drop?
(ii) Compute the depth of flow at a downstream section if the drop is 0.2 m
12. An 8 m wide rectangular channel has a flow velocity and flow depth of 4 m/s and 4 m
respectively. The channel bottom is at El. 700 m. Assuming no losses, design a transition so that
the water level downstream of the transition is at El. 703.54 m, if
(i) The channel width remains constant, and
(ii) The channel bottom level downstream of transition is at El. 700.2 m.
13. A hydraulic jump is formed in a 4 m wide outlet just downstream of the control gate. The
flow depths just upstream and downstream of the gate are 20 m and 2 m respectively. If the
outlet discharge is 40 m
3
/s determine
(i) Flow depth at downstream side
(ii) Thrust on gate
(iii) Energy losses in the jump
Assume there is no loss in the flow through gate.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
9.4 Transition in Field Example
Problem:
Malaprabha Project transition at entry to tunnel and exit from tunnel
Tunnel: Approach cut 1280.16 m long
Slope (bed) 1:2700, y=5.4864 m
Side slope 1:1
b =6.096 m, velocity =1.794 ms
1
Horse shoe shape Tunnel length 4620.77 m
Bed Slope 1: 1230
Tunnel
Diameter 6.0198 m
Velocity 2.384 m/s
Exit b =6.096 m
Depth =4.2672 m
Velocity =1.453 m/s
13.4112
6.096 m
1
1
1
1
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3.6576
6.096 m 15.24 m
0.957
4.980 m
6.096
3.6576
0.957
x
2
y
2
6.096
2
2.49
2
______
= 1
______
+
Plan of transition  Trapezoidal to Horse shoe Tunnel
2 2
o
o
x y
1
37.1612 62001
d =6.0198 m
d
3099 m
2
.
.
+ =
=
3.6576 m
15.24 m
4.93776 m
6.0198 m
Design Discharge = 62.297 m
3
/s
n = 0.018
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0.2153
1
d0
__
2
d0
d
0
d
0
0.2153
1
4.9804
Exit Tunnel
u/s
6.0198 m
33.528 m
4.2672 m
6.096 m
4.2672 m 6.096 m
1
1
1
1
4.2672 m
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
10.1 Characteristics of critical flow
The characteristics of critical flow are
(i) The specific energy and specific force are minimum for the given discharge.
(ii) The Froude number is equal to unity.
(iii) For a given specific energy the discharge is maximum at the critical flow.
(iv) The velocity head is equal to half the hydraulic depth in a channel of small slope.
(v) The velocity of flow in a channel of small slope with uniform velocity distribution, is
equal to the celerity of small gravity waves
( )
C gh = is shallow water caused by local
disturbance.
(vi) Flow at the critical state is unstable.
Critical flow may occur at a particular section or in the entire channel, then the flow in
the channel is called "Critical flow".
( )
c
y f A, D = for a given discharge.
For a prismatic channel for a given discharge the critical depth is constant at all sections
of a channel. The bed slope which sustains a given discharge at a uniform and critical
depth is called "Critical slope S
c
". A channel slope causing slower flow in sub critical
state for a given discharge is called "sub critical slope or mild slope". A slope greater
than the critical slope is called steep slope or super critical slope.
10.1.1 Critical Flow
For a given specific energy and discharge per unit width q, there are two possible (real)
depths of flow, and that transition from one depth to the other can be accomplished
under certain situations. These two depths represented on the two different limbs of the
Ey curve separated by the crest c, are characteristic of two different kinds of flow; a
rational way to understand the nature of the difference between them is to consider first
the flow represented by the point c. Here the flow is in a critical condition, poised
between two alternative flow regimes, and indeed the word critical " is used to describe
this state of flow; it may be defined as the state at which the specific energy E is a
minimum for a given q.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
10.1.2 Analytical Properties of Critical Flow
Consider the Specific energy equation
2 2
2
v q
E = y + = y +
2g
2gy
in which y is the depth of flow and q is the discharge per unit width.
Differentiating the above equation with respect to y and equating to zero it can be
written as
2
3
2
2 3
3
c c
2
c c
dE q
= 1  0
dy
gy
q
q = gy or y =
g
and V = gy
=
The subscript c indicates critical flow conditions.
Thus the critical depth y
c
is a function of discharge per unit width alone.
Further, the above equation it can be written as
2
c
c
2
c
c c c
c c
2
2
V 1
= y .
2g 2
Thus the specific energy for critical flow can be expressed as
V 3
E = y + = y
2g 2
2
or y = E
3
d E
The second derivative should be negative i.e., =  ve
dy
The above equations are established by considering the variation of specific energy with
y for a given q. Clearly the curve will be of the general form as shown in Figure.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Variation of the Discharge with depth for a given specific energy value
Maximum point
E
0
0
y
c
=
__
3
2
E
0
q
y
q
max
How q varies with y for a given E=E
o
?
When and then 0. , 0, 0, y E q Similarly when y q
2
c 0
c 0
dq
2q = 4gyE  6gy 0
dy
6gy = 4gyE
2
i.e., y = E
3
=
Differentiating again it can be established that
2
2
d E
ve
dy
= +
Alternative approach:
Show that the flow is maximum when it is critical flow for a given specific energy plot the
graph " E
0
verses q
Solution:
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( ) ( )
=
=
= =
+ =
2 2
2
2
1
2
V Q
Specific energy = E = y + = y +
2g 2gA
when " = 1.0 "
Q = 2gA Ey A . 2g E  y
dQ
For the flow to be maximum, "
dy
dQ d
A g E  y )
dy dy
dA d
= g E  y ) A 2g Ey )
dy dy
=
0"
2 ( 0
2 ( . ( 0
2
( )
=
=
=
=>
=
=
2
2
2
2
A 2g
dA
g E  y )
dy 2 E y
A 2g
dA
g Ey )
dy 2 E y
dA
2 E y A (1)
dy
But Q = A g E  y)
Q
E  y ) =
gA
dA
Substituting in eqn: (1 ) and taking T,
dy
Q T
A
gA
=>
( 0
2 (
2 (
2(
=
= =
=
=
2
3
2 2 2
3
2
2
2
Q T
gA
Q T V T V
But
gA gD gA
V
gD
V
But F
gD
1
1
=>F
2
=1
=>F =1 i.e., Flow is critical. Flow is maximum for a given specific energy, when it is in
critical state.
Which is essentially equation representing the critical flow. Thus critical flow connotes
not only minimum specific energy for a given discharge per unit width, but also
maximum discharge per unit width for given specific energy.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Any one of the above three equations may be used to define critical flow. For example:
(1) The crests of Ey curves drawn for all values of q can be joined by a straight line
having the equation y = 2E / 3, as shown in Figure.
45
y
E
q1 SubCritical flow limb
SuperCritical flow limb
q2
q3
q3 > q2 > q1
o
(2) y
c
increases with q. The curves of higher value of q are to the right of a curve with a
low value of q.
For a given q and if the slope is small than y 0, E , an asymptote.
Similarly y =E is another asymptote. The specific energy equation can be written as
( )
2
2
q
E y y a constant
2g
= =
For a given specific energy and q there are three routes for depth  two of them are real
and one imaginary. These supercritical and sub critical depths are called alternate
depths.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
10.2 The Occurrence of Critical Flow; Controls
In addition to the type of problem in which both q and E are initially prescribed; there is
a problem which is of practical interest: Given a value of q, what factors determine the
specific energy E, and hence the depth y? Conversely, if E is given, what factors
determine q?
The answer to these questions is that there are many different kinds of control
mechanism which can dictate "what depth must be for a given q, and vice versa".
Example is the sluice gate; For a given opening of the gate there is a certain
relationship between q and the upstream depth, similarly for the downstream depth.
Weirs and spillways are further examples of this kind of mechanism. The flow resistance
due to the roughness of the channel bed will have some effect.
The flow situation in any channel is substantially influenced by the control mechanisms
operating within it. The notion of a "control"  any feature which determines a depth 
discharge relationship  is of primary importance in the study of free surface flow. There
are certain features in channel which tend to produce critical flow, and are therefore
controls (see box) of a rather special kind.
Three types of controls namely
(i) downstream control
(ii) upstream control and
(iii) Artificial control.
are identified.
Normally, the sub critical flow deals with downstream control and supercritical
flow deals with the upstream control.
The nature of these features, are determined by considering the general problem of flow
without losses in a rectangular channel section of constant width, whose bed level may
vary. This is a particular situation of the transition problem. (See box).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Transition (flow basis):
1. Sub critical to Sub critical
2. Sub critical to Super critical (Hydraulic drop)
3. Super critical to Sub critical (Hydraulic J ump)
4. Super critical to Super critical
Transition Structure:
Converging Diverging
1. Rectangular cross section to Rectangular cross section
2. Rectangular cross section to Trapezoidal cross section
3. Trapezoidal cross section to Trapezoidal cross section
4. Trapezoidal cross section to Rectangular cross section
5. Trapezoidal cross section to circular cross section or Horse shoe tunnel
6. Horse shoe tunnel to Trapezoidal cross section
7. Horse shoe tunnel to Rectangular cross section etc.
Method of connection in transition (gradual):
a. By straight wall
b. By Quadrant (cylindrical)
c. By warped
The transition could be abrupt such as sudden expansion or sudden contraction.
The transition could be gradual over certain distance.
The transition can be in vertical plane such as steps, humps, drops.
The transition could be both in plan and in elevation.
Generally, transition are provided as inlet and outlet structure.
The flow in such transitions is three dimensional and complex.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Following assumptions are made
1. Constant rectangular channel
2. Short reach.
3. No frictional loss.
4. Hydrostatic pressure distribution is assumed.
b
q
V __
2g
y
T.E
z
x
z = f(x)
Plan
Longitudinal section
y
2
__
V __
2g
2
__
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The total energy H and q
Q
= discharge per unit width
b
=
are constant,
( )
2
2
2
q
H = y+z+ = E+z = constant
2gy
differentiating with respect to x, the distace along the channel
dE dz
+ =0
dx dx
which may be rewritten as
dE dy dz
+ = 0
dy dx dx
Substitutingand simplifying
dy dz
1F + = 0
dx dx
2
2
2 2
2
2
2
3
3
dE V
( =1F ; F = )
dy gy
V
E = y +
2g
dE d Q Q dA
= 1+ 1+ 2A
dy dy 2g dy
2gA
dE Q
=1 T = 1 F
dy
gA
=
2
2
3
Q T
(i.e) F
gA
=
It is to be noted that the Froude number F plays a key role in this equation. This
equation demonstrates in nutshell from a result from the Ey curve.
If there is an upward step in the channel bed, i.e., if dz/dx is positive, then the product
( )
2
dy
1F
dx
must be negative and vice versa (see box).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
If
dz
dx
is positive
Bed
z
( )
( ) ( )
( ) ( )
2
dy
1 F negative
dx
dy
F 1 Subcritical ve depth decreases along x
dx
dy
F 1 Supercritical +ve depth increases along x
dx
=
<
>
If
dz
dx
is negative
Bed
z
( )
( ) ( )
( ) ( )
2
dy
1 F positive
dx
dy
F 1 Subcritical +ve depth increases along x
dx
dy
F 1 Supercritical ve depth decreases along x
dx
=
<
>
However, if the channel bed is horizontal i.e.,
dz
0
dx
= ,. Then the product
( )
2
dy
1F
dx
is then
equal to zero. Hence, either
dy
= 0 or F = 1 (critical flow)
dx
.
The first situation occurs in the steptransition problem when
dz
0
dx
= ,
dy
0
dx
= both
upstream of the step and over the step, and in both cases F 1 .
For the second situation, the question is " Can a situation be visualized in which
dz
0
dx
=
and
dy
0
dx
?
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The answer is yes.
Consider the Free outflow from a Lake as an example of critical flow.
y = E
c
2
3
0
dz dy
dx dx
0, F = 1 = 0,
E
o
Flow
P
An example of Critical  Free Outflow from a Lake
When water is released from a lake over a short (but smooth) crest such that it flows
downstream freely. In other words either a free overfall within a short distance
downstream or a steep slope whose bed resistance imposes no effective constraint on
the flow.
At the crest P,
dz
0
dx
= the flow is accelerating at this point, resulting in
dy
0
dx
. Then the
Froude number must be equal to unity, and hence the flow would be critical. In cases of
a sharp edged (e.g., V notch weir) crest, and a completely free overfall, are considered
as pressure distribution would be non hydrostatic; for the reason the curvature will not
be large. However, even if the vertical accelerations is large, as near brink of a free
overfall, the flow is still can be approximated as the critical condition. Experimental
evidence indicates that the flow depth right at the brink of an overfall is approximately
c c c
5
y , (i.e 0.715 y ) and that y = y
7
at a distance upstream from the overall edge of weir of
infinite height, the discharge is remarkably close to that obtained by assuming critical
flow at the crest, despite the pronounced vertical curvature of the flow. Assuming that
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
the pressure distribution is hydrostatic, it can be concluded that when water is released
from a lake without any downstream constraint critical flow occurs at the section of
maximum vertical constriction: such a section is therefore a control. Similarly that critical
flow occurs at a corresponding horizontal constriction.
Free overfall (drop) Free overfall over a sharp crested weir
Free Over fall over an arch dam
10.2.1 End Depth or Brink Depth
When the channel terminates abruptly the end weir is known as The Weir of Zero
height". The flow in the end reach of the channel becomes an overfall. Measuring the
depth at the end section of the channel, the discharge can be estimated. Rouse first
identified this aspect in a horizontal rectangular channel (with sub critical approach
flow). The end depth (also called the brink depth) was 0.715 times the critical depth.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
When the canal drops suddenly, a free overfall is formed, since flow changes to
supercritical flow can be used as a measuring device.
Free overfall profile
Minimum drop distance
H
1
y
c
y
3
2 Level 1 0 +1
 0.5
 0.6
b
v __
2g
2
__
X
__
y
c
y
__
y
c
The drop distance should be more than 0.6yc. Brink depth y
b
will be different at the
centre and sides of the canal (which is higher). The roughness of the canal affects the
brink depth and hence the bed and sides should be finished smooth.
2
o
2
q
H
2gy
y = +
Differentiating w.r.t 'y' assuming Q to be constant.
2
o
3
dH q
1
dy
gy
=
o
dH
0
dy
= if the flow is critical, hence
2
3
c
q
y
g
=
3/2
c
b c
3 2
b
If =1, then Q = b g y
Rouse showed y 0 715y
y
Thus Q = b g
0.715
/
. =
This derivation is assumed for a free fall with an unconfined nappe. This value is
modified as 0.705 when the flow is two dimensional. This results in a error of 2 to 3 %
respectively for the above two cases.
The width of the canal should not be less than 3 y
c
. This is applicable to canals with
slopes upto 0.0025.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
P
Thin weir plate  Free over fall
L
Brink depth
or
End depth
y
c
Brink depth
x
(y )
b
y
c
_
_
y
b
1.4 , x = 3 to 4 y
c
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
10.2.2 Constriction in bed width
In case of a horizontal channel bed and a variable width b, the energy equation can be
written, taking z as a constant but q as a variable function of x as
=
2
2 2 2
2
V
Total energy = TE = H = z + y + ( = 1.0 )
2g
q .b Q
= > H = z + y + z + y +
2gA
2
2
g .b
( )
( )
( )
0
0 0
= =
= =
2
2
2
2
2
2
3 2
y
q x
H = y + z +
2gy
Differentiating both sides with respect to " x ",
q x
dy dH dz d
+ +
dx dx dx dx 2gy
dH dz
If and No energy loss, Horizontal channel
dx dx
dy dy q q
 +
dx dx gy gy
( )
0 =
=
2
2
dq
= 0
dx
and by continuity equation q b = a constant, Q.
Then
dq dQ db
= 0 = b + q
dx dx dx
dq db
b q
dx dx
dq
Eliminating , between above two equations then it may be written as
dx
dy q
1F 
dx gy
( )
2 2
q db
= 0
b dx
dy y db
i.e., 1F  F = 0
dx b dx
It can be concluded that critical flow occurs when
db
dx
, i.e., at a section of maximum
horizontal constriction. The critical flow will not occur at a section of maximum width, but
only at a section of minimum width.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
( )
Converging
db dy
i 0 F<1 subcritical then 0 depth decreases as x increases
dx dx
dy
F>1 supercritical then 0 depth increases as x increases
dx
Diverging
db dy
i 0 F<1 subcritical then 0
dx dx
< <
>
> > depth increases as x increases
dy
F>1 supercritical then 0 depth increases as x increases
dx
<
F<1
db
__
dx
F>1
< 0
db
__
dx
> 0
F<1 F>1
dy
__
dx
< 0
Sub critical
Super critical
dy
__
dx
> 0
Sub critical
Super critical
Horizontal constriction
Converging channel
Diverging channel
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Derive the following equation for a non prismatic channel, assuming no energy loss.
+
=
3
c
2
c
y db
S
dx dy by
dx
y
y
3
.
1
Solution:
Total energy at any section is given by
= + + >
2
2
V
H = z + y + ( = 1.0 )
2g
Differentiating wrt "x",
dy dH dz d V
dx dx dx dx 2g
(1)
=
= = => =
+ >
f
f
2
dH
But S
dx
dz dH
Similarly S But S
dx dx
Substituting in eq: ( 1 ),
dy d V
0 =  S +
dx dx 2g
0 0
(2)
= +
2
2
dy d Q
 S +
dx dx 2gA
0
Consider a rectangular channel with varying width
= =
2 2 2
2 2 2 2 2
2
3 2 2 3
2 2
3 2 2 3
d Q d Q Q d 1
dx dx g dx 2gA 2g b y b y
dy Q db
=
g dx dx b y b y
dy Q db Q
=
dx dx gb y gb y
2
2 2
2
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
:
+ =
=> + =
=
+
=
2 2
3 2 2 3
2 2
3 2
2
3
c
3
c
2
c
Substituting this expresstion in eq: ( 2 ),
dy dy Q db Q
 S
dx dx dx gb y gb y
dy q q db
 S
dx dx gy gb y
q
But y
g
y db
S
dx dy by
dx
y
y
3
0
1 0
.
1
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
11.1 Critical depth in Trapezoidal and Circular channels
Problem:
( )
3 3
' '
2 3
c c
5 '
c
y y 1
Q m
For trapezoidal channel showthat ,
gb 2y 1
' c
c
my
where y =
b
+
=
+
Solution:
( )
2
5
f Q,b,g,m
Q
Combining f ,m
gb
V
Forcritical flow F 1.
gD
V
D
g
Fromcont
2
c
c
The most important basic problem is to determine the critical depth.
From the dimensional analysis
y
y
it can be rewritten as
b
=
=
= =
=
( )
2 2
2
2
2
2
2
Q
inuity equation V
A
Q
A D .
g
Q
Sectionfactor Z
g
Consider
b
2m
T 1 b
b
A
2m T
1
b
A D b
2m
1
c c
c
c
c
c
c
c
c
c
2 c
c
A= b+my y
my
= 1+ y
b
y
my
1+ y
b
y
my
1+ y
my b
Z 1+ y
y b
=
=
=
= +
=
+
= =
+
( ) ( )( )
( )
( ) ( )( )
( )
( ) ( )
( )
( )
( )
3
3
2
2
2
2 2
2
2
3
2
b
Defining
b
Q
g
Multiplying
b
Q
g
Q
g
c
' c
c
' ' 3
c c c
2
'
c
3
5
' ' 3 ' '
3 3
c c c c c
'
c
5 5 ' '
c c
' '
3
c c
5 '
c
my
y =
b
1+y 1+y y
Z =
1+2y
m
on both sides by we get
b
1+y 1+y y 1+y 1+y
m m
y
b b 1+2y 1+2y
1+y y
m
b 1+2y
=
= =
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Problem:
Show that for circular channel
=
2
o
Q sin cos
64 sin d gd
2
In which y is the depth of flow " d
0
" is the diameter of the circular channel.
Solution:
When flow is critical,
( ) ( )
( )
=
=
=> = >
=
=
2
2
2 2
0
2
0
V
Froude number F =
gD
Q
=> V = gD
A
Q
gD
A
d r
Area of flow A = sin
2 8
substituting it can be written as
d
A= sin2
8
d
Top width = T = 2
2
1
(1)
sin
2
2
sin( ) ( )
( ) ( )
( )
= =
= =
2
d sin  d
sin2 sin d d A
Hydraulic depth D =
T 8 sin 8 sin
sin d
D=
4 sin
Q
From eq: ( 1 ) gD =
A
sin
2 2 2 cos
cos
( )
( )
( )
( )
( )
=> =
=
2
2 4
2
4
2
4
2
5
2
2
Q
d
sin
64
gd Q
gD =
4 d
sin
16
g d Q
4 d
sin
16
Q
sin gd
Q
sin d gd
2
2
2
3
3
2 2 cos
sin cos
sin
cos
sin cos
sin
cos
sin . cos 1
64
1
sin cos
64
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
11.2 Hydraulic exponent for critical flow: M
The section factor Z for critical flow in general can be expressed as
2 M
o c
Z =C y
in which M is an exponent to be determined and
0
C is a constant proportionality.
Taking logarithm on both sides
( ) ( ) ( )
( ) ( )
o c
o c
c
2 ln Z=ln C +M ln y
Differentiatingwith respect to y
d d d
2 ln Z = ln C +M ln y
dy dy dy
d M
ln Z = 1
dy 2y
But from definition
A A
Z =
T
Taking logarithm it may be expressed as
( ) ( )
( ) ( ) ( )
( ) ( )
3/2 1/2
lnZ=ln A +ln T
d 3 d 1 d
ln Z = ln A ln T
dy 2dy 2dy
d 3 1 dA 1 dT
ln Z = 2
dy 2A dy 2T dy
Comparing equations (1) and (2) it may be written as
c
c
c
3 1 dA 1 dT
M =2y
2A dy 2T dy
dA
But =T,then
dy
T 1 dT
M =y 3 
A T dy
y A dT
M = 3T
A T dy
M is known as Hydraulically exponent for critical flow.
It may be noted that no particular channel shape has been assumed.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(a) If channels of rectangular cross section,
( )
c
c
dT
=0
dy
3y T
M =
by
M =3.0 T =b
( )
( )
( ) ( )
( )
( )
'
'
c
c
' '
c c
' c
c
c
b For trapezoidal channel obtain the following expression
1 2y
2y
M 3
1 y 1 2y
y
i nwhich y m .
b
Solution:
For trapezoidal channel
A = b+my y , T =b
for critical flow
+
=
+ +
=
( )
( )
( )
( )
c
c
c c c
c
c c c
c
c
dT
+2my , =2m.
dy
y A dT
Substituting the above in the standard expression for M = 3T
A T dy
y y y
M 3 b+2my 2m
y y y
b 1+
2my 1
M = 3b 1+ 
my b
b 1+
b
b+m
b+m b+2m
=
c
c
c
c c
c
c c
c c c c
c c
my
2my
b
2my
b 1+
b
my 2my
b 1+
2my 1 b b
M = 3b 1+ 
my 2my b
b 1+ 1+
b b
2my 2my my 2my
3 1+ 1+  1+
1 b b b b
M =
my 2my
1+ 1+
b b
( )
( )( ) ( )
( )
( ) ( )
( )( )
( )
( ) ( )
c
2
' ' '
c c c
' '
c c
'
'
c
c
' '
c c
my
If
b
3 1+2 1+2  1+ 2
1
M =
1+ 1+2
3 1 2y 2 1 y y
M
1 y 1 2y
1 2y
2y
M 3
1 y 1 2y
'
c
' ' ' '
c c c c
' '
c c
y
y y y y
y y
=
+ +
=
+ +
+
=
+ +
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
2
2
2 2 2 4 5
2 M
Showfor Triangular channel M =5.0
Solution:
y
Sectionfactor Z=A D =my
2
y y m
Z = my my y
2 2 2
Z =Cy
Comparingthese two equations
= =
( )
2 1
2 1
2
1
2
1
2 M
c
c
c
M =5
Critical flow exponent for non prismatic channel Nature channel :
log Z log Z
tan =
log y log y
Z
log
Z
tan =
y
log
y
Z =Cy
2lnZ=ln C+M lny
ln Z
M =2 =2tan
ln y
M =2tan
1
2
1 2
1
2
1
2
1 c
2 c
1 2 c c
c
1
2 c
1
2
c
c
2lnZ =lnC+M lny
2lnZ =lnC+M lny
Subtracting
2lnZ  2lnZ =M lny lny
y
Z
2ln =M ln
Z y
Z
2ln
Z
M = 2tan
y
ln
y
=
c
2
0
my
(C) It may be noted that by using Vs M, a single curve can be constructed. Then this curve could be
b
identical to the curve with m=1.0
y Q
Similarly the graph for Vs can be construc
d D gD
ted.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
11.3 Problem
Derive the value of M, N for rectangular (narrow, wide), Trapezoidal, Triangular channel
by using the following expression.
2
3 5 2
3
=
y y A dT dP
M = T and N T R
A T dy A dy
Solution:
( )
( )
( )
( )
(a) Rectangular Channel
y A dT
M 3T
A T dy
y by
= 3b 0
by b
y
= 3b
by
M 3.0
For wide rectangular channel
dp
A=by, R y, p b 0
dy
2y dp
N 5T 2R
3A dy
2y by
5b 2 0
3by b 2y
2y
= 5b
3by
=10/3=3
=
=
=
=
=
+
[ ]
( )
.33
For narrow channel
b dp
R= as b 0, p=b+2y 2
2 dy
2 y b
N 5b 2 2
3by 2
2 y
= 5b 2b 2.0
3by
For Chezy relationship
y dp
N= 3T R
A dy
y by
N= 3b 0
by b 2y
N 3.0
=
=
=
+
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
2
2
2
2
2
(b)Triangular Channel:
1) Area of triangle
A =my
2) P =2y 1+m
my
3) R =
2 1+m
4) T =2my
y A dT
M= 3T
A T dy
y my*y 2my
= 32my
2y*y 2my*y y
y
= 5my 5
my*y
2y my
N 5*2my 2 2 1 m
3A
2 1 m
=
= +
+
[ ]
2
2
2
2
2
2y my
= 10my 2 1 m
3my
1 m
2 y
= 10m 2m
3
my
2 16
=8m*
3m 3
+
+
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
( ) ( )
( )
( ) ( )
( )( )
2
2
2
3
3
3
3
3
(c ) Trapezoidal Channel
y A dT
M = T
A T dy
b+my y d b+2my y
= b+2my
b+my y b+2my dy
b+my y y
= b+2my 2m
b+my y b+2my
b+2my b+my 2my y
=
b+my y b+2my
b+2my b+my 2my
=
b+my b+2my
( )
( )
2
2
2
2
3 1 2
1 2
3 1 2
1 2
3 1 2
1 2
+
+
+
+
=
+
+
' c
c
' ' '
c c c
' '
c c
my my my
b 1+ 2 b
b b b
=
my my
1+ by b
b b
my my my
1+ 2
b b b
M =
my my
1+
b b
my
if y
b
y 1+y 2y
M =
1+y y
_________________________
( )
( )
( )
( )
( )
5 2
5 2 2
8
3 1 2
8
3 1 2
+
2
2
2
2
2
'
'
2
______________________________
2y dp
N = T R
3A dy
b+my y 2y
= * b+2my 1+m
3 b+my y
b+my 1+m
y
y
10 1+2m
1+m
b
b
N =
y y
3 1+m 1+m
b b
y
1+m
10 1+2y
b
N =
y
3 1+y
1+m
b
=
' n
my
in which y
b
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
12.1 Critical flow depth computations
One of the important aspects in Hydraulic Engineering is to compute the critical depth if
discharge is given.
Following methods are used for determining the critical depth.
(i) Algebraic method.
(ii) Graphical method.
(iii) Design chart.
(iv) Numerical method. Bi section method/ Newton Raphson method.
(v) Semi empirical approach  a method has introduced by Strarb.
12.1.1 Algebraic method
In this method the algebraic equation is formulated and then solved by trial and error.
The following example illustrates the method.
1. Consider a trapezoidal channel:
2.
( )
( )
( )
( )
( )
( )
( ) ( )
c c
c c
c
c 1
1/2
c c
1 c c
c
3
2 3
1 c c c
6 5 4 3
c c c c c
A b my y
b my y
D
b 2my
Q
Z constant C known
g
b my y
C b my y (1)
b 2my
C b 2my b my y
leadsto
y py qy ry sy t 0
inwhich theconstants p,q,r,s andt areknown.
= +
+
=
+
= = = =
+
= +
+
+ = +
+ + + + + =
Solvethis by polynomial or by trial anderror method.
It wouldbeeasier to solvetheequation (1) by trial anderror procedure.
After obtainingtheanswer check for theFroudenumber which shouldbeeq
ual to1.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Example:
Consider a Rectangular channel and obtain the critical depth for a given discharge.
Solution:
1/2
3/2
c
2/3 2
2
c
A by
Area by D y
T b
Q
Z by y
g
Q
y
b g
Q q q
y
g b g g
3
= = = =
= =
=
= = =
12.1.2 Trial and error method
For a given trapezoidal channel obtain the critical depth by trial and error method.
Solution:
( )
( )
( )
( )
( )
3/2
c c
1/2
c
3
2
3
c
3
c
3
3
c c
c
Fortrapezoidal channel
b my y
A D
b 2my
b my y Q
Squaring y constant
b 2my g
For agiven b, m,Q, select avalueof y
Assume b 6m, m 2m, Q 12m /s Solvefor y
6 2y y
1
6 4y
+
=
+
+
= =
+
= = =
+
=
+
( )
3
3
c c
c
44
14.679
9.81
3 y y
36
3.6697
3 2y 9.81
=
+
= =
+
Assume a value of y
c
and compute A D and compare with the value obtained by
Q
g
.
y
c
A D
A D
Remarks
1.2 23.708 too high
0.5 1.339 low
0.8 6.170 high
0.65 3.10
0.70 3.94
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Remarks column indicate that the values are high or low when compared to the given
value. The improvement is done till it converges.
In the above table y
c
lies between 0.65 and 0.70.
This could be improved further by selecting the values in between these two.
12.1.3 Graphical method
For natural channels and complicated channels, the graphical method is adopted. A
curve is generated assuming different values of y
c
and Z. The value of
Q
g
is computed
and y
c
is obtained from the chart. A one meter diameter culvert carries a discharge of
0.7 m
3
/s. Determine the critical depth.
y d
0
T
( )
o
1.5
0.5
0
0.5
1 sin
D d
g
sin
2
2 sin
Z d
32 sin
2
=
Knowing the value of d
0
for different values of depth A and D could be obtained from the
table.
Example:
A one meter diameter pipe carries a discharge of 0.7 m
3
/s. Determine the critical depth.
c
c c
c
Q 0.7
Z 0.2235
3.132 g
Construct agraph of y Vs andobtain thevalueof y
Fromthegraph y 0.4756
Z
=
= = =
From the design chart determine the critical depth for a circular channel of 0.9 m
diameter. Discharge 0.71 m
3
/s.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Solution:
( )
2.5
0
c
c
0
0.71
Z 0.22669
9.81
Z
0.29499
d
y
0.56, y 0.49527
d
from table
m
= =
=
= =
( )
( )
2
2
0.27
c
0.75 1.25
1.0 17
Q
29.5
g 9.81
29.5 6
y 0.81 0.86
30 2
2 6
m
= = =
= =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
12.1.4 Graphical Procedure
Straub proposed several semi empirical equations to obtain the critical depth. The
advantage of this is a quick estimation of the critical depth. However, the equations are
non homogenous.
y
c
Graph showing variation of section factor
with critical depth for a given pipe of
diameter d
o
Z=A D
y
c
d
0
__
y
c
b
__
or
D A
_____
d
0
2.5
D A
_____
b
2.5
or
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Reference:
Straub W.O, Civil Engineering, ASCE, 1978 Dec, pp 70  71 and Straub 1982.
Table: Semi empirical equations for the estimation of y
c
(Straub, 1982) MKS units
Channel type
Equation for y
c
in terms of
2
Q / g =
b
Rectanglar
1/3
2
b
b
m
1
Trapezoidal
0.27
0.75 1.25
b
0.81
30m
m b
2.5
2.5
Rangeof applicability
Q
0.1 4.0
b
Q
For 0.1
b
use equation for rectangular channel
< <
<
l
m
TRIANGULAR
0.20
2
2
m
y
y =cx
2
x
Parabolic
( )
0.25
0.84c
2
y cx =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
d
0
Circular
[ ]
[ ]
0.25
0.26
0
0.52
c
0.3
0
c
3 1
0
1.01
d
Q
y 0.053
d
y m
Q ms d m ,
=
=
= =
c
o
Rangeof applicability
y
0.02 0.85
d
y
x
a
b
Elliptical
0.25
0.22
2
0.84b
a
c
Rangeof applicability
y
0.05 0.85
2b
a =major axis
b =minor axis
y
y = cx
1
m1
____
x
Exponential
( ) 1/ 2m 1
3 2m 2
m c
4
( ) 1/ m 1
y cx
=
Example:
3
c
b 6.0m, m 2, Q 17m /s determiney = = =
Solution:
From table
0.27
0.75 1.25 2.5
2
b Q
0.81 for 0.1 4.0
30m
m b b
Q
where
g
17
The value of 0.19,
c
2.5 2.5
y
Q
b 6
= < <
=
= =
It is in the range of the equation. Substituting the appropriate values,
( )
( )
2
0.27
c
0.75 1.25
117
29.5
9.8
29.5 6
y 0.81 0.86
30 2
2 6
m
= =
= =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Problem:
Non rectangular channel involves trial and error solution.
Obtain the critical depth for the trapezoidal channel of bottom width 6 m with a side
slope of 2.5: 1, which carries a discharge of 20 m
3
/s.
6 m
m
1
m
1
y
c
Solution:
Trial and error procedure
( ) ( )
( )
[ ]
( )
( )
c c
c
c c
c
c
0.5
2 2 2
c c
c c
2
c
c
c
c
y 6 2.5y y
T 6 5y
6 2.5y y
A
D
T 6 5y
Q
Z A D
g
6 2.5y y
V Q / A 20*20
6 2.5y y
2g 2g 6 5y
6 2.5y 19.62
y
gy
c
A= b+my
b+2my=
v
?
?
= +
= +
+
= =
+
= =
+
= = = +
+
+
=
= =
Solution of Algebraic or Transcendental Equations by the Bisection Method
In the algebraic expression F(x) =0, when a range of values of x is known that contains
only one root, the bisection method is a practical way to obtain it. It is best shown by an
example.
The critical depth in a trapezoidal channel is to be determined for given flow Q and
channel dimensions.
2
3
Q T
1 0
gA
=
The formula must be satisfied by some positive depth y
c
greater than 0 (a lower bound)
and less than, an arbitrarily selected upper bound say, 10 m.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
T is the free surface width
c
b 2my + . The interval is bisected and this value of y
c
tried. If
the value is positive, then the root is less than the midpoint and the upper limit is moved
to the midpoint and the remaining half bisected, etc.
This method gives the solution very quickly.
F(x)
0
100
Bisection
T
b
m
1
m
1
Trapezoidal
y
Newton Raphson Method is discussed elsewhere.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
12.2 Problems
There are three types of problems in critical flows as shown in table.
Type m Q y
c
b or d
I
?
II
?
III
?
Types I and II are easy to solve.
Type III problem requires a different approach.
Type I problem
c
3/2
c
2
my
is known.
b
Qm
Fromthegraph y Vs Z
b gb
Qcan be determined.
=
Type II problem
Here the solution is for obtaining critical depth. There are different methods. Graphically
3/2
2
c
Qm
Z can becomputedandvalueof can beobtainedfromwhich
b gb
y can becomputed.
c
my
b
=
Type III problem
This problem can be solved using simultaneous solution of two algebraic equations
which is illustrated below.
Defining
( ) ( )
c c
1
0
3/2
1
2 2
0 o
my y
Y for trapezoidal channel or for circular channel .
b d
Qm Q
andX or
b gb d gd
=
=
Then
c
1
my
b
Y
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
3/2 5/2 3/2
1
1
2 2 1/2 1/2
2 2
c c c
c
2 1/2
1 1
5/2
5/2 1
1 1
2
c c
5/ 2
1 1 1
Qm Y Qm
X
m y m y g g my
m y
Y Y
QY
M Y
my gy
X M Y
= =
= =
=
In which
( )
( )
3
1
2
c c
c
3/ 2
c
1 1
2
3
' '
2 3
c c
5 '
c
1
2
c c
5/ 2
1 1 1
5/ 2
3/ 2
c
1
2
3/ 2 2
c c
2.5 5/ 2
1
Q
M
my gy
GivenQ, y , m,
Qm my
X , Y
b b gb
y 1 y
Q m
gb 1 2y
Q
M andis known.
my gy
X M Y
Substitutingin theaboveequation
Qm my
M
b b gb
Qm my gy
b y
Q g
Solvethe
=
= =
+
=
+
=
=
=
=
equation andobtain solution for bedwidth bfor trapezoidal channel.
Similarly solvefor diameter for thepipeline.
the
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Problem
Compute the critical depth in a trapezoidal channel for flow of 30 m
3
s
1
. The channel
bottom width is 10.0 m, side slope m =2. The bottom slope is negligible and 1 =
.
T
b
m
1
m
1
Trapezoidal
y
Solution
Given
10
=
=
3 1
c
Bottom width b m
Sideslope m =2
Flow Q = 30 m s
= 1
Critical Depth y ?
For finding the critical depth,
Cross sectional area of the channel A= ( b + 2
C C
C C
y ) * y
= ( 10 + 2 y ) * y
Section factor Z = A D
10 2 2 +
C
C C C
in which D= A / T
for trapezoidal channel the top width T = ( b+ 2m y )
D = ( 10 + 2y ) * y /( * * y )
then the section factor ( ) ( ) 10 2 2
981
+
= =
C C C
Z = A 10 + 2 y * y / * * y
30
by using the equation A D Q / g , = 9.578
.
Substituting all the parameters A, P, T, D, and Q in the above equation and sol
( )
( )
( ) ( )
( ) ( )
3
2
1
2
3 1
2 2
3
10 2
9578
10 4
10 2 9578 10 4 0
10 2 91743 10 4 0
+
=
+
+ + =
+ + =
c
c c
c
c c c
c c c
ving for y one gets
y y
.
y
y y . y
y y . y
c
by trial and error, y = 0.91 m
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Problems
1. A trapezoidal channel with side slopes of 2 horizontal to 1 vertical is to carry a flow of
16.7 m
3
/s. For a bottom width of 3.65 m, calculate (a) the critical depth and, (b ) the
critical velocity.
2. A rectangular channel carries 5.60 m
3
/s. Find the critical depth y
c
and critical velocity
V
c
for
(a) a width of 3.65m and, (b) a width of 2.75m ,
(c) What slope will produce the critical velocity in (a) if n =0.020 ?
3. Find the diacharge over a broad crested weir of 5.0m length and head 1.0m above
the crest. Assume coefficient of discharge to be 0.9.
yc
H
y1
V1
___
Broad Crested Weir
P
B
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
P is the height of weir, B is the breadth of the weir. Assume the approach velocity
1
V to
be very small.
=
3/2 c
c
3 y Q
Answer: H = , y Q = 0.544L gH
2 L
2
1
,
g
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
13.1 Measuring Flumes
Measuring flumes, on the contrary, introduce a widthwise contraction in the channel to
achieve the same objective as weirs. Sometimes a small ramp hump on the bed may
also be provided in the flume. A contracted weir of finite crest width and a measuring
flume with a hump are not essentially different. These flumes are called 'Venturi
Flumes', Sometimes these are referred to as Venturi Weirs too. For measurement of
dscharge with venturi flumes two measurementsone upstream and one at the throat
(narrowest crosssection), are required, if the flow passes in a sub critical state through
the flume. If the flumes are designed so as to pass the flow from sub critical to
supercritical state while passing through the flume, a single measurement at the throat
(which in this case becomes a critical section) is sufficient for computation of discharge.
To ensure the occurrence of critical depth at the throat, the flumes are usually designed
in such a way as to form a hydraulic jump on the downstream side of the structure.
These flumes are called 'Standing Wave Flumes'.
It should be noted that the critical depth will not occur at a particular section of the
measuring structure for all discharges. It moves upstream with increasing discharge and
downstream with increasing boundary roughness for the given discharge. In order to get
the critical depth at a predetermined section, several modifications of the venturi flume
were incorporated. The developments of Parshall Flume and cutthroat flume (Fig 1.9 d)
are the results of such studies.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
FLUME LENGTH
X X
PLAN
y
c
SECTION along XX
Note: SubCritical to SuperCritical  Single Measurement
SubCritical to SubCritical  Double Gauging
Throat
Venturi Flume with an hump in the bed
GAUGE WELL
GAUGE WELL
W
THROAT
PLAN
Z
P
D
R
C
Z
PARSHALL FLUME
E
M B F G
WATER SURFACE
K
LEVEL FLOOR
SECTION ON ZZ
R
Standard dimensions are indicated
in the figure by letters
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
SECTION ON YY
STANDING WAVE FLUME
PLAN
Y Y
y
c
Throat
Long throated flumes
Cut throat flumes
Parshall Flumes
H Flumes
Throatless flumes with rounded transition
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
13.2 Critical depth flumes
A free flowing critical depth or standing wave flume is essentially a streamlined
constriction built in an open channel where a sufficient fall is available so that critical
flow occurs in the throat of the flume. The channel constriction may be formed by side
contractions only by a bottom contraction (or hump) only, or by both side and bottom
contractions.
The use of a weir is a simple method, but it causes relatively high head loss. The
hydraulic behavior of a flume is similar to the flow over broad crested weir.
In this regard the stagedischarge relations of several critical depth flumes in general
can be expressed as
n
0
Q =C h
where 'C
0'
is a coefficient depending on the breadth (b) of the throat, on the velocity
head
2
V 2g / at the head measurement section, and on those factors which influence
the discharge coefficient; 'h' is the piezometric level over the flume crest at a specified
point in the converging approach channel and n is a factor usually varies between 1.5
and 2.5 depending on the geometry of the control section.
The empirical relationships are derived from experimental observations for a particular
structure. Hence, the dimensions of the new structure should match exactly with that of
the structure for which the equation is derived.
Example of critical depth flumes that have such headdischarge relationship are the
Venturi flume, Long throated flume, Parshall flume, cutthroat flume, and Hflume.
1. The centre line of the flume matches with that of the canal. The flumes cannot be
used in structures like turnouts, controls and regulating device etc.
2. The critical depth flumes are
(i) Long throated flumes.
(ii) Throatless flumes with rounded.
(iii) Throatless flumes with broken phase transition.
(iv) Parshall flume.
(v) H flumes
(vi) Venturi flume with sub critical constriction.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.1 WeirsIntroduction
A control structure can be defined as a change in the crosssection of the flow whereby
the regime of the flow is modified. At such a section, a definite stagedischarge
relationship exists, enabling it to be used for the purpose of flow measurement. At a
control structure like a Weir or a measuring flume, etc, the flow changes from sub
critical to supercritical state. Thus, the flow passes through a critical section and flow is
independent of the tailwater conditions. In such a case, the flow can be measured by a
single depth measurement upstream of the critical section.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.2 Types of Control Structures
Thin Plate weirs
Weirs with finite crest width
Weirs with different longitudinal profiles
Short crested weirs
Terminal weirs
Brink depth
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
WeirsTypes of Control Structures
Weirs are structures which span across channels (may or may not to the full width)
introducing a contraction in the channel section are called "Suppressed Weirs"  occupy
Full width. Whereas "The Contracted Weirs" occupy a portion of the full width. "Weirs of
Finite Crest Width extend a certain length in the direction of flow.
(a) SUPPRESSED WEIR
PLAN
PLAN
(b) CONTRACTED WEIR
h
( b) SHARPCRESTED WEIR WITH SUBMERGED FLOW
ENERGY LINE
1
h
2
h
(a) SHARPCRESTED WEIR WITH FREE FLOW
ENERGY LINE
Thin Plate Weirs
While the flow passes over the weir, if the lower nappe springs clear off from the
upstream edge and does not reattach itself to the weir crest, it is called a 'Sharp
Crested Weir'. Hence, the sharpcrested weirs are fabricated using thin metal plate and
these have a very small 'absolute length' in the flow direction (equivalent to the
thickness of the plate) followed by a suitable chamfering at the top to make the weir
crest sharp. In literature and practice these are known as 'SharpCrested Weir', 'thin
plate weir' and 'notch'. In thin plate weirs the sharpness of the edge required to be
maintained properly otherwise with the change in the sharpness of the crest will affect
the characteristics of the weir considerably. These weirs are very sensitive to approach
flow conditions and the conditions of nappe of the flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Flow over rectangular thin plate weirs under free flow conditions has been very
extensively studied. Investigations on this basic weir in nonmodular range have been
carried out since Dubuat (1816).
It is indicated that the effect of tail water is felt even for depths of flow below the crest
level. In fact, French Standards recommend that the downstream level should be
300mm, below the weir crest. The free flow over thin plate weirs of non rectangular
shapes shown in figure.
RECTANGULAR
(SUPPRESSED TYPE)
RECTANGULAR
(CONTRACTED TYPE)
TRIANGULAR
PARABOLIC
TRAPEZOIDAL
CIRCULAR
DIFFERENT SHAPES OF SHARPCRESTED WEIR
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Weirs of Finite Crest Width
These Weirs possess a finite 'absolute length' (crest width) in the direction of flow.
When flow over a weir of finite crest width occurs unlike in sharp crested weirs the lower
nappe will be adhering to the weir crest. These are subdivided as NarrowCrested
Weirs, BroadCrested Weirs and LongCrested weirs depending on the flow profiles
over these weirs.
h
L
NARROWCRESTED WEIR
(0.4 <h/L <1.5)
h
L
BROADCRESTED WEIR
(0.1 <h/L <0.4)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
h
L
LONGCRESTED WEIR
h/L <0.1
Weirs with Finite Crest width
Among the Weirs of finite crest width, the rectangular broadcrested weir with horizontal
crest is the most common. Modified weirs such as, weir with slightly sloping crest and
weir with rounded entrance are also in vogue.
h
Finite Crest Width Weir
with Sloping Crest
HORIZONTAL
h
with Rounded Entry
R
Finite Crest Width Weir
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
h
Drowned Flow Condition
RECTANGULAR TYPE FINITE CREST WIDTH WEIR
1
h
2
Finite Crest Width Weir
The other shapes of broadcrested weirs used are triangular, parabolic, trapezoidal and
circular types.
h
P
L
Triangular
Parabolic
Trapezoidal Circular
Weirs with Finite Crest Width with
different crosssectional shapes
Longitudinal Section
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Weirs with Different Longitudinal Profiles
Further, broadcrested weirs with the longitudinal profile varying in the direction of flow
are triangular profile (crump) weirs, embankment shape (trapezoidal profile) weirs,
Semicircular profile weirs and streamlined profile weirs.
h
h
Crump Weir (Triangular Profile Embankment Weir
(Trapezoidal Profile)
L
1:3
0.64 m
0.53 m
1:3
1:1
1 __
2
1:1
1 __
2
0.38 m
Embankment weir
0.38 m
h
h
R
Semi Circular Profile
Streamlined Profile
1:5 Slope
Weirs with Different Longitudinal Profiles
Hydrofoil No. 2
L =443 mm, W =101.6 mm
Hydrofoil No. 4
a =152 mm, m =19.05 mm,
L =537.7 mm, W =43.81 mm
= 3.175 mm
Typical Hydrofoil weir shapes
a =152 mm, m =12.15 mm, = 9.525 mm
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
References:
J agannadha Rao M.V.,  Flow Measurement in Open Channels with Hydrofoil weirs and
with end depth measurement, PhD Thesis, Department of Civil and Hydraulic
Engineering, Indian Institute of Science, Bangalore, 1971.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Short crested weirsWeir sill with rectangular control section
Are the structures in which the curvature of the streamlines influence the head
discharge relationship.
1.25 b
b
Fully aerated nappe
Concrete appron (2.0 m)
h1
Short crested weirs
Approach channel with lining
05
1.5
d v 1
1.5
d v 1
2 2
Q = C C g bh
3 3
Q = 1.7049 C C bh
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Terminal Weirs
Weirs located at the end of the channel are known as 'Terminal Weirs' or 'End Weirs'.
The flow characteristics are affected by the nappe conditions i.e., whether the weir is
confined or not by extending the (side walls) and whether the area below the bottom
nappe is properly ventilated or otherwise. When the nappe spreads freely, a marginal
increase in discharge may be observed. Further, when a weir is located at or very close
to the end of a channel, the usual discharge relationship of weir flow within a certain
range of headweir height ratios found to be applicable only. This ratio ranges between
5 (Boss) to 10 (kandaswami and Rouse).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.2 End Depth or Brink Depth Method
When the channel terminates abruptly the end weir is known as "The Wier of Zero
height". The flow in the end reach of the channel becomes an overfall. Measuring the
depth at the end section of the channel, the discharge can be estimated. Rouse first
identified this aspect in a horizontal rectangular channel (with subcritical approach flow).
The end depth (also called the brink depth) was 0.715 times the critical depth.
When the canal drops suddenly, a free overfall is formed, since flow changes to
supercritical flow can be used as a measuring device.
Free Overfall Profile
Minimum Drop Distance
H
1
y
c
y
3
2 Level 1 0 +1
 0.5
 0.6
b
v __
2g
2
__
X
__
y
c
y
__
y
c
The drop distance should be more than 0.6y
c
. Brink depth will be different at the centre
and sides of the canal (which is higher). The roughness of the canal affects the brink
depth and hence the bed and sides should be finished smooth.
2
o
2
q
H
2gy
y = +
Differentiating w.r.t 'y' assuming Q to be constant.
2
o
3
dH q
1
dy
gy
=
o
dH
0
dy
= If the flow is critical, hence
2
3
c
q
y
g
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3/2
c
b c
3 2
b
If =1, then Q = b g y
Rouse showed y 0 715y
y
Thus Q = b g
0.715
/
. =
This derivation is assumed for a free fall with an unconfined nappe. This value is
modified as 0.705 when the flow is two dimensional. This results in an error of 2 to 3 %
respectively for the above two cases.
The width of the canal should not be less than 3 y
c
. This is applicable to canals with
slopes upto 0.0025.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.3 Proportional Weirs
'Proportional Weirs' are weir shapes designed to achieve a particular headdischarge or
headvelocity relationship. These weirs find application in the fields of Hydraulic
Engineering, Sanitary Engineering and Chemical Engineering. The study of Proportional
weirs started with the development of 'Sutroweir', which is a linear proportional notch.
For complete literature on this subject, the works of Kolupaila and Keshavamurthy may
be consulted. A general method of designing a weir notch having a base in any given
shape to a depth a, such that the discharge through it is proportional to any singular
monotonicallyincreasing function of the depth of flow measured above a datum was
proposed by Keshavamurthy and Seshagiri. Some typical examples of proportional
weirs are shown here discussed elsewhere.
X AXIS
O
2
3
__
s
s
Y AXIS
DATUM
W
X AXIS
Y=f(x)
h
s
W
O
Y AXIS
LINEAR PROPORTIONAL WEIR (SUTRO WEIR)
LOGARITHMIC WEIR
X AXIS
O
3
__
s
s
Y AXIS
DATUM
W
1
X AXIS
Y AXIS
d
s
W
QUADRATIC WEIR
(ORIFICE NOTCH)
LINEAR PROPORTIONAL
(ORIFICE NOTCH)
Some examples of Proportional Weirs
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
30 cm
15 cm
0
T = 25.4 mm
W =50.8 mm
X
Y
T = 50.88 mm
W' =76.2 mm
0 150 mm 150 mm
Profile of a typical baseless weir (NBW1) Profile of a typical baseless weir (NBW2)
15 cm
0
X
Y
Proportional
Portion
X
Y
Rectangular
orifice
Trapezoidal chamber
Bed of chamber
Rectangular
orifice
Profile of a typical Sutroparabolic weir Profile of a typical linear proportional orifice notch
p =215.9 mm
2b =1219.2 mm
a =101.6 mm
a =76.2 mm
p =228.6 mm
d =127 mm
Proportional
portion
1
1
30 cm
2b =1219.2 mm
d =228.6 mm
0
References:
1. Keshava Murthy K, "A Generalized Mathematical Theory of Proportional Weirs, PhD
Thesis, Department of Civil and Hydraulic Engineering, Indian Institute of Science,
Bangalore, 1968.
2. Keshava Murthy K, and Seshagiri N, "A Generalized Mathematical Theory and
experimental verification of Proportional notches", J ournal of the Franklin Institute,
Volume 285, Number 5, May 1968, Page 347  363.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.4 Flow Over Weirs
Flow over weirs is complicated because of sharp curvilinear streamlines besides
consisting of eddy regions, accelerating and retarding flow zones. The existing solution
are semiempirical in nature.
Discharge Coefficient:
For assessing the discharge passing over a weir, many semiempirical formulae are in
use. Among them, the formulae developed by Rehbock, Bazin, Von Mises, and
Weisback and Francis are popular.
Factors Affecting Flow over Weirs
The several factors affecting the flow are
The head
Fluid properties and Temperature Effects
Approach and tail water conditions
Weir Geometry
Measurement inaccuracies
Fluid Properties and Temperature Effects
The fluid properties which influence the discharge over the weir are viscosity and
surface tension. In case of water flowing over the weir these effects are negligible at
heads higher than 3 cm.
Temperature variations influence the fluid properties like viscosity and surface tension.
The variations in these fluid properties in turn will influence the discharge over the weir.
This type of problem is of importance to chemical Engineers and Sanitary Engineers.
Another closely associated problem is the influence of temperature variations on the
aeration at weirs. In many weirs in industrial processes substantial aeration takes place
when water falls over the weirs. The rate of absorption of atmospheric oxygen by the
water flowing over the weir increases with increasing temperature.
However, the effect of small temperature variations on the water flow over weirs is
negligible.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Approach Flow Conditions
The distribution of velocities in the approach flow has a definite influence on the
discharge over the weir. Kinetic energy correction factor can account for the variation in
the approach velocity. The value of this coefficient depends on the degree of non
uniformity of the approach velocity distribution.
A weir not normal to the approach flow is called a 'skew Weir'. In skew weirs there is a
discharge concentration towards one side. The discharge was found to be greater than
that over a normal weir.
Skew Weir
Approach flow
Plan
Tail Water Conditions
At high tail water levels, the flow over the weir passes in a sub critical state. In this case,
the discharge is dependent on both the upstream and the downstream water levels. The
Weir, in this case, is said to be submerged and the flow is nonmodular. "The ratio of the
downstream and the upstream water depths above the weir crest is defined as the
submergence ratio, . The limiting value of where the tail water also begins to
influence the rate of flow is called the submergence limit. Beyond submergence limit,
the discharge reduces.
The shape of the nappe may affect the discharge. The modification of the nappe
conditions result in small variations of the order of 1 to 2% in the discharge.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
CLINGING NAPPE
DEPRESSED NAPPE NAPPE WETTED
VENTILATED
UNDERNEATH
(FREE) NAPPE
DIFFERENT NAPPE SHAPES
Ventilated
Adhering to the crest and
downstream face of the weir
The ventilated nappe springs
clear of the crest
The coefficient of discharge is generally
above the free discharge coefficient upto
H/L < 0.8
The weir geometry influences the coefficient of discharge. It depends upon the pressure
distribution along the geometric profile, boundary layer growth and separation zones.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.5 Polygonal weirsIntroduction
Weirs and spillways with a polygonal discontinuous center line can be designed in
various manner. Figure 1 shows some of the examples such as square intake towers,
labyrinth weirs, duckbill overfalls.
97.5 m
30.0 m
Stilling basin
Duckbill overfall
corner angle: 90
148.0 m
14.4 m
Labyrinth weir
corner angle: 45.6
314.4
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
11.7 m
20.4 m
Labyrinth weir
corner angle: 117
242
Polygonal weir
corner angle: 133 152 255
, ,
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
5.0 m
5.0 m
to
stilling
basin
Square intake
corner angle: 90
Rectangular intake tower
corner angle: 90
8.0
3.0
15.0 m
3.0 m
Rectangular spillway
corner angle: 90
14.9 m
3.5 m
Polygonal intake tower
corner angle: 84.3
Layouts of Overfall structures with
Polygonal Center Line of Weir Crest
These weirs consist mainly of straight parts with corners inbetween. The points of
discontinuity are created by the intersection of two straight center lines. Closed
polygons are possible.
The length of an overfall structure can be considerably increased in case the width is
limited. In case of small overfall heads, the discharge capacity may increase compared
to straight overfalls situated orthogonally to the main flow direction. Intake towers in
reservoirs with small water depth
[ ]
300 m . and small floods
3
100 m/s
can be
designed as square shaft spillways instead of the continuous straight or circular crests
in plan, which are used very often.
The polygon is easier to construct than the circle. However, the hydraulic computation
of the discharge capacity for the polygon is more complicated than for the continuous
straight or circular crests in plan.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
It is possible, for any combination or shape of a polygonal weir or overfall, to do the
hydraulic computation with very accurate results with the help of the analysis given by
Indlekofer and Rouve (1975).
14.5.1 DISCHARGE OF " CORNER WEIR
They investigated the "corner" weir, which is symmetrical and has orthogonal boundary
conditions. The corner angle, , is formed by both the straight sides of the weir and is
measured in the downstream.
side wall
W
c
h
c
l /2
d
l /2
c
"corner"weir
Plan of Cornerweir
Disturbed area
(overlapping flow region)
l
c
/2
l
c
/2
l
d
/2
l
d
/2
T
y
b
Rectanglar
2W
h
Y
POEBING WEIR
X
R
Angle ' ' varies within the limits, Convex angle 0 180 <
Concave angle 180 360 <
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In the physical system ("corner weir ") in the range of 180 to 360
does not perform
satisfactorily. Therefore
1 2
360 + =
in physical system.
/ 2
/ 2
1
1
2
Layout of Corner Weir
Concave angle
0 < 180
o
o
1
<
Convex angle
360 <
o
2 < 180
o
The flow over the corner weir can be apportioned as (1) disturbed area near the corner
are shown in green color and
(2) With twodimensional flow.
The length of the area of disturbed flow = 2 + 2 l / l /
d d
The length of the corner weir is l
c
. Hence, 0
d c
l l
The local disturbance factor 'DF' with a distance, l, from the corner is defined as
( )
( ) C l
DF l
C
n
=
in which ( ) C l and
n
C are the coefficient of discharges for the corner weir of length 'l'
and for the normal flow condition.
The discharge over the weir is written as
3 2
3
2 2
c
m
* /
n c c,n
Q
C
C l g h
=
in which Q is the discharge, in m
3
s
1
; b is the width of the weir, in meter; 'g' is the
acceleration due to gravity 9.81 ms
2
; and 'h' is the head over the weir. The disturbance
factor cannot be greater than 1.
Because of the continuity of flow between the corner and the side walls, it may be noted
that the continuity of DF (L). At the point of transition, the following condition is required
to be satisfied.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
DF 1
2
d
l
=
For values 2 DF l l / , l
d
= . Accordingly, the mean distribution coefficient,
m
C , of the
overlapping flow zone can be written as
( )
2
2
DF
0
l /
d
c l dl
m
l
d
=
The discharge,
c
Q , of the corner " weir is
( )
3 2
2
2 1
3
/
c n c,n c m d
Q C g h l C l =
The overall head,
c,n
h , belonging to the discharge coefficient,
n
C , under normal flow
conditions (twodimensional flow) can be estimated from
2 2
2
c c,n
c,n c
h h
g
= +
in which
c
h =overfall head, assuming three dimensional flow, in meter, at the " corner "
weir ;
c
=flow velocity, assuming threedimensional flow, in meter per second, at the "
corner " weir ; and
c,n
=flow velocity under normal flow conditions (twodimensional
flow) at the " corner " weir. For the hydraulic calculation the length,
d
l , of the disturbed
area and the value of
m
C must be known.
14.5.2 LENGTH
d
l OF OVERLAPPING ZONE
With increasing overfall heads
c,n
h the length of the overlapping zone,
d
l , grows
symmetrically to the corner, as far as
d c
l l = .
In this case the mean disturbance coefficient is
3 2
3
2 2
c
m
* /
n c c,n
Q
C
C l g h
=
If the corresponding limiting value for
*
c,n c,n
h h = .
Using the length,
d
l , of the overlapping zone of flow, depending on the strength of
disturbance, from Eq. 7 one may obtain
3 2
3 1
1
2 2
c
d c
/
m
n c,n
Q
l l
C
C g h
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3 2
3
2 2
c
d
c
/
n c,n
Q
l l
C g h
=
Thus
1
1
d
d
m
l l
C
=
in which l
d
represents the length of disturbance. The independent variables
and l ,Q ,C h ,
c c n, c,n
are determined from experiments.
Indlekofer and Rouve have conducted investigations for sharpcrested corner weirs
with corner angles 4681 6208 8964 and 123.45 . , . , . , =
. The crest thickness was 2mm.
The discharge was determined by the Rehbock formula
( )
2 00011
3 2
06035 00813 2 00011
3 P
h .
/
Q . . b g h .
+
= + +
in which h is the overfall head, in meters; P is the weir height, in meters; and b is the
width, in meters and
n
C is the coefficient of discharge.
Length of Overlapping Zone Area
d
l
The length of overlapping zone area,
d
l , can be calculated using Eq. 12.
The length of disturbance,
d
l , is related to the overall head,
c,n
h , by a simple linear
function,
d
c,n
l A Bh = +
in which A is a constant, in meter; and B slope for
d
l . It must be mentioned that the
constant, A is very small, and either positive or negative.
14.5.3 Length
d
l of Overlapping Zone
Based on the laws of similitude, one can assume a linear relation between the length,
d
l , of the zone of disturbance and the overfall head. The length,
d
l , will be
1
d c,n
m
B
l h
C
=
or using the slope B, for the length,
d
l , of the overlapping zone
d c,n
l Bh =
Because of the linearity B
c
*
c,n
l
h
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Mean Disturbance Coefficient
m
C in overlapping zone  The mean disturbance
coefficient,
m
C , which considers the influence of the disturbance with a length,
d
l ,
compared with the flow normal to a straight weir, can be calculated by,
1
m
B
C
B
=
using the slope and B B .
Figures show the typical Polygonal Plan with angles 0 180 < <
.
3
(2,3)
2
(1,2)
1
(n,1)
n
n1
(n1,n)
n1
n
Plan of the typical polygonal corner weir , Corner Angles 0 < < 180 ;
lc
lc
lc
lc
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Length of Overlapping zone for Constant Overfall Head
Overlapping zones
Undisturbed zones
( )
3/2
c, n
1
2
2 h
3
n
d ,i
c n c n
i
Q C g l l
=
=
Example:
Discharge of SharpCrested Shaft Spillways with Equilateral Polygonal Plan
In case of shaft spillways, with equilateral polygonal in plan above Equation can be
simplified as
3 2
2
2 1
3
d /
c n c c,n
c
l
Q C g n l h
l
=
in which l
c
is the length of the crest between two corner points and n is the number of
corners.
Reference:
Indlekofer, Horst, and Rouve, Gerhard, "Discharge over Polygonal Weirs," J ournal of
the Hydraulics Division, ASCE, Volume 101, Number HY#, Proceeding paper 11178,
March 1975, pp. 385  401.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.6 Special Types of Weirs
Special types of weirs are designed to cater to the needs of a particular situation, where
the usual types of weirs fail. Flatvee weirs, Large Vee Weirs and Labyrinth weirs are
some examples of such special types.
Flat Vee weirs
Triangular profile flat Vee weir
Large Vee weirs
Vnotch weir sill
Triangular profile two dimensional weir
Triangular broadcrested weir
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Flow Measuring Structures
Flow measurement structures are required in irrigation canals in order to facilitate the
distribution of water through out the system and to keep account for seepage losses,
etc. However, in the smaller channels the flow measurement structures or devices are
closely associated with local water management practices of an irrigation command.
Several individuals have carried out investigation on flow measurement structures and
have developed discharge rating relationship for them, but it must be noted that both
national and international organisations are responsibility of Technical Committed TC
113 "Measurement of liquid flow in open channels. A list of standards relating to flow
measurement structures is given in Table 1.
ISO Standards
ISO 1438/1 Thin plate weirs and flumes
ISO 4360
ISO 4361
Triangular profile weirs.
Round nosed weirs
ISO 4359
Standing wave flumes for different throat section viz,
Rectangular, Ushape, Trapezoidal.
ISO 6417 Compound gauging structures.
ISO 3846
ISO 3847
Rectangular broad crested weirs.
By the brink depth method.
ISO4374 Round nose horizontal crest wiers.
ISO 4377 Flat V weirs.
ISO 748
Liquid flow measurement in open channel by velocity area
method.
ISO 1070
Liquid flow measurement in open channels by the slope
area method.
ISO 1100
Established and operation of a gauging station and
determination of the stage discharge relations.
ISO 2425 Measurement of flow in tidal channels
* International Organisation for Standardization
* On the suggestion of India, in 1954 the technical committee ISO/TC 30 of the ISO took
upon the task of standardization of flow measurements in open channels set up a
separate subcommittee.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Labyrinth Weir as HighCapacity Field Outlet
Irrigation engineers have been forced to adopt new solutions to the engineering
problems in order to ensure that irrigation projects are viable, or to deal with specially
adverse conditions. Updating of existing spate irrigation systems which have been
operating along traditional lines with reasonable success for substantial periods of time
required a novel approach to build the structures involved. The important feature of
these spate irrigation systems is that they are fed from non perennial rivers (normally
dry but occurring flash flood flows when storms occur in the upper catchments).
Diversion from these steep rivers was traditionally made by temporary earth banks
which are often washed away during the flood period.
Figure shows a plan view of a section of canal which includes a high capacity field
outlet having a cross  regulator immediately down stream. Closure of the cross 
regulator cause backwater in the canal, and the water levels and the extent of the
backwater is determined by the head required over the weir crest to discharge the
necessary flow of water. A short length of weir crest would result in high backwater
levels and a long length of crest would result in a relatively small increase of water
levels due to backwater. The importance of the increase of water levels and of the
extent of backwater requires raising of the canal banks, involving substantial
investment. The labyrinth weir is one such solution. This should be cheaper than a
straight weir having the same length as the developed length of a labyrinth weir. In
Figure a simple two  cycle labyrinth weir has been shown near outlet. The configuration
of labyrinth weir is determined by experiment. With some configurations, there is a
possibility of, the nappes meeting from two of the sloping sides of the labyrinths forming
a jet which may cause scour in the downstream.
A Labyrinth weir is characterized by a broken axis in plan, the total length thus being
compressed in concertina (Small musical instrument resembling an accordion but
having button like keys) form into the space available on site. The purpose of the
Labyrinth weir is to increase the discharge per unit width for a given operation head.
Another advantage of this weir can be raised for the same maximum elevation of water
level, thereby gaining substantial storage capacity.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
5.5 m
Note:
This can develop submeged turbulent rollers at very low discharges
Simple weir
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
5.5 m
Simple weir with splash plate
Note:
Distributes the flow over a greater surface area
1.2 m
Weir with cascade
1
2
3
Steps 1 and 2 =1.83 m
Steps 3 =1.52 m
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Note: When a simple weir is used energy dissipation would not be very effective.
Depending on the Tail water level, a hydraulic jump forms. In order to dissipate the
energy possible alternatives are shown above. A simple weir with splash plate will help
in distributing the flow over a great surface and the baffle blocks will assist in break up
of submerged jump. The turbulence level is not reduced by this combination. Further
cascade of splash plate will estimate the submerged jump.
Reference:
Don Richarad and others, Low Head Dam Safety with Hydraulic Models, Proceedings
National Hydraulic Engineering Conference, ASCE, 1987, pp528533.
5.5 m
Typical Labyrinth weir
Note:
Longer length, lower head, reduced over action
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
PLAN
UPSTREAM
CHANNEL
DOWNSTREAM
CHANNEL
X
UPSTREAM SURFACE
PROFILE
DOWNSTREAM SURFACE
PROFILE
SECTION  'XX'
LABYRINTH WEIR
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1:2
Warped masonry
wall
Curved masonry
wall
0.25
Highcapacity field outlet upstream of a cross regulator.
Plan view of V high capacity field outlet.
Design outflow 5 m /s
Scale 1:100
1:3
1:1
1.10 0.90
Curved masonry
wall
1:1
4.81
0.2
3.04
0.25
1.3
Dry rubble
pitching
Crossregulator
Stop logs or
gates
8.0
3
With the low ratios of head to crest length, the effectiveness of the labyrinth weir
configuration can be measured by a weir equation, such as
=
3 / 2
d
Q C g Lh
in which, C
d
is the coefficient of discharge, L the developed length of the weir crest and
h the head over the crest. High values of C
d
indicate an efficient structure, where as the
head h inevitable reduces as the length L increases.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
FlatVee Weirs
FlatVee weirs are suitable for measuring a wide range of flows accurately. It is
relatively easy to install them and their cost is low. This weir was designed by slightly
modifying the Crump Weir. Figure shows the typical flatvee weir. An accuracy of plus or
minus 0.5 percent was claimed for this weir, both in modular and non modular ranges of
flow. For small installations this weir may be prefabricated. Further these weirs can be
used on steep slopes. But in such cases, to avoid the generation of supercritical flow
state over the weir, a stilling pool in the upstream of the gauging station may be
provided. This pool helps in trapping the sediment. FlatVee Weirs are common in
Europe.
Flat Vee Weir
Flow
P
h
P
hP
GAUGE WELLS
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Triangular profile flat Vweir
Triangular Profile Flat V  Weir
View from upstream
( )
( )
25
2.5
d v e e b
b
25
2.5
d v e e b
b
4 b
Q = C C 2g h h h
15 h
b
Q = 1.18118 C C h h h
h
.
.
e
h is the effective less than
b
h . Effective head less than
1
h measured at the U/S. The
correction factor depends on the cross slope at weir profile (Table gives).
Table: Correction factor
f
C for Triangular profile flat V weirs
Weir profile Cross slope
U/S d/S 1:20 1:10
1:2 1:2 0.4 mm 0.6 mm
1:2 1:5 0.5 mm 0.8 mm
e 1 f
h h  C =
d
C 066 . = can be taken for both the profiles indicated in table. However is sensitive to
Tail water bed level, P
2
.
Then average value of
d
C 071 . = can be taken when
e
2
h
125
P
. < .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Large Vee Weirs
To estimate the regime characteristics of a river in relation to watershed protection and
flood prevention measures. A vee Weir with a very large apex angle and with very little
crest width is installed with weir crest slightly above the channel bed. Figure shows the
details of a Large vee weir.
Sheet Piles
Weir Crest
A
A
Weir Crest
Stilling Basin
Rip Rap Packing
Section along AA
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
760 mm
2
1 1
1
1 = 150 mm
2 = 120 mm
Loose Sand
Weir Crest Details
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Vnotch Weir Sill
2
m
1
x
x
0.15 0.1
0.15
0.05 0.05
m = 2, 3, 5
Dimensions are
in metre
V notch Weir sill
0 5
2.5
d v 1
2.5
d v 1
16 2
Q = C C g tan h
25 5
= 0.8965 C C tan h
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Triangular profile Two Dimensional Weir
This is also referred as CRUMP weir (1952)
Flat Vee Weir
Flow
P
h
P
hP
GAUGE WELLS
[ ]
1 2
1.5
d v e
1.5
d v e
2
Q = C C 2g bh
3
=2.9529 C C bh
/
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Triangular Broad Crested Weir
b
[ ]
1 b
0.5
2.50
d v 1
2.50
d v 1
2
b
0 50
b
d v 1
For H 1 25 h
16
Q = C C 0.2g tan h
25
= 0.8965 C C tan h
g = 9.81 ms
For H 1.25 h
h 2 2
Q = C C g h 
3 3 2
.
.
[ ]
1 50
1 50
d v 1 b
b
= 1.7049 b C C h 0.5h
.
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.7 The broad crested weir
A broad crested weir is an overflow structure with a horizontal crest above which non
hydrostatic pressure distribution occurs and may be neglected. In other words, the
streamlines are merely straight and parallel. To obtain this condition the length of weir
crest (L) should be related to the total energy head over the weir crest as
1
0.05 H L 0.08 . The upper limit as
1
H L 0.08 is fixed otherwise the energy
losses over the weir crest cannot be neglected and undulations may occur on the crest;
On the other hand the lower limit
1
H L 0.05 , is fixed such that hydrostatic pressure
distribution may be assumed.
Such a measuring structure will have insignificant energy losses in the zone of
acceleration upstream of the control section, accordingly (specific energy E) equation
may be written as E
1
=E
2
2 2
1 2
1 1 2 2 2
V V
h + =E = y +
2g 2g
In other words
( ) { }
0.5
0.50
1 2
V= 2g E y
In which E
1
equals the upstream specific energy over the weir crest.
y2
h2 yc
P
h1
y1
E1
L
weir
Flow over a broad crested weir under submerged condition
E2
1
V
1
2
______
2g
2
V
2
2
______
2g
( ) { }
2
1 2
substituting Q = VA and putting 1.0 gives
0.5
Q = A 2g E y
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
If the critical flow occurs at the control section (y =y
c
), a head discharge relationship for
various throat geometries can may be derived from
( ) { }
0.50
Q =A 2 E
c
1
g y
c
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.8 Different Types of Broad Crested Weirs
Triangular broad crested weir
Broad crested weir with rectangular cross section
Broad crested and short crested weir
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Triangular Broad Crested Weir
b
[ ]
1 b
0.5
2.50
d v 1
2.50
d v 1
2
b
0 50
b
d v 1
For H 1 25 h
16
Q = C C 0.2g tan h
25
= 0.8965 C C tan h
g = 9.81 ms
For H 1.25 h
h 2 2
Q = C C g h 
3 3 2
.
.
[ ]
1 50
1 50
d v 1 b
b
= 1.7049 b C C h 0.5h
.
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Broad crested rectangular weir
P
h1
y1
L
1
P2
Seperation of flow
Broad crested rectangular weir  The height of the weir
P1 and P2 are different in the upstream and downstream
face respectively
1
1
1
H
08 Subcritical flow occurs above the crest
L
H
0.08 033 This is the range the flow can be described a broad crested weir
L
H
0.33< about 1.
L
<
.
.
1
5 to 1.8 Seperation occurs
H
15 Behaves as a sharp crested weir
L
> .
05
1.50
d v 1
1.50
d v 1
1 1
d
1
Disharge is obtained by
2 2
Q =C C g bh
3 3
=1.7049 b C C h
H h
C remains nearly constant if 0.08 <0.33 and 035
L h P
.
.
+
The average values of C
d
within this limit is 0.848. Beyond this range a correction factor
greater than 1.0 requires to be applied.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Weirs
Weirs are the flow measuring devices. Two types of weirs (i) Sharp crested, (ii) Broad
crested are used for measuring the flow.
Sharp crested weirs: Rectangular notch
1.5
Q h , V notch
2.5
Q h , Cipoletti
weir
1.5
Q h , Proportional weir
1.0
Q h , Rehbock weir
1.5
Q h .
Broad crested weirs: Round nose horizontal broad crested weir
1.5
Q h , triangular
broad crested weir
1.7 to 2.5
Q h , Broad crested rectangular profile weir
1.5
Q h ,
Faiyum weir
1.6
Q h , Romijn movable / regulating weir. The broad crested weirs are
those structures over which the streamlines are parallel to each other over the crest
such that the hydrostatic pressure distribution can be assumed along the length of the
weir (L).
(i) Round nose broad crested weir (fig)
0.50
1.50
d v 1
1.50
d v 1
2 2
Q c c g bh
3 3
Q 1.7049 c c bh
g is 9.81 m/s/s.
=
=
( )
1.50
d
1
Lr
Lr
c 12 1
b h
=
in which is the factor which allows for the influence of the boundary layer on the crest.
=0.005 for concrete structure and =0.003 for laboratory broad crested weir.
h
1
h
2
P
2
L >1.75H1max
r =0.2 H
1max
P
1
2 to 3 H
1 max
Stilling well
Round nose broad crested weir
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Romijn weir was developed by the department of irrigation in Indonesia by Romijn.
Telescopic Romijn weir consists of two sliding blades and movable weir which are
mounted on a steel guide frame.
The Broad Crested Weir
A broad crested weir is an overflow structure with a horizontal crest above which non
hydrostatic pressure distribution occurs and may be neglected. In other words, the
streamlines are merely straight and parallel. To obtain this condition the length of
because weir crest (L) should be related to the total energy head over the weir crest as
1
0.05 H L 0.08 . The upper limit
1
H L 0.08 as is fixed otherwise the energy
losses over the weir crest cannot be neglected and undulations may occur on the crest;
On the otherhand the lower limit
1
H L 0.05 , is fixed such that hydrostatic pressure
distribution may be assumed.
Such a measuring structure will have insignificant energy losses in the zone of
acceleration upstream of the control section, accordingly (specific energy E) equation
may be written as
E
1
=E
2
In other words
( ) { }
2 2
1 1
1 1
1 2 2
0.5
0.50
1 2
V V
h + = E = y +
2g 2g
V= 2g E y
In which E
1
equals the upstream specific energy over the weir crest (fig)
y2
h2 yc
P
h1
y1
E1
L
Flow over a broad crested weir under submerged condition.
E2
2
V
__
2
2
_____
2g
1
V
__
1
2
_____
2g
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
P
h1
y1
L
1
P2
Seperation of flow
Broad crested rectangular weir  The height of the weir
P1 and P2 are different in the upstream and downstream
face respectively
( ) { }
2
1 2
substituting Q = VA and putting 1.0 gives
0.5
Q = A 2g E y
=
If the critical flow occurs at the control section (y=y
c
), a head discharge relationship for
various throat geometries can may be derived from
( ) { }
0.50
Q =A 2 E
c
1
g y
c
Broad  Crested Weir with Rectangular Control Section
For a rectangular control section in which the flow is critical, the area of flow A
c
=by
c
and
A
c
/ T then equation may be written as
2
c
c 1
V 1
= y
2g 2
Hence
2
y = E
3
Substitution of this relationship and A
c
=b y
c
after simplification it may be written as
1.7049
1 1
0.50
1.50 1.50
2 2
Q = g b E b E
3 3
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
14.9 Beartrap weir
Hydroelectric developments of rivers require weirs that hold the water back behind a
damming structure. Such weirs, while serving utilitarian purposes, must be so designed
as to preserve the natural amenities of the locality. The beartrap weir, developed by
Voith according to a swiss patent, meets the requirements, and even under drifting ice
conditions the Voith bear trap weir provides dependable services. As the river discharge
varies, the bear trap crest is automatically raised or lowered so as to maintain a
constant head water level.
Lowered position of beartrap weir for flood disposal
Raised position of beartrap weir for flood disposal
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.1 Flow below a Sluice Gate
Sluice
Gate
1
Hydraulic
jump
2
3
Rapidly varied flow with Hydraulic jump (1 and 3 subcritical flows,
2 Super critical flow)
y
1
H
1
=
y
l
Flow below a Sluice gate
y
2
V
1
2 ___
___
2g
W
W
( )
( )
2
1
1 1
1
d v 1
1.5
d v
1.5
v
H = y +
2g
opening w
y y
depth at vena contracta y = w or = , n =
w w
Q = C C bw 2g y y
= C C bw 2g n
= K bw 2g
The value of
1
y
n =
w
, range is 1.50 to 5.00.
ranges between 0.648 to 0.624.
C
d
ranges between 0.607 to 0.596.
As 'n' increases from 1.5, C
d
decreases upto 2.40 with a value of 0.600 to 0.596. Then
further increase in n (>2.40) the C
d
value increases from 0.596 to 0.624. For the same
range, 'K' increases from 0.614 to 1.279.
For
n =2 =0.630
n =3 =0.625
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
n =10 =0.620
Vena contra is located at a distance of
1
y
1
n
= .
The sequent depth of jump should not exceed the value given by
2 1
y H
1 16 1 1
w 2 w
= +
Figure shows the limiting tail water level for modular flow below a sluice gate.
0 1
1.5 2
3
4
5 6 7 8 9 10
1
2
3
4
5
Submerged
Flow
Limiting tailwater level for modular flow below a sluice gate
Ratio n = y
1
/w
assuming
Modular Flow
= 0.611
y
1
H
1
__
Henderson proposed an equation for the contraction coefficient for the radial (Tainter)
gate which depends on inclination angle .
2
=1 0.75 0 36
90 90
.
+
The expected error is less than 5% provided that
o
< 90 . Thus the discharge coefficient
for radial gate is given by
d
0 5
1
C
w
1+
y
.
=
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
BRINK DEPTH
When the canal drops suddenly, a free over fall is formed, since flow changes to
supercritical flow can be used as a measuring device.
L
Brink depth
or
End depth
y
c
Brink depth
x
(y )
b
y
c
_
_
y
b
1.4 , x = 3 to 4 y
c
The drop distance should be more than 0.6y
c
. Brink depth will be different at the centre
and sides of the canal (which is higher). The roughness of the canal affects the brink
depth and hence the bed and sides should be finished smooth.
2
q
H
o
2
2gy
y = +
Differentiating w.r.t 'y' assuming Q to be constant.
2
dH q
o
1
3
dy
gy
=
1
y
n =
w
if the flow is critical, hence
2
q
3
y
c
g
=
If =1, then
b
Q = b g y
Rouse showed
b
y 0 715y . =
Thus
b
y
Q = b g
0.715
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
This derivation is assumed for a free fall with an unconfined nappe. This value is
modified as 0.705 when the flow is two dimensional. This results in a error of 2 to 3 %
respectively for the above two cases.
The width of the canal should not be less than 3 y
c
. This is applicable to canals with
slopes upto 0.0025.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.3 Modern Methods of Flow Measurements
Any Engineering or natural phenomenon which deserves and attracts the attention of
Engineer needs to be studied before it can be analysed. The word study is meant to
mean 'observations and analysis'. These observations require the measurement of
phenomenon and this is where measurement techniques come into picture.
It is specially so in hydraulics considering what LEONARDO DA VINCI said about
hydraulics.
"If you have anything to do with the water, first do the experiment and then ponder
about the results".
An experiment means detail recording and measurement of phenomenon. The degree
of sophistication depending upon complexity of phenomenon and its importance. For
example measurement of discharge in a channel a gauge may be sufficient, whereas
measurement of turbulence requires hot wire anemometry and other accessory
instrumentation.
The definition of data processing is the conversion of raw data into information.
Information is such a value from which decision can be made and results inferred. Data
processing can be performed manually with the aid of simple tools as paper, pencil and
fitting cabinets or electronically with the aid of computer.
15.3.1 INSTRUMENTATION OF DATA PROCESSING
For determination of discharge one should know the relationship between stage and
discharge. A rating curve is drawn for a particular section, it is nothing but the functional
relation between stage and discharge.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
STAGE (m)
DISCHARGE m
3
/s
H
Typical Stage Discharge Relationship
Measuring stage:
Water surface
Bed
Definition of stage
H (M.S.L)
(Above Mean Sea Level)
Datum
EL 210.00 m
EL 205.00 m
EL 200.00 m
The height of a stream water surface above an arbitrary datum is called "the stage".
Stage records are also used for designing of hydraulic structures, in flood warning
systems and in planning of the use of flood plains.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Stage can be measured by any one of the methods given below:
(1) A marble column carrying a scale and grounding in a well connected with the river.
(2) Stage can be sensed by a float in a stilling well that is connected to the stream by
intake pipes.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(3) Stage can be sensed with a gas purge system known as bubble gauge. The gas is
fed through a tube and bubbled freely from an orifice mounted in the stream. The
pressure in the tube, measured with a zerodisplacement mercury monometer, is equal
to the piezometric head on the bubble orifice. It has an accuracy of about 2 mm.
Stage is recorded directly on a strip chart or may be punched on a paper tape to be fed
directly to a computer.
15.3.2 Stream flow measurement
Generally three methods are used for making stream flow measurements.
1. Current meter.
2. Dilution techniques.
3. Indirect methods.
(1) Current meter: Different types of current meter are available for measuring velocity
at points in a stream. The price current meter, consists of vertical axis rotor with six
curve shaped cups (vanes) pointing in a horizontal plane. The OH meter, widely used
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
in Germany, is a horizontal axis meter that measures the components of velocity
parallel to the meter axis. The current meter is calibrated by noting down revolutions per
minute (rpm) for different known velocity and calibration curve is plotted velocity is
measured at 0.2 to 0.8 of the flow depth.
(2) Dilution technique: Two dilution techniques are (1) the steady feed method and (2)
the instantaneous, point  source time indigenous method.
For steady feed method, a solution of tracer material with concentration C
1
is injected at
the constant injection rate Q
T
C
C
2
TIME
CONCENTRATIONTIME CURVE MEASURED AT X2
SKETCH OF THE REACH
X2
X2
C2
X1
X1
INJECTION
QT
C1 Concentration
The tracer disperses laterally into the flow and tracer concentration distribution is similar
to as shown in figure. At some point X
2
downstream, where the tracer material is
approximately uniformly mixed, the flow is sampled continuously.
By continuity ( )
T 1 T 2
Q C = Q+Q C
in which
Q is the stream discharge, C
2
is the concentration at X
2
If
T
Q << Q and if there is no tracer loss
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
T 1 2
T 1
2
Q C QC
then
Q C
Q=
C
=
If the tracer mixer has properties similar to the water (fluorescein, fluorescent, salt
solution), so that there are no density gradients, vertical mixing is very rapid due to
turbulence of the flow. Theoretically, complete lateral mixing occurs at X but practically it
occurs between 20 to 100 times the channel widths.
By instantaneous injection method, a quantity of tracer w, is injected, instantaneously at
section X and time t
0
. The cloud of tracer disperses laterally and longitudinally as it
moves downstream.
W
X0, t0
Concentration at x1
Concentration at x2
TIME
CONCENTRATION Distribution at X1 and X2
X1, t1
X2, t2
DYE CLOUD DISPERSING DOWN STREAM
Q
At the section X
2
, where the tracer is completely mixed literally, the flow is sampled
continuously. From the conservation of mass
0
W= Q C dt
in which Q is nearly constant through sampling period
0
W
Q =
C dt
The common tracers used are
(a) Salt solutions
(b) Radio active tracer are detected by its scintillation detectors
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(c) Fluorescent dyes with flourometers.
Advantage of dilution method, they condensed in closed conduits, such as penstocks,
sewers pipe lines, where currentmeter measurements are difficult, and they are fast
and accurate.
Disadvantages: Expensive for measuring large stream and the special equipments
required for the measurements of concentration.
(3) Indirect Method: Involved using various empirical formulae when it is impossible to
measure discharge such as during floods. Empirical formulae like Flaming, Manning,
Strickler formulae etc are used.
Determination of sediment concentration:
The distribution of sediment concentration is not uniform over the cross section. It varies
with particle size and with depth. It is found that 0.062 mm is distributed almost
uniformly.
0
10
20
30
40
WATER SURFACE
SEDIMENT CONCENTRATION
15.3.3 MODERN MEASURING TECHNIQUES
The two principle techniques that are used in the modern measuring instruments are
'Electroacoustics' (ultrasonic frequencies) and
'Electrooptics'.
Characteristics of sound transmission:
Sound transmission in an elastic medium has the characteristic of wave motion and its
speed is dependent upon the elasticity and density of the medium. For water, these
properties are in turn affected by the temperature, pressure, and salinity. An empirical
formula for the velocity of sound in water is given by
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
C =141000 + 421 T  3.7 T + 110 S + 0.018 y
in which C is the velocity of transmission on cm/s, T is the temperature of the water in
C, S is the Salinity in pairs per thousand, y is the depth below the surface in cm.
Refraction: occurs around the solid obstacles placed in the path of transmission, or by
temperature or density stratified layers.
Reflection: Any body immersed in water everywhere to reflect sound in. In particular the
bottom and the surface of the water can reflect sound in. In particular the bottom and
the surface of the water can reflect acoustic waves. If a body is perfect reflector, all the
energy intercepted is reflected. If however, the body is an imperfect reflector part of
interrupted energy is absorbed by the body as heat and only part of the energy is
reflected.
The Ultrasonic method:
Principle: Of the ultrasonic method is to measure the velocity of flow at certain depth in
the channel by simultaneously transmitting sound pulses through the water from
transducers located in the banks on either side of the river. The transducers, which are
designed to both transmit and receive sound pulses, are not located directly but are
staggered so that angle between the pulse path and the direction of flow is between 30
to 60 . The difference between the time of travel of the pulses in two different directions
is directly related to the average velocity of the water at the depth of the transducer.
This velocity can then be related to the average velocity of flow of the whole cross
section and, if desirable, by incorporating an area factor in the electronic processor, the
system can give an output of discharge.
FLOW
B
A
rp
v
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Notation:
L Path length path angle
(usually 30 to
60)
V
Average velocity of flow of river
at depth y.
p
r path velocity at
depth y
V Average velocity of flow of river C Velocity of sound
in water
d actual depth of flow a area of flow
1
t time taken for a pulse to travel
from A to B.
2
t time taken for a
pulse to travel
from B to A.
1
F output frequency corresponding
to
1
1 / t
2
F output frequency
corresponding to
2
1 / t
c
F output frequency corresponding
to
1
1 / t minus
2
1 / t
M Multiplication fig.
of variable
frequency
oscillator
T measuring period N no. of
coincidences in
dif.fre.store =FcT
The time taken for a pulse to travel from A to B
1
P
L
t =
C + V
Similarly time taken for a wave front to travel in the opposite direction is
2
P
p
1 2
p
1 2
L
t =
C V
2V
1 1
t t L
L 1 1
V =
2 t t
Average velocity of river flow at depth y is given by
1 2
Vp
V =
cos
L 1 1
V =
2cos t t
Transducers may be so positioned in the vertical plane so as to make average velocity
V at depth of equal to the average velocity of flow V then
1 2
L 1 1
V =
2cos t t
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
If 'Q' is discharge
Q = aV
= V dsin
2
1 2
L 1 1
Q = d tan
2 t t
There are two methods of obtaining discharge in use at present, the first where the
transducers are fixed in position and station calibrated by current meter and second
case where transducers are designed to slide on either a vertical plane or an inclined
assembly. In this case no current meter measurements are necessary, self calibrating
one. By nothing transducers through number of paths in the vertical, velocity readings
are obtained along these paths. From each set of readings vertical velocity curves are
established over as large a range in stage possible. It is then possible to estimate first, a
suitable position for the fixing of the transducers into vertical and second to establish a
curve stage the coefficient of discharge as in first method.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Projector
Receiver
Projector
Detector
Recorder Computer
Receiver
A
Clock
Receiver
B
Pulser
L
Schematic illustration for the Arrangement of
Sound Projector and Receiver
V
VP
Flow
2. ELECTROACOUSTIC MEASUREMENT OF FLOW DEPTH
A single transducer is used as the projector and receiver of sound energy for
measurement of flow depth, using either the channel bed or water surface to reflect the
projected sound wave back toward the receiver along the same acoustic axis. This
instrument generally referred to as a depth sounder, the velocity of the water does not
affect the signal as the direction of interrogation is generally perpendicular to the flow
direction downward to the bed or upward to the water surface. Hence, the distance from
transducer to the reflecting surface can be determined directly from the time lapse
between projection and acceptor and the velocity of sound in to the water. The
schematic arrangement shown in fig.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Schematic arrangement for a Depth Sounder
CLOCK
PULSER RECORDER
COMPUTER
RECEIVER
TRANSDUCER
ACCOUSTIC AXIS
TARGET AREA
d
The clock provides the time base with which the other components are synchronise.
The pulses provide regulated bursts of voltage to generate short bursts of sound energy
with the transducer at a selected frequency. The choice of frequency depends on the
power available and depth to be interrogated. The sound energy is directed to a
relatively small target area on the channel bed (or water surface). The sound energy is
reflected back toward the transducer, and receiver monitors. The return echo. Since the
same transducer is used to generate the sound wave and receive the return signal, the
frequency of interrogation (repetition) depends entirely upon the depth and velocity of
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
the sound waves. The computer simply determines the time lapse between the clock
pulse and return echo, or signal and converts the information to voltage which can
subsequently be interpreted in time of depth.
Factors affecting the reliability of an ultrasonic depth so under.
The most notable among them is the suspended sediment concentration of the stream.
15.3.4 ELECTROOPTICAL INSTRUMENTS
The Principle on which these instruments work is scattering of light by particles in the
medium is the basic principle. Instruments are available which can measure 'in situ'
suspended particle concentration in a dynamic flow fluid and multi dimensional
component measurements of fluid velocity and turbulence. The first of these
instruments uses a widefrequency band, visible light source which are the second
utilizes a narrow band coherent laser beam.
ELECTROOPTICAL MEASUREMENTS OF PARTICLE CONCENTRATION:
A wide frequency band electro optical instrument for measuring point concentration of a
particulate matter in a flow field has been developed. The principle of operation is based
on the forward scattering of light by particle.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
To
Photomultiplier
Tube
Focal
Volume
1 cm. Gap
Mirror
Dark
Zone
Opaque
Coating
From
Light
Source
Lens 2
Lens 1
Mirror
Optic Probe details
The Photo Multiplier tube in this arrangement operates in essentially a dark field which
minimizes the ambient noise and enables detection of small concentration of particular
matter in the flow field, hence of concentration in the flow field.
15.3.5 ELECTROOPTICAL MEASUREMENT OF FLUID VELOCITY
The principle of an electro optical instrument for measuring fluid velocity is based on the
Doppler effect of making particular matter in a coherent light beam and determination of
frequency shifts by an optical heterodyning technique. Since only a beam of light enters
the flow field, there is no measurable disturbance as there is with other instruments
point measurements of velocity are possible as the light beam may be focused to as
small as a few microns. There is no need for prior calibration of the instrument as there
is with standard velocity measuring instruments and the response is linear over the
entire velocity range of interest. Velocities as small as a friction of a centimeter/sec. can
be measured. Electrooptical velocity instrument is also called, Laser Velocimeter, Laser
Doppler Velocimeter and laser anemometer.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.3.6 NEW METHODS OF RIVER GAUGING
The measurement of river flow is required for river management purposes including
water resources planning, pollution prevention and flood control.
The following new methods of river gauging are designed.
(a) The moving boat method
(b) The ultrasonic method
(c) The electromagnetic method
15.3.7 THE ELECTROMAGNETIC METHOD
Faraday (1832) was the first person to notice that when the motion of water flowing in a
river cuts the vertical components of earth's magnetic field an EMF is induced. In the
water, which can be picked up by two electrodes. The EMF, which is directly
proportional to the average velocity in the river, is induced along each transverse
filament of water as the water cuts the lines of earth's vertical magnetic field. This
method was used in 195354 to measure the tidal flow through Dover strait. The result
of these experiments and others are both illuminating and encouraging, and the
application of this technique for gauging the flow in river was considered. However, the
relatively small unidirectional potentials induced in small rivers cannot be detected
during the presence of interfacing potentials.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Basic principle of fluid flow
measured in pipes by electromagnetic
induction
OUTPUT E
FIELD N
v
Supply
Velcoity of
flow = V
Measuring System
Velocity
of Water
Electrical
Potential
Generated
Induced Magnetic Field
Principle of electromagenetic
river gauging
Probes
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Diagrammatic view of an electromagnetic river gauging station
Noise cancellation
probes
Cell for producing
magnetic field
Shelter for
Instrumentation
Bed
Conductivity
probe
Signal cable duct
Noise cancellation probes
Bed Conductivity probe
Signal
probes
Typical block diagram of an electromagnetic river gauging station
Noise cancellation probes
Buried
Coil
Coil
Drive
Timing
Signals
Flow of water
Timing
Signals
Stage
Water
Conductivity
Bed
Conductivity
Telemetry
system
Noise cancellation probes
Signal
Recovery
Signal Probes
Data Processor
Data Recording
Devices
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
THEORY: The basic principle of the electromagnetic method of river gauging is the
Faraday generator effect where an electrical conductor in motion in a magnetic field
induces an electrical potentials. In the case of river, the conductor is the flowing water
and electrical potential induced is proportional to the average velocity of flow. Faraday's
law of electromagnetic induction relates the length of the conductor, moving in a
magnetic field, to the EMF generated by the equation
E =H v b
in which E is the EMF generated in volts; H is the Magnetic field in tests; v is the
average velocity of the river in m/s; b is the river width in meter.
An electromagnetic gauging station consists of the following
(1) The coils,
(2) The probes,
(3) The coil derive unit,
(4) The signal measuring unit,
(5) The stage sensor,
(6) The water conductivity sensors,
(7) The bed conductivity sensor,
(8) The data processor and
(9) The display unit.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The probes: Eight probes made in high grade stainless steel rod or slips is used. These
consist of two signal probes placed in the magnetic field generated by the coil and
located in the banks on opposite sides of the river. These probes are used to detect the
induced potentials and to define precisely the C/s of measuring the section. Weeds and
bed sediment do not cause interference since their velocity is zero they generate zero
potentials. Thus they are considered being stationary water. The stage sensor: It is
capable of providing a digital signal to the data processor is employed to define the
measurement of C/s. The water conductivity sensor: A conventional conductivity sensor
is located into the river.
The bed conductivity: In the form of bed resistance is measured.
Information relating to the stage and discharge is recorded on punched paper tape at 16
m interval and may also be displayed visually along with time.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.4 Outlets and Modules
The success of any irrigation enterprise depends on the efficiency of distributing
sufficient supply of water to the irrigator. Each irrigator has to receive certain quantity of
water proportionate to his extent in a canal system at the proper time to ensure him a
good crop. This distribution of water is carried out by means of outlets otherwise called
modules. Hence, proper design of an outlet, is of most importance not only to the canal
engineer but to the irrigator also.
In Punjab and Maharashtra, a number of outlet structures were evolved, designed to
allow into the cultivator's watercourses a constant discharge irrespective of the supply
(level) in the distributing channel (module) or discharges proportional to the supply
(level) in the channel (semimodule). A few of the structures in common use in India are
(i) Standing wave flume.
(ii) Crump's adjustable proportional Semi module.
(iii) Lindley type standing wave flume.
(iv) Orifice type standing wave pipe outlet.
(v) Gibb's module.
There are various types of modules:
a. Rigid Modules
These modules allow constant discharge within reasonable working limits of head
irrespective of water levels in the distributory and the water course of the main canal.
b. Flexible Modules or Semi Modules
This type of module gives discharge in some characteristic manner with surface level in
the supply channel but independent of the variation of the water level in the delivery
channel.
c. NonModular Outlets
Nonmodular outlets are those whose discharge is a function of the difference in levels
between the water surface in the distributing channel and the water course.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.4.1 Standing Wave Flume
The standing wave flume is a semimodule measuring discharge with a high degree of
accuracy (viz., 1.5 percent) besides having the advantage that a single gauge reading
upstream is all that required. In the standard standing wave flume evolved at Poona 
the head required can vary from 8 to 15 percent of the upstream depth of water over the
sill without affecting the discharge; the modular ratio (i.e., the ratio of the downstream
water depth to the total upstream depth, measured above the sill level) can be as high
as 85 percent in small flumes and 92 percent in large flumes. It can be best used when
variable discharge needs to be measured accurately and also when facilities for
supervision or for automatic recording for gauges are available. This flume was evolved
by Crump (Punjab) and Inglis (Bombay) after carrying out intensive model
investigations.
Standard standing wave flume design
Longitudinal Section
Plan
B
1
B
3
L
1
L
2
Glacis
B2
R
1
=H
1.5
2.5H
1.5
60
0
60
0
y
h
v
H H
1
R
hump
L
1 L
2
2:1
R = 2H
y +25%
H
L
3
y 1
3
3
y
3
Gauge chamber
h = hump height
The flume comprises of
(1) An approach channel of suitable design,
(2) A bell mouth entrance,
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(3) A throat with a horizontal bottom and vertical sides,
(4) A downstream glacis, and
(5) An expansion in which the flow is redistributed before it passes into the downstream
channel and head is recovered.
It is essentially a broadcrested weir and its discharge is given by formula
15
.
o
Q C CBH =
in which, B is the width of the throat, H is the total head (depth of water upstream +
head due to velocity of approach ) on the upstream side sill level, and C is a
coefficient to allow for losses due to friction, eddies, impact shock, etc.
1
y
v
h
Values of C and adjusted values of the constant for properly designed flumes without
piers are given in Table.
Discharge in m
3
/s C C
0
0.06  0.28 0.97 3.00
0.30  1.40 0.98 3.03
1.40  14.0 0.99 3.06
over 14.0 1.00 3.09
More abrupt curves than in the standard design will slightly lower the coefficient. The
coefficient C (=0.99) for discharges from (1.4 to 14 m
3
s
1
) was confirmed by actual
observations carried out on the prototype in Sind.
With piers, loss of energy due to shock which lowers the value of C. In Sind, falls and
fall regulators were designed using the values shown in Table above, but observations
showed that C was much lower, the average value of C for discharges 110 to 280 m
3
s
1
on the Rohri Canal being about 6 percent lower. Based on the experimental
investigations carried out at the Central Water and Power Research Station, Poona, in
1933, the following formula is suggested with the piers:
( )
15
.
Q C B k n H H =
in which, 'k' is the coefficient of contraction due to piers, (0.82 with standard piers), n is
the number of piers, B is the waterway, C =3.088, and H =total head (including velocity
head).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.4.2 Crump's Adjustable Proportional SemiModule
This semimodule can be either of the orifice type or of the open type and fixed at the
head of the outlet. These have been used extensively in Punjab.
F.S.L
Bed R.L
y
b
Brick pitching
Top of bank
Roof of block
h
H
0.76
FSL in water course
0.14
19.80
22.86
Crump's Adjustable Proportional Semimodule
H
Bed of WC
Longitudinal Section
15.4.3 Lindley Type Standing Wave Flume
This is a short throated flume with one side straight and the other curved. This is
normally used as an outlet for water courses taking off at right angles from the
distributary.
GIBB Module:
The main disadvantage of a nonmodular outlet is that cultivators can draw more water
by tampering in large numbers on a canal system.
Gibb module was found to be the only module which has no moving parts. As against
modules whose working depends on floats or other moving mechanisms there are a few
devices in which the discharge is automatically regulated by the velocity of the water
itself without the necessity of any moving parts. Gibb an Executive Engineer of the
irrigation Department, Punjab devised a module form of outlet, which was built for the
first time on the Melay distributory of the Lower Thelam Canal. This module is named as
Gibb module after its inventor and it gives an almost constant discharge over a
considerable range, irrespective of the upstream and downstream water levels. It is one
of the rigid modules without any moving parts. It does not need any supervision and
cannot be easily tampered with.
Water is led through an inlet pipe (See Figure) into a spiral rectangular trough (eddy
chamber) in which free vortex flow is developed. The water on the outside of the curve
rises in level and the water surface slopes towards the inner wall. A number of baffles
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
are inserted in the eddy chamber with their lower edges sloping at the required height
above the bottom. As the head increases, the water banks up at the outer
circumference of the eddy chamber and impinges against the baffles imparting an
upward rotational direction of flow to the water, which spins round in the compartment
between two successive baffles and finally drops on the oncoming stream of water,
thus, dissipating excess energy and keeping the discharge constant. The degree of turn
of the spiral depends on the volume of discharge and the working range required and
generally varies from one semicircle to one and a half complete circles.
Though this module gives constant discharge, it has the following disadvantages.
1. This module could be easily tampered with by breaking the baffles and eddy
chamber.
2. It is costlier than other types of outlets.
3. Construction of this module is a very difficult process and needs higher technical
skills.
4. It is said to have a lot of trouble regarding silt drawal. The vent is likely to be choked
by the silt and floating materials coming in the channel and periodical cleaning may be
difficult.
Under the circumstances stated above this module can be used in places where small
drawals are required for small plots from main channels. For e.g. in channel having 0.5
m
3
/s flow a plot of 40 hectares will be requiring 0.03 m
3
/s and the depth of flow in the
main channel will less than 0.4 m. Under such circumstances this will ensure minimum
losses due to the small branches taking off from main canal.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Baffles
Bellmouth entrance
Inlet pipe
Dia 30.48
Spout
1 in 10
1 in 10
d = 30.48 cm
0.85 m
0.34 m 0.41 m
Curved rising pipe
Outlet channel bed
Longitudinal Section
0.85 m
12.7 cm slab
Gibb module
Plan
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Side view of Gibb's Module Eddychamber Gibbs Module (Eddy chamber in action)
Gibb's Module (Side View)
Gibb's Module (Downstream View)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
15.5 Errors in Measurements
Errors in measurements are now referred as Uncertainties in measurements.
There are three types of errors which must be considered
Spurious errors (human mistakes and instrument malfunctions)
Random errors (experimental and reading errors)
Systematic errors (which may be either constant or variable).
Spurious
error
Random
Systematic errors
True value
Mean
measured
value
Random uncertainity
assessed with specific
confidence
t
1
t
2
Duration of
measurement of
single value
time
Spurious errors are errors which invalidate a measurement. They are like outliers. They
cannot be incorporated into a statistical analysis.
Random errors are error that affects the reproducibility of measurement. The mean
random error of a summarized discharge over a period is expected to decrease when
the number of discharge measurements during the period increases. Mean random
error approaches zero over a long period of measurement.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Systematic errors are errors which cannot be reduced by increasing the number of
measurements. Whenever there is an evidence of a systematic error of a known sign,
the mean error should be added or subtracted from the measured results.
15.5.1 Sources of errors
Consider discharge equation
n
d v 1
Q = W C C f g b h in which W and n are constants.
The different errors are
The error in product of
d v
C C .
The error in submerged flow reduction factor.
Error in width or angle (dimensional measurement).
Error in measurement of
1
h or h .
It may be noted that the product is
d v
C C also a function of
1
h . However the influence of
1
h on
d
C and
v
C is small and hence can be neglected.
The error in measurement of
1
h (or h ) can be divided into random part and systematic
part of the error.
Possible sources of these contributory errors are
Internal friction of the recording system.
Inertia of the indication mechanism.
Instrument error.
Setting of the structure causing changes in dimensions and asymmetry, change
in levels.
The crest not being level (zero setting) in view of the poor construction. There
may be other errors caused due to construction.
Reading and recording errors.
The overall error in the flow Q is the resultant of various contributory errors which
themselves may be composite errors. The propagation of errors depends on the
standard deviation . Error analysis has to be carried out. Often there is attendency to
over look a underestimate this.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Reference:
1. Bureau of Indian Standards 14,371, 1966.
2. Boss M.G. (Ed) Discharge Measurement Structure, Oxford and IBH Publishing
Company, NewDelhi, India, 1975.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
INTERNATIONAL STANDARDS FOR FLOW MEASUREMENT
IN OPEN CHANNELS
ISO: 555/ I 1973 Liquid flow measurement in open channels 
dilution methods, Part I Constant rate injection
method.
ISO: 555/ II 1974
Liquid flow measurement in open channels 
dilution methods, Part II Sudden injection method.
ISO: 748 1973
Liquid flow measurement in open channels by
velocityarea methods.
ISO: 772 1973 Vocabulary and symbols.
ISO: 1100 1973
Liquid flow measurement in open channels 
establishment and operation of a gauging station
and determination of the stagedischarge relation.
ISO: 1070 1973
Liquid flow measurement in open channels  slope
 area method.
ISO: 1088 1973
Collection of data for determination of errors in
measurement by velocity area methods.
ISO: 2425 1974 Measurement of flow in tidal channels
ISO: 2537 1974
Liquid flow measurement in open channels  cup 
type and propeller  type current meters.
ISO: 1438 1975 Thin plate weirs and flumes.
ISO: 1438 1979 Thin plate weirs. (Revision of ISO: 1438, 1975).
ISO: 3454 1975
Liquid flow measurements in open channels 
sounding and suspension equipment.
ISO: 3455 1976
Liquid flow measurement in open channels 
calibration of current meters in straight open tanks.
ISO: 4363 1977
Methods of measurement of suspended sediment
in open channels.
ISO: 4364 1977 Bed material sampling.
ISO: 3716 1977
Functional requirements and characteristics of
suspended sediment load samplers.
ISO: 3846 1977
Liquid flow measurement in open channels by
weirs and flumes  rectangular broad crested weirs.
ISO: 3847 1977
Liquid flow measurements in open channels by
weirs and flumes  end depth method.
ISO: 4359 1978
Liquid flow measurement in open channels 
flumes.
ISO: 4360 1978
Liquid flow measurement in open channels by
weirs and flumes  triangular profile weirs.
ISO: 4361 1978
Liquid flow measurement in open channels by
weirs and flumes  round nosed broad crested
weirs.
ISO: 4373 1978 Water level measuring devices.
ISO: 4369 1978 The moving boat method
ISO: 5168 1978
Calculation of the uncertainty of a measurement of
flow rate.
ISO: 4377 1978 Flat V weirs.
ISO: 4375 1978 Cableway system.
ISO: 6418 1978 Ultrasonic (acoustic) velocity meters.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
ISO: 4366 1978 Echo Sounders.
Reports
ISO: Data 1978
Investigation of the total error in measurement of
velocityarea methods.
Note: In 1973 all existing recommendations become standards.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
16.1 Concept of Uniform Flow
1. The depth average flow velocity (integrated over depth), area of flow crosssections
are every where constant along the channel.
2. The energy grade line S
f
, water surface slope S
w
and channel bed slope S
0
are all
parallel, i.e.
f w o
S S S = =
Figure shows Boundary layer growth in open channel with an ideal entry condition.
Development of uniform flow in a long channel
yc
When the flow enters into a channel, the boundary layer grows up to free surface. The
region for a mild channel can be divided into three zones viz., initial transitory zone in
the entrance. Flow changes from the uniform flow to critical flow in the transitory zone at
exit in mild channel. The boundary layer as it grows along the channel at the entrance
emerges to the free surface at a certain distance from the entry point. This zone is
called entry transition zone.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
If the bed slope is critical slope, then the transitory zone in the entrance only exists. The
uniform flow extends till the flow terminates and exits as a jet at critical depth. This flow
is known as critical uniform flow. The free surface will be undulating with waves moving
at C= gy.
In the case of steep channel, the flow enters either through a hydraulic drop or at
uniform flow depth. This has an initial transitory zone with an S2 type of varied flow
curve. The flow emerges from the steep slope at uniform flow depth (y
n
>y
c
).
Uniform flow
Steady
Unsteady
Turbulent
Laminar
Prismatic Channel Non prismatic Channel
Likely uniform flow situations
Ultra rapid or hyper rapid flow occurs when flow surface becomes instable and air is
insufflated into it.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
16.2 Derivation of Uniform Flow Equations
The mean velocity of a turbulent uniform open channel flow is obtained using the
following concept.
Gravitational force =Shear force
The uniform flow equations are in the following format
x y
V = CR S in which
x and y are components, and vary depending on uniform formula.
0 0
1
V = v dx dy
A
y b
Momentum Equation:
(1)
(2)
y
1
v
1
W sin
y
2
v
2
0
_
_
l
P
A
Datum
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
2 2 1 1 1 2 f
2 1 1 2 1 2
f
f
Q
V V P P W sin  P
g
If V V P =P then
W sin = P (1)
P shear force acting on boundary = Shear stress * Area
, ,
= +
= =
=
o
o
o
= * Area
= PL
P is the wetted perimeter, Sin = S
Weight W = g AL
W Sin = g AL Sin
Substituti
( )
o o
o
o o
o
o
2
o f
2
o f
ng in equation 1
g AL S PL
g AL S
RS (2)
PL
=
= =
= =
=
1 2
o
f
o
f
f
o
1 2 1 2
2
V RS
c
2g
V RS
c
2g
If C then
c
V C RS
This is known as Chezy equation. The coefficient C is either estimated or
determined experimentally. C has dimension of L T
/
/ /
.
=
=
=
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
f
o
2
f f
o o f
2
f
2. Consider Darcy Weisbach equation for loss in pipe due to friction
L V
h f
d 2g
h h 1
V 2g d S S
f L L
1
V 4R*2g *S
f
,
,
=
= = =
=
2
0
0
f
d
d R
4
P D 4
8gRS
V
f
Comparing with Chezy equation:
8g
C =
f
C 1
=
8g f
Manning formula is an emprical relation based on field obse
= =
=
[ ]
2/3 1/2
o
1

3
o
rvations and is given by
1
V= R S
n
in which V in m/s, R in m. Thus 'n' has dimensions of L T
If R=15 cm, n = 0.015, S 0 0004 then V = 0.376 m/s
The hydraulic enginee
. ,
=
rs use the n or C without bothering about dimension even though
it is very important. The treatment here is only for channels with plane bed.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
16.3.1 Resistance in Open Channel Hydraulics
If Manning and Chezy equations are compared
2
3
1
1
1 1 1
2 2 2
0 0
2 1 1

3 2 6
1
6
e
e
2 2
e1
2 2 2
2
e1
e1
1
R S CR S
n
R R
C=
n n
R
C=
n
For laminar flow:
K
f =
R
VR
R
VR
K= f
=
=
=
=
=
=
2
8g
f =
C
14
f = For triangular Smooth Channel (Refer: Chow)
R
e1
24
f = For Rectangular Smooth Channel (Refer: Chow)
R
e1
Sand Roughness Fixed to Flume Bed (Photograph  Thandaveswara)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
16.3.2 Laminar Flow with Roughness
e1
e1
60
f = for a 90 V shape channel. Roughness 0.3023 mm
R
33
f =
R
10
Laminar
Transitional Turbulent
10
2
10
3
10
4
10
5
10
6
2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8
0.004
0.006
0.008
0.01
0.02
0.04
0.06
0.08
0.1
0.2
0.4
0.6
0.8
1.0
R
e1
f
f =
14
___
R
e1
f =
24
___
R
e1
Reference:
"Chow Ven Te Open Channel
Hydraulics", Mc Graw Hill Company,
International student edition, 1959, page  10
Variation of friction coefficient f with Reynolds number Re1
in smooth channels
=
vR
__
( )
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
f =
14
___
Re1
f =
33
___
Re1
f =
60
Re1
___
10 10
2
10
3 10
4
10
5
10
6
2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8 2 4 6 8
Re1
Variation of friction coefficient f with Reynolds number Re1
in rough channels
=
vR
__
( )
10
7
2 4 6 8
0.006
0.008
0.01
0.02
0.04
0.06
0.08
0.1
0.2
0.4
0.6
0.8
1.0
f
37.5 cm
25 cm
Varwick
Varwick
1
1
20 cm
Laminar Transitional Turbulent
10
3 10
4
10
5
10
6
2 4 6 8 2 4 6 8 2 4 6 8
10
7
2 4 6 8
0.02
0.04
0.06
0.08
0.1
Reference:
"Chow Ven Te Open Channel Hydraulics",
Mc Graw Hill Company, International
student edition, 1959, page  11
Rectangular Channel  Rough flow (Roughness =0.7188)
Bazin conducted experiment using (500 measurements were made at greatest care)
(1) Gravel embedded in cement.
(2) Unpolished wood roughened by transverse wooden strip
(i) 27 mm long * 10 mm high * 10 mm spacing.
(ii) 27 mm * 10 mm at 50 mm spacing.
3) Cement lining
4) Unpolished wood
If the behavior of n and C is to be investigated then a number of basic definitions
regarding the types of hydrodynamic flow must be recalled.
Flow can be divided into
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(i) Hydro dynamically smooth turbulent flow
(ii) Hydro dynamically Rough turbulent flow
(iii) Hydro dynamically transition turbulent flow.
1 2
o
1 5
7 o
e
The boundary layer for flow past a flat plate is given by
V x
5 Laminar
x
V x
0 38 turbulent R 2 10 logarthmic velocity law holds
x
/
/
. *
=
= >
Velocity
V
99% V
y
Velocity distribution
0
0 0
k
kc kc
k
kc
Different surface roughness
(c) rough
k
k
c
=
v*
__
100
kc
=
5
v
*
__
Smooth
for average condition
kc is critical roughness height
k is roughness height
(b) wavy
(a) Smooth
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Viscous sublayer
k
s
(i) Hydrodynamically smooth
turbulent flow f = f(R
e
)
Viscous sublayer
k
s
Viscous sublayer
k
s
(ii) Hydrodynamically transition
flow f = f (R
e
, k
s
/y)
(iii) Hydrodynamically rough
turbulent flow f = f (k
s
/y)
For hydro dynamically smooth condition, viscous sub layer submerges the roughness
elements.
For hydro dynamically transitional case the roughness element are partly exposed with
reference to viscous sub layer.
For hydro dynamically rough turbulent flow the roughness elements are completely
exposed above the viscous sub layer.
For hydro dynamically rough turbulent flow resistance is a function of Reynolds number
and the roughness height.
If we define
e*
R =shear Reynolds number
* s
v K
. ; and
o
* f
v gRS
= = .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The flow is classified as follows:
* s
* s
* s
v K
4 Hydrodynamically smooth
v K
4 100 Hydrodynamically transition
v K
100 Hydrodynamically fully developed turbulent flow
<
< <
>
Summary of VelocityProfile Equations for Boundary layers with
dp
0
dx
=
Zone Smooth Walls Rough Walls
Law of the
wall
Universal equations
Laminar
sub layer
( y )
*
v y
4
<
*
*
v y v
v
=

Buffer
zone
*
v y
4 30 to 70
< <


Logarithmi
c zone
(also
called turb
ulent layer)
*
v y
30 to 70
y
0.15
>
<
*
*
*
*
v y v
A log B
v
v y v
5.6 log 4.9
v
= +
= +
*
*
v k
A log B
v y
v k
5.6 log B
v y
B f
= +
= +
=
(roughness size, shape and
distribution)
Velocitydefect law
Inner
region
(overlaps
with
logarithmi
c wall law)
y
0.15
<
Outer
region
(approxim
ate
formula)
y
0.15
<
*
*
V v y
A log B
v
V v y
5.6 log 2.5
v
= +
= +
*
*
V v y
A log
v
V v y
8.6 log
v
=
(3000
<
e
R <
70,000)
outer
region

Power Law
1
7
*
*
v v y
8.74
v
=

A and B are constants.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Table shows velocity distributions for different conditions
Pipe flow equation
e
VR
R
=
Blasius
equation
for
smooth
flow
5
e
0.25
e
e
5
e
0.3164
f = upto R <10
R
R f 1
=2log
2 51 f
R 10
.
>
e
e
1/8 2
0.25
e
e
e
C=18.755 R mks units for g = 9.806 m/sec
0 223
f
R
R 8
C = 4 2g log
2 51 C
R 8g
C = 17.72 log
2 51 C
3.5294R
C = 17.72 log
C
.
g
.
.
=
Smooth
pipe flow
Nikurads
e Rough
pipe
Nikurads
e
( )
o
1
= 0.86 ln Re f  0.8
f
1
= 1.14  0.86 ln
d f
e
s
R 8g C
= 2 log
C 2 51 8g
C 12R
= 2 log
k 8g
*
.
White
and
Colebroo
k formula
o
/d 1 2.51
= 0.86 ln
3 7 f Re f .
+
s
e
2.52 8g k C
= 2 log
14.83R 8g R f
+
Suggested modification to equation is
s
e
k C 2.5
= 2 log
12R 8g R f
+
[ASCE Task Force Committee 1963]. R is hydraulic mean radius, 4R =Diameter of
pipe.
In open channel flow following aspects come into picture
( )
e
f = f R K, C,N, F,U
(1) (2) (3)
,
In which R
e
is the Reynolds number, K is the Relative Roughness, C Shape factor of the
crosssection, N is the Non uniformity of the channel both in profile and in plan, F is the
Froude number, U is the degree of unsteadiness.
In the above equation, the first term corresponds to, Surface Resistance (Friction), the
second term corresponds to wave resistance and the third term corresponds to Non
uniformity due to acceleration/ deceleration in flow.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Surface Resistance: To be accounted based on Karman  Prandtl  velocity distribution.
The constant in resistance equation is due to the numerical integration, and is a function
of shape of the crosssection.
C 1 R
=A log +B
y' 2g f
For circular section A = 2.0, B = 0.62
For rectangular section: A = 2, B = 0.79 (for large ratio of width/depth)
=
It has remained customary to delineate roughness in terms of the equivalent sand grain
dimensions k
s
. For its proper description, however, a statistical characteristic such as
surface texture requires a series of lengths or length derivatives, though the significance
of successive terms in the series rapidly approach a minimum. Morris classified the flow
into three categories namely (1) isolated roughness flow, (2) Wake interference flow,
and (3) Quasi smooth flow. The figure provides the necessary details.
s
y
k
Isolated  roughness flow (k/s)  Form drag dominates
s
The wake and the vortex are dissipated before the next element
is reached. The ratio of (k/s) is a significant parameter for
this type of flow
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
s s
Wake interference flow (y/s)
j
j
j
Quasi smooth flow  k/s or j/s becomes significant acts as Pseudo wall
s
y
k
y
k
s s
s
j
k is surface roughness height
s is the spacing of the elements
j is the groove width
y is the depth of flow
Concept of three basic types of rough surface flow
When the roughness elements are placed closer, the wake and the vortex
at each element will interfere with those developed by the following
element and results in complex vorticity and turbulent mixing. The height
of the roughness is not important, but the spacing becomes an important
parameter. The depth 'y' controls the vertical extent of the surface region of
high level turbulence. (y/s) is an important correlating parameter.
Quasi smooth flow is also known as skimming flow. The roughness elements
are so closed placed. The fluid that fills in the groove acts as a pseudo wall
and hence flow essentially skims the surface of roughness elements. In such
a flow (k/s) or (j/s) play a significant role.
k, j, s should describe the characteristics of roughness in one dimensional situations is
Areal concentration of or density distribution of roughness elements. (after Moris).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
16.3.3 Areal concentration or Density Distribution Roughness
Elements
Spheres
Spatial distribution of roughness
Schlichting, 1936
Koloseus (1958) and Koloseus and Davidian (1965)
conducted experiments using Cubical Roughness
Symmetrical diamond shaped pattern.
O'Loughlin and Mcdonald (1964) Cubes arranged
as in (1) abd (2) also sand grains (2.5 mm dia)cemented
to the bed .
1 2
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Logarithmic plot of data from figure at low
concentration
Effective roughness as a function of form pattern, and
concentration of roughness elements. (Assuming high
Reynolds number)
Open channel resistance (after H. Rouse, 1965)
1.0 0.1 0.01
0.001
0.1
1
10
0
1
2
3
4
0 0.2 0.4 0.6 0.8 1.0
Spheres
Sand
Cubes
Nikuradse
Sand
Schlichting (1936)  Sphere spacing
Koloseus (1958)
Koloseus and Davidian (1965)
Cubical Roughness
Symmetrical diamond shaped pattern
O'Loughlin and Mcdonald (1964)
k
s
___
y
Areal concentration
Cubes arranged as in 1 and in 2.
Also sand grains centered to the sand grains
(2.5 m diameter)
Areal concentration
k
s
___
y
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Resistance of a bridge pier in a wide channel, after Kobus and Newsham
F = 1.5
1.0
0.5
b
3b
d = 3b
V
0 0.5 1.0 1.5 2.0 2.5
0
0.5
1.0
1.5
Froude number, F
Variation of pier resistance with lateral spacing "S"
0 0.5 1.0 1.5 2.0
Froude number, F
0
0.5
1.0
1.5
C
D
S
D
__
= 5
7.5
30
D
S
D
d = 30
V d
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Loss at one of a series of channel bends after Hayet
0.2 0.4 0.6 0.8 1.0 2.0 4.0
0.1
0.2
0.4
0.6
y/b = 1/16
y/b = 1/8
y/b = 1/4
2b
4b
90
0
y
b
Froude number, F
Some of the important References:
(i) Task force on friction factors in open channels Proc. ASCE J I. of Hyd. Dn. Vol. 89.,
No. Hy2, March 1963, pp 97  143.
(ii) Rouse Hunter, "Critical analysis of open channel resistance" , Proceedings of ASCE
J ournal of Hydraulic division, Vol.91, Hyd 4, pp 1  25, J uly 1965 and discussion pp 247
 248, Nov. 1965, March 1966, pp 387 to 409.
Schlichting, "Boundary layer theory", Mc Graw Hill Publication.
16.3.4 Open Channel Resistance
There is an optimal area concentration 15% to 25% which produces greater relative
resistance.
1 R
A log B
DhS f
= +
h is the roughness height , S is the areal concentration (<15%), D is the constant which
depends on shape and arrangement of the roughness elements.
For sanded surface: D =21 and B =2.17
The existence of free surface makes it difficult to assume logarthmic velocity distribution
and to integrate over the entire area of flow for different crosssectional shapes. The
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
lograthmic velocity distribution can be integrated only for the wide rectangular and
circular sections.
Effect of boundary nonuniformity is normally ignored and particularly so for gradually
varied flow profile computation.
The dependence on Froude number is clearly seen in case of pier.
In case of unsteady flows such as floods, it is assumed that the inertial effects are small
in comparison with resistance. Hence, the resistance of steady uniform flow at the same
depths and velocity is taken to be valid.
Where the Froude number exceeds unity, the surface has instability in the form of roll
waves.
Earlier formulae for determining C (for details refer to Historical development of
Empirical relationships)
1. G.K. Formula (MKS)
2. Bazins Formula 1897 (MKS)
3. Powell Formula (1950) FPS while using Powell formula C must be multiplied by
0.5521 to get C in m
1/2
s
1
4. Pavlovskii Formula (1925)
Manning equation is applicable to fully developed turbulent rough flow.
Slope of the straight line is 1:3
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1/ 6
1/ 3
s
2
s
1/ 6
s
1/ 3
s
s
1/ 3
k g g
f C
R k
C
n k
k
f = 0.113
R
If we replace k by diameter of the grain size (d)
d
f = 0.113
R
8g 8g R
C =
f 0.113 d
1/ 6
2
1/ 6 1/ 6
1/ 6
1/ 6
1/ 6 1/ 6
for MKS units g = 9.806 m/s
8 * 9.806 R R
C = 26.3482
0.113 d d
R
or C = 26.34
d
R
n
C
1
n = *d 0.0379d
26.34
=
=
=
A number of empirical methods to relate n diameter of the particle are advanced.
1 Strickler
(1923)
[ ]
1/6
n = 0.02789 d d in m
This is not applicable to mobile
bed
2 Henderson's
interpretation
of Strickler's
formula
[ ]
1/6
50
n = 0.034 d d in feet
3a Raudkivi
(1976)
[ ]
1/6
n = 0.047 d d in m
3b Raudkivi
(1976)
[ ]
1/6
65
n = 0.013 d d in mm
d
65
=65 % of the material by
weight smaller.
3c Raudkivi
(1976)
[ ]
1/6
65
n = 0.034 d d in feet
4 Garde and
Ranga Raju
[ ]
1/6
50
n = 0.039 d d in feet
( )
( )
( )
1/6
1 6
50
0 039 0 3048
0 039 0 82036 0 03199
n = 0.03199 d , d is in 'm'
/
. * .
. . . = =
5 Subramanya
[ ]
1/6
50
n = 0.0475 d d in m
6 Meyer and
Peter and
Muller
[ ]
1/6
90
n = 0.038 d d in m
(Significant proportion of coarse
grained material)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
7 Simons and
Sentrvrk
(1976)
[ ]
1/6
n = 0.047 d d in mm
8 Lane and
Carbon
(1953)
1/6
75
n= 0.026 d
(d in inches and d
75
=75% of the
material by weight is smaller)
( )
* f
* s
1/6 1/6
1/6 1/6
1/6 1/6
1/6
8) Consider
v g R S
k
4 < 100 Transition flow
v
R R
n = but C = 26.35
C d
R d 1
n = d 0 03795 d (d in m)
26 35
R 26 35
Conditon for fully develop
.
.
.
=
<
= =
( )
( )
6
8 6 * s
6
6
f
6 2 2
6
f
6 6
6
ed rough flow
v k n
100 d = 3 3458 10 n
0.03795
n 1
g R S
0.03795
Assuming
= 1.01 * 10 m /s g = 9.806 m/s
9 806 1
n R S 100
1 01 10
0.03795
n
. *
.
. *
= =
=
14
f
RS 9 635 10
. *
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Laminar flow
Smooth surfaces
Fully rough zone
Transition zone
Commercial surfaces
Sand coated surface (Nikuradse)
Reynolds number Re = 4 V R/v
Modified Moody Diagram showing the Behavior of the Chezy C after Henderson
10
3
10
4
f =
0.316 _____
Re
0.25
(C = , mks) 15.746 Re
1 __
8
1 __
f
= 2.0 log ( )
Re f _____
2.51
C = 4 2g log ( )
Re 8g
2.51C
_____
Blasius equation ( ) Re <10
5
______
= 100
v* ks
1 __
f
=
C __
8g
= 2.0 log ( )
12R
___
ks
10
5
10
6
10
7
10
8
2 4 6
30
40
50
60
70
80
90
100
110
120
130
140
150
180
or Manning
507
252
126
60
30.6
15
10
do _____
2ks
2R _____
ks
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
16.4 History of Uniform Flow Velocity and Resistance Factor
The design of the crosssection of the Roman aqueducts was based on structural rather
than hydraulic requirements. Though the importance of the downward slope of the
channel was realized, the aqueducts were laid at slopes governed by the topographic
considerations alone.
HERO of Greece (after 150 B.C.) has clearly indicated that the rate of flow depended
upon the overall change in the elevation on one hand, and upon the velocity as well as
the crosssectional area on the other.
LEONARDO DA VINCI (14521519):
The water of straight rivers is the swifter the farther away it is from the walls, because
of resistances.
Water has higher speed on the surface than at the bottom. This happens because water
on the surface borders on air which is of little resistance, because lighter than water,
and water at the bottom is touching the earth which is of higher resistance, because
heavier than water and not moving. From this follows that the part which is more distant
from the bottom has less resistance than that below.
As regards the basic law of continuity of flow, he has clearly stated that:
A river in each part of its length in an equal time gives passage to an equal quantity of
water, whatever the width, the depth, the slope, the roughness, the tortuosity.
Each movement of water of equal surface width will run the swifter the smaller the
depth.
The law of continuity was explained in more certain terms and popularized by
BENEDETTO CASTELLI (c 1577 c 1644), became widely known in Italy as Castelli s
law.
The 18th century witnessed the advent of hydrodynamics LEONHARD EULER (1707
1783) giving the equations of motion of an ideal fluid and DANIEL BERNOULLI (1700
1782) enunciating the famous energy equation which goes by his name.
HENRI DE PITOT (1695 1771) devised the velocity measuring device which carries
his name the Pitot tube.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
CORNELIUS VELSON (1749), at Amsterdam, came to the conclusion that the velocity
of flow should be proportional to the square root of the slope.
ALBERT BRAHMS, in 1757, considered the resistance thus set up to be proportional to
the area of crosssection divided by the length of the wetted perimeter. Thus, resulted
the expression R =A / P where R is the hydraulic radius; A the area of cross section of
the flow, and P the wetted perimeter.
16.4.1 Development of the Empirical Formulae
Credit for the first as well as the most lasting equation of resistance in uniform open
channel flow goes to ANTOINE CHEZY, (1718 1798), a French Hydraulician, who
was assigned the project of determining the crosssection of a canal to supply water to
the city of Paris from the river Yvette.
Chezy put forth that
0
2
V / RS would be the same for all streams having similar
characteristics; where V is the mean velocity of flow and S the bed slope of the
channel. Chezy, however, did not assume that the value
0
2
V / RS was a constant for all
streams, as he found this value to vary from one stream to another.
The presentday Chezyformula is written as
0
V = C RS where C is known as Chezy
Coefficient. On the basis of a few observations of the flow made on an earthen channel,
the Courpalet Canal and the Seine River, Chezy arrived at the value of C equal to 31.
However, it should be noted that this formula, empirical in nature, is not dimensionally
homogeneous. The Chezy coefficient C is not a pure number, but has a dimension of
[ ] [ ]
1
1
2 L T
, where
[ ]
L and
[ ]
T are units of length and time of any measuring system.
PIERRE LOUIS GEORGES DU BUAT (17341809): He proposed a formula for average
velocity.
0 0
48.85 R0.8
V =  0.05 R
1/S  ln (1/S ) + 1.6
[in metric units].
The surface roughness of the boundaries was ignored in the formulation of the above
equation.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
J OHANN ALBERT EYTELWEIN (17641848), published, at Berlin in 1801, a formula for
open channel flow, namely
0
V = 50.9 RS (in metric units)
A firm proponent of non dimenisonal quantities in the analysis of any problem, J ULIUS
WEISBACH (18061871), was the first to write a formula for resistance to flow through
closed pipes as
2
L
0
L V
h = f
d 2g
in which f is a non dimensional friction coefficient, which is commonly known at present
as the DarcyWeisbach friction factor,
L
h the head lost due to the frictional resistance, L
the length of pipe in which the head loss
L
h has occurred and d
0
the diameter of the
pipe. Weisbach reported that f is a function of the Reynolds number
e
R and the relative
roughness, for a given shape of crosssection.
By this period, the general form of the resistance equation for the uniform flow in rigid
bed open channels was accepted to be given by the triplefactor formula
y x
0
V = CR S
which represented the interdependence between the mean velocity of flow, hydraulic
radius and the slope of the channel. Values of the coefficient C and the exponents x and
y were chosen to make the formula conform to the experimental data obtained by each
investigator. The various investigators, in choosing different values of C, x and y which
they believed to be the most probable values, have deduced a large number of
empirical flow formulae.
The first systematic and extensive series of experiments on open channel flow, to
discover how the coefficient C varied with different kinds of roughness of the
boundaries, were first begun by HENRY PHILIBERT GASPARD DARCY (18031858)
in 1855 in France, and were continued after his death by his worthy assistant HENRI
EMILE BAZIN (18291917).
DARCY conducted his studies in a wooden flume, 600 m long, drawing its supply from
the Bourgogne Canal through a specially constructed head reservoir and discharging
into the river Ouche. The flume was 2 m wide and 1 meter deep and has the feasibility
of its inclination and crosssection could be changed. Rectangular, trapezoidal,
triangular and semicircular crosssections were tested. The different surfaces tested
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
included cement, wood, brick, fine and coarse gravel, rock, and surface with artificial
roughness in the form of wooden strips fixed transverse to the flow. Measurements on
some earthen channels, which formed branches of the Bourgogne Canal, were also
made.
Bazin observed that the value of C increased with an increase in slope, but concluded
that this increase is of too small moment to be provided for in the equation.
Two Swiss engineers, EMILE OSCAR GANGGUILLET (18181894) and WILHELM
RUDOLPH KUTTER (18181888) concluded that the two formulae proposed by Bazin
stood for two extreme conditions, and none of the two could be applied for general
application. They published results in 1869.
0
0
m
a + +
n S
C =
m n
1 + a+
S R
l
A detailed account of the development of the above formula was given by LINDQUIST.
The values of the constants a, l and m arrived at by GANGUILLET AND KUTTER from
the analysis of their data were
Constants in metric units
a 23.00
l 1.00
m
0.00155
PHILIPPEGASPARD GAUCKLER (18261905) made a proposal of two formulae for
use in different slope ranges, as follows:
4/3
1 0
V = R S or
5/6 1/2
1 0
C = R S for S
0
less than 0.0007
and
2/3 1/2
2 0
V = R S or
1
6
2
C = R for S
0
greater than 0.0007
in which
1
and
2
are coefficients to be determined experimentally.
In 1889, ROBERT MANNING (18161897), an Irish engineer, presented a paper
containing several formulae for the velocity of flow in open channels, at a meeting of the
Institution of Civil Engineers of Ireland. This paper was later published in the
Transactions of the above Institution in 1891. In this paper, Manning proposed an
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
equation similar to the above equation to be in better agreement with the available
experimental data of flow in open channels than any other formula used till that time.
Manning found that the average value of the exponent of R varied from 0.6499 to
0.8395 on the basis of the experiments on artificial channels by DARCY and Bazin. He
adapted an approximate value of 2 / 3 for this exponent. MANNING finally proposed, for
earth channels in good condition, the formula.
In metric units,
1
1/2
2
0
R
V = 34 S R +  0.03
4
Or
0 03
34 1
4
R .
C
R
= +
The chronology of the present day Manning formula is given in detail in the discussions
made by KING, CHOW, ROUSE, ROBERTSON, DOOGE, POWELL, POSEY.
By 1889, it was discovered that the reciprocal of , expressed in metric units,
corresponded very closely to the roughness coefficient n associated with Ganguillet
Kutter formula. Thus, in 1891, FLAMANT gave the formula
2/3 1/2
0
1
V = R S
n
(in metric
units) as Manning equation.
Later in 1923, STRICKLER supported the same formula, independently and chiefly
based on his own observations in Switzerland. His analysis resulted in the equation.
2/3 1/2
0
V = MR S
Manning formula reads as
2/3 1/2
0
1
V = R S
n
and the coefficient C turns out to be
1 6
1
/
C R
n
=
It is to be noted that the same numerical value of n can be used both in English and
metric systems.
The coefficient C has one and the same value for all channels of very large dimensions.
Thus, Bazin proposed a new formula
2
2
0 0115 1
V
B
RS
.
R
= +
(in metric units)
which can be reduced to the form
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
86 96
1
B
.
C
R
=
+
(in metric units)
The term
B
in the above equation is a roughness factor. However, Bazins
B
exhibits
a thirty fold variation for a threefold variation in Kutters n.
As the slope of the channel is, once again, not considered in the above equation,
Bazins C is considered to be a function of R alone and not S
0
.
Another empirical formula for the Chezy coefficient C was given by PAVLOVSKII , in
1925.
The formula is
i
R
C
n
= (in metric units)
in which
( )
2 5 0 13 0 75 0 10 i . n . . R n . =
The values of n in the above formula are the same as those in the case of Manning
formula. The use of this formula is limited to the ranges of hydraulic radius between
0.10 and 3.0 m and n between 0.011 and 0.040. For practical purposes, PAVLOVSKII
also offered two approximate formulae for the exponent i, VIZ.,
1 5 i . n = for R less than 1 meter and
1 3 i . n = for R greater than 1 meter.
But it is the original formula of PAVLOVSKII which, in spite of its cumbersome form, is
generally used in preference to the above simplified formulae.
1. CHEZY FORMULA (1775):
V = C RS
2. DU BUAT FORMULA (1779):
0 0
48 85 0 8
V 0 05
1 1
1 6
. R .
. R
ln .
s s
=
+
3. GIRARD FORMULA (1803):
4. DE PRONY FORMULA (1804):
2
0
RS = 0.00004445 V + 0.00030931 V (in metric units)
5. EYTELWEIN FORMULA (18141815):
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
0
RS = 0.0000243 V + 0.000336 V (in metric units)
6. LAHMEYER FORMULA (1845)
This is based on 616 gaugings on the river Weser in Germany, and takes into
consideration the effect due to bends in a river.
0
c
RS W
= 0.0004021 + 0.0002881
r V V
(in metric units)
in which W is the width of the river and
c
R the radius of curvature of the river. For a
straight reach of the river, the term containing
c
R should be dropped out. It is to be
noted that the term
c
W / r is reported under the root sign by LELIAVSKY.
7. ST. VENANT FORMULA (1851):
( )
11/21
0
V = 60 RS (in metric units)
8. TADINI FORMULA (1850): (in metric units)
0
V = 50 RS
The same formula is also attributed to COURTOIS.
9. HUMPHREYS and ABBOT FORMULA (1861)
2
'
0
0.0025 * 0.933 0.933
V= + 68.72 R S  0.05
R + 0.457 R + 0.457
(in metric units)
10. GANGUILLET AND KUTTER FORMULA (1869):
In metric units, C is given by
0
0
1 0.00155
23.0+ +
n S
C =
0.00155 n
1+ 23.0+
S R
11. REDUCED FORM OF GANGUILLET KUTTER FORMULA:
100 R
C
n R
=
+
(in metric units)
12. GIBSON FORMULA:
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1
24 55
1 24 55
.
n
C
n
.
R
+
=
+
(in metric units)
13. MANNING FORMULA (1889):
2/3 1/2
0
1
V = R S
n
(in metric units)
14. BAZIN FORMULA (1897):
86 96
1
B
.
C
R
=
+
(in metric units)
15. SIEDEK FORMULA (1901):
This formula was given, in metric units, for the case of natural streams and rivers.
( )
mean 0
1/20
y 1000 S
V =
W
where
mean
y is the mean depth of flow. This formula was stated to be applicable to
normal channels was classified, with the corresponding correction to the basic formula
given above and is expressed in terms of tables and involved formula.
16. VELLUT FORMULA (1902):
1
23 0
25 0
1
V
V
.
C
.
R
+
=
+
(in metric units)
Where
V
is the roughness coefficient.
17. HERMANEK FORMULA (1905): (in metric units)
This formula is proposed for rivers and streams. Forcheimer, modified the formula and
presented the same as follows.
( )
( )
( )
0.5
mean
0.75 0.5
mean
0.60 0.5
mean
i V =30.7 y S for y < 1.5 m
ii V = 34.0 y S for 1.5 < y < 6 m
iii V = 44.5 y S for y > 6 m
18. MATAKIEWICZ FORMULA (1911): (in metric units) reported by STRICKLER
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0.7 0.493 10S
0.7
(i) V = 35.4 y S
(ii) V = 35.4 R S
in which is a variable exponent dependant on boundary roughness
+
19. KOCHLIN FORMULA (1913): (in metric units)
( ) K
V = C 1 + 0.6 R RS
where
K
C is the roughness parameter.
20. BARNES FORMULA (1916):
V = C R S
in which C, and vary depending on the type of the channel boundaries
21. STRICKLER FORMULA (1923):
2/3 1/2
V = M R S
22. FORCHHEIMER FORMULA (1923):
0.7 0.5
V = C R S
where the value of the coefficient C varied from 143 to 43 (in English units)
23. PAVLOVSKII FORMULA (1925): (in metric units)
i
1
C = R
n
in which
( )
i = 2.5 n  0.13  0.75 R n  0.10
Manning formula has the main advantage of being simple, easily remembered and least
laborious in computations. Also, it was found from the analysis of several tests under
wide ranges of flow conditions as regards roughnesses of the boundaries, and shape,
size and types of channels, that this formula yields results accurate enough for all
practical purposes, when the values of roughness coefficient "n" already standardized
for Ganguillet Kutter formula themselves were adapted. This formula was more
accurate for small slopes. The change over to the use of Manning formula was thus
made convenient for there was no need to get familiarized with a new set of roughness
coefficients.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Another advantage of the simple form of Manning formula is that a very simple relation
exists between any given value of n and the corresponding value of velocity or slope. If
a certain error be made in selecting n, then the computed value of velocity, and also the
discharge in its turn, will involve the same percentage error but in the opposite direction.
Likewise the value of slope computed to give a certain velocity will contain twice the
same percentage error. The importance of this knowledge is of immense help to the
designers.
REYNOLDS who, by his classical experiments with dyes, demonstrated clearly the
difference between the two types of flows viz; laminar and turbulent and indicated the
presence of a critical velocity. REYNOLDS also showed the physical significance of his
dimensionless number.
e
V L V L
R = =
he showed that a corresponding change in the law of resistance
occurred with the change in the type of motion.
By this time, the DarcyWeisbach equation for head loss through circular pipes
2
f
L V
h = f
d 2g
was well established.
A set of very comprehensive and carefully conducted tests on the flow of water in
circular pipes of different materials and of different diameters, by DARCY, revealed the
following important phenomena.
(a) The coefficient of friction f is dependent on the Reynolds number
e
R and the relative
roughness of the pipe
0
k
d
, where k is the average depth of pipe wall roughness and
0
d
is the diameter of the pipe.
(b) The coefficient f decreases with an increasing Reynolds number, the rate of
decrease being smaller for greater relative roughness.
(c) The coefficient f is independent of the Reynolds number for certain relative
roughness, and
(d) The coefficient f increases with an increasing relative roughness for any particular
value of Reynolds number.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
From dimensional analysis also, the same result is obtained, ( )
e o
f = f R , k / d
In 1932  33, NIKURADSE conducted a series of wellplanned tests on flow through
circular pipes, artificially roughening the inside walls of the pipes by cementing layers of
sand grains of uniform diameter.
Together with the theoretical work of PRANDTL and von KARMAN, Nikuradses
experimental findings have led to the establishment of semi rational formulae for
velocity distribution and hydraulic resistance for turbulent flows in circular pipes.
The HagenPoiseuille equation can be written as
f
2
0
32 V L
h
d
=
Where
V
is the specific weight of the liquid.
In 1913, BLASIUS, drawing on the boundary layer theory, developed an empirical
expression for the coefficient of friction f
( )
0 25 0 25
0
0 3164 0 3164
f =
V d
. .
e
. .
R
/
= .
This result was based on the experimental data of flow in smooth circular pipes with the
Reynolds numbers up to 100,000.
For the range, 4 000 100 000
e
, R , , an almost perfect agreement between this equation
and the experimental curve of NIKURADSE was observed. However, BLASIUS
equation deviated considerably from the experimental curve when the Reynolds number
exceeded 1,00,000.
COLEBROOK and WHITE carried out their investigations using commercial pipes and
found significant difference in the value of f from those of NIKURADSE in the transition
region from smooth turbulent to completely rough flow.
0
1 18 7
1 74 2 0
s
e
k .
. . log
r f R f
= +
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
MOODY has plotted the above equation to appear in the form of a family of
e
log f vs log R curves for various
0
1 18 7
1 74 2 0
s
e
k .
. . log
r f R f
= +
values.
Application of the semirational formulae to open channel flows:
Analysing Bazins experimental data in this connection, KEULEGAN arrived at the
equation ( ) 1 2 034 2 211
s
/ f . log R / k . = + in the case of turbulent flow in roughwalled
channels.
Thus, the logarithmic formulae for rough walled channels were expressed as follows:
*
*
Rv V
A 5 75 log
v
S
.
= +
for smooth channels and
r
* S
V R
= A + 5.75 log
v k
for rough
channels.
in which the characteristics
S
A and
r
A are functions of the Froude number.
16.4.2 Exponential Formulae
STRICKLER expressed the Manning 'n' in terms of roughness
s
k as
1/6
s
n = 0.00106 k
(
s
k in cm)
But he started with the numerical value of 1.476 instead of 1.486 in the Manning
formula. Stricklers formula for n is given by
1/6
m
n = 0.0342 d
in which
m
d is the median sieve size of the sand grains and in feet.
1/6
m
n = 0.02789 d
in which
m
d is in "m".
WILLIAMSON from his experimental data and also with some suggested corrections to
Nikuradses data, gave the formula
1/6
s
n = 0.031 k (in English units)
Bretting stated that the logarithmic equation for the rough turbulent flow could be
replaced by three exponential formulae each valid for a particular range of values of
relative roughness. He found that exponential law equivalent to Manning formula was
valid when
s
4.32 < R / k < 276 requires Manning formula to be as given below.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1/6
s
n = 0.0387 k (
s
k in meters)
Manning formula is an exponential equation applicable to a particular range. In the first
place, Mannings formula, in which V is associated with square root of S0, is there by
limited in its application to the fully developed rough turbulent flow.
1/6
C = 8g / f = R / n or
1/6
n = R f / 8g
For fully developed flow at high Reynolds number, f is found to be independent of
Reynolds number, and nearly proportional to
1/3
1 / R . Thus, in the fully developed
regions of flow, a nearly constant value of n is realized.
HENDERSON gives the criterion, for the satisfactory application of Manning equation,
to be
( )
6 14 1/6
f
6 2
n RS 3.0755 * 10 with the assumption n = 0.03795 d
1 01 10 m /s and g = 9.81 m / s / s . *
=
Significant differences were observed between the discharge computed using a
constant value of n and the actual discharge in the case of channels which gradually
closed at the top, during the experimental investigations.
Moreover, it has also been observed that the value of the coefficient 'n' varies
considerably, even in prismatic channels (without gradually closing tops, (i) with age; (ii)
in the presence of of algae and vegetation and (iii) when the water carries sediment. A
deposit of slimy silt on the bottom and sides of the channel was found to greatly reduce
the frictional resistance to flow.
In the case of silt carrying waters, the lower layers of the moving water which are
heavily siltladen will form a kind of slurry which produces a lubricating effect in damping
the vortices created at the surface of contact between the boundaries of the channels
and the flowing water. The presence of large boulders on the bed also contributes to the
varying nature of the coefficient 'n' with the stage of flow.
The variation of the Manning coefficient 'n' with the curvature of the channel was
investigated by EDDY and SCOBEY. The results, in general, indicated that while
relatively low values of n were obtained for channels having smooth curvature with large
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
radius, sharp curvatures of the channels resulted in increased values of n. The effect of
channel irregularity, nonlinear alignment of the channel and obstructions to the flow on
the flow characteristics and the roughness coefficient. Further, the value of Manning n
was observed to vary with the stage and discharge in the natural streams and rivers,
depending upon the existing conditions of the particular channel. In 1956, COWAN
developed a procedure to select the value of n applicable to natural streams, floodways
and similar channels. This method involved the selection of the basic
0
'
n value for a
straight, uniform, smooth channel in the natural material and of the modifying values for
each of the five primary affecting factors; viz.
(i)
1
'
n due to the surface irregularities;
(ii)
2
'
n due to the variation in the shape and size of the channel crosssections;
(iii)
3
'
n due to the presence of obstructions in the flow;
(iv)
4
'
n because of growth of vegetation, algae or weeds; and,
(v)
5
'
n due to the meandering of the channel.
COWAN presented the values of the correction factors for various conditions. The value
of n may be computed by the equation,
( ) 0 1 2 3 4 5
' ' ' ' ' '
n n n n n n n = + + + +
The factors affecting the Manning coefficient are summarized in an excellent manner by
CHOW and he has stated that there is no evidence about the size and shape of a
channel as an important factor affecting the value of n.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
17.1 Friction
The Chezy and Manning equations have a long history in hydraulics. These empirical
relationships are being used for more than two centuries since their development.
Chezy equation was proposed by Antoine Chezy about 1769. Similarly, Gauckler in
1868 proposed the Manning formula. These equations do not account for turbulent
processes. Both these equations require estimation of a resistance coefficient. The
assumptions made while deriving the equations are steady uniform flow. However,
these equations are being used in nonuniform as well for unsteady flows. Only a few
attempts have made to investigate the validity of these assumptions s for nonuniform
flow. However, it is believed that these equations work for these cases also. The
primary difficulty in predicting the frictional resistance still lies in estimating the
resistance coefficient for a natural Situation. Figure shows a variation of Manning n
obtained by Baltzer and Lai for a natural channel. The large scatter at low Reynolds
numbers is to be expected since the accuracy requirements on data become extremely
high under this condition.
Reynolds number, n =
UH
v
__
Manning n vs. Reynolds number for threemile Slough near
Rio Vista, California. (After Baltzer and Lai)
0
1 X 10
6
2 X 10
6
3 X 10
6
4 X 10
6
5 X 10
6
6 X 10
6
0.01
0.02
0.03
0.09
0.08
0.07
0.06
0.05
0.04
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Reference Mahmood. K and Yevjevich. V (Ed) Unsteady flow in Open Channels,
Volume  I Water Resources publications, Fort Collins, Colorado 1975,
Page No. 47
Variation of flow rate with n for the Detroit River
Qo=Calculated flow for no = 0.025
CORRECTION FUNCTION
FOR MANNING'S COEFFICIENT
0.7 0.8 0.9 1.0 1.1 1.2
1.3 1.4 1.5 1.6 1.7
1.8
0.7
0.8
0.9
1.0
1.1
1.2
n/n0
0.6
One may note that, the relative error in the resistance coefficient leads to a relative error
in velocity or flow of the same magnitude. The above figure shows the effect of over
estimating the value of n leading to underestimating the flow rate and vice versa. This
figure is based on computations made for the Detroit River. The width of the line
indicates the variation of n with the normalizing flow Q
0
. Situations such as a portion of
the flow occupying a flood plain also complicate the assessment of frictional resistance.
In general, the estimation of frictional resistance in natural channels is yet to be
resolved.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
17.2 Ganguillet and Kutter (G & K) formula (River Mississippi)
Note: Kutter's C (for MKS units; for converting it into FPS multiply suitably; see Chow
pp.98).
+
+
2 / 3 1 / 2
1 / 6
0.00155
23 +
S n
C =
n
1+
S R
Manning's n ( Irish Engineer,1889)
1
V= R S
n
Chezy's C ( French Engineer,1768 )
R
C =
n
1
0.00155
23
Type of channel boundary surface Value of n
Very smooth surface such as glass, plastic or brass 0.010
Very smooth concrete and planned timber 0.011
Smooth concrete 0.012
Ordinary concrete lining 0.013
Glazed brick work 0.014
Vitrified clay 0.014
Brick surface lined with cement mortar 0.015
Cement concrete finish 0.015
Unfinished cement surface 0.017
Earth channel in best condition 0.017
Neatly excavation earth canals in good condition 0.017
Straight unlined earth canals in good condition 0.020
Rubble masonry 0.020
Corrugated metal surface 0.020
River and earth channels in fair condition 0.025
Earth channel with gravel bottom 0.025
Earth channel with dense weed 0.035
Mountain stream with rock beds and rivers with variables section &
some vegetation along banks
0.045
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
17.3 Conveyance
The conveyance of a Channel Section is a measure of the carrying capacity of the
channel section. The discharge of uniform flow in a channel may be expressed as W
Q=V A =C
0
A R
x
S
y
=K S
y
in which K =C
0
A R
x
.The term K is known as the conveyance of the channel section as
it is directly proportional to discharge (Q). When either the Chezy formula or the
Manning formula is used as the uniform  flow formula, then the exponent y is equal to
1 / 2, hence the discharge is equal to Q = K S
and hence the conveyance is equal to
Q
K =
S
This equation can be used for computing the conveyance when the discharge and slope
of the channel are given. When the Chezy and Manning formulae are used the
conveyance can be written as K=CAR
1/2
and
2
1
3
K = A R
n
respectively and in which C
is Chezy resistance factor and, n is the Manning roughness coefficient.
Generally, these two uniform flow equations are made use off to compute the
conveyance when the geometry of the water area and the resistance factor or
roughness coefficient are given. Since the Manning formula is used extensively, most of
the following discussions and computations will be based on
2
1
3
K = A R
n
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
17.4 The Section Factor for UniformFlow Computation
The term AR
2/3
is known as the section factor for uniform  flow computation; in case of
Manning formula. This would be AR
1/2
for Chezy formula. It is an important parameter
in the computation of uniform flow. From the equations given above, this factor may be
written as
0
For Manning formula
2
n Q
3
AR =
S
2
3
AR = n K
0
1
Q
2
AR =
C
1
K
2
C
For Chezy formula
S
AR =
Primarily, above equation applies to a channel section when the flow is uniform. The
right side of the equation contains the values of n or C, Q and S; but the left side
depends only on the geometry of the water area. Therefore, for a given condition of n or
C, Q, and S
0
, there is only one possible depth for maintaining a uniform flow, provided
that the value of A R
2 / 3
(or AR
1 / 2
) always increases with the increase in depth, which
is true in most cases. This depth is the normal depth y
n
. When (n or C) and S0are
known at a channel section, it may be seen from above equation that there can be only
one discharge for maintaining a uniform flow through the section, provided that A R
2 / 3
(
or A R
1 / 2
) always increases with increase of depth. This discharge is the normal
discharge.
An exponential Channel is defined to be that channel for which the relationship
between depth y and area of cross section A may be expressed in the form
A =k y i
in which k is a coefficient, different values for the exponent viz.; i =1, 1.5, 2.0,
represent rectangular, parabolic and triangular channels.
The above equation is a very useful tool for the computation and analysis of uniform
flow. When the discharge, slope, and roughness are known, this equation gives the
section factor and hence the normal depth y
n
can be computed. On the other hand,
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
when n or C, S
0
, and the depth (hence the section factor), are given, the normal
discharge Q
n
can be computed from this equation in the following form:
This is essentially the product of the water area and the velocity defined by the Manning
or Chezy formula. Sometimes the subscript n is used to indicate the condition of uniform
flow.
In order to simplify the computation, dimensionless curves showing the relation between
depth and section factor have been prepared for rectangular, trapezoidal, and circular
channel sections for Manning formula. These curves aid in determining the depth for a
given section factor, and vice versa. The A R
2 / 3
values for a circular section are given
in the table in Appendix. With the advent of numerical methods the usage of the
dimensionless graph is limited.
0.0001 0.001 0.01 0.1 1 10
0.01
0.02
0.04
0.06
0.08
0.1
0.2
0.4
0.6
0.8
1
2
4
6
8
10
d0
y
1
m
y
b
m = 1.5
m = 2.0
m = 2.5
m = 3.0
m = 4.0
Values of
2/3
_____
b
8/3
and
AR
2/3
_____
do
8/3
AR
Curves for determining the normal depth
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Problem: Calculate conveyance factor K using Manning equation for a trapezoidal
channel.
Solution:
( ) ( )
( )
[ ]
2/3
2/3
2/3
2/3
2
2/3
2/3 2/3
2/3
2 2/3
5/3
5/3 5/3
2/3
2 2/3
5/3
5/3
AR
K
n
b my y b my y
n b 2 1 m y
my my
b 1 y 1 b y
b b
y
n 1 2 1 m b
b
my
1 y b
b
y
n 1 2 1 m b
b
my
1 by
1 b
K
n
1
=
=
=
=
+ +
+ +
+ +
+ +
+
+ +
+
=
2/3
2 2/3
5/3
1/3
5 5 3
2/3 2 3
2
5/3
8/3
5/3
2/3
2
y
2 1 m b
b
my
1
1 y b b b
K
n
b b
y
1 2 1 m
b
my
b 1
y b
K
b
y
n 1 2 1 m
b
+ +
+
=
+ +
+
=
+ +
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
18.1 The Hydraulic Exponent for Uniform Flow
Computation.*
Assuming the conveyance K as a function of the depth of flow y*, it may be expressed
as
2 N
0
K =C y (1)
in Which C
0
is a coefficient and N is known as the hydraulic exponent for uniform
flow".
*This is strictly applicable to sections which are wide and are described by the
exponential equation
Taking logarithms on both sides of above equation and then differentiating with respect
to y, it may be written as
( ) (2)
d N
ln K =
dy 2y
Now, taking logarithms on both sides of Eq. (2) and then differentiating this equation
with respect to y under the assumption that Resistance factor is independent of y, the
expression for N may be obtained. [See Box]
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
K
2
=C
0
y
N
K
2
=C
0
y
N
Taking logarithm on both sides
( )
0
2 ln K = ln C +N ln y
Differentiating with respect to y
d N
(ln K) =
dy 2y
Consider Manning formula
2
2
1
2
3
K = AR
n
1 2
2 ln K= 2 ln +ln A+ ln R
n 3
Differentiating with respect to y
d d 1
2 ln K = 2 ln +2
dy dy n
( )
4
ln A+ ln R
3
d 1 dA 2 1 dR
ln K = +
dy A dy 3 R dy
equating the right hand side
N T 2 T 2 A dP
= + 
2y A 3 A 3 p dy
T 2 T 2 dP
= +  R
A 3 A 3 dy
2y dP
N= 5T 2R
3A dy
( )
2 ln K = ln C +N ln y
2 dA 1 dR
+
A dy R dy
Differentiating with respect to y
d N
(ln K) =
dy 2y
Consider Chezy formula
2
1
2
2
K = C AR
1
2 ln K= 2 ln C+ln A+ ln R
2
Differentiating with respect to y
d
2 ln K =
dy
d
ln
dy
( )
2
dA A
dy
equating the R.H.S
3 T 1 dP

2y 2 A 2P dy
y A dP
N= 3T 
A P dy
y dP
N= 3T  R
A dy
1 dA 1 1 dR
K = +
A dy 2 R dy
p T 1 dP
= + 
A 2 R dy
2RP
T T 1 dP
= + 
A 2A 2P dy
N
=
*
dA
T
dy
* These are the general equation for the hydraulic exponent N.
* This is strictly applicable only to section which are wide and are described by
the exponential equation
For a trapezoidal channel section having a bottom width b and side slopes 1 on m, the
expression for A, T, P and R may be obtained from Table. Substituting them in equation
in the Box and simplifying, it results in
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
n
my
in which y =
b
y
2
1+m
b 1+2y 10 8
N = 
y 3 1+y 3 2
1+2 1+m
b
This equation indicates that the value of N for the trapezoidal section is a function of m
and y / b. For values of m =0, 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, and 4.0, a family of curves for
N versus y /b may be constructed (Fig). These curves indicate that the value of N varies
within a range of 2.0 to 5.0.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Logarithmic plot of "K" as ordinate against the depth as abscissa will appear as straight
line then
( )
( )
1 2
1 2
log K /K
N=2
log y /y
The hydraulic exponent is equal to twice the slope of the tangent to the curve at the
given depth. When the cross section of a channel changes abruptly with respect to
depth, the hydraulic exponent will change accordingly.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
18.2 Establish the conditions for maximum discharge and
maximum velocity  a relation between depth and diameter
using chezy equation and Manning equation as shown in
table for a Circular Channel.
Show that
Manning's equation Chezy's equation
Maximum conveyance
0
0938
y
.
d
= or 302 22
'
0
095
y
.
d
= or 308
Maximum velocity
0
081
y
.
d
256 2756
' ''
0
081
y
.
d
=
257 27
'
=
Solution
Chezy equations
(a) Circular section (Maximum discharge)
( )
( )
( )
( )
( )
( )
( )
2
2
0
2
0
2
3 2
5 2
0
1 2
2
2
2 2
2
2
2 2
/
/
/
r
A sin
p r , Q AC RS
r
Q sin C RS
sin
r r
Q sin C S
r
sin
A r
R s
P r
sin
r
Q S C
in
=
= =
=
= = =
52
y d
0
T
1 2
3
3 0
/
d A dA dP
P A
d P d d
=
=
Let ; ( )
3 2
05
/
.
x sin
=
5 2
0
2 2
/
xr
Q S = C
Differentiating x w.r.t to and equating to zero.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( ) ( ) ( ) ( )
1 2 05 3 2
3 2
3 1
1 0
2 2
/ .
/
dx
sin cos sin
d
= =
/
( )
( )
1 2 3 2
3 1
1
2 2
/ /
sin
sin
cos
= =
( ) 3 1
sin
cos
=
( ) 3 2 sin cos = ;
308 =
Radians.
Then the depth for maximum discharge.
( )
180 1 26 1899
2
y r r cos r cos . r
= + = + =
128
308 180 64 ,
2
90 64 26
= =
=
095
o
y . d =
1899
095
2
.
.
=
(a) Manning Equation  Maximum Discharge
2 3
0
/
d
AR
d
=
1 3
5
2
0
/
d A
d
P
=
5 2
dA dP
P A
d d
= 0
( ) ( )
2 2
1
2 2
dA r r
cos A sin
d
= =
[ ]
dP
r P r
d
= =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( ) ( ) ( )
2 2
5 1 2
2 2
r r
r cos sin
0 r =
( ) (
3
3
5 1
2
r
cos r sin ) =
( ) ( )
1
1
25
cos sin
.
=
( ) ( 5 1 2 cos sin ) =
5 5 2 2 cos sin =
3 5 2 cos sin =
0
302 22
1876 1876
2 2
'
d
y r r cos . r .
=
= = =
0
0938 y . d =
(b) Circular section (Maximum velocity)
Using Manning equation
( )
2
1
2 2
r r
A sin ; R
sin
= =
V R
2 3
0
/
d
R
d
=
2 3
2 3
0
/
/
d A
d
P
=
0
dA dP
P A
d d
=
( ) ( ) ( )
2 2
1 0
2 2
r r
r cos sin
r =
( ) ( )
3 3
1
2 2
r r
cos sin =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0 cos sin + =
tan =
257 2756 25730
' '
'' =
The depth of water for maximum velocity is
2575
180 5125
2
.
y r r cos r r cos .
= + = +
=
0.81 diameter =0.81d
0
Problem
What would be the difference in discharge when it is running full and when it is
under
0
0938
n
y
.
d
=
Solution
0
0938
n
y
.
d
=
2 3
8 3
0
03353
/
/
AR
.
d
=
0
10
n
y
.
d
=
2 3
8 3
0
03117
/
/
AR
.
d
=
03353
10757
03117
max
full
Q .
.
Q .
= =
i.e. Maximum discharge is 76 higher than discharge in pipe when flowing full. . %
If Manning's equation is used.
If Chezy's equation is used,
0
095
n
y
.
d
=
2
0
077072
A
.
d
=
0
269057
P
.
d
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
0
0
0
028645
A
d R
.
P
d
d
= =
1 2
5 2
0
041249
/
/
AR
.
d
=
When full
2 1 2
0 0
5 2
0
1
039269
4 4 4 4
/
/
d d AR
.
d
= = =
041249
10504
039269
max
full
Q .
.
Q .
= =
504 . % excess.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
18.3 The channel bed slopes may be classified into the
following five categories
Sustainable Slopes
Mild Slope (M)
Critical Slope (C)
Steep Slope (S)
Non Sustainable Slopes
Horizontal H
Adverse Slope A
The slope that can uniform flow is called sustainable slopes. The mild slope sustains
sub critical ( ) 1
r
F < uniform flow, denoted as M. The critical slope sustains uniform flow
at critical depth ( ) 1
r
F = denoted as C. steep slope sustain the supercritical uniform flow
( ) 1
r
F > denoted as S.
When the slope is zero (Horizontal) then
2
3
1
o
V R S zero
n
= =
*
n n n
Q V y y indicated as y =
2
3
1
o
V R S
n
=
Thus y
n
is imaginary or negative (from chezy's equation).
Slopes are also classified as
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Sustaining slope
(i) Mild always subcritical uniform flow is sustained F<1
(ii) Critical always sustains critical uniform flow F=1
(iii) Steep always sustains super critical uniform flow F>1
NonSustaining slope
(i) Horizontal Normal depth y
n
*
(ii) Adverse
Slope
Positive slope
Negative slope
So = 
dz
dx
__
So =
dz
dx
__
Energy slope
Friction slope
Bed slope
Water surface slope
S
e
S
f
S
0
S
w
=
=
=
=
dz
dx
__
dH
dx
__
dy
dx
__
dh
f
dx
__
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
18.4 Types of Problem in Uniform Flow
The computation of uniform flow in carried out by using the continuity equation and
uniform  flow formula. When the Manning formula is used, the six variables involved in
computations are:
The normal discharge Q
n
, The mean velocity of flow V , The normal depth y
n
, The
coefficient of roughness n, The bed slope of channel
O
S and, the geometric elements
that depend on the shape of the channel section, such as area.
When any four of the above six variables are known, the remaining two unknowns can
be determined by the two equations. The following are some types of problems
associated with uniform  flow computations.
1. To compute the normal discharge Q
n
: this is required to be computed for the
termination of the capacity of a given channel or for developing a synthetic rating curve
of the channel.
2. To determine the velocity of flow V : It plays an important role in many applications
such as. It is often essential to study the scouring and silting effects in a given channel.
3. To compute the normal depth y
n
: Determination of the stage discharge relationship in
a given channel requires the computation of the depth of flow.
4. To determine the channel roughness n: This is used to ascertain the roughness
coefficient in a given channel; the coefficient thus determined is useful for other similar
channels.
5. To compute the channel slope S
o
: This is required to be computed for adjusting the
slope of a given channel eg: irrigation channel, power channel.
6 .To determine the dimensions of the channel section: This computation is required for
designing the channel.
Table 1 lists the known and unknown variables involved in each of the above six types
of problem. The known variables are indicated by a tick mark whereas the unknowns
required in the problem are indicated by a question mark (?). The unknown variables
that can be determined from the known variables are indicated by an *.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
TABLE 1: TYPE OF PROBLEMS OF UNIFORM  FLOW COMPUTATIONS
Type of
problem
Discharge
Q
Velocity
V
Depth
y
Roughness
n
Slope
So
Geometric
elements
1 ? *
2 * ?
3
* ?
4
*
?
5
*
?
6
*
?
By varying combinations of various known and unknown variables, more types of
problems can be generated. In design problems include the use of the hydraulically
efficient section and economical section.
Computation of the Uniform Flow
The normal depth and velocity may be computed by using a uniform  flow formula. In
the following computations, the Manning formula is used
Example: A trapezoidal channel (Fig), with b (m), m, S
o
, and n carries a discharge of Q
m
3
/ sec. Compute the normal depth and velocity.
Following are the different methods used for determining the uniform flow depth.
1. Algebraic method
2 Graphical method
3Trial error method
4Numerical method
Newton Raphson method
Bi section method
Secant method
Algebraic Method: For geometrically simple channel sections, the uniform flow
condition may be determined by an algebraic solution, as illustrated below.
Solution 1: The Analytical Approach. The hydraulic radius and water area of the given
section are expressed in terms of the depth y
n
as
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Substituting the given quantities in the above expressions in the Manning formula and
simplifying, one gets an algebraic expression as follows
a
o
+a
1
y
n
=[y (b+y)] 2.5
in which a
o
and a
1
are constants, b is the bed width of the channel in meter. This
equation is to be solved by trial and error for y
n
in meter. Then area of the flow an
square meter and velocity V
n
in m s
1
can be determined. Froude number is computed
to check whether the flow is sub critical or supercritical?
Solution 2: The Trail  and  error Approach .Some engineers prefer to solve this type of
problem by trial and error. Using the given data, the right side of equation for section
factor for uniform flow. Compute nQ/S
0
. Then, assume a value of y and compute the
section factor A R
2/3
. Make several such trails until the computed value of AR
2/3
is very
closely equal to x; then the assumed y for the closest trail is the normal depth. This trail
and error computation is shown as follows.
y A R R
2/3
A R
2/3
Remarks
B. Graphical Method. For channel of complicated cross section and variable flow
conditions, a graphical solution of the problem is found to be convenient. By this
procedure, a curve of y against the section factor A R 2/3 is first constructed and the
value of is computed. According to Eq. ( ), it is evident that the normal depth may be
found from the y  A R
2/3
curve where the coordinate of A R
2/3
equals the computed
value of. When the discharge changes, new values of are then computed and the
corresponding new normal depths can be found from the same curve.
C. Method of Design Chart. The design chart for determining the normal depth (Fig) can
be used with great expediency.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
19.1 Problem: Maximum Discharge
In partially full channel having an equilateral triangular cross section, the rate of
discharge is Q =KAR2/3 in which K is a constant, A flow area, R is the hydraulic mean
radius. Determine the depth at which the discharge is maximum, for a triangular
channel.
( ) ( )
A = b  y/ 3 * y and P = b + 4 y / 3
60
0
60
0
y
b
Triangular
Solution:
( )
( )
2
Area A = b  y / 3 * y
= by 0.58y
Perimeter P = b + 4 y / 3
= b + 2.31 y
For a given depth, the discharge is to be maximum.
According to Manning formula,
( )
( )
2 3 1 2
/ /
2
1
Q = * AR S
n
Where n and S are constants
by  0.58 y
Now, R = A / P =
b + 2.31 y
Substituting the values in the above expression, then,
Q = by
( )
( )
{ }
( )
( )
( )
( )
( )
1 3
5
2
4 5
2
2 2
4
5 0 58 2 31 2 0 58
0
2 31
+
=
+
/
2
 0.58 y / b + 2.31 y
differentiating the right hand side of the equation, then,
dA dP
b + 2.31y * * by . y * b . y * * by . y
dy dy
b . y
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( ) 2 31 =
2 2
2
2
But dA / dy = b  1.16 y and dP / dy . .
Then the above equation reduces to
5b 10.72 y  10.37 =0
5b 10.37
y =
10.72
y = 0.46
( )
0 97
1 / 2
2
6b .
is the depth at which the discharge is maximum.
Computation of uniform flow depth in CBI&P channel
( )
( )
0
3 1
0
2 2
2
2
2/3 1/2
2/3
2
1
Q = 30m s , S = , m = 1.25, n = 0.015
1600
1 1
A= 2 y Cot + y 2
2 2
= y +Cot
P = 2y +Cot
A y
R= =
P 2
Cot =1.25
= 38.6 0.644 radians
A = 1.894y
1
Q = AR S
n
1.894 y
30 = y 1/1600
0.015 2
=
( )
2
3
8/3
3/8
0.01530 1
y =
2 1.894 1/1600
y = 14.1 = 2.70m
y
1
m
Problems:
1. A trapezoidal channel has a bottom width of 6.00 m, side slopes of 1 to 1, and water
flows upto a depth of 915 mm. For n =0.015, and a discharge of 10.20 m
3
/s, calculate
(a) the normal slope,
(b) the critical slope and critical depth for 10.20 m
3
/s, and
(c) the critical slope at the normal depth of 915mm.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2. A cement lined laboratory channel of rectangular shape is laid at a bed slope of
0.0004. If it is 150 cm wide, what discharge can be expected at uniform depth of 60 cm
in the channel? Take n to be 0.011. [Answer: 0.78 m
3
s
1
]
3. Water flows at a uniform depth of 1.25m in a rectangular channel 4 m wide laid at a
slope of 1 in 1000. Compute the average shear stress on bottom of the channel.
( )
=
f
R S
4. A discharge of 40.0 m
3
/s flows in a trapezoidal channel with bottom width 4.0 m and
side slopes 2 H: 1 V. If the normal depth at a bottom slope of 0.0016 is 2.0 m, determine
the va1ues of n and C. Is the flow sub critical or supercritical? (Answer: n =0.0184,
C=56.22, F<1).
5. A circular pipe of reinforced concrete is to be used as a storm drainage conduit. It has
to carry a discharge of 2.5 m
3
/s at a slope of 1 in 2500, when running 0.9 full, determine
the required size. Assume n =0.013. ((Answer: do =1.81 m)
6. A trapezoidal channel excavated in earth has to carry a discharge of 5.0 m
3
/s at a
velocity of 0.75 m/s. The channel is 5.0 m wide at the base and has side slope of 1:2. At
what slope should it be laid? Take n to be 0.02.
7. The normal depth of flow in a trapezoidal concrete lined channel is 2 m. The channel
bed width is 5 m and has side slopes of 1:2. Manning n is 0.015 and the bed slope is
0.001. Determine the discharge Q, mean velocity, V and Reynolds Number, R
e
.
Classify the flow according to Froude Number.
8. In the previous problem if the discharge is changed to 30 m
3
/s, what would be the
normal depth of flow?
9. During large floods, the water level in the channel shown in figure exceeds the bank
level of 2.5 m. The flood banks are 10m wide and are grassed with side slope of 3: 1 (H:
V). The estimates of Manning n for these flood banks is 0.035. Estimate the discharge
for a maximum flood level of 4m and the velocity coefficient. Draw a stage discharge
relationship curve, given So =0.001, and n for main section as 0.015
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
10 m
5 m
2.5 m
10 m
3
1
2 2
1 1
3
1
10. A concrete pipe 750mm diameter is laid on a gradient of 1:200. The estimated value
of Manning n is 0.012 and pipe full discharge is estimated to be 0.85 m
3
/s.
(a) Calculate the discharge for a proportional depth of 0.938 d
o
using Manning equation.
Explain why the discharge in (a) is larger than the pipe full discharge?
11. A trapezoidal channel having bottom width 6 m and side slope of 2.5 horizontal to 1
vertical is laid on a bottom slope of 0.0025. If it carries a uniform flow of water at the rate
of 10 m
3
/s, compute the normal depth and the mean velocity of flow. Take Manning n as
0.025. What is the value of Froude Number?
12. The figure shows a highway gutter having one side vertica1, one side sloped at 1 on
m.
T
y
y
1
m
1
Manning n, depth of flow y, and longitudinal slope S
o
are given. Express discharge as a
function of side slope, depth, n and bed slope.
+ +
8 / 3
5 /3
2
Answer
Q = ( l / n ) f ( m ) y
0.3142 m
f ( m ) =
m
2/3
(1 1 )
(a) Compute the discharge when n =0.017, y =6.5 cm and S
o
=0.03, m =24.
(b) Compute the discharge when m =24, n =0.015, y =8.00 cm, S
o
=0.04.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
13. Water flows at a velocity of 1 m /s in a rectangular channel 1.0 m wide. The bed
slope is 2 x 10
3
and n =0.015.
Find the depth of flow under uniform flow condition.
14. Find the discharge in a trapezoidal channel with a bed width of 10 m. Side slopes
1:1 and depth of flow of 2.0 m under uniform flow conditions. So =10
4
and n =0.02.
Also find Chezy coefficient at this depth.
15. A sewer pipe is proposed to be laid on a slope of 1 in 2500 and is required to carry
1.5 m
3
/s. What size of a circular pipe should be used if the pipe has to flow half full and
n =0.015?
16. Design an earthen trapezoidal channel for water having a velocity of 0.6 m / s. Side
slope of the channel is 1.5 : 1 and quantity of water flowing is 3 m
3
/s. Assume Chezy
coefficient as 65
1/2 1
m s
.
17. Design a trapezoidal channel for Carrying 30 m3/s of water. Bed slope of the
channel is 1:18,000 and side slope of 2 horizontal to 1 vertical Assume C in Chezy
formula as 50
1/2 1
m s
.
18. A trapezoidal channel has slope 1.5 horizontal to 1 vertical. It is to discharge 20
m
3
/s of water with a grade of 0.5 m per km. Find the depth of channel for its best form.
Use Manning formula taking n =0.01.
19. A circular pipe of 2.5 m diameter is laid at a slope of 1 in 1200. Find the maximum
discharge that can be secured at atmospheric pressure, if the value of C is 50
1/2 1
m s
.
What would be the depth of the flow for this discharge?
20. Determine the hydraulically efficient section of a trapezoidal channel, given n =
0.025, to carry 12.75 m
3
/s.
To prevent scouring, the maximum velocity is to be 920 mm/s and the side slopes of the
trapezoidal channel are 1 vertical to 2 horizontal.
What slope S
0
, of the channel is required?
21. Show a correlation between roughness factor f and roughness factor n.
(a) What is the average shear stress at the sides and bottom of a rectangular flume 3.65
m wide, flowing 1.20 m deep and laid on a slope of 1.60 m / 1000 m?
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
22. What flow can be expected in a 1.20 m wide rectangular cementlined channel laid
on a slope of 4 m in 10,000 m, if the water flows 600 mm deep? Use both Kutter's C and
Manning n.
23. In a hydraulic laboratory, a flow of 0.412 m
3
/s was measured from a rectangular
channel flowing 1.20 m wide and 600 mm deep. If the slope of the channel was
0.00040, what is the roughness factor for the lining of the channel?
24. On what slope should a 600 m long vitrified sewer pipe be laid in order that 0.17
m
3
/s will flow when the sewer is half full? What slope is required if the sewer flows full?
25. A trapezoidal channel, bottom width 6.1 m and side slopes 1 to 1 flows 1220 mm
deep on a slope of 0.0009. For a value of n =0.025, what is the uniform discharge?
26. Two concrete pipes (C =55) must carry the flow from an open channel of half
square section 1.83 m wide and 0.915 m deep ( C =66). The slope of both structures is
0.00090. (a) Determine the diameter of the pipes.
(b) Find the depth of water in the rectangular channel after it has become stabilised, if
the slope is changed to 0.00160, using C =66.
27. An average vitrified sewer pipe is laid on a slope of 0.00020 and is to carry 2.36
m
3
/s when the pipe flows 0.90 ful1. What size pipe should be used?
28. How deep will water flow at the rate of 6.80 m
3
/s in a rectangular channel 6.0 m
wide, laid on a slope of 0.00010? Use n =0.0149.
29. How wide must be rectangular channel be constructed in order to carry 14.15 m
3
/s
at a depth of 1.83 m on a slope of 0.00040? Use n =0.010.
30. A channel with a trapezoidal cross section is to carry 25 m
3
/s. If slope S
0
=
0.000144, n =0.015, base width b =6.0 m and the side slopes are 1 vertical to 1.5
horizontal, determine the normal depth of flow y
n
31. Compute the normal depth in a trapezoidal channel having a bottom width of 10 m
and side slope of 2H to 1V and carrying a flow of 30 m
3
/s. The slope of the channel
bottom is 0.001 and n =0.013.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
19.2 Problem: Irregular Channel
Compute the uniform flow depth for the C.S. when flow is 283 m
3
/s. n =0.024,
S
0
=0.0001
River stage (in m) above
an arbitrary Datum
Distance to first perimeter
intersection from south
Bank
Distance to second perimeter
intersection from south Bank
4.6 100 100
6.1 73 140
7.6 61 160
9.1 52 180
11.0 46 220
12.0 40 260
14.0 34 365
15.0 24 370
17.0 6.1 375
0
60 120 365
Distance from South Bank
River bed
elevation
15
20 (xi, yi)
(x, y)
River bed elevation has a function of the distance (after French)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
19.3 Solution of Algebraic or Transcendental Equation by the
Bisection Method
In the algebraic expression F (x) =0, when a range of values of x is known that contains
only one root, the bisection method is a practical way to obtain it. It is best shown by an
example.
The critical depth in a trapezoidal channel is to be computed for given discharge Q and
the dimensions of the channel. The corresponding equation is
2
3
Q T
1 =0
gA
must be
satisfied by some positive depth y
c
greater than 0 and less than an upper bound say
100 m. T is the top width given by (b +2 my
c
). The interval is bisected and this value of
y
c
tried. If the value of F is positive, as with the solid line shown in figure, then the root is
less than the midpoint and the upper limit is moved to the midpoint and the remaining
half bisected and the procedure is repeated.
F(x)
0
100
Bisection
T
b
m
1
m
1
Trapezoidal
y
Similarly it could be used for obtaining uniform flow depth using the following equation.
2
3
0
2
3
0
2
3
0
1
Q = AR S
n
1
Q  AR S 0
n
nQ
 AR 0
S
=
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
19.4 Solution of Manning Equation by Newton Raphson
Method
There is no general analytical solution to manning equation for determining the flow
depth given the flow rate because the area A and hydraulic radius R may be
complicated functions of the depth. Newton Raphson method can be applied iteratively
to obtain a numerical solution. Suppose that at iteration k the depth y
k
is selected and
the flow rate Q
n
, is computed using manning formula using the area and hydraulic
radius corresponding to y
k
. This Q
k
is compared with actual flow Q
n
; then the objective is
to chose y such that the error.
f (y
k
) =Q
k
 Q
n
is within the tolerance limit. The gradient of f with respect to y is
k
n
2 1
3 2
o k
dQ
df(y )
k
=
dy dy
k k
because Q is constant. Hence, assuming manning roughness is constant,
df 1
= S A R
k
dy n
k
1
2
o
2 1
3 2
o k
k
k k
df
dy
1

2
3
1 2A R dR dA
3
= S +R
n 3 dy dy
k
1 2 dR 1 dA
= S A R +
k
n 3R dy A dy
k
2 dR 1 dA
=Q +
3R dy A dy
in which the subscript k out side the
bracket indicates that the
quantities in the bracket computed for y =y
k.
In Newton's method,
given a choice of y , y is chosen to satisfy
k k+1
0 f (y)
df
k
=
dy y +y
k
k k+1
This
( )
k
y is the value of y ,
k+1
f (y )
k
y =y 
k+1 k
df / dy
k
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Which is the fundamental equation of the Newton's method. Iterations are continued
until there is no significant change in y
n
; this will happen when the error is nearly zero or
an acceptable prescribed tolerance.
Thus for manning equation it may be written as
( )
k
1 Q / Q
k
y = y 
k+1 k
2 dR 1 dA
+
3R dy A dy
For rectangular channel A = b y and R = b y / b +2y where b is the
o o o o
channel width; The quantity in denominator can be for rectangula
( )
( )
2
1
=
=
2 1
3
2
3
2 1
3
d d
dy dy
dA A dP
P dy dy
P
T R dP
P P dy
consider
dR dA
R dy A dy
P T R dP T
A P P dy A
dP
T R
A d
=
+
+
A
R
P
r channel
( )
( )
( )
( ) ( )
2 2
3 3
5 2 1
3 3
For rectangular channel
5 2 1
2
3 3 2
51 4 1
3 3 2
5 2 12 5 10 4
3 2 3 2
5 6
=
3
o
o o
o
o o
o o
o
T
y A
T R dP T
A A dy A
T dP
A P dy
b
b y b y
y b y
b y y b y y
y b y y b y
b y
y b
+
+
+
+ +
=
+ +
+
( )
( )
( )( ) ( )
2
o
y +
o
o
1  Q /Q
k
y = y 
k+1 k
5 b + 6 y
k
3y b + 2 y
k k
Similarly the channel shape function 2/3R dR/dy + 1/A (dA/dy)
for other cross sections can be derived.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
( ) ( )
( )
( )
( )
( )( )
2 2 2
o
2
o o
b 2 6 1 4 1
Trapezoided Channel
3y b b 2 1
8
Triangular Channel
3y
4 2sin 3 5 cos
Circular Conduit
3 sin sin
2
in which
o
my y m my m
my y m
d
+ + + + +
+ + +
+
1
2
2cos 1
o
y
d
=
Example:
Compute the flow depth in a 0.6 m wide rectangular channel having n=0.015, S
0
=
0.025, and Q =0.25 m
3
s
1
.
B
y
Solution:
o
3 1
2 1
3 2
2
1
3
2
Let wide b 0.6
Manning coefficient 0.015
bed slope 0.025
discharge 0.25
normal depth ?
A
Hyraulic mean radius R =
p 2
1
1
2
o
o
k
o
k
m
n
S
Q m s
y
b
b y
Q AR S
n
by
Q b y S
n b y
=
=
=
=
=
=
+
=
=
+
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
( )
( )
( )
( )
( ) ( )
( )
5
1
3
2
2
3
5
1
3
2
2
3
1
2
0.6
1
0.025
0.015
0.6 2
5
53 53
3
0.6 4.4993
10.5409* (1)
2 3 2 3
0.6 2 0.6 2
5 6
Shape function =
3 2
5 0.
=
k
k
k
k
k
o k
k k
by
Q S
n
b y
y
Q
y
y y
k k
Q
k
y y
k k
b y
y b y
=
+
=
+
= =
+ +
+
+
( )
( ) ( ) ( )
( )
( )
k+1
6 6 3 6 1 2
3 0.6 2 3 0.6 2 0.6 2
0.25
1 0.6 2
y (2)
1 2
k k k
k k k k k k
k k
k
k
k
y y y
y y y y y y
y y
Q
y
y
+ + +
= =
+ + +
+
=
+
Iteration (k) 1 2 3
( )
k
y m
0.100 0.1815 0.1727
3 1
Q(ms ) 0.1125 0.2684 0.2488
( )
( )
V
Froude number F =
gy gy
02488 06 01727
1844
9.81*0.1727
super critical flow
Q / A
. / . * .
F .
=
= =
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
20.1 Slope Area Method
Slope area method is used for estimating the flood discharge.
Assumptions made are
(i) High Flood Level (HFL) mark is known.
(ii) Total area is effective in transporting the flow.
(iii) No water falls.
(iv) Long reach.
Also the slopearea approach is justified if the change in conveyance in the reach is less
than 30 percent.
Although a straight, uniform reach is preferred, a contracting reach should be chosen
over an expanding reach if there is a choice.
One or more of the following criteria should be met in determining the reach length:
(a) The length should be greater than or equal to 75 times the mean depth of flow,
(b) The fall of the water surface should be equal to or greater than the velocity head, (If
velocity =1 m/s,
2
V
005 m
2g
. = and if the velocity =2 m/s,
2
V
020 m
2g
. = ) and,
(c) The fall should be equal to or greater than 0.15 m.
When the reach is contracting
( )
V <V , k=1.0
u d
. When the reach is expanding
( )
V >V , k=0.5
u d
. The 50% decrease in the value of k for an expanding reach is
customarily assumed for the recovery of the velocity head due to the expansion of the
flow.
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
20.2 Normal and Critical Slopes
When discharge and roughness are given, the Manning formula can be used for
determining the slope of the prismatic channel in which the flow is uniform at a given
depth y
n
. The slope thus determined is called Normal Slope S
n
.
Now by changing this slope, the normal depth could be made equal to critical uniform
flow for a given Q and n. This slope is called Critical slope S
c
.
The smallest critical slope that sustains a given normal depth is called limiting slope S
L
for a given shape and roughness.
By adjusting the slope and discharge if critical uniform flow is obtained that it is called
Critical slope of normal depth S
cn
.
These definitions will be illustrated in the following examples.
1. Normal and Critical Slopes:
Example: Rectangular open channel has a bottom width of 6.0 m, n =0.02.
a) For y
n
=1.0 m, Q =11 m
3
/s, find normal slope.
b) Find the limiting critical slope and normal depth of flow for Q =11 m
3
/s.
c) Find the critical normal slope given y
n
=1.0 m and determine the discharge for this
depth and slope.
Solution:
( )
( )
( )
2 2
2
2
n
2 4/3 2/3
A 6
a A by 6*1 6m , P b 2y 6 2*1 8m , R 0.75m
P 8
Qn
11*0.02
S 0.001972
A R
6* 0.75
V Q 11
Froude 0.5853
A 6 9.81*1
subcritica
n n
number =
gy gy
= = = = + = + = = = =
= = =
= = =
( ) l Hencemildslope
n c
y y . . >
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
( )
2
2
n c
c n
1/3
c c
V D A by
For critical flow, D y
2g 2 T b
Q
by y
but y y for theuniformcritical flow.
2*9.81 2
11*11
y y 0.7
6*6*9.81
b
= = = =
= =
= =
( )
( )
2
2
2
c
2/3 2/3
c 0
2
n
0
A 6*0.7 4.2m P 6 1.4 7.4
4.2
R 0.57m
7.4
Critical slope:
nQ 0.02*11
S 0.0058
AR
4.2* 0.57
S S
c If y 1.0 A 6m P 8.0m, R 0.75 F 1
m
= = = + =
= =
= = =
>
= = = = =
( )
( )
2/3
1/2
cn
2
cn
2/3
3
V
gy
V 9.81*1 3.1m/s
1
3.1 0.75 S
0.02
3.1*0.02
S 0.00564
0.75
Q 3.1*6 18.6m /s
=
= =
=
= =
= =
Problem: A trapezoidal channel has a bottom width of 6 m, side slopes of 2: 1 (H: V)
and n =0.025.
(a) Determine the normal slope at a normal depth of 1.00 m and the discharge is 11
m
3
/s.
(b) Determine the normal slope and corresponding normal depth when the discharge is
11 m
3
/s.
(c) Determine the critical slope at the normal depth of 1.00 m and calculate the
corresponding Q.
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
( ) ( ) ( )
( )
2
2
2/3
2
2/3 1/2 3
n
2/3
n
c
a A b 2y y 6 2 1 8m
P b 2y 1 m 6 2 5 10.472m
R 0.7639m AR 6.685
1 nQ
Q AR S 1.692*10
n
AR
S 0.001692
V
b S ? F 1
gD
n
S
= + = + =
= + + = + =
= =
= = =
=
= = =
( )
( )
( )
( )
( )
( )
( )
( )
( )
( ) ( )
( )
3/2
3
3
V gD 9.81D
y 6 2y
D A 6 2y y
6 2my
y 6 2y
D
6 4y
y 6 2y
Q 11
V 9.81*
A 6 2y y 6 4y
11 6 4y 6 2y y g
Squaring
121 6 4y g 6 2y y
121 6 4y 9.81
,
=
= =
+
= = +
+
+
=
+
+
= =
+ +
+ = +
+ = +
+ = ( )
3
3
6 2y y
Bytrial anderror
+
( ) ( )
c
2
1 2/3
2
c
2/3
Say y 0.648m
A 6 2 0.648 0.648 4.7278m
P b 2y 5 8.8979m
R 5.313*10 AR 3.1016
nQ
S 0.007861
AR
=
= + =
= + =
= =
= =
R.H.S L.H.S
y =0.65 121 (6+2.6) =1040.6 1048.039
y =0.648 1039.63 1036.689
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
( ) ( )
( )
( )
( )
c
2
1 2/3
2
c
2/3
2
Say y 0.648m
A 6 2 0.648 0.648 4.7278m
P b 2y 5 8.8979m
R 5.313*10 AR 3.1016
nQ
S 0.007861
AR
c
A 6 2 1 8m
P 6 2 5 10.472m
R
Given normal depth = 1 m
=
=
= + =
= + =
= =
= =
= + =
= +
( )
( )
( )
c
2/3
1/2
cn
1/2 2
cn
2/3
1
cn
0.7639m
T b 2my 6 2*2*1 10m
A
D 0.8
T
V gD 9.81*0.8 2.801m/s
1
2.801 * 0.7639 S
0.025
2.801*0.025
S 8.3809*10
0.7639
S
=
= + = + =
= =
= = =
=
= =
/2
3
0.007024
Q 2.801*8 22.408m /s
=
= =
Example:
A trapezoidal channel has a bottom width of 6 m, side slopes of 2:1 and, n =0.025.
(a) Determine the normal slope at a normal depth of 1.2 m when the discharge is
10 m
3
/s.
(b) Determine the critical slope and the corresponding normal depth when the
Q =10 m
3
/s.
(c) Determine the critical slope at y
n
=1.2 m and compute the corresponding Q.
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
Solution:
6
2
1
Trapezoidal
y
( )
( ) ( )
( )
( )
2
2
2 1
3 2
n
2 2
2
n
4/3 4/3
3
n
c
12
A 10.08m V 1.1905m/s,
10.08
P 6 2 1.2 5 11.36m R 0.8868
1
V R S
n
0.025 1.1905
n V
S
R
0.8868
S 1.039*10 0.00104
b Critical depth y ?
Q
Z
= = =
= + = =
=
= =
= =
=
=
( )
n
2
2 2
2/3 1/2 c
c
n n
4/3 2/3
n c
n
12
A D Z 3.8313
g g
Q 12
V 2.3565m/s
A 5.0922
n V 1 0.025*2.3565
V R S S
n
R R
S 0.014718 y 0.69
c y 1.2m R 0.8
= = =
= = =
= = =
= =
= =
2
c
3
A
868 A 10.08m D 0.9333m
T
V gD 3.0259m/s
Thereforethedischarge Area * Velocity 10.08 * 3.0259 30.50m /s
Solve by trial and error.
= = =
= =
= = =
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
( )
c
c
2/3
c
c
Graphical approach
Limit slopeis thesmallest critical slopefor agiven shapeandroughness
a Determine S
Q K S
1
Q AR S
n
Q
Q Z g A D g
g
Forrectangular channel
c
Z or
=
=
= = =
2/3
c
1.5
2/3
1.5
: It can bewritten as
1 by
Q by S
n b y
also
by
Q by g by g
b
Rewriting the equation
Q Q
y or y
b g b g
Substituting the above value in Manning formula for discharge it may b
=
+
= =
= =
2/3
2/3
2/3
c
2/3
2/3
e written as
Q
b
b g
1 Q
Q b S
n b g
Q
b 2
b g
This is an Implicit function and solution is by trial and error approach.
=
+
Q
S
c
Alternatively
( )
( )
2/3
2/3
2/3
2/3
2/3
c
2/3
2/3
b
Q
b g
b 1
Q Q S
n b g
b b g 2Q
=
+
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
( )
2/3
1
2/3
2
*3
3
c
2/3
1 1
1
2/3
2/3
1
2/3
1 c 1
c c c
If b g C
b 1 b 1
Q Q S
nC C
bC 2Q
nC b 1
Q
C b S bC 2Q
Anequation in terms of S is obtained. So chooseQandobtain S , plot QVs S .
=
=
+
=
+
Example 3: Determine the limit slope of rectangular channel of 3 m width and
roughness of 0.02. Consider the following cases
For depths (i) y =0.5 m, (ii) y =2 m.
Do we have limit slopes for these conditions?
Solution:
Section factor for critical flow
( )
3/2 1/ 2 3/2
c
3/2 3/2 3/2
2/3 1/2
c
2
2 3/2
2 2
c
2 4/3 4/3
2
c
Q
Z A D A T by for rectangular channel
g
Q by g 3* 9.81*y 9.3962 y
1
But Q AR S
n
n by g
Q n
S
A R
by
by
b 2y
Simplifying
S g
c
=
=
S
= = = =
= =
= =
+
=
( )
4/3
2 1/3
3
n c
2y
n y 1
b
Case i y 2m S 9.6399 * 10
+
= =
Note: There could be a situation where limit slope is not possible in expected range of
flow depths.
Graphical approach:
Table: To plot Q Vs S
c
y (m) Q (m
3
/s) S
c
0 0 0
0.1 0.2971 0.008807
0.2 0.8405 0.007975
0.3 1.5440 0.007475
0.4 2.3770 0.007298
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
0.5 3.3322 0.007255
0.6 4.3670 0.0072866
0.7 5.5030 0.007364
0.8 6.7234 0.007474
0.9 8.0227 0.0076057
1.0 9.3963 0.007754
1.25 13.1317 0.008173
1.50 17.2621 0.008638
1.75 21.7527 0.009129
2.0 26.5767 0.009639
3.0 48.8245 0.011177
4.0 75.1702 0.01397
Limit slope computations:
Q (m
3
/s) y =0.5 m, S
c
y =2 m, S
c
1 6.5741 * 10
4
1.3646 * 10
5
2 2.6296 * 10
4
5.4585 * 10
5
3 5.9166 * 10
3
1.22817 * 10
4
4 0.0105185 2.1834 * 10
4
5 0.01643 3.4116 * 10
4
6 0.02367 4.9124 * 10
4
8 0.04207 6.6867 * 10
4
10 0.06574 1.10535 * 10
3
15 0.147917 3.0704 * 10
3
20 0.262963 5.4585 * 10
3
30 0.59166 0.0122817
40 1.5018 0.2183
60 2.3667 0.04913
80 4.2074 0.08733
100 6.5741
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
0
0
5
10
15
20
25
30
35
40
45
50
Supercritical flow
Subcritical flow
critical slope
From analytical solutions:
( )
( )
2
2
L
1/3 1/3
4/3
3 3 3/ 2
L
26.16* 0.02
y 0.5 1 n
, S 26.16 0.007255799
b 3.0 6
b 3
S 0.007255, 9.0694*10 y 3 2y ,Q 3 gy
c
S
= = = = =
= = + =
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
Analytical approach for obtaining limit slope:
Rectangular channel:
Consider a rectangular channel of width b and depth of flow y with Manning roughness
coefficient n.
Then
( )
( )
( )
( )
( )
( )
( )
( )
2/3 1/2
c c c c
4/3 4/3 2 2
c c c c
4/3 4/3
4/3
c 2
c c
c
2
4/3
c c
c
4/3
c
4/3
2
c
4/3 1
c
1
V gy V R S
n
n b 2y n b 2y
S gy S gy
by by
b 2y
S n gy
by
By definition of limit slope,
dS n gy d
b 2y
dy dy
by
b 2y d n g
dy b y
= =
+ +
= =
+
=
= +
+
=
( ) ( )
( )
( )
( )
( )
/3
1 4
2 4 4
1
3 3
3 3
c c c c
4/3
4
1 4
1
3
c
3 3
3
c c c
4
1
3
3
c c
4 1
3 3
c
c
1
1
3
c 3
1
c
4
4
3
c 3
4 1
3 3
c
n g 8 1
b 2y y b 2y y 0
3 3 b
b 2y 4
2* b 2y y y
3 3
b 2y y
2*4
b 2y
y
y
1 2 b
b
2*4 y
y
1 2 b
b
y
2*4 1 2
b
+
+ + + =
+
+ =
+
=
+
+
=
+
+ =
( )
1
c
1
c
c
c
c
c c
c
c
2
c c
c
4/3
c
by
y b
2*4 1 2
b y
y b
8 1 2
y b
b b
8 2 6
y y
y 1
S is maximum, when
b 6
Substituting into equation wecan get theexpression of limiting slope
n gy y
S 1 2
b
by
+
.
+ =
= +
= =
=
= +
4/3
4/3
4/3 2
4/3 c
c
4/3 4/3
c
4/3 2
1/3 1/3
c
4/3
2 1/3 4/3 2
L
1/3 4/3 1/3
2 2
L
1/3 1/3
b
n gy 1
S 1 2* b
6 b y
n g 1 1
1
3 b y
b
n 6 *4 n
S 9.81* 26.157
b 3 b
n n g
S 26.157 or 2.67 in which bis in meter
b b
=
.
= +
+
= =
=
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
20.2.1 Froude Criteria for Sub Critical and Super Critical Flow
( )
( )
( )
2/3 1/2
0
0
0 L 0 L
L
2/3
0
2/3
4
2
0
R S V
F
gD n gD
S
S S subcritical, S S super critical, 1critical
S
For rectangular channel.
by
A by, R , D y
b 2y
by S
F
n gy b 2y
If flow is critical uniformflow then
n gy b 2y
S
= =
< > =
= = =
+
=
+
+
=
( )
( )
( )
/3
4/3
4/3
2 1/3
0
4/3
2
L
4/3
4/3 1/3
0
4
1
L
4/3
3
4/3
1/3
0
1/3
L
4/3
0
1/3
L
4/3 1/3
0
L
4
by
n gy b 2y b S
S
2.67n g by
2y
g 1 b b
S b
S
26.16 y b
2y
1 b
S 9.81 b
S 26.16
y
2y
1
S b
2.667
S
y
b
S 2y y
1 2.667
b b S
2y
1
b
+
=
+
=
+
=
+
=
+ =
+
3
0
L
3
4
0
L
3
4
0
L
S y
2.667
b S
S 2y y
1 2.667 0
b b S
S b 2y
1 2.667 0
y b S
=
+ =
+ =
c 0
L
y S
Therearetwo solutions of for 1 and
b S
y 1 y 1
onesolution has a . Theother solution .
b 6 b 6
y
Theflowis super critical between thesetwo values andit is subcritical for all theother values of
b
>
> <
0
L
y
b
S y 1 y 1
If 1; , Henceflowis critical andalsotheFroudenumber will be maximum .
S b 6 b 6
at = = =
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
10.0
.01 .02 .03 .05 .07 .10 .20
.30 .50
.70
1.0
2.0
3.0 5.0 7.0
10.0
0.001
0.002
0.003
0.004
0.006
0.008
0.01
0.02
0.03
0.04
0.06
0.08
0.1
0.2
0.3
0.4
0.6
0.8
1.0
2
3
4
With variation of
1
___
6
__
y
b
__
y
b
with for uniform flow in rectangular channel
__
S0
SL
__
S0
SL
Q
___
b
5/2
g
0
L
0
L
2/3
0
S y
For 1.0for all only subcritical exists.
S b
S y
For 2.0theflowis super critical for most of thepractical rangeof
S b
Maximumvalueof Froudenumber:
S
1 by
Froudenumber F
n b 2y gy
<
>
=
+
2
L
1/3
n
S 26.16
b
=
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
( )
( )
2
2 4/3 2/3
0 2
2 4/3
4/3
4/3
2 0
2 4/3
L
1/3
0
2 4/3
2/3
2
0
1/3 4/3
L
1/6
1/2
0
2/3
L
b y
S 1
F
gy n 2y
b 1
b
S 1 y
F
S n 2y
1
b
1 y
S
n g 2y
1
b
y
26.16
F S
b S g 2y
1
b
y
S 26.16 b
F
9.81 S
2y
1
b
F 1.63
=
=
+
=
+
+
=
+
=
+
=
1/6
1/2
0
2/3
L
0
max
L
0
y
S b
2
S
2y
1
b
dF y 1
0 condition for maximumandoccurs at
dy b 6
S
F
S
Given n, S channel width could beestimated such that theFroudenumber will never exceed
+
= =
=
0
a predetermined valueirrespectiveof thevalueof discharge.
Problem:
Given S 0.0025, n 0.25estimatethewidth of thechannel such that
maximumFroudenumber is 0.5irrespectiveof thedisc
= =
( )
( )
2
L
1/3
2/3 1/2
0
5/3 1/ 2
0
4/3
1/2
0
5/ 2
L
harge.
Sridharan and Lakshmana Rao havepresented thedesign chart for rectangular channel and the
details areas follows
n
S 26.16
b
1
Q AR S
n
by S 1
Q
n
b 2y
y
S Q b
1.632
S b g
=
=
=
+
=
5/3
2/3
0
5/2
L
2y
1
b
S y Q
A design chart is created Vs for different values of for
b S gb
different constant values of Froudenumber.
+
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
Reference:
1. J ones L.E. and Tripathy B.N. "Critical slopes for Trapezoidal channels ", ASCE HY1,
4202, Vol. 91, pp 85  91.
2. Nagar S. Lakshmana Rao and Kalambar Sridharan, "Limit slope in uniform flow
computations", Proceedings ASCE J l. Vol. 96, No. Hy1, J an. 1970, p 7011, pp 95 to
102.
Problem:
1. Show that for Trapezoidal channel that there does not exist any limit slope for when
m
0.5
2. Show that the limit slope for trapezoidal channel is given by the following equation.
( ) ( )
3 2
2 2 2 2 2
y y y
4m m 1 10m 4m m 1 10m 6 m 1 1 0
b b b
inwhich m is the side slope.
+ + + + + + =
3. Show that for circular channel the limit slope is given by
2
L
1
3
0
0
n
S 33.06
d
inwhich d is the diameter of the circleinfeet andthesubtendedangleby thefree
surfaceat thecentrecorresponds to13206'
=
4. Establish that for triangular channel the limit slope will be zero.
5. For trapezoidal channel, show that
( )
4/3
2
1/3
0
2 2 1/3 1/3
y
1 2 m 1
S b b
(i)
F gn
y y y
1 2m 1 m
b b b
y
f m, in which .
b
x
*
S
S
+ +
= =
+ +
= =
Hydraulics Prof. B.S. Thandaveswara Rao
Indian Institute of Technology Madras
( )
( )
( ) ( )
[ ]
( )
( )( )
4
2
3
*
L
2
1 2 1 m
(ii)
1 2m 1 2m
afifth degreeequation in m 0, then 4th degree
Fiveroots: atleast one vereal root, two roots areimaginary.
dS
S 0,
dy
8 1 m 1 m 1 2m 1 2 1
3
*
S
if
+ +
=
+ +
=
+
= =
+ + + +
( )
( )
2 2 2
m 1 10m 10m 0 + + + =
Example:
A Trapezoidal channel with a bottom of 6.2 m and side slope of 0:5; 1, n =0.02 develop
a graph Q Vs S
c
and obtain the limiting critical slope.
Hint:
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )
( )[ ]
( )
( )
2
2
2
2
3
3 2
c
c
3
3
c
Cn
4
2
c 2
c c c
c
V D
Critical flow
2g 2
A 6.2 0.5y y
P 6.2 2*0.5y y
6.2 0.5y y
R
6.2 2*0.5y y
Q
V
6.2 0.5y y
6.2 0.5y y Q 1
*
2g 2 6.2 y
6.2 0.5y y
6.2 0.5y y g Q
g 6.2 y
6.2 0.5y y g
S
6.2 0.5y
6.2 y 6.2 0.5y y
6.2 2*0.5y
=
= +
= +
+
=
+
=
+
+
=
+
+
+
=
+
+
=
+
+ +
+
/3
c Cn
Select different values of y andcalculateS andQ
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
21.1 Design of Canals
Many procedures have been developed over the years for the hydraulic design of open
channel sections. The complexity of these procedures vary according to flow conditions
as well as the level of assumption implied while developing the given equation. The
Chezy equation is one of the procedures that was developed by a French engineer in
1768 (Henderson, 1966). The development of this equation was based on the
dimensional analysis of the friction equation under the assumption that the condition of
flow is uniform. A more practical procedure was presented in 1889 by the Irish engineer
Robert Manning (Chow, 1959). The Manning equation has proved to be very reliable in
practice.
The Manning equation invokes the determination of flow velocity based on the slope of
channel bed, surface roughness of the channel, crosssectional area of flow, and wetted
perimeter of flow. Using this equation, the solution procedures are direct for
determination of flow velocity, slope of channel bed, and surface roughness. However,
the solution for any unknown related to the crosssectional area of flow and wetted
perimeter involves the implementation of an implicit recursive solution procedure which
cannot be achieved analytically. Many implicit solution procedures such as the Newton
Raphson, RegulaFalsi (false position), secant, and the Van WijngaardenDekkerBrent
Methods (Press et al., 1992).
One of the important topics in the area of Free surface flows is the design of channels
capable of transporting water between two locations in a safe, cost  effective manner.
Even though economics, safety, and aesthetics must always be considered, in this unit
thrust is given only to the hydraulic aspects of channel design. For that discussion is
confined to the design of channels for uniform flow. The two types of channels
considered are
(1) lined or nonerodible;
(2) unlined, earthen, or erodible.
There are some basic issues common to both the types and are presented in the
following paragraphs.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1. Shape of the cross section of the canal.
2. Side slope of the canal.
3. Longitudinal bed slope.
4. Permissible velocities  Maximum and Minimum.
5. Roughness coefficient.
6. Free board.
1. Shape of cross section
From the Manning and Chezy equation, it is obvious that the conveyance of a channel
increases as the hydraulic radius increases or as the wetted perimeter decreases. Thus,
there is among all channel cross sections of a specified geometric shape and ares an
optimum set of dimensions for that shape from the viewpoint of hydraulics. Among all
possible channel cross sections, the hydraulically efficient section is a semicircle since,
for a given area, it has the minimum wetted perimeter. The proportions of the
hydraulically efficient section of a specified geometric shape can be easily derived. The
geometric elements of these sections are summarized in Table. It should be noted that ,
the hydraulically efficient section is not necessarily the most economic section.
In practice the following factors are to be kept in mind:
a. The hydraulically efficient section minimizes the area required to convey a specified
discharge. however, the area which required to be excavated to achieve the flow area
required by the hydraulically efficient section may be much larger if one considers the
removal of the over burden.
b. It may not be possible to construct a hydraulically efficient stable section in the
available natural condition. If the channel is to be lined, the cost of the lining may be
comparable with the cost of excavation.
c. The cost of excavation depends on the amount of material that is to removed, in
addition to. Further Topography of the land access to the site also influence the cost of
disposal of the material removed.
d. The slope of the channel bed must be considered also as a variable since it is not
necessarily completely defined by topographic consideration. For example, a reduced
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
channel slope may require a larger flow area to convey the flow, on the other hand the
cost of excavation of the overburden may be reduced.
2. Side slopes
The side slopes of a channel depend primarily on the engineering properties of the
material through which the channel is excavated. From a practical viewpoint, the
side slopes should be suitable for prelimianary purposes. However, in deep cuts, side
slopes are often steeper above the water surface than they would be in an irrigation
canal excavated in the same material.In many cases, side slopes are determined by the
economics of construction. In this regard following observations are made:
a. In many unlined earthen canals, side slopes are usually 1.5 : 1; However,
side slopes as steep as 1:1 have been used when the channel runs through cohesive
materials.
b. In lined canals, the side slopes are generally steeper than in an unlined canal. If
concrete is the lining material, side slopes greater than 1 : 1 usually require the use of
forms, and with side slopes greater than 0 .75 : 1 the linings must be designed to
withstand earth pressures. Some types of lining require side slopes as flat as those
used for unlined channels.
c. Side slopes through cuts in rock can be vertical if this is desirable.
Table: Suitable side slopes for channels built in various types of materials (chow, 1959)
Material Side slope
Rock Nearly vertical
Muck and peat soils 1 / 4 : 1
Stiff clay or earth with concrete lining 1 / 2 : 1 to 1 : 1
Earth with stone lining or each for large channels 1 : 1
Firm clay or earth for small ditches 1 1/2 : 1
Loose,sandy earth 2 : 1
Sandy loam or porous clay 3 : 1
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Indian standards for canal in cutting and embankment
Side slope (Horizontal to Vertical m:1)
Material (soil) Cutting Embankment
Hard clay or gravel 0.75 : 1 1.5 to 1.0
Soft Clay and alluvial
soils
1.0 to 1.0 2.0 to 1.0
Sandy loam 1.5 to 1.0 2.0 to 1.0
Light sand 2.0 to 1.0 2.0 to 1.0 to 3.0 to 1.0
Soft rock 0.25 to 1.0 to 0.5 to 1.0 
Hard rock 0.125 to 1 to 0.25 to 1.0 
3. Longitudinal slope
The longitudinal slope of the channel is influenced by topography, the head required to
carry the design flow, and the purpose of the channel. For example, in a hydroelectric
power canal, a high head at the point of delivery is desirable, and a minimum
longitudinal channel slope should be used. The slopes adopted in the irrigation channel
should be as minimum as possible inorder to achieve the highest command. Generally,
the slopes vary from 1 : 4000 to 1 : 20000 in canal. However, the longitudinal slopes in
the natural river may be very steep (1/10).
Slope of the channels in Western Ghats
Gentle slope 10 m / km S
0
=0.01
Moderate
slope
10 to 20 m /
km
S
0
= 0.01 to
0.02
Steep slope 20 m / km S
0
0.02
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0.1
0.05
0.02
0.01
0.005
0.002
0.001
0.0005
0.0002
0.0001
Median (d50)
Grain Size in mm
F = 0.85
F = 1.0
Bank Full Discharge, m
3
/s
4. Permissible Velocities: Minimum and Maximum
It may be noted that canals carrying water with higher velocities may scour the bed and
the sides of the channel leading to the collapse of the canal. On the other hand the
weeds and plants grow in the channel when the nutrients are available in the water.
Therefore, the minimum permissible velocity should not allow the growth of vegetation
such as weed, hycinth as well you should not be permitting the settlement of suspended
material (non silting velocity). The designer should look into these aspects before
finalizing the minimum permissible velocity.
"Minimum permissible velocity" refers to the smallest velocity which will prevent both
sedimentation and vegetative growth in general. an average velocity of
(0.60 to 0.90 m/s) will prevent sedimentation when the silt load of the flow is low.
A velocity of 0.75 m /s is usually sufficient to prevent the growth of vegetation which
significantly affects the conveyance of the channel. It should be noted that these values
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
are only general guidelines. Maximum permissible velocities entirely depend on the
material that is used and the bed slope of the channel. For example: in case of chutes,
spillways the velocity may reach as high as 25 m/s. As the dam heights are increasing
the expected velocities of the flows are also increasing and it can reach as high as 70
m/s in exceptional cases. Thus, when one refers to maximum permissible velocity, it is
for the normal canals built for irrigation purposes and Power canals in which the energy
loss must be minimised. Hence, following table gives the maximum permissible velocity
for some selected materials.
Maximum permissible velocities and n values for different materials
material V (m / s) n
Fine sand 0.5 0.020
vertical Sandy loam 0.58 0.020
Silt loam 0.67 0.020
Firm loam 0.83 0.020
Stiff clay 1.25 0.025
Fine gravel 0.83 0.020
Coarse gravel 1.33 0.025
Gravel 1.2
Disintegrated Rock 1.5
Hard Rock 4.0
Brick masonry with cement pointing 2.5
Brick masonry with cement plaster 4.0
Concrete 6.0
Steel lining 10.0
5. Resistance to the flow
In a given channel the rate of flow is inversely proportional to the surface roughness.
The recommended values for a different types of lining are given below:
Manning roughness for the design of several types of linings is as follows
Surface Characteristics Value of n
Concrete with surface as indicated below
(a) Trowel finish 0.012  0.014
(b) Flat finish 0.013  0.015
(c) Float finish some gravel on bottom 0.015  0.017
(d) Gunite, good section 0.016  0.017
Concrete bottom float finished sides as indicated below
(a) Dressed stone in mortar 0.015  0.017
(b) Random stone in mortar 0.017  0.020
(c) Cement rubble masonry plastered 0.016  0.020
Brick lining 0.014  0.017
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Asphalt lining
(a) Smooth 0.013
(b) Rough 0.016
Concrete lined excavated rock with
(a) Good section 0.017  0.020
(b) Irregular section 0.022  0.027
These values should, however, be adopted only where the channel has flushing
velocity. In case the channel has nonflushing velocity the value of n may increase due
to deposition of silt in coarse of time and should in such cases be taken as that for
earthen channel. The actual value of n in Manning formula evaluated on the basis of
observations taken on Yamuna Power Channel in November 1971 ranged between
0.0175 and 0.0229 at km 0.60 and between 0.0164 and 0.0175 at km 2.05. The higher
value of n evaluated at km 0.60 could be attributed to the deposition of silt in head
reaches of the channel.
Table: Manning Roughness Coefficients
nvalue different depth ranges
Depth ranges
Lining
Category
Lining Type
0 15 cm 15 60 cm >60 cm
Concrete 0.015 0.013 0.013
Grouted Riprap 0.040 0.030 0.028
Stone Masonry 0.042 0.032 0.030
Soil Cement 0.025 0.022 0.020
Rigid
Asphalt 0.018 0.016 0.016
Bare Soil 0.023 0.020 0.020
Unlined
Rock Cut 0.045 0.035 0.025
Woven Paper Net 0.016 0.015 0.015
J ute Net 0.028 0.022 0.019
Fiberglass Roving 0.028 0.021 0.019
Straw with Net 0.065 0.033 0.025
Cured Wood Mat 0.066 0.035 0.028
Temporary
Synthetic Mat 0.036 0.025 0.021
2.5cm (d
50
) 0.044 0.033 0.030
Gravel
Riprap
5 cm (d
50
) 0.066 0.041 0.034
15cm (d
50
) 0.104 0.069 0.035
Rock
Riprap
30cm (d
50
)  0.078 0.040
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6. Freeboard
The term freeboard refers to the vertical distance between either the top of the channel
or the top of the channel is carrying the design flow at normal depth. The purpose of
freeboard is to prevent the overtopping of either the lining or the top of the channel
fluctuations in the water surface caused by
(1) wind  driven waves,
(2) tidal action,
(3) hydraulic jumps,
(4) superelevation of the water surface as the flow goes round curves at high velocities,
(5) the interception of storm runoff by the channel,
(6) the occurrence of greater than design depths of flow caused by canal sedimentation
or an increased coefficient of friction, or
(7) temporary misoperation of the canal system.
There is no universally accepted role for the determination of free board since, waves,
unsteady flow condition, curves etc., influence the free board. Free boards varying from
less than 5% to 30% of the depth are commonly used in design. In semicircular
channels, when the velocities are less than 0.8 times the critical velocity then 6% of the
diameter as free board have been proved to be adequate.
The freeboard associated with channel linings and the absolute top of the canal above
the water surface can be estimated from the empirical curves. In general, those curves
apply to a channel lined with either a hard surface, a membrane, or compacted earth
with a low coefficient of permeability. For unlined channels, freeboard generally ranges
from 0.3m for small laterals with shallow depths of flow to 1.2m for channels carrying 85
m
3
/s at relatively large depths of flow. A prelimimary estimate of freeboard for an
unlined channel can be obtained from USBR formula.
B
B
1/ 2
3
F Cy
in which F
C is a coefficient. However, it may be noted that C has dimensions of L .
C var ies from 1.5 at Q 0.57 m / s to
is the freeboard in feet, y is the design depth of flow in feet,
=
=
3
2.5 for canal
capacity equal to and more than 85 m / s.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The free board recommended by USBR for channels are given below
Q m
3
/s Free board F
B
in m
<0.75 0.45
0.75  1.5 0.60
1.5  85.0 0.75
>85 0.90
The free board (measured from full supply level to the top of lining) depends upon the
size of canal, velocity of water, curvature of alignment, wind and wave action and
method of operation. The normal free board is 15 cm for small canals and may range up
to 1.0 m for large canals. The U.S.B.R. practice for the minimum permissible free board
for various sizes of canal is given in Figure. Indian Standard IS : 4745 recommends a
free board of 0.75 m for canal carrying a discharge of more than 10 m
3
/sec.
Free board as per Indian Standards (IS 4745  1968), (IS 7112  1973)
Free board (m) Discharge Q (m
3
/s)
Unlined Lined
<10.0 0.50 0.60
>10.0 0.75 0.75
HEIGHT OF BANK ABOVE W.S
HEIGHT OF HARD
SURFACE OR BURIED
MEMBRANE LINING
ABOVE W.S.
HEIGHT OF EARTH
LINING ABOVE W.S
DISCHARGE CAPACITY IN m
3
/s
Bank height for canals and free board for hard
surface or buried membrane and, earth lining
1
2
0.1
.2 .4 .6 .8
1
2 4 6 8
10
2 4 6 8
1000
0
1
100
2 4 6 8
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Free boards provided in some of the major lined canals in India are given below
Sl.No. Name of Canal Free Board F
B
in m
1 Yamuna Power Channel 0.75
2 Nangal Hydel Channel 0.76
3 Gandak Canal 0.45
4 Lower Ganga Canal (Link Canal) 0.30
5 Rajasthan Feeder Channel 0.76
6 Tungabhadra Canal 0.30
7 Mannaru Canal 0.30
8 Sunder Nagar Hydel Channel 0.91
9 Sarda Sahayak Feeder Channel 1.25
Actually adopted Free board for different ranges of discharge in India are below
Q (m
3
/s) <0.15 0.15  0.75 0.75  1.50 1.50  9.00 >9.00
Free board
(m)
0.30 0.45 0.60 0.75 0.90
References
1. IS: 4745  1968, Code of practice for Design of Cross Section for Lined Canals,
Indian Standards Institution, New Delhi, 1968.
2. IS: 7112  1973, Criteria for Design of Cross Section for Unlined Canals in Alluvial
Soil, Indian Standards Institution, New Delhi, 1974.
When flow moves around a curve, a rise in the water surface occurs at the outer bank
with a corresponding lowering of the water surface at the inner bank. In the design of a
channel, it is important that this difference in water levels be estimated. If all the flow is
assumed to move around the curve at the subcritical average velocity , then super
elevation is given by
2
mb
max
c
V 2T
y =
2g r
In India, the minimum radii of curvature are often longer than those used in the United
States. For example, Some Indian engineers recommend a minimum radius of 91m for
canals carrying more than 85 m
3
/s ( Houk, 1956 ). Suggested radii for different
discharges are given in table below.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Radius of curves for lined canals
Discharge (m
3
/s) Radius (minimum) in m
280 and above 900
Less than 280 to 200 760
Less than 200 to 140 600
Less than 140 to 70 450
Less than 70 to 40 300
Note: Where the above radii cannot be provided, proper super elevation in bed shall be
provided.
The width of the banks along a canal are usually governed by a number of
considerations which include the size of the need for maintenance roads. Where roads
are needed, the top widths for both lined and unlined canals are designed so that
precipitation will not fall in to the canal water and, to keep percolating water below the
ground level beyond the banks.
21.1.1 Hydraulically Efficient Channel
It is well known that the conveyance of a channel section increases with increases in
the hydraulic radius or with decrease in the wetted perimeter. Therefore, from the point
of hydraulic aspects, the channel section having the least wetted perimeter for a given
area has the maximum conveyance; such a section is known as the Hydraulically
efficient channel. But this is popularily referred as Best Hydraulic section. The semicircle
has the least perimeter among all sections with the same area; hence it is the most
hydraulically efficient of all sections.
The geometric elements of six best hydraulic section are given in Table. It may be noted
that it may not be possible to implement in the field due to difficulties in construction and
use of different materials. In general, a channel section should be designed for the best
hydraulic efficiency but should be modified for practicability. From a practical point of
view, it should be noted that a best hydraulic section is the section that gives the
minimum area of flow for a given discharge but it need not be the minimum excavation.
The section of minimum excavation is possible only if the water surface is at the level of
the top of the bank. When the water surface is below the bank top of the bank (which is
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
very common in practice), channels smaller than those of the best hydraulic section will
give minimum excavation. If the water surface overtops the banks and these are even
with the ground level, wider channels will provide minimum excavation. Generally,
hydraulically efficient channel is adopted for lined canals. It may also be noted that
hydraulically efficient channel need not be economical channel (least cost).
Geometric elements of best hydraulically efficient section (figures)
Cross
Section
A P R T D
Z = A D
Rectangular 2y
2
4y 0.5 y 2y y 2y
2.5
Trapezoidal
2
3y
( )
2
1 732 . y
2 3y
(3.464y)
0.5 y
4 3
3
y
(2.3094y)
3
4
y
(0.75y)
2 5
3
2
.
y ( )
2 5
1 5
.
. y
Triangular
2
y
2 2y
2.828y
2
4
y
0.3535y
2y
2
y
0.5y
2 5
2
2
.
y
2 5
0 707
.
. y
Semi
Circular
2
2
y
y
0 5 . y 2y
4
y
2 5
4
.
y
2 5
0 25
.
. y
Parabola
2
4
2 y
3
2
4
2 y
3
2
1 89 . y
8
2
3
y
3.77y
y/2
0.5y
2 2y
2.83y
2
y
3
0.667y
2.5
8
3y
9
2 5
1 5396
.
. y
Hydrostatic
Catenary
1.40 y
2
2.98 y 0.468 y 1.917 y 0.728y 1.91 y
2.5
** Hydrostatic Caternary (Linteria)
Flexible sheet: Filled with water upto rim, and held firmly at the top ends without any
effect of fixation on shape. Shape assumed under self height of water is called
Hydrostatic Catenary.
21.1.2 Selection of Lining
Introduction
The need for lining channels in alluvium has long been identified to conserve every bit
of water for more and more utilisation. Lining of an irrigation channel is restored to
achieve all or some of the following objectives keeping in view the overall economy of
the project.
The major advantages of rigid impermeable linings are as follows:
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
(a) Reduction of seepage losses resulting in a saving of water which can be utilised for
additional irrigation.
(b) Prevention of water logging by reducing seepage to watertable.
(c) Reduction in area of crosssection (and there by saving in land) due to increase in
permissible velocity by reduction in the value of rugosity and availing of steeper slope,
where available. Minimize excavation costs
(d) Improvement of discharging capacity of existing channels.
(e) Improvement of operational efficiency.
(f) Prevention of weed growth.
(g) Reduction of maintenance cost.
(h) Long economic life
(i) Insure Cross section stability from scour, low flow conditions etc.
Canal Lining
The lining commonly adopted for irrigation channels can be classified into three groups
1. Rigidimpermeable Lining,
2. Flexible and Permeable Permanent Linings and
3. Flexible Temporary Linings.
Example for the same are indicated in the box.
Rigid Impermeable Linings
Rubble Masonry
Castinplace Concrete
Grouted Riprap or Grouted
Precast Concrete
Soil Cement
Flexible and Permeable Permanent Linings
Riprap or Stone Blocks
Gabions
Interlocking Precast Concrete
Interlocking Synthetic Units
Vegetation and Grasses
Flexible Temporary
Bare Soil
Straw with Netting
Hemp or Jute Mats
Synthetic Matting
Canal Lining
There are different types of lining like Cement Concrete, Shotcrete, Soil cement,
Asphaltic Concrete, etc.
Advantages of Flexible and Permeable Linings:Lining easily fits to cross section shape.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Allows infiltration into channel bed, hence loss of water. Partial failure can occur and still
can resist erosion.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
21.1.3 Design of Lined Channels
Lined channels are built for five primary reasons:
1. To permit the transmission of water at high velocities through areas of deep or
difficult excavation in a cost  effective fashion.
2. To permit the transmission of water at high velocity at a reduced construction cost.
3. To decrease canal seepage, thus conserving water and reducing the waterlogging of
lands adjacent to the canal.
4. To reduce the annual cost of operation and maintenance.
5. To ensure the stability of the channel section.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The design of lined channels from the view point of hydraulic engineering is a rather
elementary process which generally consists of proportioning an assumed channel
cross section. Details of some typical cross section of lined channels used on irrigation
projects in the India are given elsewhere. A recommended procedure for proportioning a
lined section is summarized in table given below. In this table, it is assumed that the
design flow Q
D
, the longitudinal slope of the channel S
0
, the type of channel cross
section e.g., trapezoidal, and the lining material have all been selected prior to the
initiation of the channel design process.
Step Process
1 Estimate n or C for specified lining material and S
0
2
Compute the value of section factor
( )
2/3 1/2
o o
AR nQ/ S or AR Q/ C S = =
3
Solve section factor equation for yn given appropriate expressions for A
and R ( Table ) Note: This step may be required with assumptions
regarding side slopes, bottom widths, etc. (As a thumb rule for quick
computation y can be taken as 0.5 A and for trapezoidal section it can be
shown as
b
4 m
y
= . In India, y for the trapezoidal channel can be taken as
0.577 A which corresponds to
b
3 m
y
= for earth canals).
4 If hydraulically efficient section is required, then the standard geometric
characteristics (click) are used and yn is to be computed.
5
Check for
1. Minimum permissible velocity if water carries silt and for vegetation
(Check whether the velocity is adequate to prevent sedimentation (V=0.6
to 0.9 m / s). Check whether velocity is adequate to prevent vegetation
growth (V =0.75 m/s)).
2. Froude number
(Check Froude number and other velocity constraints such as ( for non
reinforced concrete linings V 2.1 m/s and Froude number . For
reinforced linings )).
0.8
V 5.5 m/s
Generally, Froude number should be as small as possible for Irrigation
canals and should be less than 0.35. Higher Froude numbers is permitted
as in the case of super critical flows such as in chutes, flumes. Decide the
dimensions based on practicability.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
6 Estimate
1. Required height of lining above water surface,
2. Required freeboard, Figure.
Balance excavations costs, costs of channel lining and assess the needs to
modify "Hydraulically efficient section".
7 Summarize the results with dimensioned sketch.
Example of Rigid Lined Channel Design: Design a concrete lined channel (rough finish
n =0.015) to carry 20 m
3
/s on a slope of 0.0015. Consider the hydraulically efficient
trapezoidal shape.
Solution
For hydraulically efficient trapezoidal channel
( )
2
2/3 1 2
0
2
1
3
2
2
8
3
1 73 3 46
2
0 015
1
Q = AR S
n
1
20 1 73 0 0015
0 015 2
7 107
2 086
/
y
A . y , P . y, R
n . ,
y
. y .
.
y .
y . m
= = =
=
=
=
=
For Trapezoidal channel width is given by
( )
o
2
b = y,
3
b =1.15y = 2.409 m
3
m = 0.5773 i.e., = 60
3
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
2
Q 20
Velocity = = = 2.656 m/s
A
1.73y
A 1.73y
Hydraulic mean depth D = = = 0.749y = 1.563 m
4
T
y
3
3
V
Froude Number = 0 678
gD
Freeboard for discharge Q = 20m /s is 0.75 mto nearest convenient elevation.
Freeboard may be modified to 0.764 m.
Hence, the total depth of the channel 2.086 + 0.764 = 2.850 m
. =
Hence the total depth of the channel is 2.850 m. The designed cross section is shown in
the figure.
0.764 m
2.086
0.58
1.0
60
o
2.85 m
Free board =
b = 2.4 m
Design a trapezoidal channel to carry Q =20.25 m
3
/s, V =1.5 m
3
/s, n =0.025, S
0
=
0.0016, side slope m =2. Assume a bed width of 6 m.
Solution
Step 1: Q, n, S
0
and m have been given
( )
( )
2
2
2 1
2 1
A b my
P b y m
b my
R
b y m
= +
= + +
+
=
+ +
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
2/ 3
0
2
nQ 20.25
AR 0.025* 12.656
S 0.0016
Discharge 20.25
Area = 13.5m
Velocity 1.5
13.5 6 2y y
Solving for y, weget y 1.5 m
b
4
y
Add a free board of
0.75 m.
= = =
= =
= +
=
=
Designed channel is shown in figure.
y = 1.5
b = 6 m
F
b
=0.75
2
1
21.1.4 Design of Stable Unlined Channels
Erodible Channels which Scour but do not silt. The behaviour of flow in erodible
channels is influenced by several parameters and precise knowledge is not available on
various aspects. Unlined channels with channel bed and banks composed of earth,
sand or gravel must be designed so that they maintain a stable configuration. There are
three procedures.
1. Velocity based Method of maximum permissible velocity.
2. Regime Theory  Emprical equations for channels with equilibrium sediment
throughput ("Live  Bed" equations).
3. Shear Based  Tractive force methods, Shield analysis.
Method of maximum permissible velocity also known as non erodible velocity:
It is the highest mean velocity that will cause no erosion in the channel body.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
When compared with the design process typically used for lined channels, the design of
stable, unlined or erodible, earthen channels is a complex process involving numerous
parameters, most of which cannot be accurately quantified. The complexity of the
erodible channel design process results from the fact that in such channels stability is
dependent not only on hydraulic parameters but also on the properties of the material
which composes the bed and sides of the channel.
A stable channel section is one in which neither objectionable scour nor deposition
occurs. There are three types of unstable sections: (USBR).
The pioneering work of Fortier and Scobey ( 1926 ) was the basis of channel design.
1. The banks and bed of the channel are scoured but no deposition occurs.
Example: When the channel conveys sediment free water (or water with only a very
small amount of sediment) but with adequate energy to erode the channel.
2. Unstable channel with deposition but no scour.
Example: When the water being conveyed carries a large sediment load at a velocity
that permits sedimentation.
3. Unstable channel with both scour and deposition occur.
Example: When the material through which the channel is excavated is susceptible to
erosion and the water being conveyed carries a significant sediment load.
These types of channels can be designed using the method of maximum permissible
velocity.
The following important points are to be noted.
1. First, the maximum permissible velocity is recommended for canals with a sinuous
alignment.
2. Second, these data are for depths of flow less than 0.91 m . For greater depths of
flow, the maximum permissible velocity should be increased by 0.15 m/s.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3. Third, the velocity of the in canals carrying abrasives, such as basalt raveling, should
be reduced by 0.15 m /s.
4. Fourth, channels diverting water from silt  laden river such as Ganga River should be
designed for mean design velocities 0.3 to 0.61 m/s greater than would be allowed for
the same perimeter material if the water were transporting no sediment.
USSR Data
Sand Gravel Pebbles
Silt
F
M C F M C F M C L
Average particle size, mm
U.S. and U.S.S.R. data on permissible velocities for noncohesive soils.
Legend: V.F.  very fine; F fine; Mmedium; Ccoarse; Llarge
U.S. Dept. Agriculture, Bureau of Soils Classification
Clay Silt
Sand Gravel
V.F. F M C F M
0.1
0.2
0.3
0.4
0.5
0.6
0.8
1
2
3
4
5
6
8
10
20
30
40
50
60
80
100
0.005 0.01 0.02 0.05 0.1 0.2 0.5 5 10 20 50 100 200 2 1
L
U.S. standard mesh sieve sizes
Following Steps are used for Designing
Given a particular soil type, the channel is designed so that the design velocity does not
exceed V
max
for that soil and the channel side walls are with appropriate side slopes.
General guidelines: Froude number should be less than 0.35
Step 1: For the given kind of material estimate the roughness coefficient n, side slope
m, and the maximum permissible velocity.
Step 2: Hydraulic mean radius is computed by using Manning formula.
Step 3: Area of flow is obtained using continuity equation.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Step 4: The wetted perimeter is computed using the information obtained in steps 2
and 3.
Step 5: Solve simultaneously for b and y.
Step 6: Add a proper free board. Modify the section for practicality.
Example
A trapezoidal channel with bottom width of 6m, side slopes of 3H:1V carries a flow of 50
m
3
s
1
on a channel slope, So of 0.0015. The uniform flow of depth for the channel is 1.3
m with n =0.025. This channel is to be excavated in stiff clay. Check whether the
channel will be susceptible to erosion or not.
y =1.3 m
b = 6 m
1
3
( ) ( )
1
A= b+my y = 6+ 3*1.3 *1.3 = 12.87 Sq.m
Q 20
V= = =1.554 m s
A 12.87
which is higher than the permissible velocity (of V =1.25 ms
1
)
From graph
( )
( )
o
S 0 0015 0 0065 0 65
Side slope adopted 3:1 which is < 1 : 1
. . . % = <
Suggestion : Increase width, b, to reduce velocity:
( )
2
n
For b = 8.4 m, y = 1.3 m Corresponding area of flow A = 8.4+3 *1.3 = 15.99 m
Q 20
V = 1.251 m/s which is equal to the permissible velocity
A 15 99 .
= =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
21.1.5 Method of Tractive Force
However, a design methodology based primarily on experience and observation rather
than physical principles. The first step in developing a rational design process for
unlined, stable, earthen channels is to examine the forces which cause scour. Scour on
the perimeter of a channel occurs when the particles on the perimeter are subjected to
forces of sufficient magnitude to cause particle movement. when a partical rests on the
level bottom of a channel, the force acting to cause movement is the result on the flow
of water past the particle. A particle rests on the slope side of a channel is acted on not
only by the flow  generated forces, but also by a gravitational component which tends
to make the particle roll or slide down the slope. If the resultant of those two forces is
larger than the forces resisting movement, gravity, and cohesion, then erosion of the
channel perimeter occurs. By definition, the tractive force is the force acting on the
partical composing the perimeter of the channel and is the result of the flow of water
past these particles. In parctice, the tractive force is not the force acting on a single
particle, but the force exerted over a certain area of the channel perimeter. This concept
was first stated by duBoys( 1879 ) and restated by Lane ( 1955 ).
In most channels, the tractive force is not uniformly distributed over the perimeter.
0.750ySo
0.750ySo
0.970ySo
y
4y
Side Slope, m: 1 = 1.5 : 1
Tractive force distribution obtained using membrane analogy
This distribution varies depending on the cross section and material
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Oslon  Florey
Cruff
Simon
Normal
Preston Tube
0
0.9 m
0.3 m
Boundary shear distribution,
Central Water Power Research Station
(August, 1968)
Discharge: 0.06 and 0.11 m
3
/s
Cross section of the flume: 0.9 m wide , 0.3 m deep
Normal's Method: Based on the concept of zero momentum
Simon's Method: Based on the following equation assuming Karmann constant to be 0.4
2
2 1
0
2
1
u u
y 2.3
log
k y
=
Cruff's Method: Uses the above equation but k value is obtained from velocity profiles.
Oslon and Florey Method: Membrane analogy.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Prototype Rough
Trapezoidal Boundary
Bed
0 5
10 15 20 25 30
0.5
1.0
1.5
2.0
b/y or b/R
max
RS
______
Prototype Rough Trapezoidal
Boundary (Left)
Prototype Rough Trapezoidal
Boundary (Right)
0 5
10 15 20 25 30
Maximum shear on bed and sides for alluvial channel
based on Normal's Method.
(U.P. Irrigation Research Institute Roorkee,
Annual Review, 1971)
0.5
1.0
1.5
2.0
Sides
b/y or b/R
max
RS
______
max
RS
______
= 1.2 for bed and 0.6 for the sides when exceeds 10
b
__
y
when compared to Lanes values of 0.98, 0.78 respectively
The maximum net tractive force on the sides and bottoms of various channels as
determined by mathematical studies are shown as a function of the ratio of the bottom
width to the depth of flow. It may be noted that for the trapezoidal section, the maximum
tractive force on the bottom is approximately
0 0
ys ys and on the sides 0.76 .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The figures show the maximum unit tractive forces in terms of
0
ys for different
b
y
ratios.
Maximum unit tractive forces in terms of
yS
On sides of channels On bottoms of channels
b/y
b/y
Rectangle
Trapezoidal, m = 1
Trapezoidal, m = 2
Trapezoidal, m = 1.5
Trapezoidal,
m = 2 and 1.5
Rectangle
0 1 3 2 4 5 6 7 8 9 10
0 1 3 2 4 5 6 7
8 9 10
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0
When a particle on the perimeter of a channel is in a state of impending motion, the
forces acting to cause motion are in equilibrium with the forces resisting motion. A
particle on the level bottom of a channel is subject to the unit tractive force on a level
surface and effective area.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0.1 0.2 0.3 0.4 0.6 0.8 1.0 2.0 3.0 4.0
20
22
24
26
28
30
32
34
36
38
40
42
Particles size in inches (1 inch = 25.4 mm)
Angles repose for non cohesive material
inches
mm
In the above figure the particle size is the diameter of the particle of which 25 percent of
all the particals, measured by weight, are larger.
Lane ( 1955 ) also recognized that sinuous canals scour more easily than canals with
straight alignments. To account for this observation in the tractive force design method,
Lane developed the following definitions.
Straight canals have straight or slightly curved alinments and are typical of canals built
in flat plains.Slightly undulating topography.
Moderately sinuous canals have a degree of curvature which is typical of moderately
rolling topography.
Very sinuous canals have a degree of curvature which is typical of canals in foothills or
mountainous topography. Then, with these definitions, correction factors can be defined
as in Table.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Degree of sinuousness (stream length/ valley length)
Correction Factor
Straight Channels 1.00
Slightly Sinuous Channels 0.90
Moderately Sinuous Channels 0.75
Very Sinuous Channels 0.60
Reference
Craig Fischenich "Stability Thresholds for Stream Restoration Materials", May 2001.
10
20
30
40
50
60
70
80
90
100
200
300
400
500
600
700
800
900
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
20000
0.1 0.2 0.3 0.4 0.5 1 2 3 4 5 6 7 8 9 10 20 30 40 50 60 70 80 90
0.003
0.004
0.005
0.006
0.007
0.008
0.009
0.01
4
3
2
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.09
0.
0.
08
07
0.06
0.05
0.04
0.03
0.02
Fortier & Scobey  Recommended for canals in fine sand
with water containing colloids
Line representing relations of tractive forces b/ft
2
= 0.5
Diameter in inches
Tractive force kg/m
2
= diameter in centimeters (approx)
Recommended value for
canals with high content of
fine sediment in the water
Schoklltach  Recommended
for canals in sand
U.S.S.R. Canals with
2.5 % colloids in water
NUERNBURG KULTURAMPT (NK)
Fortier & Scobey  Recommended for canals, in fine sand and clear water
Recommended values for canals with clear water
Stroub values of critical force
U.S.S.R. Canals with clear water
Recommended values for canals
with low content of fine sediment
in the water
Recommended values for canals in
coarse, noncohesive material size
25% or larger
U.S.S.R.
Canals with
0.1% colloids
in water
NK
MEAN DIAMETER, MILLIMETERS
DESIGN OF SMALL DAMS USBR' 87
NK
Conversion Factor
2 2
1 lb/ ft 47.87 N/ m =
Plasticity index (PI) is the difference in percentage of moisture between plastic limit and
liquid limit in Atterberg soil tests. For canal design PI can be taken as 7 as the critical
value. In this figure, for the fine non cohesive , i.e.,average diameters less than 5mm ,
the size specified is the median size of the diameter of a partical of which 50 percent
were larger by weight. Lacey developed the following equations based on the analysis
of large amount of data collected on several irrigation canals in the India.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1/ 2
s
1/ 3
s
4 5/ 3 1/ 6
0 s
P 4.75 Q
f 1.76 d
Q
R 0.47
f
S 3*10 f Q
=
=
=
=
In which P is the wetted perimeter (m), R is the hydraulic mean radius (m), Q is the flow
in m
3
/s, d is the diameter of the sediment in mm, f
s
is the silt factor, S
0
is the bed slope.
Table: particle size and silt factors for various materials
Material Size (mm) Silt factor
Small boulders, cobbles,
shingles
64  256 6.12 to 9.75
Coarse gravel 864 4.68
Fine gravel 48 2.0
Coarse sand 0.52.0 1.44  1.56
Medium sand 0.250.5 1.31
Fine sand 0.060.25 1.1  1.3
Silt (colloidal) 1.0
Fine silt (colloidal) 0.4  0.9
Taken from Gupta (1989)
Combining the above equations the following resistance equations similar to the
Manning equation based on the regime theory is obtained.
2/ 3 1/ 3
0
V 10.8 R S = in which V is the velocity in m/s.
21.1.6 The Tractive Force Method
When water flows in a channel, a force that acts in the direction of flow on the channel
bed is developed. This force, which is nothing but the drag of water on the wetted area
and is known as the tractive force. A particle on the sloping side of a channel is subject
to both a tractive force and a downslope gravitational component. It is noted that the
tractive force ratio is a function of both the side slope angle and the angle of repose of
the material composing the channel perimeter.In the case of cohesive materials and fine
noncohesive materials, the angle of repose is small and can be assumed to be zero;
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
i.e.. for these materials the forces of cohesion are significantly larger than the
gravitational component tending to make the particles roll downslope.
Consider the shear stress at incipient motion (which just begins to move particles) for
uniform flow.
The tractive force is equal to the gravity force component acting on the body of water,
parallel to the channel bed.
Gravity component of weight of water in the direction of flow is equal to
0
ALS in
which, is the unit weight of water, A is the wetted area, L is the length of the channel
reach, and S
0
is the slope. Thus, the average value of the tractive force per unit wetted
area, is equal to
0
0
ALS
RS
PL
0
s
A
p
bAp
my b/2
Plan View
L
C
Flow
W
s
= submerged weight of the particle
0
Ws
On the surface of the side slope
0 0
y
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
From Force diagram, resultant Force, R:
( ) ( )
2
2
s 0 p s
R W sin + A =
Resisting Force, F
s
:
( )
s 0
s s 0
Wcos is the weight component normal to side slope
tan is the coefficient of friction due to angle of internal friction
F W cos tan =
Therefore
s
2 2 2
s 0 s 0 p
2
s 0
s 0
2
p
R F at incipient motion.
W cos tan W sin A
for the unit tractive force that causes impending motion on a sloping surface
W tan
tan cos 1
A
tan
s
s
Solving
=
= +
=
On the channel bed, with being zero it reduces to
0
s
p b s b
p
W tan
A W tan
A
= =
Tractive Force Ratio
2 2
s
2 2
b
0 0
sin
K 1 1
sin
tan
cos
tan
= = =
K is the reduction factor of critical stress on the channel side.
Thus the ratio is a function of only side slope angle and angle of repose of the
material
0
.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Example:
Canal cross section: World's largest canal
Full supply level at Head Regulator 91.44 m
Length upto Rajasthan border 458.00 km
Length in Rajasthan 74.00 km
3 1
Total 532 km
Bed width at head reach 73.1 m
Fully supply depth at head reach 7.60 m
Design discharge(head reach) 1133 m s
Gujar
3 1
ath  Rajasthan border 71m s
No. of branches 42
Length of distribution Network 66000 km
concrete lining of 100 mm to 125 mm concrete
Total Lining
Phase
5
3 1
I 150.58 + 93.93 + 39.26 = 283.77
Phase II 126.14 + 1.08 +22.60 = 149.82
Total = 435.59 x 10 Sq.m
2) Sardar Sarovar Project
design disharge 86937.2 m s will be the 3rd largest in the world.
Gaz
5 3 1
5 3 1
enba, china 1.13 x 10 m s
Tucurri Brazil 1.0 x 10 m s
Radial gates of chute spillway 7 nos 18.3m x 18.3m
For sertvice spillway 23 radial gates of 18.3m x 16.75.
Dam is 12.0 m concrete gravity dam
Hei
3
ght of dam from foundary 163.00 m
Gross storage 9497.07 m
The design procedure for flexible lining channel consists of following steps:
1. Channels are usually trapezoidal or triangular (with rounded corners) or parabolic.
2. If lined with riprap, m >3, no need to check for blank stability.
3. Channel slopes can be steep depending on application.
4. Most flexible linings give adequate protection upto
0
S 0.01 .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The Limiting shear stress or limiting velocity procedure is also commonly used. In this
approach, the uniform depth is calculated for the maximum discharge Q and this value
is to be compared either , and if they satisfy their
add the freeboard and the design is complete. Table below lists the values for various
lining types.
max permissible max permissible
vs. or V vs. V
Table : Permissible shear stresses for lining materials
Lining Category Lining Type Permissible Unit Shear Stress (kg/m
2
)
Woven Paper Net 0.73
J ute Net 2.20
Fiberglass Roving
Single 2.93
Double 4.15
Stream with Net 7.08
Cured wood Mat 7.57
Temporary
Synthetic Mat 9.76
Class A 18.06
Class B 10.25
Class C 4.88
Class D 2.93
Vegetative
Class E 1.71
2.5 cm 1.61 Gravel Riprap
5 cm 3.22
15 cm 9.76 Rock Riprap
30 cm 19.52
Channel Design using Tractive Force
Procedure:
1. Estimate the roughness in the channel
2. Estimate angle of repose of candidate material.
3. Estimate channel sinuosity and tractive force correction factor.
4. Specify side slope angles.
5. Estimate "tractive force ratio", K, between the sides and the bottom of the channel.
6. Determine the maximum permissible tractive force for the canditate material.
7. Assume that side channel shear stress limits design and determine the uniform flow
depth in channel.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
8. Calculate the required bottom width.
9. Check that the permissible tractive force is not exceeded on channel bed.
10. Check that the design velocity exceeds the minimum permitted velocity (usually 0.6
to 0.9 m/s) and check the Froude number of the flow (F=subcritical).
11. Estimate the required freeboard.
Example:
1. Design a trapezoidal channel to carry 20 m
3
/s through a slightly sinuous channel on a
slope of 0.0015. The channel is to be excavated in coarse alluvium with a 75 percentile
diameter of 2 cm of moderately rounded particles.
1. Manning n:
( )
1 6
75
n for gravel ranges: 0.020  0.030
Assume n = 0.025
n = 0.038 d 0 020
/
. =
2. Angle of repose:
o
75
d = 2cm = 0.8 in = 32
3. Slightly sinuous channel: Cs =0.90
4. Side Channel slope: Try 2H:1V
1
1
= tan 26 6
2
.
=
5. Tractive force ratio:
2 2
s
2 2
b
sin sin 26 6
1 1 0 53
sin sin 32
.
K . = = = =
6. Permissible Tractive Force:
( )
( )
2 2
b s
2
s b
Bed: = C 16 N/m 14.4 N/m
Side: = K = 0.53 14.4 7 6 N/m .
=
=
7. Assume incipient motion on side wall:
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )( )
2
s o o
s
n
o
0 76 y S = 7.6 N/m
7 6
y 0.68 m
0 76 S 0 76 9790 0 0015
.
.
. . .
=
= = =
8. Solve for bottom width b:
( )
5/3
2/3 1 2 1 2
0 0
2/3
2 2
1 1 A
Q = AR S S
n n
P
where A=by+my , P= b+2y 1+m
b = 2.42m smallest positive real solution
/ /
=
9. Tractive force on bed:
( )( )( )
2
b 0 0
2 2
0 97 y S 0 97 0 68 0 0015 9 7 N/m
1 7 N/m 14 4 N/m
. . . . .
. .
= = =
<
10. Check velocities:
( )( ) ( )
2
2 2
Area = by + my = 24.2 0.68 2 0.68 17 4 m
Q 20
V = 1 1 m/s
A 17.4
V V
F=
gD A
g
B
T = Top width = T+2 my = 26.92 m
D = A/T = 0.65 m
Froude number = 0.44
.
.
+ =
= =
=
11. Free board:
3
For Q = 20 m /s the freeboard will be 0.75 m
Total depth = 0.68 + 0.75=1.43 m
2. Design a straight trapezoidal channel for a design discharge of 20 m3/s. The bed
slope 0.00025 and channel is excavated through the fine grave having particle size of 8
mm. Assume the material to be rounded moderately and water has low concentration of
sediment. Q = 10 m3/s, S0 = 0.00025, moderately rounded. Diameter = 8 mm =
8
0 3149
25 4
. "
.
= .
For fine gravel n =0.025 is assumed
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Side slope (assume) =2.5 : 1 =21.80 =2148'
1
1
= tan
2.5
From fig for 8 mm diameter moderately rounded angle of repose
2 2
0
2 2
sin sin (21.80) 0.1379
K= 1 1 1
0 1654
sin sin 24
0 1663 0 4077
.
. .
= =
=
Critical Shear Stress =0.13 * 47.87 =6.2231 N/m
2
No correction for alignment.
Maximum unit Tractive force =0.785 y S0 =0.75 * 9806 * y * 0.00025 =1.8386 y.
1.8386 y =6.2231
6.2231
y = 3 385 m
1.8386
. =
width required to carry the flow of 20 m
3
/s
1
( )
( )
( )
( ) ( ) { }
( ) { }
( ) { }
( ) { }
2
3
0
2
5
3
0
2
3
0
5
3
0
2
3
0
B+my y
1
b+my y S Q
n
B+2 1+m y
b +2.5 3.385 3.385
1
0 00025 20
0.025
b +2 5 3.385
b +8.4625 3.385
31 6227
b +15.138
.
.
=
=
=
Solve by trial and error for b.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
21.1.7 Economic Aspects of Canal Design
( )
( )
( )
5/ 3
2
2/ 3 n
2/ 3
2
0
1/ 4
2
3/ 8
n 0
5/ 8
by my
Q
(i) AR
S
b 2y 1 m
solve
b / y 2 1 m
y = Q S
b / y m
+
=
+ +
+ +
+
the above equation for y
(ii) /
if b/y, z are specified the equation can be solved explicitly for y and b.
The cost of materials used in lining a channel can be specified interms of the value of
material used. This may be expressed as
( ) ( )
( ) ( )
( )
B b
2
s s B
2
s B
iii t b 2b' Bb k
t 2E E' 2 y F 1 m
Therefore C bB k 2 y F 1 m
Notation :
C
b
s
b
Cost of bed material C = per unit length
(iv) Cost of side material C
C=C
= total material cost per un
+ = +
= + = + +
+ = + + + +
b
s
b
it length,
C = material cost per channel base per unit length,
C = material cost of sides per unit channel length,
b' = bottom corner width,
t = thickness of the base materi
B
s
al,
t = channel side lining thickness,
= cost of base lining material per unit volume,
= cost of side lining material per unit volume,
B = cost of base material for specified
B
'
thickness per unit area,
= cost of side lining material for specified thickness per unit area,
F = vertical free board requirement,
E = wetted length of the side,
E = side length of the free board.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Minimum Cost Trapezoidal Section by Optimisation Technique
Lagrange Multiplier technique can be used. Ratio of marginal changes in section factor
are equal to the marginal changes in the costs i.e.
( )
( )
2/ 3
2/ 3
AR
C
b b
C
AR
y
y
The above equation represents the minimum cost of the optimal cost subject to the
equation. Substituting, then the optimal solution of the above is given by,
( )
2
1 2 3
2 2
1
2
2
3
1
1/ 2
2
2 2 1
y y
K K K 0
b b
B
K 20 z 1 1 4 4z z 1
B B
K 1 6z z 1 10z
B
K 5 then,
2K b
y
B
K K 20 K
whichis a function of z and the ratio of the unit costs of the bas
+ + =
= + + +
= +
=
=
+ +
0
e to side slope material viz;
B Unit Cost of Base Material
Unit Cost of Side Material
SolutionSteps
1. S , B, , n, z and Q are given. Determine K1, K2, and K3.
2. Estimate b/y for minimum cost using equati
=
on.
3. Estimate the minimum cost depth of flow using equation.
4. Obtain the minimum cost bottom width by multiplying y times the ratio of b/y.
5. Generate the graphs for y Vs b for different values of B
0
n/Q
/ and for a given value of z.
S
6. Also study the sensitivity of lining cost to variations of side slope (or side slope ratios).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Sample Run
Data
s B
Q in cumecs, B, and k in R , F in m
0.08, 0.001, 0.014, 0.50, 105.0, 65.0, 15.0, 0.15
Result
Minimum Lining cost per unit Length = Rs. 109.51
Minimum cost bottom width = 0.186 m
Minimum cost depth of flo
( ) ( )
( ) ( )( )
( )
( )
( )
1
2
3
2
w = 0.366 m
105
K 20 0.25 1 1 4 4 0.50 1 0.25 8.3189
65
105
K 1 6 1 0.25 10 0.50 1.615 12.2005
65
K 5 1.615 8.075
2 8.315
b
y
12.20 12.25 20 1.615 8.315
= + + + =
= + =
= =
=
+ +
References
1. Hager, W.H. 1985, Modified venturi channel. J ournal of the Irrigation and Drainage
Engineering, ASCE, 3(1): 1935.
2. Hager, W.H. and P.U. Volkart, 1986, Distribution channels, J ournal of Hydraulic
Engineering, ASCE, 112(10): 935952.
3. Trout T.J ., "Channel Design to minimise lining material cost" J . of Irrigation and
Drainage Division Division, ASCE Vol. 105, Dec 1982, pp 242  245.
21.1.8 Seepage in Canal
Introduction
Seepage is one of the most serious forms of water loss in an irrigation canal network.
Water lost by seepage cannot be recovered without the use of costly pumping plant. In
addition excessive seepage losses can cause low lying areas of land to become
unworkable. As the water table rises, water logging and soil salinisation can occur,
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
necessitating the installation of elaborate and costly drainage systems. Furthermore the
cultivable area is reduced, resulting in a loss of potential crop production.
The accurate measurement of seepage in existing irrigation canals enables very
previous reaches to be identified and lined to conserve water; losses amounting to as
much as 40% of the total inflow to a scheme have been recorded. Moreover valuable
information about the long term performance of different types of canal linings in general
use can be obtained, enabling conveying efficiencies to be improved in the future.
Three methods of seepage measurement are in common use at the present, namely:
ponding; inflow/outflow; seepage meter. Other methods of seepage detection are also
used, such as for example, chemical tracers, radioactive tracers, piezometric surveys,
electrical borehole logging, surface resistivity measurements, and remote sensing.
These methods suffer from the disadvantage that they are either more difficult to use or
interpret.
Ponding Method
Ponding is considered to be the most accurate method of seepage measurement. It is
frequently used as standard with which to compare other methods. The procedure, in
principle is simple, a stretch of canal under investigation is isolated and filled with water.
The rate of seepage is determined by one of two methods. In the first, which is the one
usually employed, the rate of fall of the water level is recorded (falling level method).
Alternatively, the rate at which the water must be added to keep the water level constant
is recorded, (constant level method).
In practice the ponding method has certain advantages:
1. The accuracy of measurement is not dependent on the length of the test reach
provided it is sufficient to compensate for normal errors.
2. The requirement for trained manpower is small.
3. Sophisticated equipment is not required for the test.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
The disadvantages of the method are
1. Costly bulkheads must be built at each end of the reach if existing structures are not
available.
2. The normal flow through the canal must be stopped for the duration of the test.
Hence the methods is usually restricted to smaller canals.
3. The rate of seepage loss from the test section can vary with time because of the
sealing effect of fine material settling out in the water, or in the case of a canal which is
initially dry, because of the time taken to re saturate, or a combination of both.
4. The rate of seepage loss can be very different from that measured in the canal in
flowing water because of 3.
5. Large quantities of water are required if the canal under test is initially dry.
Inflow / Outflow Method
Next to ponding, inflow/ outflow, is the most commonly used method for the
measurement of seepage. The discharges into, and out of a selection reach of a river or
canal are measured. the rate of seepage is derived from the difference. In comparison
with the ponding method, the inflow/ outflow method has certain advantages:
1. Any impedance to the normal operation of the canal os minimised.
2. No costly bulkheads are required.
3. Seepage is measured with the canal in its normal discharge state, thus eliminating
the effects of silting, algae and fungoidal growth, and distortion of the local seepage
flow.
4. Measurements can be made even when numerous off takes are spread without too
great an increase in overall cost.
The disadvantage of the method on the other hand are
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1. The errors in the flow measurement tend to overshadow the seepage losses,
especially in large canals greatly reducing its accuracy.
2. Measurement becomes very labour intensive if a large number of off takes are
present.
3. Only the bulk measurement of seepage, over the test reach is obtained, which can
attain a considerable length because of 4. In large canals very large reaches are
required to improve the accuracy of an individual measurement because of 1. Various
methods are available for the measurement of a canal or river discharge. These can be
divided into two classes: Continuous methods; Occasional methods. Only gauging
structures, ultrasonic, and electromagnetic, among the Continuous methods, and
velocity area, and dilution gauging among the Occasional methods are considered to be
potentially accurate enough for the estimation of seepage. Each of these techniques is
outlined briefly below in the context of the inflow/ outflow method.
Velocity Area Method
This method is the mostly used of all discharge measurement techniques. The area of
flow is determined by sounding, and the mean velocity by current metering. The product
of the two giving the discharge. Some care must be taken when selecting a site on a
canal or river however. Ideally the test reach should be straight and free from
obstructions, weeds, or off takes, and have a stable bed. Before beginning a discharge
measurement, a preliminary survey should be carried out to determine the bed profile,
and to ensure that a welldeveloped velocity distribution exists along the channel. All
soundings should be related to an established datum.
The method of current metering depends on the depth of flow and velocity, ranging from
the use of wading rods to a cable car suspended across the channel. For most gauging
work on irrigation canals however the current metering is usually carried out either with
wading rods of from boat. The accuracy of the measurement depends firstly on a
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
number of verticals at which velocity readings are taken and to a lesser extent on the
number of levels velocities are measured at on each vertical.
The achievable accuracy can be optimised with the available equipment, time, and
manpower. The length of time given to each current meter reading depends very much
on flow conditions, but during the preliminary tests it is advisable to record for the
recommended 3 minutes while taking readings after each minute for comparison. If very
accurate results are required it is essential that the survey is carried out by an
experienced, welltrained team.
The inflow/ outflow method is very sensitive to canal size.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
21.2 Typical canal cross sections
Water enters the conveyance system through the intake structure located at the dam.
Depending on the topography of the terrain, this conveyance system may take the
shape of the tunnels, canals, flumes or pipes. Geological factors do influence the type of
the system to be adopted.
Some of the shapes of canals adopted are shown in following figures.
Original ground level Canal in filling
Canal in cutting
Embankment
Original grand slope
Retaining wall
Part in cutting and in filling
Original ground level
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
braced type
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Typical cross sections of Power canals
Covered duct R.C.C
on piles when there is
land slide problem
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Side slopes and other salient features
Width
'b' in
m
Length
in km
Side
slope
m : 1
Depth y
(m)
Discharge
Q (m/s)
3
Average
Velocity
(m/s)
S
0
Nangal   1.25 : 1  354  
Sutlej 
Beas link
9.45 11.8 1.5 : 1 6.26 255.0 2.1 1 / 6666
Lower
Sileru
11.9 15.6 1.5 :1 3.97 127.4 0.665 
Yamuna
Hydel
Stage I
11.0  1.5 : 1  200.0  
Hirakud 51.0  1 : 1 6.3 707 1.97 
Maximum permissible velocity (safe against erosion)
(1) Stenbergs formula
b
b
V = 4.43 2d
in which V in m/s, d is the diameter of the particle in m
(2) Bogardi and Yens formula
4/9
m
V = 22.9 d 1
In which V is the velocity in cms
1
, d
m
is the effective size of particle in cm, is the
specific gravity of the particle.
= +
=
=
= +
= + = =
=
or 2 b y =
b = 2y
y
Hydraulically efficient rectangular
channel is half of a square.
2. Trapezoidal Section:
1
m
2
2 1 P b y m = + +
( ) A b my y = + or
A
b m
y
= = y
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
2
P 2 1
A
my y m
y
= + + (For a given area of flow)
Differentiate with reference to y assuming A and m to be constant.
2 2
p
2 1 0
d
Ay m m
dy
= + + =
Substituting for area, the above equation can be rewritten as
2 2
2 1 0 y ( b my )y m m
+ + + =
2
2 1 0
b my my
m
y
+ +
+ = + + =
2
2
1
2
b my
y m
+
= +
Half the top width =side slope distance (for given side slope)
2
2 1 2 2 1 b y m my y m m
= + = +
Substitute this value of b into the equation A and P and simplifying
2
2 2
05
2
P 2 2 1
2 1
2 1
.
y m
A y m m
A
y
m m
m
= +
= +
=
+
Substitute the value of y into P
( )
( )
1 2
2
05
2
P 2 2 1
2 1
/
.
A
m m
m m
= +
+
( )
2
P 2 2 1 A m m
= +
which is the m value that makes P least?
D.w.r to m and equate it to zero
P
0
d
dm
=
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
3
3
3
m = =
1 3
3
y
tan
my m
= = =
60 =
2
3
b y =
This means section is a half hexagon. If a semi circle is drawn with radius equal to
depth y then sides of this section are tangent to the circle.
2
__
y
3
1
m =
60
60
Half Hexagon  inscribed circle
of radius equal to depth is
tangential as shown in figure
b =
3
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Triangular Section: If is half angle
2
2
2
2 2 4
a sin
Area ay sin y ay sin
ay sin acos sin a
R sin
a
y
cos
a
= =
= = =
=
y
a
a
a sin
Hydraulically efficient channel
Half a square on vertex
R should be max.
2
0 2 0
4
dR a
, cos
d
= =
45 =
45
o
45
o
Half Square on its apex
Free surface width is equal to the diagonal
It is a half square resting on its apex and maximum width is equal to diagonal.
Alternative derivation for Triangular Section:
2
2
2 1
A y tan
y A/ tan
P y m
=
=
= +
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
l
m
y
2y sec =
2
A
sec
tan
=
( )
( )
3
3 2
3
3
2
2 0
2
0
2
/
dp d A
sec
d d tan
dp sec tan sec
A
d tan
tan
sec
sec tan
tan
=
=
=
=
3
2 0 sec tan sec =
Solve for
45 =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Hydrostatic Caternary (Linteria)
2 4 2 4 4
1
1
1 1
3 15 3 5 5
1 2
2 4 864 8 32 256
= + +
=
=
Equation for h
4
ydrostatic catenary is given by
y
x k k k k sin k sin
k
y ycos
x , y are measured from mid point of the surface
k= sin ; slope angle at
2
1
2
1 1
the point x y . varies from 0 at the
bottom of the curve to at the ends.
sin
sin , is slope at any point ( x, y)
k
For the hydraulically efficient channel
= 35 37' 7" , y= 3.5 m. Fin
35 37 7
17 48 335
030585
= =
=
d A, R, D, Z at full depth.
Also plot the cross section of the channel
' "
' . "
2 2
k= sin .
2
si
sol
si
ut
n
ion:
=
1
1
2
2 2
2 2 2
1
2 2
90 0707
2
170992
= =
=
= = = =
=
2
n sin
2 2
or k =
k sin
sin
k
sin cos cos
or cos
k
y ycos cos cos 45 k .
A . m P= 10.443 m
5 2
16374
67114 22293
=
= =
/
A
R = . m
P
T . m D = 2.5478 m Z = A D . m
y
1
(m) 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
( in
deg )
90 81.78 73.39 64.62 55.15 44.4 31.0 0
k 0.7071 0.6614 0.6235 0.5916 0.564 0.5400 0.51887 0.5
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Exercise: Plot the graph using the above data
x
1
x
y
1
y
5 2
170992 16374
67114 22293
= =
= =
2
/
A
A . m P= 10.443 m R = . m
P
T . m D = 2.5478 m Z = A D . m
Flexible Sheet: Filled with water upto rim, and held firmly at the top ends without any
effect of fixation on shape. Shape assumed under self weight of water is called
Hydrostatic Catenary.
Rounded bottom triangular section
0
m
1
cot
0
=
m
__
1
0
=
cot
1
m
T
r
1
m y
0
r
( )
( )
( )
2 2 1
1
2
2
2 2
2 1
A y m cot m
P y m cot m
A r y
R
P
T r m
=
= +
= = =
= +
Hydraulically efficient sections could be derived using Lagrange Multiplier approach.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
22.5 Design of Unlined Channels
In 1895 Kennedy evolved an equation for , nonsilting and nonscouring velocity and y
depth of flow after studying dimensions of stable alluvial canals ranging over 30 sites in
a strech of 144 km on upper Bari Doab Canal System in Punjab.
0
V
( )
064
0
0546
.
V . y in mks =
After a study of hydraulic dimensions of 4,345 km of channels in the Punjab, Lindley
suggested, in 1919, the following relations connecting the velocity V (m
3
s
1
), the surface
width B (m) and the depth of flow y (m):
( )
( )
( )
057
0335
161
057
02646
786
.
.
.
V . y in mks
V . B in mks
B . y in mks
=
=
=
Lindley's theory was further advanced by Lacey in 1929, but he adopted P, the wetted
perimeter and R, the hydraulic mean radius, as the flow parameters instead of the
surface width B, and the depth of flow, y, and in addition introduced a 'silt factor', f.
Lacey's formulae in their final form is as follows:
( )
1 2
4825
/
P . Q in mks =
1 3 1 3
04725
/ /
R . Q / f = (Same for both units)
( )
5 3 1 6
00003015
/ /
S . f / Q in mks =
( )
3 4 1 2
1
/ /
o
a
V R S in m
N
= ks
1 4
00225
/
a
N . f = (Same for both units)
Lacey's general flow equation is similar to that of Manning; but Na in Lacey's relation is
an absolute rugosity coefficient which, in addition to boundary friction, allows for shock
losses in the channel due to irregularities or bends. The silt factor f was correlated
approximately to the silt grade m (in mm) by the relation 176 f . m = on the implicit
assumption of the 'regime charge' being carried by a channel in regime. Chitale
analysed the data subsequently in 1966.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Though the Lacey equations have been in common use in India and also in other
countries, it has been long realised that these equations were not perfect and suffer
from certain shortcomings. Perhaps the major difficulty experienced in application of
Lacey's equations is in the choice of appropriate value of the silt factor 'f' occurring in
the Lacey formulae for depth and slope. It is also found that canals designed according
to Lacey formulae give a somewhat wider and shallower section. Moreover divergence
from dimensions given by Lacey equation in existing stable channels with those given
by Lacey formulae errors resulting from adoption of Lacey equations for P, R, and S0
were worked out which ranged from 11.27 to 83.47 percent.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
22.6 Examples
Design a triangular lined channel section with rounded bottom as prescribed by CBI&P
given Q =30 m
3
s
1
, n=0.015, bed slope S
0
1:1800, side slope m =1.25.
Solution :
The side slope of the channel m =1.25,
2
2
i.e., cot = 1.25,
= 0.6747 radian = 38.6598
Area of the channel A = y ( + cot )
= 1.9247 y
Perimeter of the c
8
3 8
2 2 3
3
2
3
19247
19247 2
2
=
2 / 3 /
hannel P = 2y ( + cot ) = 2y(0.6747+1.25)
Hydraulic radius R= A / P
= y / 2
. y
therefore AR . y *(y / ) = =1.2125y >1
0015 30
1
1800
=
o
nQ . *
and = 19.092 >2
S
( )
( )
3
8
3
8
12125
157458
+ +
+
2
2
i.e., cot = 1.25,
therefore = 0.6747 radian
Area of the channel A = y cot by
= 1.9247 y by
2
2
>1
and Area A = Q / V
= 125 /2.5= 50 m
= 50 m . >
( )
+
2
2
Equating equations 1 and 2
50 = 1.9247 y by >A
Perimeter of the channel P = 2y +cot + b >3
2
19247
38494
+
=
+
2/3
= 3.8494y + b
. y by A
R =
P . y b
1
1
2.5 = R
2500
0.015
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
3
2
15
05
00375
002236
.
.
0.015*2.5 .
R= =
.
1
2000
= 2.1719
230213 =
A 50
P= = . m
R 2.1719
Equating equations 3 and 4
23.0213 = 3.8494y + b
therefore B = 23.0213  3.8494y >5
put equation 5
230213 38494 0 + =
2
in equation A, then
50  1.9247 y ( . . y) * y >B
Solving the equation B one obtains
b = 12.04 m and
y = 2.8527 m
add free board = 0.6473 m
therefore the total height of the channel H =2.8527 + 0.6473 = 3.5 m and
the top width of the channel T= b + 2* m * H
= 12.04 + 2(1.25*3.5) = 20.79 m
T = 20.8 m (approximately)
3. Design the lined canal to carry Q =100 m
3
/s, with n =0.013, bed slope S
0
=1:2500,
V =2 m/s, side slope m =1.25, and hydraulic radius R =1.48.
4. Design a triangular lined channel section with rounded bottom given Q =300 m
3
/s, n
=0.014, bed slope S
0
=0.0016, side slope m =1.25.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
GRADUALLY VARIED FLOW
23.1 Introduction
The flow in an openchannel is termed as gradually varied flow (GVF) when the depth of
flow varies gradually with longitudinal distance. Such flows are encountered both on
upstream and downstream sides of control sections. Analysis and computation of
gradually varied flow profiles in openchannels are important from the point of view of
safe and optimal design and operation of any hydraulic structure.
23.2 Basic Assumptions in GVF Analysis
1. The gradually varied flow to be discussed here considers only steady flows. This
implies that (i) flow characteristics do not change with time, and (ii) pressure distribution
is hydrostatic over the channel section.
2. The head loss in a reach may be computed using an equation applicable to uniform
flow having the same velocity and hydraulic mean radius of the section. This implies
that the slope of energy grade line may be evaluated using a uniform flow formula such
as Manning equation and Chezy equation, with the corresponding roughness coefficient
applicable primarily for uniform flow.
3. Channel bottom slope is small. This implies that the depth of flow measured vertically
is same as depth of flow measured perpendicular to channel bottom.
4. There is no air entrainment. Advanced text books may be referred to study the effects
of air entrainment.
5. The velocity distribution in the channel section is invariant. This implies that the
energy correction factor, , is a constant and does not vary with distance.
6. The resistance coefficient is not a function of flow characteristics or depth of flow. It
does not vary with distance.
7. Channel is prismatic.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
23.3 Dynamic Equation for Steady Gradually Varied Flow
Consider the flow profile in an elementary length dx of an open channel as shown in
Figure 23.1.
Figure: 23.1 Derivation of the gradually varied flow equation
Channel bottom
dx
Slope = S
0
Datum
dH
Energy grade line
Water Surface
Flow
v
2
___
2g
Z
The total head above the datum at a section is
2
v
H = z + y +
2g
(23.1)
where H is the total head; z is the elevation of the channel bottom; y is the depth of flow;
is the energy coefficient; g is the acceleration due to gravity; and V is the average
velocity of flow through section. Here, bottom of the channel is considered on the X
axis. Equation (23.1) is differentiated with respect to x to obtain.
2
dH dz dy d V
dx dx dx dx 2g
= + +
(23.2)
As the slope of the channel bottom is assumed small, Sin tan , in which is the
angle of the channel bottom with horizontal. Slope is considered positive if it depends in
the direction of flow. Therefore, referring to Figure 23.1, slope of the energy grade line,
f
dH
S
dx
= , and slope of the channel bottom,
0
dz
S
dx
= .
Equation (23.2) becomes
2
0 f
dy d V
S S
dx dx 2g
+ =
(23.3)
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Velocity, V can be expressed in terms of the flow rate, Q and area of the cross
section, A.
Q
V=
A
(23.4)
Noting that flow rate, Q remains constant with respect to x (no lateral inflow or outflow),
but area, A changes, differentiating Equation (23.4) with respect to x and subsequent
substitution in Equation (23.3) leads to
2
3
dy Q 2 dA
+
dx 2g A dx
(23.5)
However, for a prismatic channel
dA dA dy dy
= T
dx dy dx dx
= (23.6)
where, T =free surface width. Substitution of Equations (23.4) and (23.6) in Equation
(23.5) and subsequent simplification results in the following gradually varied flow
equation,
o f
2
3
S S dy
=
Q T dx
1
gA
(23.7)
Equation (23.7) is a nonlinear firstorder differential equation. In this equation, slope of
the energy grade line,
f
S may be estimated using the Manning's equation.
2 2
f 2 4/3
n Q
S
A R
= (23.8)
where n is the Manning roughness coefficient; and R is the hydraulic mean radius.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
23.4 Classification of Gradually Varied Flow Profiles
It is important to systematically classify the water surface profiles in a channel before
computation of flow profiles is carried out. Such classification helps to get an overall
understanding of how the flow depth varies in a channel. It also helps to detect any
mistakes made in the flow computation.
It may be recalled here that
( )
2
2
3
Q T
F = 23.9
gA
where F =Froude number. Substitution of Equations (23.8) and (23.9) in Equation
(23.7) leads to
( )
2 2
0 2 4/3
2
n Q
S 
dy
A R
= 23.10
dx 1 F
For a specified value of Q, both F and
f
S are functions of the depth, y. In fact, both F
and
f
S will decrease as y increases. Recalling the definitions for the normal depth,
n
y ,
and the critical depth,
c
y , the following inequalities can be stated
( )
f 0 n
f 0 n
S > S when y < y
23.11
S < S when y > y
( )
c
c
F > 1 when y < y
23.12
F < 1 when y > y
A gradually varied flow profile is classified based on the channel slope, and the
magnitude of flow depth, y in relation to
n
y and
c
y . The channel slope is classified
based on the relative magnitudes of the normal depth,
n
y and the critical depth,
c
y .
n c
n c
n c
0
0
y > y : "Mild slope" (M)
y < y : "Steep slope" (S)
y = y : "Critical slope" (C)
S =0 : "Horizontal slope" (H)
S <0 : "Adverse slo
pe" (A)
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
It may be noted here that slope is termed as "sustainable" slope when
0
S 0 > because
flow under uniform conditions can occur for such a channel. Slope is termed as
"unsustainable" when
0
S 0 since uniform flow conditions can never occur in such a
channel. Flow profiles associated with mild, steep, critical, horizontal, and adverse
slopes are designated as M, S, C, H and A profiles, respectively.
The space above the channel bed can be divided into three zones depending upon the
inequality defined by equations (23.11) and (23.12). Figure 23.2 shows these zones for
a mild and a steep channel.
(a) Mild Channel
(b) Steep Channel
NDL: Normal depth line
CDL: Critical depth line
Figure 23.2: Profile Classification
Zone  1
Y
c Y
n
2
3
NDL
CDL
Bed
Y
c
Y
n
2
3
NDL
CDL
Bed
1
The space above both the CDL and the NDL is designated as zone1. The space
between the CDL and the NDL is designated as zone2. The space between the
channel bed and CDL/NDL (whichever is lower) is designated as zone3. Flow profiles
are finally classified based on (i) the channel slope and (ii) the zone in which they occur.
For example, if the water surface lies in zone1 in a channel with mild slope (Figure
23.3), it is designated as M1 profile. Here, M stands for a mild channel and 1 stands for
zone1.
It may be noted that an M1 profile indicates a subcritical flow since flow depth, y is
greater than the critical depth,
c
y .
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Figure 23.3: M1 Profile
M1
NDL
CDL
Bed
Water Surface
Similarly, an S2 profile (Figure 23.4) indicates the water surface lies in zone2 in a steep
channel. It may be noted that a S2 profile indicates a supercritical flow since flow depth,
y is lower than
c
y .
S2
Figure 23.4: S2 Profile
CDL
Water Surface
NDL
Bed
Table 23.1 presents types of flow profiles in prismatic channels. In this table, a channel
slope is described as critical slope when critical conditions occur for uniform flow i.e.
when
n c
y y = .
Table 23.1: Types of Flow Profiles (S
c
: Critical Slope)
Profile Designation Slope
zone  1 zone  2 zone  3
Relative
position of y
Type of Flow
Adverse S
0
=0 None
A2
A3
y >y
c
y <yc
Subcritical
Supercritical
Horizontal S
0
=0
None
H
2
H
3
y >y
c
y <y
c
Subcritical
Supercritical
Mild 0<S
0
<S
c
=0
M1
M2
M3
y >y
n
>y
c
y
n
>y >yc
y
n
>y
c
>y
Subcritical
Subcritical
Supercritical
Critical S
0
=S
c
>0
C1
C
2
y >y
c
=y
n
Subcritical
uniform 
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
C
3
y =y
c
=y
n
y
c
=y
n
>y
critical
Supercritical
Steep S
0
>S
c
>0
S1
S2
S
3
y >y
c
>y
n
y
c
>y >y
n
y
c
>y
n
>y
Subcritical
Supercritical
Supercritical
23.5 Variation of Flow Depth
Qualitative observations about various types of water surface profiles can be made and
the profile can be sketched without performing any computations. This is achieved by
considering the signs of the numerator and the denominator in Equation (23.10). The
following analysis helps to know (i) whether the depth increases or decreases with
distance; and (ii) how the profile approaches the upstream and downstream limits. First,
consider the following general points:
c
y > y ; flow is subcritical; F<1 ; denominator is positive.
c
y < y ; flow is supercritical; F>1; denominator is negative.
n
y = y
;
flow is uniform;
f 0
S = S ; numerator is zero.
n f 0
y > y ; S < S ; numerator is positive.
n f 0
y < y ; S > S ; numerator is negative.
As
n
y y (y tends to
n
y );
f 0
S S ;
f 0
S S ; numerator approaches zero;
dy
0;
dx
the surface profile appraches normal depth asymptotically.
As
c
y y ; Flow tends to critical conditions; F 1; denominator tends to zero;
dy
;
dx
water surface profile approaches the critical depth vertically.
It is not possible to have a vertical watersurface profile. Therefore, it is assumed that
the water surface profile approaches the CDL at a very steep slope. It may be noted
that when the water surface slope is very steep, it cannot be assumed that
accelerations in the vertical direction are negligible. This means that the theory of
gradually varied flow should breakdown in such a situation because pressure is no
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
longer hydrostatic in those regions. Thus equation (23.10) is not valid whenever flow
depth is close to the critical depth.
As
f 0
dy
y ; S 0; F 0; S ;
dx
Water surface profile becomes horizontal as flow
depth becomes very large.
For a wide channel, hydraulic mean radius R h and
2
2
3
q
F
gy
= . Equation (23.10) can be
simplified to
( )
( )
3 10/3 2 2
0
10/3 3 2
gy S y q n
dy
=
dx y gy q
where q =flow rate per unit width. It can be seen from the above equation that
dy
dx
as y 0 . In other words, water surface profile tends to become vertical as the flow
depth tends to zero.
The qualitative characteristic of any type of watersurface profile may be studied using
the points discussed earlier. For example, consider an M1 profile. For an M1 profile,
n c
y>y >y .
c
y > y implies that F<1 and
n
y > y implies that
f 0
S < S .
Therefore,
0 f
2
S S dy
dx 1F
+
= = = +
+
This means that flow depth increases with distance x. On the downstream side, as y
keeps increasing
dy
dx
tends to
0
S and the water surface becomes horizontal. On the
upstream side, as y approaches the normal depth,
n
y , it approaches asymptotically. The
sketch of an M1 profile is shown in Figure 23.5.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
ApproachesNDL
asymptotically
becomes horizontal
Water Surface
CDL
Bed
x
Figure 23.5: Sketch of an M1 profile
NDL
Similarly, consider an M2 profile. In an M2 profile,
n c
y >y>y .
c
y > y implies that F<1 and
the denominator is positive. On the other hand,
n
y<y implies that
f 0
S > S . Therefore,
0 f
2
S S dy Ve
Ve
dx 1F Ve
= = =
+
This means that flow depth decreases with distance x. On the downstream side, as the
flow depth decreases and approaches the CDL, it approaches vertically. On the
upstream side as the depth increases and approaches the normal depth, it approaches
asymptotically. The sketch of an M2 profile is shown in Figure 23.6.
Figure 23.6: Sketch of an M2 profile
Water Surface
NDL
CDL
Bed
Now, Consider an S2 profile. In an S2 profile,
c n
y > y > y .
c
y < y implies that F>1 and
the denominator is negative.
n
y > y implies that
f 0
S < S . Therefore,
0 f
2
S S dy Ve
Ve
dx 1F Ve
+
= = =
This means that flow depth decreases with distance x. On the downstream side, as y
decreases towards
n
y it approaches NDL asymptotically. On the upstream side, as y
increases toward
C
y , it approaches CDL almost vertically. The sketch of an S2 profile is
shown in Figure 23.7.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
CDL
NDL
Bed
Figure 23.7: Sketch of an S2 profile
Water Surface
Proceeding in a similar manner, other water surface profiles can be sketched. These
sketches are shown in Figure 23.8. The profiles are shown in dashed lines as they
approach the CDL and the channel bed to indicate that gradually varied flow
assumption is not valid in those regions.
Zone 1
M1
NDL
CDL
M2
NDL
CDL
M3
NDL
CDL
Zone 2
Zone 3 MILD
CRITICAL
NDL /
CDL
C1
C2
C3
S1
S2
S3
CDL
NDL
NDL
STEEP
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
HORIZONTAL
CDL
NONE
CDL
H2
H3
A3
A2
NONE
CDL
Bed
Figure 23.8: Water Surface Profiles
ADVERSE
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
24.1 Real Life Cases of Water Surface Profiles
It is interesting to see how water surface profiles discussed in the unit 23 occur in real
life. For example, an M1 profile occurs behind a dam or a sluice gate located on a mild
channel. The dam or a sluice gate piles up water behind it such that the flow depth is
greater than the normal depth. Far away from the dam or sluice gate on the upstream
side, the flow would be occurring under uniform conditions and the flow depth would be
normal. In a similar manner, S1 and C1 profiles occur on the upstream side of a sluice
gate located on a channel with steep and critical slopes, respectively. An M2 profile
occurs on the upstream side of a free over fall at the downstream end of a mild channel
since a critical depth occurs in the vicinity of a free over fall. Similarly, a H2 profile
occurs on the upstream side of a free over fall at the downstream end of a horizontal
channel.
Critical flow conditions occur at the entrance to a steep channel from a lake or a
reservoir. However, flow should tend towards uniform flow conditions far away from the
entrance if the channel is long. Therefore, a S2 profile occurs in steep channels, on the
downstream side of the entrance. A few real life cases of water surface profiles are
shown in Fig 24.1.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Sluice gate
M1
M2
NDL
CDL
Bed (mild slope)
M3
Hydraulic J ump
Free over fall
(a) M1, M2 and M3 profiles
S2
S1
S3
CDL
NDL
CDL
Bed (steep slope)
Hydraulic J ump
Sluice gate
(b) S1, S2 and S3 profiles
Hydraulic J ump
Sluice gate
H2
H3
Free over fall
(c) H2 and H3 profiles
Figure 24.1: Reallife cases of watersurface profiles
Bed (horizontal)
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
24.2 Sketching of Composite Water  Surface Profiles
Characteristics of watersurface profiles in prismatic channels have been discussed in
the previous sections. However, in real life, a channel system may have variable cross
section or bottom slope. Also, it may have several control sections. A control section is
a section at which there is a unique relationship between the depth and discharge. For
example, weirs, sluice gates and spillways are control sections. They create sub critical
flows on the upstream side when they are performing under free flow conditions.
However, sub critical flow conditions occur on the downstream side also, if the control
structure is submerged or drowned. Similarly, critical depth occurs in the vicinity of a
free overfall in a mild channel. This acts as a downstream control for sub critical flows
since there is a unique relationship between the flow depth and the discharge when the
flow is critical. Critical flow conditions also occur at the entrance to a steep channel if
the water level in the lake or reservoir which is feeding the channel is higher than the
level of CDL at that point.
Steps outlined below are followed to sketch the composite water surface profiles in a
series Channel system.
Compute normal and critical depths for each reach of the channel system based
on specified flow rate, roughness coefficient, slope of the reach, and the channel
cross section.
Plot the channel bed, the normal depth line (NDL) and the critical depth line
(CDL) for each reach in the system.
Mark the control sections i.e., identify the sections where (i) the flow passes
through a critical depth (ii) the flow is expected to occur under uniform conditions,
and (iii) there is a control structures such as a weir, a sluice gate, and a spillway.
It may be noted that uniform flow conditions occur in long prismatic channels, far
away from control sections. Critical depth occurs at (i) the free overfall, and (ii)
the entrance to a steep channel from a lake, when the water level in the lake is
above the the elevation of the CDL at the entrance. Critical depth also occurs
when channel bed slope changes from mild to steep.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Starting from each control point, sketch the appropriate water surface profile
depending on the zone in which the depth at the control section falls and the
nature of the slope.
Qualitatively locate the hydraulic jumps wherever the flow changes from
supercritical to sub critical. For example, if there is a sluice gate at the
downstream end of a steep channel, the flow is sub critical on the upstream side
of the gate. However, if the channel is long, flow is supercritical far away from the
gate on the upstream side. Therefore, a hydraulic jump occurs in such a channel
(Figure 24.1 b). Also, on the downstream side of a sluice gate on a long mild
channel, the flow is supercritical immediately downstream of the gate. However,
far away from the gate on the downstream side, flow is subcritical. Therefore, a
hydraulic jump occurs in such a case also (Figure 24.1 a).
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
24.3 Examples
24.3.1 Example 24.1
Sketch and label type of water surface profiles in the Channel shown in Figure 25.1. All
the channels are long.
Figure 24.2: Channel for Example 24.1
Reservoir
NDL
CDL
CDL
NDL
CDL
Channel1
Channel2
Channel3
Solution
Channel1 is a MILD channel since NDL is above CDL.
Channel2 is a STEEP channel since NDL is below CDL.
Channel3 is a HORIZONTAL channel since NDL does not exist.
Critical flow conditions occur at the downstream end of Channel3 since it is not a
steep channel and there is a free overfall.
Critical flow conditions occur at the junction of Channel1 and Channel 2 since
the uniform flow in Channel1 is sub critical while uniform flow in Channel2 is
supercritical.
Flow is uniform in both Channel1 and Channel2 far away from the junction
point, since the channels are long. Thus flow depths in Channel1 and Channel2
fall between NDL and CDL. Therefore, flow profile in Channel1 is M2 type, while
flow profile in channel2 is S2 type
In Channel3, downstream portion would have sub critical flow conditions (critical
depth occurs at the downstream end), while in the entrance region, flow would be
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
supercritical as it enters from the steep Channel2. Therefore, a hydraulic jump
should occur in Channel3. The composite profile is shown in Figure 24.3.
Reservoir
NDL
CDL
Channel1
Channel2
M2
S2
Control Point
CDL
NDL
Hydraulic J ump
Control Point
CDL
Channel2
Figure 24.3: Solution for Example 24.1
H2
H3
24.3.2 Example 24.2
Sketch and label the types of water surface profiles in the Channel shown in Figure
24.4. All Channels are long.
Reservoir
CDL
NDL
NDL
CDL
Sluice Gate
CDL
NDL
Overfall
Channel1
Steep
Channel2
Mild
Figure 24.4: Channel for Example 24.2
Solution
Critical flow conditions occur at the entrance to Channel1 because Channel1 is
steep and the reservoir water level is above the CDL.
Flow depth in the Channel1 varies from critical depth at the entrance to the
uniform flow depth far downstream. This is an S2 profile.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
At the entrance to Channel2, flow depth is equal to the normal depth in Channel
1. This depth is below the CDL for Channel2. Therefore, flow in the upper
reaches of Channel2 is supercritical. An M3 profile occurs in this region.
The sluice gate in Channel2 creates subcritical flow conditions on the upstream
side and supercritical flow conditions on the downstream side. This acts as a
control.
On the upstream side of the sluice gate in Channel2, flow has to change from
supercritical because flow is supercritical in the upper reaches. Therefore, a
hydraulic jump occurs at some distance on the upstream side of sluice gate in
Channel2.
The sluice gate opening is such that the flow depth on the downstream side of
the gate is below CDL. Therefore, flow is supercritical here.
The Channel2 is long on the downstream side of the gate also, and it is mild.
Therefore, it cannot sustain supercritical flow conditions at distances far from the
sluice gate. Flow changes from supercritical to subcritical at some distance
downstream of the gate. This is accompanied by the formation of a hydraulic
jump.
There is a free over fall at the downstream end of the Channel2. Therefore,
critical depth occurs at this location. Note that the Channel2 is mild and the flow
is subcritical on the upstream side of the free over fall.
All the channels are long. Therefore, uniform flow conditions are realized in all
the channels far away from the control sections.
Keeping in mind the above points, the composite water surface profile can be
drawn as shown in Figure.24.5.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Reservoir
Control Point
NDL
CDL
Steep
Mild
NDL
CDL
CDL
NDL
Mild
Control Point
Over fall
Control Point
Control Point
Sluice Gate
Figure 24.5: Solution for Example 24.2
S2
M3
HJ
M1
M3
HJ
M2
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
25.1 Computation of Gradually Varied Flow
Qualitative sketching of water surface profiles in channels having gradually varied flow
has been discussed elsewhere. However, quantitative information on the variation of the
flow depth and flow velocity along a channel is required in many engineering
applications. For example, construction of a dam across a river raises water levels on
the upstream side of the dam. Estimation of the extent of inundation in such a case is
possible only by performing computations to determine the flow depths. Impounding of
water behind a dam also changes the self cleansing ability of the river to assimilate the
municipal waste discharged into it. Thus quantitative knowledge of flow depths and
velocities is essential while conducting the Environmental Impact Assessment (EIA)
studies also. These computations, generally known as GVF (gradually varied flow)
computations determine the watersurface elevations along the channel length for
specified (i) discharge, (ii) flow depth at any one location, (iii) the Manning roughness
coefficient, (iv) longitudinal profile of the channel, and (v) channel crosssectional
parameters. Generally, systematic numerical procedures are used for this purpose. All
these numerical procedures are either based on the numerical solution of the nonlinear
first  order ordinary differential equation for GVF (Eq. 23.7)
( )
o f
2
3
S S dy
= 23.7
Q T dx
1
gA
or on the direct application of the algebraic energy equation, using certain
approximations. These methods for the GVF computation are presented in the following
sections.
25.2 Direct Step method
In the Direct Step method, the location where the specified depth, y
d
occurs is
determined, given the location for the occurrence of depth, y
u
. Consider the channel
shown in figure 25.1. In this channel, say depth y
u
occurs at a distance x
u
from the
reference point. Discharge, Q, channel bottom slope, S
0
, the roughness coefficient, n
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
and cross  sectional shape parameters (which relate A, P and R to y) are also known.
The problem now is to determine the location x
d
(fig. 25.1).
25.1 Definition sketch for Direct Step Method
Datum
Channel Bed
Flow
Water Surface
(known) yu
Flow
(known) yd
d
d
(to be determined)
(known)
xu
Zu
Zd
xd
u
u
Energy equation between sections u and d can be written as follows
( )
2 2
u u d d
f
u u d d d u
V V
z y z y S x x
2g 2g
25.1
+ + = + + +
where subscripts "u" and "d" denote the values at the corresponding sections, and
f
S is
the average slope of the energy grade line between sections u and d. It may be noted
that the slope of the energy grade line, S
f
can be determined using Equation 23.8.
( )
2 2
f 2 4/3
n Q
S 23.8
A R
=
S
f
varies between sections u and d since the flow depth, and consequently A and R vary
between these two sections. S
f
may also due to variation in the roughness between the
two sections. Following equations may be used to determine
f
S .
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
( )
( )
u d
u d
u d
u d
f f f
f f f
f f
f
f f
1
S S S (25.2 )
2
S S * S (25.2 b)
2S S
S (25.2 c
S S
Arithmetic mean
a
Geometric mean
Harmonic mean
= +
=
=
+
)
Experience has indicated that the arithmetic mean (Eq. 25.2 a) gives the lowest
maximum error, although it is not always the smallest error. Also, it is the simplest of the
three approximations. Therefore, its use is generally recommended, and is used herein.
Noting that the bed elevations Z
u
and Z
d
are related through the bed slope, S
0
and the
distance between the sections, (x
d
 x
u
), Eq. 25.1 can be written as
( ) ( ) ( )
2 2
d u
f
d d u u d u 0 d u
V V
y y S x x S x x 25.3
2g 2g
+ + + =
However,
2 2
u u
u u u u
2
u
2 2
d d
d d d d
2
d
V Q
y y E
2g
2gA
V Q
y y E
2g
2gA
+ = + =
+ = + =
u d
25.4
where E and E are specific energies at section u and d, respectively.
Eq
d
uation 25.3 can now be used to determine x .
( )
u d
d u
d u
0 f f
E E
x x
1
S S S
2
25.5
= +
+
In equation 25.5, specific energies
u d
E and E , and the friction slopes,
u d
f f
S and S , can
be computed using the known values of (i) flow depths
u d
y and y , (ii) the flow rate Q,
(iii) the roughness coefficient, n and (iv) the cross sectional shape parameters.
Therefore, x
d
can be computed easily. For example, for a wide rectangular channel
(assuming
u d
1.0 = = ).
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
( )
( )
2
u u
2
u
2
d d
2
d
2 2
10/ 3
u
2 2
10/ 3
d
q
E y
2gy
q
E y
2gy
and
n q
y
n q
y
Substitution
= +
= +
=
=
u
d
f
f
25.6
S
25.7
S
of Eqs (25.6) and 2 ( )
( )
( )
2
u d u
2 2
d u
2 2
0
10/ 3 10/ 3
d u
q 1 1
x y y
2g
y y
q n 1 1
S
2
y y
+ +
=
+
d
5.7 in Equation 25.5 yields
x 25.8
In Equation 25.8, , , q,
u d 0 u
y , y S n and x are known, and
d
x can be determined easily.
Now that the location of section d is known, it is used as the starting value for the next
step. The water surface profile in the entire channel may be computed by increasing or
decreasing the flow depth, and determining the locations where these depths occur. For
example, say one is interested in determining the changes in flow depths in a mildly
sloping river due to the construction of a dam. The flow depth just behind the dam, y
dam
is known for the specified discharge, Q, the spillway length and the spillway
configuration. Flow depth far upstream of the dam is equal to the normal depth, y
n
since
uniform flow conditions exist there, assuming that the channel is prismatic. By varying
the flow depth value between y
dam
and y
n
in a systematic stepwise manner, and
applying Eq. (25.5) recursively, the extent to which the dam affects the water levels can
be determined. This is illustrated in example 25.1.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Direct step method has the following disadvantages:
Interpolations become necessary if the flow depths are required at specified
locations.
It is inconvenient to apply this method to non prismatic channels because the
crosssectional shape at the unknown location should be known a priori.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
25.3 Example 25.1
A wide rectangular channel having a bottom slope of 0.001 is carrying a flow of
3 m
3
/s/m. A control structure is built at the downstream end which raises the water
depth at the downstream end to 4.5 m. Determine the distance from the control
structure at which the flow depth is equal to 3.8 m. Manning n for the channel is 0.012.
3 m
2
/s
3.8 m
4.5 m
x =?
Control
Structure
Fig. 25.2: Definition sketch for Example 25.1
Solution
Divide the distance into two reaches as shown in Fig. 25.3. The flow depths at
sections 1, 2 and 3 are 3.8 m, 4.0 m and 4.5 m, respectively. Distances
1 2
x x and are determined as follows.
1
2
3
3.8 m
4.0 m
4.5 m
1
2
3
Fig. 25.3: Solution for Example 25.1
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Apply Equation 25.8 between sections 2 and 3. Here,
2 3 2
X X X ; 4.0 0.001;
2
3 2 0
Y = 4.5 m; Y m; S n =0.012 and q = 3.0 m /s = = = . Y
2
is
taken as upstream depth while Y
3
is taken as downstream depth.
2
X 499.3m =
Apply Equation 25.8 between sections 1 and 2. Here
1 2 1 2 1
X X X ;Y 4.0 Y 3.8 m; m = = = ; Y
1
and Y
2
are taken as upstream and
downstream depths, respectively.
1
X 199.7m =
Either Equation 25.8 for a wide channel or Equation 25.5 for any general
prismatic channel can be applied in the above step by step manner to determine
the entire water surface profile behind a control structure.
In this example, Y
n
=Normal depth =1.08 m. Therefore, flow depth varies from
4.5 m (Behind the structure) to 1.08 m (far upstream of the structure). Procedure
given in the earlier steps can be used to determine the extent to which the control
structure is affecting the water surface profile, by computing the location where a
flow depth of 1.08 m occurs.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
26.1 Standard Step Method
In the standard step method, flow depth at a specified location, y
d
is determined, given
the flow depth, Y
u
at another specified location. Consider the channel shown in Figure
26.1. In this channel, say Y
u
occurs at a distance X
u
from the reference point.
Discharge, Q, Channel bottom slope, S
0
, the roughness coefficient, n and cross
sectional shape parameters (which relate A, P and R to y) are also known. The problem
now is to determine the flow depth, Y
d
at the specified location X
d
(figure 26.1).
Fig. 26.1: Definition sketch for standard step method
y
u
(known)
y
d
(unknown)
Water surface
Flow
Datum
Channel Bed
X
d
(known)
Z
u
Z
d
u
d
u
d
Equation (25.3) ( ) ( ) ( )
2 2
d u
f
d d u u d u 0 d u
V V
y y S x x S x x 25.3
2g 2g
+ + + =
can be rewritten as
( )
( ) ( ) ( )
d u
2 2
f d u f
d u
d u d u 0 d u
2 2
d u
S x x S
Q Q
y y x x S x x
2.0 2
2gA 2gA
+ + = + + 26.1
In Equation 26.1, the flow rate (Q), the roughness coefficient (n), distances X
d
and X
u
,
the channel slope (S
0
), the flow conditions at section u (
u u u
y , and A ) are known.
Therefore the right hand side of Eq. (26.1) can be determined. On the left hand side, the
area, A
d
and the friction slope,
d
f
S are functions of the flow depth Y
d
. Thus we have
one equation (Eq. 26.1) in one unknown Y
d
. Therefore, Y
d
can be determined by solving
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Equation (26.1). Equation (26.1) is a nonlinear equation. Either trial and error or
numerical techniques such as bisection, Newton Raphson techniques etc. can be used
for solving Eq. (26.1).
For example, for a wide rectangular channel (assuming
u d
1.0 = = ), Eq. (26.1)
becomes
( )
( ) ( )
2 2
2 2 2 2
d u
d u d u 0 d u 2 10/3 2 10/3
d d u u
n q x x
q q n q
y y x x S x x
2gy 2y 2gy 2y
+ + = + + 26.2
In Eq. (26.2),
u d u 0
q, n, y , x , x , S andg are known, and so Y
d
can be determined by
solving this equation. Note that Eq. (26.2) is nonlinear.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
26.2 Example 26.1
A wide rectangular channel having a bottom slope of 0.001 is carrying a flow of 3
m
3
/s/m. A control structure is built at the downstream end which raises the water depth
at the downstream end to 4.5 m. Determine the flow depth at a distance of 1000 m
upstream of the control structure. Manning n for the channel is 0.012.
Y = ?
4.5 m
Control
structure
3 m
2
/s
1000 m
Fig. 26.2: Definition sketch for Example 26.1
Solution
Divide the distance into two reaches as shown in Fig. 26.3. Section 2 is 500 m
upstream of section 3, while section 1 is 500 m upstream of section 2. Flow
depths Y
1
and Y
2
are determined as follows.
1
Fig. 26.3: Solution for Example 26.1
2
3
1
2
3
y
1
=3.50 m
y
2
=4.0 m
4.5 m
500 m
500 m
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Apply Eq.(26.2) between sections 2 and 3. Here,
( )
d u
X X 500 0.001, = =
2
0 3
m, S n = 0.012, q = 3.0 m /s and y =4.5 m. Y
2
is taken as
upstream depth while Y
3
is taken as downstream depth. In this example, we
know Y
d
and we are determining Y
u
by solving Eq. (26.2). Equation (26.2)
simplifies to the following:
2
2 3.333
2 2
0.459 0.324
4.025 y
y y
= +
For different values of Y
2
, R.H.S. of the above equation is as shown below
y
2
(m) R.H.S (m)
3.99 4.0156
3.90 3.9267
3.999 4.0245
4.000 4.0255
Therefore y
2
=4.000 m.
Apply Eq. (26.2) between sections 2 and 3. Here,
( )
d u
X X 500m, 0.001,
2
0 2
S n=0.012, q = 3.0 m /s and y =4.0 m. = = Y
1
is taken as
upstream depth while Y
2
is taken as downstream depth. Here, we know Y
d
and
we are determining Y
u
. Equation (26.2) simplifies to the following:
1
2 3.333
1 1
0.459 0.324
3.532 y
y y
= +
For different values of Y
1
, R.H.S of the above equation is as shown below,
y
1
(m) R.H.S (m)
3.49 3.5227
3.50 3.5325
3.499 3.5315
Therefore, y
1
=3.50 m.
Either Eq. (26.1) for any general prismatic channel or Eq. (26.2) for a wide
channel can be applied in the above step by step manner to determine the flow
depth at any given location upstream of the control structure. Thus the entire
water surface profile behind the control structure can be determined.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
In this example, a trial and error method is used for solving the nonlinear
algebraic equation. One can use NewtonRaphson method for the same
purpose.
In this, as well as in the example 25.1, more accurate results for water surface
profile can be obtained by performing computations over more number of
reaches for the same distance.
Example 26.2
A rectangular channel of 6.0 m width carries a discharge of 12.0 m
3
/s. The channel
slope is 0.0001 and the Mannings n =0.018. There is a free over fall at the downstream
end of the channel. Determine the flow depth at a section 500 m upstream of the free
over fall. Use Standard Step method and one reach.
Solution
Unit discharge, q =12.0 / 6.0 =2 m
2
/s
Critical depth,
1/3
2
c
q
y 0.742 m
g
= =
Normal depth, y
n
:
2/3
n
n 0
n
By 1
q y S
n B 2y
=
+
B =6.0 m
S
0
=0.0001
n =0.018
Solving by trial and error; y
n
=2.8 m
y
n >
y
c;
therefore slope is MILD
Therefore, Critical depth occurs at the downstream end. Denoting section2 as the
section at the downstream end, section1 on the section at 500 m upstream of the free
over fall,
2/3
n
n
n
6y
y 3.6
6 2y
=
+
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
2 2 2 2
1 1 2 2
2 2 2 4/3 2 4/3
1 2 1 1 2 2
2
2 1 0
q q Q n x 1 1
Z y Z y
2
2gy 2gy A R A R
y 0.742 m
Z Z S x 0.0001*500 0.05
+ + = + + + +
=
= = =
1
2 2 4/3
1 1 1
1
1
1
0.2038 0.324
y 2.238
y y (R )
6y
R
6 2y
+ =
=
+
Solving by trial and error,
y
1
=depth of flow =2.25 m
Example 26.3
Solve the problem in Example 26.2 using two reaches
Solution
Section 1: Located at 500 m upstream of free over fall
Section 2: Located at 250 m upstream of free over fall
Section 3: Located at downstream end.
Consider sections 2 and 3
2 2 2 2
2 2 3 3
2 2 2 4/3 2 4/3
2 3 3 3 2 2
3
3 2
q q x Q n 1 1
Z y Z y
2
2gy 2gy A R A R
y 0.742M, X=250 m
Z Z 0.0001*250 0.025
+ + = + + + +
=
= =
With the above values, equation for y
2
is given as
( )
2
2 2
2 4/3
2
2
2
2 2
6y 0.2038 0.162
y 1.675 m; R =
6 2y
y
y R
+ =
+
Solving by trial and error,
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
y
2
=1.655 m
Consider Sections 1 and 2
2 2 2 2
1 1 2 2
2 2 2 4/3 2 4/3
1 2 2 2 1 1
2
2 1
q q x Q n 1 1
Z y Z y
2
2gy 2gy A R A R
y 1.655 m
x 250 m
Z Z 0.0001*250 0.025
+ + = + + + +
=
=
= =
With the above values, Equation for y
1
is given as
( )
1
2 4/3
2
1
1 1
1
1
1
0.2038 0.162
y 1.759
y
y R
6y
R
6 2y
+ =
=
+
Solving by trial and error,
y
1
=1.740 m
Flow depth at 500 m upstream of the d/s end =1.740 m. The above value is very much
different from the value of 2.25 m obtained when only one reach is considered. This
illustrates the importance of choosing a small value of x to obtain accurate results.
However, Computational effort increases if a very small value of x is chosen. Above
points should be kept in mind while performing GVF computations.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
27. Integration of Differential Equation
Computation of water surface profile can be done by numerically solving the nonlinear
ordinary differential equation (Eq. 23.7). Three methods: (i) Euler method, (ii) Improved
Euler method, and (iii) Fourth  order Runge  Kutta method, are presented here.
27.1 Euler Method
Referring to Fig. 27.1, say flow depth Y
i
at a distance X
i
from the reference point is
known. We also know the flow rate, Q, the roughness coefficient, n, the channel slope,
S
0
and the channel cross sectional shape parameters. We want to determine the flow
depth, Y
i+1
at a distance X
i+1
.
tangent line
y = f(x)
true value of
flow depth
y
y
i+1
y
i
X
i
X
i+1 x
Fig. 27.1: Schematic representation of the Euler method
Rate of variation of flow depth, y with a distance, x can be evaluated as follows:
( ) ( )
dy
f x,y 27.1
dx
=
Considering the point, i
( ) ( )
i i
i
dy
f x ,y 27.2
dx
=
From the governing equation (Eq. 23.7),
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
( )
( )
( )
i
0 f
i i
2
i
3
i
S S
f x ,y 27.3
Q T
1
g A
In Eq. (27.3), subscript i indicates the value of variable evaluated at distance X
i
.
i
f i i
S ,T andA are dependent on the flow depth Y
i
, and therefore, they can be
determined explicitly. Thus at any location ( )
i i
i, f x , y can be evaluated. This is nothing
but the slope of the line drawn tangent to the curve y =f(x) at x =x
i
(Fig 27.1). If we
assume that this value does not change in the interval from x
i
to x
i+1
, flow depth at x =
x
i+1
can be determined as follows.
( ) ( )
i 1 i i i
y y f x ,y x 27.4
+
= +
The above method is known as Euler's Method.
Once Y
i+1
is known, we can determine Y
i+2
at location x
i+2
by repeating the above
procedure. Referring to Fig. 27.1, it can be seen that there is a difference between the
estimated value of flow depth Y
i+1
and its true value. Taylor series expansion would
show that Euler's method is only first  order accurate. This error in the computation of
flow depth at each step may get magnified as the value of x increases, and therefore,
this method is usually unstable. Very small values of x may be required to obtain
satisfactory results.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Example
A wide rectangular channel having a bottom slope of 0.001 is carrying a flow of 3
m
3
/s/m. Flow depth at a particular location is 2.0 m. Determine the flow depth at a
distance 500 m downstream of this point. Manning n for the channel is 0.012. Use
Euler's method.
Solution
i
i
2
i i
2
i
y 2.0 m
T 1.0m(unit width)
A y * 1 2.0m
q 3m /s
1.0 ass
=
=
= =
=
= ( )
( )
( )
( )
( )
i
2 2
4
f
10/3
i
4
i i
2
3
4
umed
q n
S 1.2858*10
y
0.0011.2858 * 10
f x ,y
3
1
9.81* 2
9.843* 10
i+1
=
=
y
= =
( )
i
4
f x , x
2 9.843*10 *500
2.492m
Flowdepth at 500mdownstream 2.492m
i i
= y y
=
=
+
+
=
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
27.2 Improved Euler Method
We can improve the accuracy of Euler's method by using the slope at more than one
point. For example, say
*
i+1
y is the flow depth at
i+1
x obtained using Euler's method.
( ) ( )
i i i
y f x ,y x 27.5 = +
*
i+1
y
We can compute the slope of the curve at X =Xi+1 using the above value of
*
i+1
y .
( )
( )
( )
i 1
*
0 f
*
i 1 i 1
2 *
i 1
3
*
i 1
S S
f x , y 27.6
Q T
1
g A
+
+ +
+
+
where superscript * indicates the values obtained using the flow depth
*
i+1
y . Now, we
can determine the value of y
i+1
from the equation given below.
( )
( )
( )
*
i 1 i i i i 1 i 1
1
y y f x ,y f x ,y x 27.8
2
+ + +
= + +
This method is known as the "Improved Euler" method. It is secondorder accurate.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Example
A wide rectangular channel having a bottom slope of 0.001 is carrying a flow of 3
m
3
/s/m. Flow depth at a particular location is 2.0 m. Determine the flow depth at a
distance 500 m downstream of this point. Manning n for the channel is 0.012. Use
Improved Euler method.
Solution
( )
( )
i
i
2
i i
2
y 2.0 m
T 1.0 m (unit width)
A y 1 2.0 m
q 3 m /s
1.0 assumed
=
=
= =
=
=
( )
i
4
f
4
i i
*
i 1
*
i 1
*
i 1
S 1.2858*10
f x ,y = 9.843* 10
y 2.492 m
T 1.0 m
A
+
+
+
=
=
=
( )
[ ]
i 1
2
* 5
f
* 4
i i 1
4
i 1
2.492 m
S 6.1772 * 10
f x ,y = 9.9735* 10
1
y =2.0 + 9.843 9.9735 *10 *500
2
+
+
+
=
=
+
=2.495 m
Flow depth at 500 m downstream =2.495 m.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
27.3 Fourth  Order  Runge  Kutta Method
In the fourth order Runge  Kutta method, the slope of the water surface profile, f(x,y) is
determined as a weighted mean of four slopes as given below.
( )
( )
[ ]
[ ] ( )
1 i i
2 i i 1
3 i i 2
4 i i 3
i 1 i 1 2 3 4
s f x , y
x 1
s f x , y s x
2 2
x 1
s f x , y s x 27.9
2 2
s f x x, y s x
and
1
y y s 2s 2s s x 27.10
6
+
=
+ +
+ +
+ +
= + + + +
=
=
=
This method is fourthorder accurate.
Example 27.3.1
A rectangular channel of 5.0 m width carries a discharge of 10 m
3
/s. The channel slope
is 0.0001 and the Mannings n =0.018. Flow depth at a particular section in this channel
is 2.5 m. Determine the flow depth at a distance of 1000 m downstream of this section.
Solution
q =unit discharge =10 / 5 =2 m
2
/s
critical depth,
1/3
2
c
q
y 0.742 m
g
= =
Normal depth, y
n
2/3
n
n
n
2/3
n
n
n
5y 1
q y 0.0001
n 5 2y
5y
or y 3.6
5 2y
=
+
=
+
Solving by trial and error,
y
n
=2.945 m
In this case,
n c
y y > MILD SLOPE
c c 2
y y and y y M Profile > <
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Thus the water surface profile is of M2 type. Therefore, the flow depth decreases in the
downstream direction.
0 f
2
3
2 2
f
2 4/3
s s dy
s Slope
dx
q
1
gy
q n
S =
y R
= = =
Calculations for slope s
1
(based on y =y
i
=2.5 m)
q =2 m
2
/s
s
0
=0.0001
n =0.018
y =2.5 m; A =2.5 * 5 =12.5 m
2
P=5+2*2.5=10 m; R =1.25 m
2 2
4
f
2 4/3
4
5
1
2
3
q n
S = 1.54*10
y R
0.0001 1.54*10
s 5.544*10
q
1
gy
= =
Calculations for slope s
2
Depth for slope s
2
is given by
i 1
x
y s 2.472m
2
= =
y =2.472 m; A =12.36 m
2
P =9.944 m; R =1.243 m
Based on this; s
2
=6.032*10
5
Calculations for slope s
3
Depth for slope s
3
is given by
i 2
x
y s
2
+ =2.470 m
y =2.470 m; A =12.35 m
2
P =9.94 m; R =1.242 m
Based on this ; s
3
= 6.076 * 10
5
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
Calculations for slope s
4
Depth for slope s
4
is given by
i 3
y s x + =2.439 m
y =2.439 m; A =12.195 m
2
P =9.878 m; R =1.235 m
Based on this ; s
4
= 6.6286 * 10
5
( )
i 1 i 1 2 3 4
i
x
y y s 2 s s s
6
y 2.5 m
x 1000m
+
= + + + +
=
=
i 1
y
+
=Flow depth at 1000 m distance =2.439 m
Let us consider the solution for the above problem using Eulers method
5
i 1 i 1
y y s x =2.51000*5.544*10 2.445 m
+
= + =
Thus there is an error of 6 mm if Eulers method is used instead of Runge Kutta
method. The resulting error is significant in case the flow depth is close to the critical
depth. Consider the solution to the problem in Example 27.3.1 using the Eulers method.
y 0.78 m
[A slightly higher value is taken so that the singularity in GVF equation is avoided.
0 f
2
S S dy
dx
1 F
dy
dx
becomes infinity when F tends to one or when flow approaches Critical Conditions]
A =0.78 * 6 =4.68 m
2
P =7.56 m
R =0.619 m
3
0 f
2
3
S S dy 3.937*10
0.028
dx 0.1408
q
1
gy
= = =
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
[ ]
2 3
y y 250.0 0.028 7.77 m = =
This is obviously a wrong answer since flow depth in this case cannot exceed the
normal depth value of 2.8 m. Therefore, one has to watch out for numerical errors while
applying these schemes for GVF computation. These numerical errors can be reduced
by taking small values of x and by using higherorder methods such as Fourthorder
RungeKutta method.
Hydraulics Prof. B.S. Murty
Indian Institute of Technology Madras
27.4 HEC  2
HEC  2 is a very popular computer program developed by the Hydrologic Engineering
Center of the U.S. Army Corps of Engineering Center of the U.S. Army Corps of
Engineers, for the purpose of computation of water surface profiles in natural and man
made channels. This software is based on the application of standard step method for
solving the onedimensional, steady gradually varied flow equation. HEC RAS model is
a newer version of HEC  2 model. This model can simulate flow through single, tree
type, and fully looped systems. It can simulate sub critical flow, supercritical flow, and a
mixture of both within a system. Effect of bridges, weirs, gates and culverts on water
surface profile can be computed. An excellent Graphic User Interface for input and
output handling makes this software very simple to use even for complex river systems,
with many different hydraulic structures, and with irregular crosssections. More details
about this software package can be found in "www.hec.usace.army.mil".
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Normal Hydraulic Jumps NHJ
Click for Video link
28.1 Introduction
When the depth of flow changes rapidly from a low stage to a high stage, it results in an
abrupt rise of water surface. This local phenomenon is known as 'hydraulic jump'. It
occurs in a canal below a regulating sluice, at the toe of a spillway or at the place where
a steep channel slope suddenly turns flat. It is well known that a large amount of
air entrains in the roller portion of the jump due to the breaking of the water surface.
Consequently a large amount of energy loss occurs in the jump through dissipation in
the turbulent body of water. A considerable amount of investigations, both theoretical
and experimental, have been carried out on the jump (See box  History).
Sluice
Gate
1
Hydraulic
jump
2
3
Fig. 28.1  Rapidly varied flow with Hydraulic jump (1 and 3 subcritical flows,
2 Super critical flow)
Fig. 28.2  Formation of Hydraulic jump at the toe of the spillway
Hydraulic jump
Toe
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
History
The earliest description of the hydraulic jump appears to be by Leonardo da Vinci
in 1452 1519. Bidone was the pioneer to conduct investigations on the hydraulic
jump in 1818 1819. Belanger in 1828 developed the momentum equation
connecting the sequent depths. Then onwards innumerable contributions have
been made towards the understanding of the basic mechanism of the hydraulic
jump. The following are some of the significant contributors amongst several
investigators responsible for present state of knowledge of the jump: Bresse (
1860 ), Darcy and Bazin ( 1865 ), Uniwin ( 1875 ), Ferriday and Merriman ( 1894
), Gibson ( 1913 ), Kennison ( 1916 ), Woodward and Riegel Beebe ( 1917 ),
Koch and Cartstanjen ( 1926 ), Lindquist ( 1927 ), Safranez ( 1917 ), Einwachter
( 1933 ), Smetana ( 1934 ), Bakhmeteff and Matzke ( 1936 ) , Escande ( 1938 ),
Citrini ( 1939 ), Nebbia ( 1940 ), Kindsvater (1944 ) , Blaisdell (1948 ), Forster
and Skrinde (1950 ), Moore and Morgan ( 1957 ), and Rouse et al. (1958 ). A
detailed mathematical treatment of hydraulic jump was made by Flores (1954 ) .
Rajaratnam's contributions to the knowledge of hydraulic jumps during 1960s are
outstanding. For a comprehensive bibliography on the jump, reference may be
made to the following references: 'The standing wave or hydraulic jump ( 1950 ),
(Central Board of Irrigation and Power)' ; ' A bibliography on hydraulic jump
(Central Board of Irrigation and Power) ( 1955 )'; ' Hydraulic energy dissipators (
1959 )' ( Elevatorski  The Hydraulic jump, May 1955); Hydraulic Energy
dissipators Elevatorski  Mcgraw Hill, 1959) 'Open Channel Hydraulics Chow
V.T., McGraw Hill ( 1959 )';'Advances in Hydroscience (Hydraulic jump by
Rajaratnam.N edited by Chow.V.T. Vol.  4 , Academic Press New york and
London, Page 197 to 280 ( 1967 )' ; Self Aerated flow characteristics in
developing zones and in Hydraulic jumps, (Thandaveswara Phd Thesis, Indian
Institute of Science, Bangalore, J une 1974).
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Hydraulic J ump has many practical applications, for example (a) to dissipate the high
kinetic energy of water near the toe of the spillway and to protect the bed and banks of
a river near a hydraulic structure (b) to increase the head in the power channel (c) to
remove air pockets from pipes (d) for mixing of chemical in water supply system.
Figure below shows a schematic view of the classical hydraulic jump on a horizontal
floor. The details in this unit are confined to the case of the hydraulic jumps on level
floors in rectangular channels and this type of jump is referred to as the Normal
Hydraulic J ump (NHJ ). The supercritical Froude number of the approach flow is the
major parameter that influences the characteristics of the hydraulic jump.
L
j
L
rj
Toe
1
2
y
2
V
2
F
1
y
1 V
1
y
28.3 SCHEMATIC VIEW OF THE HYDRAULIC J UMP
1
2
x
Hydrostatic pressure distribution
Roller zone
y
r
Super critical to sub critical
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
28.2 Classification of Jumps
As mentioned earlier, the supercritical flow Froude number influences the
characteristics of the hydraulic jump. Bradley and Peterka , after extensive experimental
investigations, have classified the hydraulic jump into five categories as shown in
Figure 28.4.
The hydraulic jump is the phenomenon that occurs where there is an abrupt transition
from supercritical (inertia dominated) flow to sub critical (gravity dominated) flow. The
most important factor that affects the hydraulic jump is the initial Froude number
1
F .
1
1
V
F
gD
=
in which
1
V is the longitudinal average velocity at the initial section, g is the acceleration
due to gravity and D is the hydraulic mean depth (ratio of area of flow at free surface
width).
As mentioned above, it occurs in a straight prismatic horizontal channel of rectangular
shape in which boundary friction is negligible (NHJ ).
The hydraulic jump can be classified based on initial Froude number as
Undular( )
1
1 17 F . = , weak( )
1
17 25 F . . = , oscillating jet( )
1
25 45 F . . = , steady
( )
1
45 90 F . . = and strong( )
1
90 F . > .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Undular jump 1.0 < F
1
< 1.7
Weak jump 1.7 < F
1
< 2.5
Oscillating jump 2.8 < F
1
< 4.8
Steady jump 4.5 < F
1
< 9.0
Strong jump F
1
> 9.0
Fig. 28.4  Classification of the J ump
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Type of J ump Froude Number Remarks
Critical flow
1
F 1 = Wavy surface , celerity
c gy =
Undular jump
1
1 < F 17 . < Undulations on the
surface
Weak jump
1
1.7 < F 25 . <
Small rollers, No baffles.
Oscillating J ump
1
2.5 < F 45 . <
No periodicity. Rip rap
may get damaged
Canal drops, difficult to
handle.
Baffle blocks or
appurtenances are of
little value. Wave
suppressors may be
designed.
Steady jump
1
45 < F 90 . . < Position, is sensitive to
variation of Tail Water,
Efficiency is 45 to 70 %.
Strong J ump
1
F 90 . >
Efficiency is 85 %
The jumps can also occur on horizontal bed or sloping bed. The jump can take place in
radially diverging, radially converging, rectangular, sudden convergence or expansions
in plan. The jump can occur in different shape of the cross section of the channel such
as rectangular, trapezoidal, parabolic, circular channels. The jump can occur in the
conduit either at the free surface or fully flowing downstream condition. The annular
jump is yet another type. The jump can be either free (unsubmerged) or submerged
condition such as in the downstream of sluice gates. The jump can be a forced one with
the appurtenances (such as baffles, sills, chute blocks) or free (i.e. either without any
appurtenances). J ump could be either stationary or moving (hydraulic bore).
The jump can be in stratified flows such as warm and cold water (flowing over each
other), air and water (classical jump) or in case of gas to gas (internal jump).
The Important macroscopic parameters are initial depth
1
y , sequent depth
2
y , Initial
mean velocity
1
V , mean velocity at the end of the jump (exit velocity)
2
V , length of jump
( )
j
L and the roller
( )
rj
L .
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Classifications of Jumps
I. Based on
Froude
Number
Undular, weak, oscillating, steady and strong (See the figure  2 above)
II. Based on
Bed Slope
Horizontal
,
Sloping
III. plan shape
of boundary
Rectangular
Radial Diverging Channel
Radial Converging Channel
Sudden Expansion
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
IV. Shape
y
1
y
2
Rectangular
T
b
m
1
m
1
Trapezoidal
y
1
y
2
T
m
1
m
1
y
1
y
2
Triangular
T
d
o
Circular free surface flow
y
1
d
o
y
1
Circular free surface to Pressure flow
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
V. Fluid Status
Water
liquid
Gas
Air
Classical Hydraulic jump
Gas to Gas Internal jump
liquid (warm)
2
liquid (Cold)
Liquid to Liquid (
2
>
1
)
VI.
Submergence
Sluice
Gate
1
Hydraulic
jump
2
3
Rapidly varied flow with Hydraulic jump (1 and 3 subcritical flows,
2 Super critical flow)
Free J ump (Not Submerged)
Sluice
Gate
1
2
3
Rapidly varied flow with Hydraulic jump (1 and 3 subcritical flows,
2 Super critical flow)
Submerged J ump
Tail water depth is greater
than the sequent depth
VII. Motion
Stationary J ump
Moving
J ump
Example:
Hydraulic
Bore
VIII.
Appurtenances
unforced or Free
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Free J ump (Unforced)
(without Appurtenances)
Forced
Chute block
Baffles or Piers or floor blocks End Sill
Forced J ump (with Appurtenances)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
J ump in gradual expanison  looking downstream
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
28.3 Momentum equation
The ratio of sequent depth
2
0
1
=
y
y
y
may be computed assuming
hydrostatic pressure distributions
uniform velocity distributions
air entrainment is negligible and
timeaveraged quantities
in sections 1 and 2.
Belanger's momentum equation for sequent depths of a hydraulic jump on a level floor
in a rectangular channel can be derived by applying momentum equation between
sections 1 and 2 as given below.
2 2
1 2
1 2
1 1 1 2 1 2 1 1 2 2
1 1 1 1 1 2
1 2
2 2
2 2
2 2 2
2 1
1
1 2
2 2 2 2
1 1 2
2 2
1 1 2
2 2 2 2 2
2
1 1
1 2 3
1 1 1
Q Q
ZA ZA
gA gA
For a rectangular channel A by , A by , Q VA V A ,
VA V y y y
V ,V , Z , Z ,
A y 2 2
Q y Q y
by b
gby 2 gby 2
Q y Q y 1 y
gyb 2 gb y y 2
V Q b y Q
F
gy gy gb y
di
+ = +
= = = =
= = = =
+ = +
+ = +
= = =
2
1
2
2 2 2
1 1 2
3 2 2 2 3
1 1 1 2 1
2
2 2
1 2
1 1
2 1
vided by y
Q y Q y 1 y 1
0
gy b 2y gb y y y 2
1 y y 1
F F 0
2 y y 2
+ =
+ + =
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
( )
( )
2
2 2 2
1 1 2
3 2 2 2 3
1 1 1 2 1
2
2 2
1 2
1 1
2 1
2
2 2
1 2
1 1
2 1
3
2 2
2 2
1 1
1 1
3
2 2
2 2
1 1
1 1
Q y Q y 1 y 1
0
gy b 2y gb y y y 2
1 y y 1
F F 0
2 y y 2
y y
2F 1 2F 0
y y
y y
2F 1 2F 0
y y
y y
2F 1 2F 0
y y
This can be r
+ =
+ + =
+ =
+ =
+ + =
2
2
2 2 2
1
1 1 1
2
2 1
1
2
2
2 2
1
1 1
2
2 1 2
1
1
ewritten as
y y y
2F 1 0
y y y
y
1 0 y y uniform flow.
y
y y
2F 0 a quadratic equation.
y y
1 1 8F y 1
Hence 1 8F 1
y 2 2
+ =
= =
+ =
+ +
= = +
(28.1)
= +
y
1
2 2
1 8F 1
1
y 2
1
in which y
2
, y
1
are sequent and initial depths respectively and
1
1
1
V
F =
gy
is the initial
Froude number. Equation 28.1 has been verified by many investigators experimentally
and often a ratio lower than the one calculated by the equation has been recorded.
Belanger , did not consider the bed shear force while deriving Eq. 28.1. Rajaratnam in
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
1965, proposed the following momentum equation taking into consideration the
integrated shear force.
(28.2)
+ +
3
y y
2 2 2 2
1 2F 2F = 0
1 1
y y
1 1
In which is the non dimensional integrated shear force, given by
f
2
1
P
y
2
and is a
function of Froude number. P
f
is the integrated shear force.
He used the data of Rouse et al. , Harleman, Bakhmeteff ,Safranez , Bradley  Peterka ,
along with his own. Figure 2 shows the effect of shear force on sequent depth ratio.
Belanger
Rajaratnam
Sarma and Newnham
Eq. 28.1
Eq. 28.3
Eq. 28.2
0
2
4 6 8
10
2
4
6
8
10
12
14
F
1
Fig. 28.5  Variation of sequent depth ratio
y
2
y
1
___
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Sarma and Newnham 1975 introducing the momentum coefficient (
j
1.045 = ) for the
non uniform velocity distribution obtained the following modified momentum equation
(28.3)
= +
y
1
2 2
1 10.4 F 1
1
y 2
1
In Eqn. 28.3, a value of
j
was used by them based on the assumption of a similarity
profile for the velocity distribution. Eq. 28.3 gives a higher value for the sequent depth
ratio, compared to the value computed from Eq.28.1. Their analysis was carried out
upto a Froude number value of 4.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
28.4 General Hydraulic Jump Equation
The internal equation of motion for a hydraulic jump can be integrated in two ways
whether the control volume is defined on a microscopic scale. The hydraulic jump takes
place over a short distance (of the order of five times the sequent depth) the transition is
dominated by initial momentum flux and pressure forces due to sequent depth.
Boundary shear forces are secondary in nature. Consider the situation shown in
Figure 28.1 (unsubmerged, forced hydraulic jump in a radially diverging sloping
channel). The macroscopic approach is as follows:
Ps /2
V2
V1
Ps /2
Plan
__
__
z
_
y
2
y
1
+
+
1
z
__
2
Section 1
Section 2
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
y1
y
Wcos
Ff
V1
P1 V2
x
y2
FD
1
2
Figure 28.1  Schematic diagram of a hydraulic jump in free surface flows
P2
Longitudinal Section
__
__
Using the Greens theorem, the Reynolds equation for turbulent flow can be integrated
over the control volume V to obtain.
x
x x
j
i i
d + d =  p d + x d V
i
i j i j
u u u u
( ) ( ) ( ) ( ) ( )
u j
i
+ d (28.1)
j
1 2 3 4 5
x
x
in which is outwardly directed normal.
The first term represents the net flux of momentum through the boundary , due to
mean flow. The second term represents the net momentum transfer through the
boundary due to turbulence. The third term represents the pressure force (resultant
mean normal) exerted on the fluid boundary . The fourth term represents the net
weight of the fluid within the control volume V and the fifth term represents mean
tangential force exerted on the boundary .
A macroscopic momentum equation is obtained if the above equation is applied to the
control volume V , shown in Fig 1.
Consider the momentum in the x direction, then it can be written as
{ } ( ) 2 1
2 1 2 1 1 2
V V (28.2)
2
+ = +
s D f
Q Q I I P P P sin F W sin F
in which, pressure force, is force on the side wall.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
2
dA/ QV , I = v' dA, P = ' gAz cos v =
, pressure force
s
P , is force on the side
wall.
The continuity equation is
1 2
1 2
Q = V A = V A (28.3)
Equations 2 and 3 are to be solved simultaneously to determine the sequent depth,
velocity
( )
2 2
v , y for given initial condition ( )
1 1
v , y . If
0
0 S = , rectangular channel without
baffles, and no side thrust, then it simplifies to the standard format (equation 4)
2 2
2 2 2 2
1 2
1 1 2 2
V V
2 2
y y y y
g g
+ = + (28.4)
1 2
1 2
V V y y = (28.5)
When solved results in
( )
3 2 2
0 1 0 1
2 1 2 0 y F y F + + = or
( )
2
0 1
1
1 8 1
2
y F = + (28.6)
in which
0
y is the sequent depth ratio
2
1
y
y
.
Bed friction decreases the ratio by about 4% at
1
10 0 F . = . It is to be noted that the
macroscopic approach yields only sequent depth ratio and no information regarding
surface profile or the length of the jump. In radial stilling basins, sloping basins, forced
hydraulic jump even the sequent depth ratio depends on the internal flow and hence the
physical model is used for determining.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
29.1 Energy loss in the Jump
The loss of energy E in the normal hydraulic jump is equal to the difference in specific
energies before ( E1 ) and after ( E2 ) the jump and can be shown to be equal to
( )
3
y y
2 1
E = (29.1)
4 y y
1 2
Show that
( )
3
2 1
1 2
y y
E
4y y
( )
( )
( )( )
2
1 2
2
1 2 1 2 2
y y
Q 1
y y y y (1)
2gb 4
= +
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Specific Energy Equation
( )
( )
( )
( )
2 2
1 2
1 2
2 2
1 2 2 2 2 2
1 2
2
1 2 2 2 2
1 2
2 2 2
2 1
1 2 2 2 2
1 2
2
2 1
1 2 2 2 2
1 2
V V
y y E
2g 2g
Q Q
y y E
2gy b 2gy b
Q 1 1
y y E
2gb y y
Q y y
y y E
2gb y y
y y Q
E y y 1
2gb y y
Substituting equatio
+ = + +
+ = + +
+ =
+ =
+
=
( ) ( )
( )
( )
( )
( )
( )
( )
( )
2 1
1 2 2 1 1 2 2 2
1 2
2
2 1
1 2
1 2
2 2
1 2 1 2 1 2
1 2
1 2
3
2 2 1 2 1 2
1 2 1 2
1 2 1 2
n (1)
y y 1
E y y 1 y y y y
4 y y
y y 1
y y 1
4 y y
4y y y y 2y y
y y
4 y y
y y y y
E y y 2y y
4y y 4y y
+
= +
+
=
=
= + + =
or the relative energy loss can be written as
+
=
+ +
3
2
1 8F 3
1
E 1
(29.2)
E 8 2 2
1 (2 F ) [ 1 8F 1]
1 1
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
( )
{ }
( )
{ }
( )
3
2 1
1 2 1 2
2
1 1 1
1
3 3
3 2
2 1 2 1
1 1 2 1 2
2
2
1
1 1
1
1
3
2
2 1
1
2
2 1 1
3
2
1
2
1 2
y y
E E E 4y y
V E E
y
2g
y y y y
1 1
y 4y y y 4y
=
y
y V
2 F
2
2
2 2gy
1
1 8F 1 1
y 2
2
= . *
4y y 2 F
1
1 8F 1 1
1
2
= *
2 F y
2
y
= =
+
=
+
+
+
+
+
+
( )
( )
( )
( )
( )
( )
1
3
2
1
2
2 2
1 1
3
2
1
2 2
1
1 1
1
1 8F 1
2
=
1
2 F 1 8F 1
2
1+ 8F 3
E
=
E
8 2 F 1 8F 1
+
+ +
+ +
It can be seen from Eq.29.2 that when F1 is 20, the relative loss would be 86% but to
dissipate 99.6 % of the initial energy a Froude number of 1000 is required !. The energy
in the jump is dissipated through the conversion of mean kinetic energy into turbulence
and through viscous action. Rajaratnam ( 1967 ) presented the following equation to
predict the energy profile along the jump.
+
=
2 "
y / y 0.632 F f ( )
1 1
y
E
1
(29.3)
2
E
F
1
1
1+
2
x
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
He evaluated the value of ( x / y
1
) using experimental results. Equation 29.3 shows
that the energy profile falls rapidly in the initial part of the jump and approaches the
downstream energy line asymtotically. This analysis includes the turbulent kinetic
energy and the work done by the Reynolds stresses.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
29.2 Turbulent Characteristics of the Jump
To understand the turbulent characteristics of hydraulic jumps, one of the major
difficulties is the number of entrained air bubbles that form the fluid discontinuities.
Rouse et al. simulated the jump in an air duct shaped to match the profile of the jump.
This investigation was conducted based on the assumption that if the mean flow
patterns were similar and the energy changes and the Reynolds number are
comparable than the patterns of turbulence would also be similar. The mean velocity
distribution was similar to Rajaratnam's finding except that the scale of the maximum
velocity , V
max
, was very much lower.
Their analysis of the turbulent intensities and the product of the mean components,
indicates that the fluctuating velocity is small near the toe and large in the regions of the
maximal velocity gradient in the longitudinal distance x / y2 between 1.0 and 2.0 and
again becomes small towards the end of the jump. The maximum values of
2 2
' '
and
v u
V V
were about 27 % and 20 % respectively for a Froude number value of
6.0. The turbulent shear stress also exhibits maximum values in the zones of maximal
velocity gradient.
Along the jump turbulence production, dissipation and convection occurs. The
turbulence production is greatest in the regions of greatest velocity gradient, being at a
maximum at x / y
2
= 1.0 and the maximum dissipation occurs at a later section.
Production of turbulence become very small at x / y
2
>5.0.
Using the momentum equation, Rouse et al. further computed the mean momentum
flux, integrated shear stress, turbulent momentum flux, and pressure assuming
hydrostatic pressure distribution and negligible turbulence level at the toe.
Hubbard in 1959 conducted some investigations in a hydraulic jump in an open channel
regarding the velocity fluctuating component. He chose the section just downstream of
the roller where the longitudinal turbulent velocity component was measured along a
vertical line. He observed that the fluctuations were so erratic that only a rough
indication of the root mean square value or the instantaneous mean velocity could be
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
obtained. He concludes that due to the lower velocity of water the frequency spectrum
of turbulence shifted downwards. Figure shows a typical record obtained by him. It can
be seen that bursts of high intensity turbulence are interspersed with a period of
relatively tranquil flow.
) a ( (b)
0.1 sec t
u
'
(a) :
y
y2
=0.65 ,
=0.4 ,
__
y
y2
__
y
2
=0.9
x
__
y2
=1.4
x
__
Fig. 29.1  A TYPICAL TURBULENCE FLUCTUATION RECORD
(b) :
(Hubbard, Tr. ASCE, V 124, pp 962  964, 1959)
Resch and Leutheuser measured turbulent intensities in jumps for Froude numbers 2.85
and 6.0 with two different approach flow conditions. They concluded that the turbulent
structure in the jump strongly depends on the supercritical flow characteristics. Fully
developed upstream flow leads to a state of jump turbulence underlaid by a wall
turbulence layer. They found that the wall turbulent layer is totally absent in case of
undeveloped upstream flow cases. They concluded further that the length of the jump
directly depends on these flow features. Later they investigated the Reynolds stress
characteristics in the jump. They used the earlier technique of eliminating the effect of
air bubbles which consisted of digitising the turbulent signal to permit using a computer
programme by means of which liquid phase points are separated from air phase points,
after choosing a characteristic length. This confirmed their earlier results that the
turbulent structure in the jump is influenced by the upstream flow conditions. In their
investigations they have assumed that the air bubble do not play any significant role and
do not affect the turbulence structure.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
29.3 Pressure Distribution in the Jump
Usually the pressure distribution in the hydraulic jump is assumed to be hydrostatic.
Rajaratnam in 1965 showed, however, that the pressure distribution is non hydrostatic.
Also he pointed out from his result that there exists a narrow region near the wall
(around 5 % of the depth) exhibiting hydrostatic pressure distribution and the deviation
from the hydrostatic pressure distribution is more in the earlier reaches of the jump and
increases as the Froude number increases. The pressure profiles indicate adverse
pressure gradient and vary in a complex manner with the distance. The profile
configuration is controlled by the supercritical Froude number.
Vasiliev and Bukreyev found that the frequency of the pressure fluctuation distribution
greatly differs from the normal distribution in the initial portion of the jump including the
roller and just downstream of it. They found that the most intensive fluctuations and the
widest spectra take place in
2 2
' '
and
v u
V V
, and the coefficient of pressure variation
is 17% in this zone. The diminution of the spectra ordinate may be approximated by a
power law with an exponent of 1.5 to 2.0.
King in 1967 showed the largest pressure fluctuations occur at a frequency of about 15
Hz in a model. He cautions that the fluctuation of 5 Hz to 1 Hz should be avoided. Later
Bowers and Tsai in 1969 found that most of the energy is contained in frequencies less
than 1 Hz. From the pressure records, Lesleighter from that the peak to peak
fluctuations, during the interval of less than 0.5 sec . was of the order of 50% of the
maximum velocity head in the prototype. His results indicate larger uplift pressures with
Skewness.
In 1971, Sadasivan found that the maximum pressure fluctuation occurs between
1
x
8 12
y
from the toe and the magnitude increases with Froude number. Correlation
of non dimensional spectral density with nondimensional frequency showed a power
law variation with an exponent of 1.7.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
29.3.1 Pressure fluctuations in the hydraulic jump
Pressure fluctuations in the hydraulic jump have been studied by many investigators by
mounting the pressure transducers on the stilling basin floor. The fluctuating pressure
can be characterized by rms value. Figure shows the dimensionless rms value as a
function of distance relative to the length of roller
( )
07 =
rj r
L . L . The peak
dimensionless rms pressure of 0.05 to 0.082 occur at 30 % to 40 % from the roller
length or at about 25 % the length of the jump.
Typical rms pressure versus X/L, at bed of hydraulic jumps
X/L
r
2
1
v
1
2 __
2
Khader and Elango
Free jump
4.7 < F
1
< 6.6
Bourkov
Free jump
Schiebe
and
Bowers
Vasiuev
and
Bykreyev
Wisner
(pipe)
0.02
0.04
0.06
0.08
0.10
0
0.5
1.0
Note: The fluctuating pressures in the Hydraulic jump formed on the stilling basin are
measured using Pressure Transducer. The figure shows the results obtained by Khader
and Elango in 1974. The above figure shows the variation of dimensionless rms peak
fluctuations obtained by various investigators analysed by them.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
29.4 Velocity Distribution in Hydraulic Jump
The approaching uniform flow velocity imparts some amount of energy to the ambient
fluid which changes the velocity distribution. After the jump, the variation of depth, flow
pattern, and air entrainment influence the velocity distribution. There has not been much
work on the velocity measurements particularly in the roller zone. Experiments by the
Miami conservancy District in 1917, clearly show that the approaching high velocity of
the water gradually diminishes through the jump.
Later Hubbard recorded the turbulent fluctuation in a hydraulic jump just downstream of
the roller, to compare the results of the air model investigated by Rouse et al.
However, it appears that it is Rajaratnam in 1965 who rationalised the analysis. He
conducted an extensive investigation of the mean velocity distribution in the jump
formed just below a sluice gate in a smooth channel in a Froude number range of 2.68
to 9.78. His measurements were confined to forward flow. He compared his results with
the wall jet and he was able to show the existence of similarity law of velocity
distribution. further, he concluded that the velocity in the boundary layer follows the
defect law and hydraulic jumps, the pressure gradient is adverse and its effect must be
felt as observed by Clauser.
Resch and Leutheusser in 19711972, have measured the turbulent velocity fluctuations
both in forward and backward flow of the jump.
An understanding of the velocity distribution is necessary when energy loss is to be
computed. However, it is usual to assume a uniform velocity distribution. Till recently
there had not been much work on velocity measurements in hydraulic jumps and
particularly in the roller zone. Miami Conservancy district conservancy report shows that
the velocity of the water gradually diminishes through the hydraulic jumps.
Hubbard and Rajaratnam investigated about the velocity distribution in jumps. The latter
conducted extensive investigations on the velocity distribution. His measuremets were
confined to forward flows in a Froude number range of 2.68 to 9.78. His analysis
followed the analogy of a wall jet.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
In the following paragraphs the results of the Thandaveswara are reported in which an
attempt is is made to measure the velocity in the roller zone also.
In Figures a to f the normalised velocity ( V/ V
1
max ) distribution along the jump have
been plotted against the normalsied depth ( y / y
1
) in which V
1
max is the maximal
velocity of the appraching flow and y
1
is the depth of the approaching flow. The velocity
profile rises sharply up to the maximum velocity of the flow and then decreases
gradually as to the depth increases and finally becomes zero. These velocity profiles
exhibits similarity with the wall jet velocity profile as discussed by Rajaratnam. In
backflow the roller zone, is shown in dotted lines in Figures a to f. The negative sign
indicates only the direction. These components are only approximate, as the roller is full
of eddies and even the presence of a pitot tube will cause disturbances and affect their
characteristics. J ust downstream of the roller, the velocity profiles begin at a higher
level. In this region the flow becomes almost static and full of vortices. The presence of
vortices is discussed elsewhere. Farther downstream of this region the flow reverts to
nearly uniform flow.
In Figures g to h. the variation of the velocity profile along the jump is presented for the
PHJ . This also exhibits a sharp rise up to the wall turbulent zone. As observed in the
NHJ , there exists a zone near the bed where the flow is anticlockwise and the velocity
profile shown is only for the main flow direction. Later, flow returns to the normal
condition.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0 0 0 0 0 0 0 0 0
y
___
y1
0
4
6
8
10
0 0.4 0.8 1.2
Run R1
v
v1
_____
max
(a)
0 0.4 0.8 1.2
v
v1
_____
max
0
2
4
6
8
10
12
0 0 0 0 0 0 0 0 0 0
Run R2
y
___
y1
(b)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
0 0.2 0.8
0 0 0 0 0 0 0 0 0
0
2
4
6
8
10
12
14
16
Run R3
y
y
1
___
v
v
1
_____
max
Velocity Distribution in the Jump (NHJ)
(c)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
y
y
1
v
v
1
_____
max
0
0.4 0.8
Velocity Distribution in the Jump (NHJ)
0
2
4
6
8
10
12
14
16
18
Run R4
___
(d)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Velocity Distribution in the Jumo (NHJ)
0 0.4 0.8
0
2
4
6
8
10
12
14
16
18
20
0 0 0 0 0 0 0 0 0 0
y
y
1
v
v
1
_____
max
___
Run R5
0
1.2
(e)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Velocity Distribution in the Jump (NHJ)
0 0.4 0.8
0
2
4
6
8
10
12
14
16
0 0 0 0 0 0 0 0 0 0
y
y1
v
v
1
_____
max
___
1.2
0 0
Run R6
(f)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Velocity Distribution in the Jump (PHJ)
0 0.4 0.8
0
2
4
6
8
10
0 0 0 0 0 0 0 0 0
y
y
1
v
v
1
_____
max
___
1.2
*
Run B0
(g)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Velocity Distribution in the Jump (PHJ)
0 0.4 0.8
0
2
4
6
8
0 0 0 0 0 0 0 0 0
y
y
1
v
v
1
_____
max
___
1.2
*
0
Run B2
(h)
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
29.5 Length of the Jump
The length of the jump is an important factor in the design of stilling basins. The
beginning of the jump or the toe of the jump can easily be fixed as the mean position of
the oscillation at the abrupt rise of the water surface. But there has not been any
general accord as to the location of the end of the jump and has become a controversial
issue . Riegel Beebe ( 1917 ), Ludin ( 1927 ), Woycicki ( 1931 ), Knapp (1932 ),
Safranez ( 1933 39 ), Aravin ( 1935 ), Kinney ( 1935 ), Iranchenko, Chertoussou, Page
(1935), Bakhmeteff and Matzke (1936), Douma ( 1934 ), Posey ( 1941 ), Moore ( 1943 )
, Wu ( 1949 ), and Bradley  Peterka ( 1955 57 ) are some of the investigators , who
have proposed definitions for the length of jump. In the following paragraphs the relative
merits and demerits of some of the definitions are discussed.
Bakhmetoff and Metzke who were the first to investigate systematically the longitudinal
elements of the jump, took the end of the jump as the section of maximum surface
elevation before the drop off caused by the channel conditions downstream. In fact,
because of the flat nature of the water surface , they could only mark out a region in
which the end of the jump could be arbitarily fixed. The jump lengths as given by
Bakhmeteff and Matzke are somewhat shorter than the jump lengths produced in wider
channels probably because they are affected by the friction of the narrow width of the
flume.
Stevens while discussing the paper by Bakhmeteff and Matzke propossed that the
length of the jump is a result of two motions : first the translatory motion of the water
prism downward and secondly the vertical motion due to the rate of conversion of
kinetic to potential energy .
Another definition which seems to have found favour with earlier investigators is that the
end of the jump may be taken as the end of the surface roller. But it has been
confirmed, firstly by the experimental results reported by Mavis and Luksch (1936) and
later by Rouse et al. that the lengh of the jump is always greater than the length of the
roller.
Hydraulics Prof. B.S. Thandaveswara
Indian Institute of Technology Madras
Behera and Qureshy, and Qureshy ( 1947 ) defined the length of the jump as the
distance between the well defined toe and the section at which a cylinder placed in the
flow on the floor of the channel will just topple. At first the cylinder should be placed far
downstream and moved back upstream until it is toppled by the flow. The shape and
size and weight of the cylinder influences the fixing of the length in addition to the forces
exerted on the cylinder being affected by the boundary layer near the channel. This
definition is from personal error but of little use for designing purposes.
Bradley and peterka ( 1957 ) in their investigations on the stilling basins, have defined
the end of the jump as the section at which the high velocity jet begins to leave the floor
or immediately downstream of the roller, whichever is farther away from the toe of the
jump. Instead of defining the end of the jump as the section at which the high velocity jet
begins to leave the floor ( which does not eliminate the personal error completely ) it
would have been better if the bed velocity at the end in the downstream channel had
been chosen as a certain percentage of the appeoach velocity as suggested by
Rajaratnam ( 1961 ).
Elevatorski's ( 1955 ) definition for the end of the jump is also not much different from
the previous roller end definition. Rama Muthy ( 1960 ) defined the length of the jump as
the distance from the toe of the jump to the section where the flow depth reached a
value of 98% of the tailwater level. This definition is free from personal error and agrees
with the findings of earlier investigators. However, it is not theoretically sound as the
jump is not an asymptotic phenomenon. Nevertheless, it is very useful because of its
simplicity. However, this again suffers from the problem of finding the sequent depths
accurately.
Rajaratnam in 1961 suggested two criteria for the length of the jump. The first criterion
is based on the fact that the mean energy is first transformed into turbulence which later
decays through viscous shear. Based on the results of Rouse et al. it can be concluded
that the turbulent velocity components become uniform through the depth and decay in
the longitudinal dire
Much more than documents.
Discover everything Scribd has to offer, including books and audiobooks from major publishers.
Cancel anytime.