Nptel PDF

Hydraulics Prof. B.S.
Thandaveswara
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" - JOHN.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 (1863-1886) 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 1874-79, 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.
Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara
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. John'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

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 July 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 North-Western Province and director of the proposed Ganga Canal. In
1838, Cautley submitted to the government the first proposal to take a canal from

Ganga at Haridwar.
Governor General Lord Hardinge visited the site personally and authorized the
construction of canal in 1842.
James 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 Cot-ton’s 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 Cotton’s 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

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 January 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
June.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

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 irrigation-cum-navigation facilities in the Delta by building an anicut
across the wide river.
Reference
Madras Musings-September 16-30, 1999.
Outstanding contributions to sub-surface and surface flow research came from Col.
Clibborn and Kennedy during 1890's. Col. Clibborn carried out the historic experiments
(1895-97) 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

reliable design of hydraulic structures, design of stable channels, efficient distribution
devices, weed control, anti-water 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 1929-39,
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
(1895-97) with Khanki sand, Khosla propounded (1929-36) 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 1929-36 on the Panjnad Weir. This
was followed by laboratory research on models of Rasul Weir (1930-34) and Panjnad
Weir (1934-35) 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

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 semi-modular type was in 1902 by Kennedy-the sill outlet. Kennedy's gauge
outlet was introduced in 1906which was further improved in 1915. By 1922 Kirkpatrick
on the Jamras (Sind) and Crump in Punjab developed semi-modules 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), Joshi'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 broad-crested weirs were utilised by Burkitt in developing the 'Head-less
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 all-round 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

presentation of the country's highest award - Bharat Ratna - in 1955.
To perpetuate the memory of this great engineer-statesman, 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 Joglekar (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

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 1916-1928, 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 ease-water 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.

1.3.2 CO-ORDINATION 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 co-ordinates
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

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
1953-54. 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, pre-irrigation 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

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 1946-47, 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.

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. Joglekar.
(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 head-quarters 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
echo-sounding.

(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. T-shaped 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.

(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 June 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.

(11) Karnataka Engineering Research Station, Krishnarajasagar
Though it started as a small section attached to Gauging Sub-Division 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.

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 short-term and long-term 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 samples-later 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) small-scale 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.

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 co-operative 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.

Another field of study related to the engineering works connected with the control and
training of rivers. This required comparatively large-scale 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 large-scale 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, Jammu 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 Link-Part 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.

(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 Bhagirathi-Hoogly 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.

(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.

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 three-dimensional 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.

1.4 CLASSIFICATION OF FLOW
Fluid flow
Uniform flow Non-uniform flow
Subcritical Gradually varied flow

Critical
Supercritical Rapidly varied flow
Spatially varied flow

Steady Unsteady
One dimensional
Two dimensional
Froude number Pressure Flow Three dimensional
Reynolds number
Fluid flow Spatial
Free Surface Flow
Temporal
Compressible / incompressible Single phase
Two phase
Reciprocating upstream flow Multi phase
Unidirectional upstream flow
Highly irregular
Highly variable upstream flow
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

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 mild-sloped 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.

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.

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).
θ1 θ1 A = by + y2 ( θ1+ Cotθ1)
θ1 θ1 y
1
y
y P = b + 2y( θ1+ Cotθ1)
y 1
m m
R by + y2 ( θ1+ Cotθ1)
=__________________
b + 2y ( θ1+ Cotθ1)
b
Lined channel section for Q > 55 m3/s
0
θ1 θ1 1 y22θ1
1 y 2θ1 y 1 A = 2(1+y2Cotθ1) +__
2
m
y m = y2(θ1+Cotθ1)
P=2yCotθ1+2yθ1 = 2y(θ1+Cotθ1)
A ____________
__ y
y2(θ1+Cotθ1) __
R= = =2
P 2y(θ1+Cotθ1)
Lined channel section
for Q < 55 m3 / s
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 egg-shaped, ovoid,
Semi-elliptical, U-shaped, catenary, horseshoe, basket-handle, etc. The complete
rectangular and square are also common for large sewers.

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.
Cartesian equation: y = a cosh(x/a)

Hydrostatic Catenary
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.

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 θ , 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.
y
horizontal
900
x
Normal and vertical depths

Box
θ = 10ο , cosθ = 0.9848,thus there would be an error of 1.51%.

y = d cos θ
If x is measured along the horizontal direction instead of the sloping bed, then the
2% error occurs at about θ = 11° or Sο = 0.20 . On the other hand if x is measured
along the sloping bed instead of the horizontal 2% error occurs at
about θ = 16ο or Sο = 0.29 , which is an extremely steep slope in open channels.
However, there is exception in cases such as spill ways, falls, chutes.

m=2
m=2
1 y 1
b θ0
Stage
P
Datum
A = Area of flow
T = free surface width (m)
m = side slope defined in horizontal to 1 vertical; m:1
m = cot θ
l
θ0
m
P = Wetted perimeter is the boundary which is in
contact with the flow (m)
b = bed width in (m)
y = depth of flow
A Channel cross section
Water surface EL 210.00 m
Bed EL 205.00 m
H (M.S.L)
(Above Mean Sea Level)
EL 200.00 m Datum
Definition of stage
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

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 cross-sectional 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 b
R= When a shallow channel of b is used and y → ∞ then R → .
P 2
b
__
2
R
b
b
__
y then R
2
Hydraulic mean radius
Wide Rectangular
b R y R
dR
__
dy
R
Trapezoidal

The hydraulic mean depth D is the ratio of the water area to the free surface
A
width, D = . The section factor for critical- flow computation m is the product of the
T
A
water area and the square root of the hydraulic depth, Z = A D = A . The section
T
factor S.F for uniform-flow computation in case of Manning formula is the product of the
2
water area and the two-thirds power of the hydraulic radius S.F = AR 3 other wise for
2
chezy's formula it is i.e., AR 3 . 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
dy
distance
River bed elevation as a fuction of distance from
the river bank

2.1.4 Circular channel

Normalised geometric characteristics are shown in figure. When the flow is full the
⎛ π dο 2 ⎞
⎜ d
⎟
1 A
hydraulic mean radius is dο (i.e) ⎜ = 4 = ο ⎟ which is less than the maximum
4 ⎜P π dο 4 ⎟
⎜ ⎟
⎝ ⎠
hydraulic mean radius which occurs at 0.81dο when relative velocity of the flow is
considered for constant Manning roughness coefficient. similarly it is 0.938d (click) for
ο
2
maximum value of AR 3 when the discharge is maximum.
1.0
0.9
0.8
0.7
Z
___ A
___
y 0.6
__ 2.5
do ___
D Ao
do d00.5 do
0.4
___
R
y
0.3
___
P Ro
0.2 Po ___
T
0.1 do
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3
Po = π do π___
do
2
do
___
A0 = R0 =
4 4
Subscript zero indicates full flow condtion
Normalized geometric elements for a circular section
Problem: Write a computer program to obtain the geometrical elements of a circular
y AR 2 / 3
shape channel and obtain the Vs
d0 A0R 02 / 3
Compute the geometric elements, area, hydraulic mean radius, hydraulic mean depth
for the following cases:

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
Stage of flow (m) Distance of the embankments form reference

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

Solution:
20
* *
* *
16 * *
* *
* *
* *
12 * *
* *
* *
* *
8 * *
* *
* *
* *
4 * *
* *
*
0.0 6.0 12.0 18.0 24.0 30.0

Distance from reference (m)
Natural channel

600.0 *
*
500.0
*
400.0 *
*
300.0 *
*
200.0 *
*
*
100.0 *
*
*
*
*
0.0 3.0 6.0 9.0 12.0 15.0
Depth of flow (m)
Variation of area with depth of flow
60.0 *
50.0
40.0
30.0
*
20.0
* *
* *
** Natural
10.0 ** **
**
* ** ** *
* * * ** ** **
** *
** *** *** ** Triangular
** ** *
* * *
0.0 3.0 6.0 9.0 12.0 15.0
Depth of flow (m)
Variation of hydraulic mean depth with depth of flow

8.0
*
*
6.0 *
*
*
*
*
4.0 *
*
*
*
2.0 *
*
3.0 6.0 9.0 12.0 15.0

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.

d0
Horse shoe tunnel

If d0 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.

2.2 Verify the geometrical elements for Circular channel

Top width of water T
2 2
⎛T⎞ ⎛ dο ⎞ 2
⎜ ⎟ =⎜y - ⎟ =r
⎝2⎠ ⎝ 2 ⎠
2
T2 ⎛ d ⎞
or = r2 − ⎜ y - ο ⎟
4 ⎝ 2 ⎠
2 2
⎛d ⎞ ⎛ d ⎞
T = 2 ⎜ 0 ⎟ −⎜y - ο ⎟
⎝ 2 ⎠ ⎝ 2 ⎠
dο2 ⎛ 2 dο2 d ⎞
T=2 − ⎜⎜ y + - 2y ο ⎟⎟
4 ⎝ 4 2 ⎠
dο2 d 2 2ydο
=2 − y2 − ο +
4 4 2
T = 2 y ( -y +dο ) or T = 2 y ( dο − y )
θ T
sin ( 180 - ) = r
2 2
⎛ θ⎞ θ
or T = dο sin ⎜180 − ⎟ = dο sin
⎝ 2⎠ 2
Area of flow = Area of circle - Area above the chord
T ⎛ d ⎞
Area of triangle = x ⎜ y - ο ⎟
2 ⎝ 2 ⎠
d ⎛ θ ⎞⎛ d ⎞
= ο sin ⎜180 − ⎟ ⎜ y - ο ⎟
2 ⎝ 2 ⎠⎝ 2 ⎠
⎡ d ⎤
⎢ y- ο
θ 2 or y - ο = ο cos(180 − ) = − ο cos ⎥
d d θ d θ
⎢ cos(180 − ) = ⎥
⎢ 2 r 2 2 2 2 2⎥
⎣ ⎦
Area of full circle
Area for θ = xθ
2π
π dο2 d2
= θ = οθ
4.2π 8
dο2 dο2
θ − sin θ
8 8
dο2
Area of flow = (θ − sin θ )
8

dο 1 dθ
P = 2π x θ= ο
2 2A 2
1
A 8(
θ − sin θ ) dο2 d θ sin θ
⎛ ⎞
R= = = ο⎜ − ⎟
P dο2
4 ⎝θ θ ⎠
2
dο ⎛ sin θ ⎞
R= ⎜1 − ⎟
4 ⎝ θ ⎠
Z =A D = A A T
1
A 8(
θ − sin θ ) dο2
D= =
T θ
dο sin
2
⎛ ⎞
A dο ⎜ θ − sin θ ⎟
D= = ⎜ ⎟
T 8 ⎜ sin θ ⎟
⎝ 2 ⎠
⎛ ⎞
dο ⎜ θ − sin θ ⎟
D= ⎜ ⎟
8 ⎜ sin θ ⎟
⎝ 2 ⎠
⎛ ⎞
1 dο ⎜ θ − sin θ ⎟
Z = (θ − sin θ ) dο2 ⎜ ⎟
8 8 ⎜ sin θ ⎟
⎝ 2 ⎠
2 (θ − sin θ )
1 .5
Z= 5/2
0.5 d 0
32 ⎛ θ⎞
⎜ sin ⎟
⎝ 2⎠

2.3 Difference between Pipe Flow and Free Surface Flow

OPEN CHANNEL FLOW PIPE FLOW
Defines as a passage in which liquid A pipe is a closed conduit which is
flows with its upper surface exposed to used for carrying fluids under pressure.
atmosphere. The flow in a pipe is termed as pipe
The flow is due to gravity flow only when the fluid completely fills
Flow conditions are greatly influenced the cross section & there is no free
by slope of the channel. surface of fluid.
Hydraulic grade line coincides with the Hydraulic grade line does not coincides
water surface with the water surface.
The maximum velocity occurs at a little The maximum velocity occurring at the
distance below the water surface. pipe centre.
The shape of the velocity profile is Velocity Distribution is symmetrical
dependent on the channel roughness. about the pipe axis.
Horizontal
2
V1
____ TEL hf
2g 2
Velocity V2
____
P1
____ head 2g
HGL
Piezometer
Piezometric
head P2
____
(1)
V
Z1 PIPE AXIS
(2)
Datum Z2
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.

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

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.
Inflow
Inflow Outflow
x
Outflow
Elemental volume
Generally, the mass balance is written in all the three directions namely x, y and z.

∂ρ u ∂ρ v ∂ρ w
+ + =0
∂x ∂x ∂x
in which
u, v and w are the velocity components in x, y, z directions respectively,
ρ is the mass density of the fluid. If the mass density is cons tan t the above
equation can be rewritten as
∂u ∂v ∂w
+ + =0
∂x ∂x ∂x
If v=0, w=0 i.e., for one dim ensional flow it reduces to
∂ρ u
=0
∂x
Mass
Mass density ρ =
Volume
∂ρ u
*elemental area = constant
∂x
Integrating one gets
UA = constant
∴ Volume rate could be exp ressed as m3 / s. This is generally known as
flow rate or discharge ( Q ) and expressed as cubic meter / second and is
abbreviated as cumec (m3 / s).
Q = Area * Velocity = AV

3.2 ENERGY IN FREE SURFACE FLOW

It is known in basic fluid mechanics that the total energy in ( Newton-meter 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
Vx
H = d cos θ + α + zx
x x 2g
Total Energy Line

__2
α1 v1
hf 2g
__2
α v H
y1 = d1 cos θ
__2 2g
α1 v2
__ _
2g x v1 z
y = d cos θ y
Streamlines z1
y2 = d2 cos θ 900
Y Section YY
o z
z2 Datum
1
__ 2
H=z+y+ α v
2g
Energy in gradually varied open channel flow
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
2
V
channel section, θ is the angle of the channel bottom with horizontal and is the
2g
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
A
section V = ∫ v dA . In order to account for the variation of the velocity due to non uniform
0
pattern of velocity distribution, an energy correction factor α is used.

depth
of flow
y
Logarithmic
Theoretical uniform flow Linear Power velocity
velocity distribution velocity Law distribution
(Ideal) distribution
Typical velocity distribution
Inner wall C
L Outer wall Inner wall C
L Outer wall
1.25
1.30 1.30
1.25
1.15 1.15
1.10 1.10
1.00
0.77 0.78
0.93
STATION A STATION B
C Inner wall C
L Outer wall
Inner wall L Outer wall
1.20
1.15
1.10
1.15
1.05
1.00 1.08
1.00
0.95 0.83
0.80
STATION C STATION D
ISOVELS in open channel bend [Normalised with Vmax ]
Q = 83.5 lps, F = 0.41, Re= 103460
0.998
0.963
0.905
0.888
0.895
0.860
0.85
0.825
0.813
0.791
0.775
0.740
0.765
0.722
0.665
0.687
0.628
Section A Section 4
Q = 33.61 l/s, F = 0.2457 Re = 179574, n = 0.009834
Non-Dimensionalised isovels (v/vmax)

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 ( α ) may be used for correcting for the
over-all effect of the non-uniform velocity distribution. Thus, the total energy at the
2
V
channel section may be written as H = d cos θ + α +z.
2g
y
d
θ
θ
900
Normal and Vertical depths

for channel of small slope, θ ≈ 0 thus, the total energy at the channel section is
2
V
H= y + α +z .
2g
The slope of the energy line is denoted by Sf, the slope of water surface is denoted by
Sw and, the slope of the channel bottom by Sο = sinθ with an assumption that
sinθ tan θ
≈ =1 (See box).
θ θ

If the value of θ is taken as

°
(i) 6 ; sinθ = 0.1045, tanθ = 0.1051
tanθ sinθ
the difference is 0.0006 then ≈ =1
θ θ
(ii) ∴ If θ =10° , sin θ = 0.1736, tan θ = 0.1763; difference is 0.0027
10° = 0.9848
cos θ = cos
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 θ = 11° or S = 0.20 which is an extremely steep slope in open
ο
channels. However, there is exception in cases such as spillways, falls, and
chutes.
Spillways will have slopes of θ = 45D to 60D .
In the uniform flow, Sf = S w = Sο . According to the law of conservation of energy, the total
energy head at the upstream section should be equal to the total energy head at the
downstream section plus the loss of energy hf between the two sections. In other words
2 2
V1 V2
d1 cos θ + α1 + z1 = d 2 cos θ + α 2 + z2 + hf
2g 2g 1-2
This equation applies to parallel or gradually varied flow. For a channel of small slope, it
2 2
V1 V2
may be simplified as z1 + y1 + α1 = z 2 + y2 + α 2 + hf
2g 2g 1-2
The above equations are known as the energy equation. If α1 = α 2 = 1 and, hf = 0
2 2
V1 V2
then the above equation reduces to y1 + + z1 = y2 + + z 2 = constant
2g 2g
This is the well known Bernoulli (energy) equation.
Problem
(This may be attempted after learning about Hydraulic Jumps).
The reservoir level upstream of 50 m wide spillway for a flow of 1350 m3/s is at elevation
250 m. The downstream river level for this flow is at El. 120. Determine the invert level
of the stilling basin having the same width as the spillway so that the hydraulic jump is
formed in the stilling basin. Assume that the losses in the spillway are negligible and
also find downstream depth, Froude number, y1, y2, F1, F2 and ∆E and power
dissipated in this system.

3.3 Basic Momentum Equation

The momentum of the flow passing a channel section per unit time is expressed
βγ Q V
by , in which β is the momentum coefficient,
g
⎛ kg m ⎞
γ ⎜ = ρ g = 1000 3
*9.806 = 9.806 kN ⎟ is the specific weight of water, Q is the
2
⎝ m s ⎠
discharge in m3s-1, and V is the mean velocity in m3s-1.
As per Newton's second law of motion, the rate of change of momentum in the body of
water in a flowing channel is equal to the resultant of all the external forces that are
acting on the body. Applying this to a channel of large slope (Figure), the following
expression for the rate of change of momentum in the body of water confined between
sections 1 and 2 can be written as
1 2
_
V1
_
W sinθ V2
P1 y1
y2 P2
Wcosθ
γ y1 Pf
θ
Z1
w
L Z2
Datum
Figure - Momentum equation

γQ
g
( )
β 2 V 2 -β1 V1 =P1 -P2 +Wsinθ-Pf
in which subscripts refer to sections 1 and 2; P1 and P2 are the resultants of pressure
forces acting on the two sections; W is the weight of water bounded between the
sections; and Pf 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.

For gradually varied flow, the values of P1 and P2 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 P1 and P2 cannot be so computed but are to be corrected. For simplicity, P1
and P2 may be replaced, respectively, by β1′P1 and β 2′ P2 in which β1′ and β 2′ are correction
coefficients at the two sections. These coefficients are called pressure distribution
coefficients. Since P1 and P2 are forces, the coefficients may be specifically called force
coefficients. It can be shown that the force coefficient may be expressed as
1 A 1 A
β′ =
Az ∫0 h dA = 1 +
Az ∫0 c dA
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.

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
Q
continuity equation namely v = . The velocity varies locally and spatially depending on
A
the type of channel (straight, steep, bends, meandering, etc.,) and the flow (uniform,
non-uniform, 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:

2.0
1.0
0.0
47427 mm
0
1.524 m
3.05 m
Q = 78.96 m3/s
Q = 70.68 m3/s
Q = 49.21 m3/s
Godavari Western Main Canal - Mile 5/2
1.5
1.0
0.5
0.0
45897.1 mm
0.0FT
1.524 m
3.05 m
Q = 145.96 m3/s
Q = 135.67 m3/s
Q = 129.73 m3/s

0.883
0.898
0.213
0.93
0.96
0.990
0.998
0.805
0.775
0.742
0.62
SECTION 7
0.948
0.960
0.970
0.98
0.99
O
SET III SECTION 0
Q = 1.187 CFS, F = 0.2457, Re = 179574
NON-DIMENSIONALISED ISOVELS (v/vmax)

Inner wall C
L Outer wall Inner wall C
L Outer wall
0.85 1.30
1.25
1.15
1.10
1.15 1.30
1.25
1.10
1.0
0.77
Station A Station B
C
L
Inner wall C
L Outer wall Inner wall Outer wall
1.30
1.30 0.92 1.25
1.23
1.27 1.220
1.00 1.30
1.08 1.00 0.95
0.95 1.00
0.75 0.80
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.
Isovels
Secondary
currents
(a) Open channel

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.

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 cross-section 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 cross-section normal to the flow direction.
The cross-section 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 cross-section for
channels with different widths is given in Table.
Description of Channel Number of Observation Maximum width of

(m) verticals segments (m)
Width less than 15 15 1.50
Width between 15 and 90 15 6.0
Width between 90 and
15 15.0
150
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.

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
Approximate surface Number of verticals for

Channel capacity m3/s
width (m) depth and velocity
(a) Above 85 Above 35 11
(b) Between 15-85 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 : 1192-1959.
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
hand-lines 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 two-point method the velocity observations are

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 two-point and one point methods are in
common use in India, though sub-surface 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 1911-1920 (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 cross-section was obtained by the current meter
whereas the surface velocity was measured using floats.
The following relationship was obtained
V (m/s) = 0.8529 Vs + 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:
Vs
=1 + 2.5 g / C
V
V
The usual assumption made in practice is that s = 0.85 which corresponds to 'C' value
V
of 52.4 m0.5 s-1.

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 Slope-Area 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
cross-sectional 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 non-uniform
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 Stage-Discharge 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).

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
Instructions for collection of data for
IS: 2915
analysis of errors
IS: 3918 Use of current meter
IS: 6059 Weirs of finite crest width
IS: 6062 Standing wave flume-falls
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, June 1974. Page 13 to 24.

4.3 Radio-active 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 cross-section 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 radio-active 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 m3s-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) radio-active tracer
technique on River Mutha, in the recirculation system of the CWPRS and in River Tapi
and
(ii) Chemical Salt dilution method downstream of tail-race 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.,
dc d m
=
c m
The mixing length depends upon many factors such as: (i) degree of turbulence, (ii)
geometry of the cross-section, (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 m3s-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 ( lmix ) in mountainous rivers varies linearly with the

average river width Bo in the experimental reach and is governed by the relation
lmix = KBo + 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.
Average Observed
River
Name of top water mixing
Discharge m3s-1 slope Remarks
the River surface length
(m/km)
depth (m) (km)
as per
as per
area
dilution
Ganga velocity
method
method
(m3s-1)
(m3s-1)
Ganga 296.00 319.00 1.21 70.00 4.30
Power
mixing
Ganga 136.00 - 1.21 57.00 -
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

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 cross-section 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
V (m/s) = 0.3048 (1.004 ν 0.6 + 0.041)

in which ν 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
1936-38 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
V (m/s) = 0.3048 (1.010 ν 0.6 - 0.059 )
in which ν 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 (Vr )
( )
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 (Vr ) has been used.
⎛ ⎛ y −l ⎞⎞
V (m/s) = 0.3048 ⎜ V ⎜⎜1.012 − 0.116 ⎟⎟ ⎟⎟
⎜ r y
⎝ ⎝ ⎠⎠
in which V is in FPS.
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:
V ( m / s ) = 0.3048 ⎛⎜ 2ν − ν ⎞
⎟
⎝ 0 . 80 y 0 . 4 y ⎠
in which ν and ν in FPS at 0.8 and 0.4 depth of flow respectievely.
0.80 y 0. 4 y

0.990
Manning, n = 0.020
X = 0.90
0.980
X = 0.85
0.970
X = 0.80
0.960
__ X = 0.75
V
____
Vr
X = 0.70
0.950
X = 0.65
0.940
0.930
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)

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 cross-section.
4.4.5 Error in Width
In the Pivot-point 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 cross-section line pegs and
prefixed pivot-point 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

Range of width Absolute error (m) Mean standard Remarks

observed (m) deviation (%)
300 to 600 2.34 ± 0.386 From one bank
From each of two
600 to 1200 6.72 ± 0.564
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 Absolute error (m) Mean standard deviation

(m) (%)
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 six-point 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

V=
1
10
(
Vs + 2V0.2 + 2V0.4 + 2V0.6 + 2V0.8 + Vbottom )
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
50b 50
X m (b) = or
T m +1
in which, X m ( b ) is systematic error in discharge due to 'm' number of verticals. Random
errors X v2 and X 2 due to velocity and depth after analyzing 43 sets of observations
y
having approximately 100-200 number of velocity observations on the Maharashtra
2 103
canals, Gole et al. (1973) have suggested the following two equations: X v =
m2
28
and X 2 y =
m4 / 3
The average coefficient of variation for horizontal distribution of velocity was obtained as
32 and for mean depth as 5.3 percent.

__ 2
100.0 x __+_ 2 σ ( x2__ )
v v
50.0 __ 2 10
__ 3
x __ = 2
v m
10.0
5.0
__ 2
x __
v
1.0
0.5
0.1
1 10 100 200
m
__ 2 T
__
Variation of x __ with m = [ b -1]
v 1

10.0
m>70
5.0
2.0
_2 1 2
x _ = _ X_ 1.0
y m y m<70
0.5
0.1
1 5 10 50 100
m is the Number of verticals)
_2
Variation of x_ with the number of verticals
y
The total error could thus be worked out as
0.5
⎡ 2 ⎤
( b ) + ⎛⎜ X v2 + X 2y ⎞⎟ ⎥
1
XQ = ⎢ X m
⎣ 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.

8.0
7.0
ISO data
6.0 Carter and Anderson
Delft data
CWPRS data
5.0
4.0
3.0
2.0
1.0
0
0 10 20 30 40 50 60
w
Number of verticals m = __ - 1
b
Showing theoretical curve for Xm and experimental
data as a function of number of verticals
4.4.10 Effect of Deployment on Random Errors X 2 and X 2

v y
When the 'm' number of verticals are deployed according to a particular scheme for
precisely locating the depth (cross-section) profile, it is expected that it would have
some effect on the estimate of X 2 and X 2 . Since deployment help in getting the
v y
nearest estimate of the representative profile and hence coefficient of variation in
velocity and depth over a cross-section, 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

2
significant unless 'm' is very small. To verify the above hypothesis X and X 2 were
v y
estimated for different deployment of vertical and the total contribution against the mean
function for
1
m
(
X 2 v and X 2 y ) (See Figure). It is found that the difference in the X 2
Q
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.
3.0
100 1.0
m>30
50 0.5
20 __2 10 3 0.2
X = m + m 4/3 )
( __ 28
__
__ 2
1 2 2 2 q
__
m
X __ 1 x
__
=m ( __y + x __ )
q v
10 0.1
5
m<30
0 5 10 50 100
m number of verticals
__2
Variation of X __ with m
q

Table Shows computation of X m from experimental data (CBI&P)
Sl.No % error with Number of X m Intrinsic error

reduced number verticals for in standard
of verticals, Em = 8 standard discharge due to
discharge 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
Computation done by CBI & P (Central Board of Irrigation and Power)

Average E = 3.30
( Set Error ) σ = 2.93
2
⎛1 ⎞
⎜ E ⎟ + σ = (1.15 ) + ( 2.93)
2 2 2
⎝2 ⎠
( )
2
or l
Xm −Xm = 11.3125
( lX m )
− X m = 3.36 but X m = 3.06
∴ l
X m = 3.36 + 3.06 = 6.42 percent
in which, ( lX m −Xm ) is the estimate of the error due to reduced number of verticals,
relative to the standard discharge.
4.4.11 Total Error X Q

2
Total error could be worked out using Equations (9) and (10) for X and X 2 and the
v y
theoretical value X m 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 semi-theoretical 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

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.
16
14
12 Deift data
Carter and Anderson data
ISO specification (CWPRS) data
10 MERS data
0 5 10 15 20 25 30 35 40 45 50 55
m (Number of verticals)
Variation of total error XQ with number of verticals

12.0
10.0
8.0
6.0
Deift data average
CWPRS data
MERS data
4.0
APERL data
(assuming Xm as error
in standard discharge)
2.0
0
20 40 60 80 100 120 140
Theoretical mean error (percent)
α 100 32 19 15 11 9 7 5 4 3
(w
Number of vertical or __
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.

(d) The value of X m could be obtained from figure for a given number of verticals with
50b
equal spacing. The formula could be used only when end segment spacings are
W
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
⎛ 50b ⎞ 1 ⎛ 2⎞
XA = ⎜ ⎟ + ⎜Xy ⎟
⎝ W ⎠ m⎝ ⎠
Since the systematic error (50b/W) is dominant one in X A 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 qi' 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 X A .

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 Maximum space

a cross-section of verticals in m
Width of channel not
1 exceeding 15 m where the 15 1.5
river bed changes abruptly
Width of channel from 15 to
2 15 6.0
90 m
Width of channel from 90 to
3 15 15
180 m
Width of channels greater
4 25 -
than 180 m
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 m3s-1 Approximate No. of verticals for

surface width (m) 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.

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 14m3/s to 85 m3/s , 5 or 7 verticals
gave discharge values within a range of ±5% error, as compared to 15 verticals. For 85
m3/s to 225 m3/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.

1.5
1.0
0.5
0.0
45897.1 mm
0.0FT
1.524 m
3.05 m
Q = 145.96 m3/s
Q = 135.67 m3/s
Q = 129.73 m3/s
2.0
1.0
0.0
47427 mm
0
1.524 m
3.05 m
Q = 78.96 m3/s
Q = 70.68 m3/s
Q = 49.21 m3/s

1.5
1.0
0.5
0
4.0233.6 mm
Distance
Q = 28.49 m3/s
Q = 52.78 m3/s
Q = 73.60 m3/s
1.524
Q = 82.5 m3/s
3.05
Nizamsagar canal M 2/0
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

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 m3/s to 145.96 m3/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.

4.5 Errors in Depth Measurement in high velocity flows in

laboratory using Point Gauges: (Jayaraman and Sethuraman,
1973)
Water depth measurements in high-velocity 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 good-old point-gauge. Although more sophisticated
electrical probes are available for depth measurement, these invariably give rise to flow-
disturbance when immersed in a high velocity channel-flow. In order to improve the
accuracy of point-gauge 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 Contact-Time Indicator that can be
used with any point-gauge to improve the accuracy of depth measurement.

7.8
7.6 7.56 (Mean)
7.49
7.4
32% 50%
7.2
7.0
0 20 40 60 80 100
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
point-gauge measurement in high-velocity flows:
1. Point-gauge measurement in high-velocity 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 flow-depth 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

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
point-gauge measurement in high velocity-flows. 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 test-run,
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. JAYARAMAN and V. SETHURAMAN "IMPROVING THE ACCURACY OF POINT-
GAUGE MEASUREMENT IN HIGH-VELOCITY FLOWS", Journal of Hydraulic
Research, Volume 11, Number 4, 1973, Page Number 317 to 323.

5. Secondary Current and Spiral Flow

The curve of constant velocity for rectangular and triangular cross-section 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 non-circular
cross-section 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 cross-section 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.

Fig. 1. Curves of constant velocity for pipe of rectangular cross-section,

after Nikuradse
Fig. 2. Curves of constant velocity for a pipe of equilateral triangular

cross-section after Nikuradse
a b
Fig. 3. Schematic representation of Secondary flows in pipes
of triangular and rectangular (open channel) cross-section

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, non-uniformity 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

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 non-uniformity 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 straight-non-circular channel sections and in pipes
due to boundary resistance (figure 5).
2. Secondary flow developed due to non-uniform bed configuration as in case of alluvial
channels.
3. The ' strong ' currents caused in bends due to centrifugal force.

SPIRAL FLOW
y
INSIDE
OUTSIDE WALL
SECTION ON A-A
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.

Isovels
Secondary
currents
(a) Open channel (b) Equivalent closed conduit
Comparison of Open Channel Flow with Closed-Conduit 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.
Side Slope, m: 1 = 1.5 : 1
y
4y
0.750 γ ySo 0.750 γ ySo

0.970γ ySo
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, Jacob, 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

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.
1 V
10d [m/s)
10d 6.00
2 3
10d
5.75
measuring
station
5.50
5.25
2 1
3 5.00
4.75
4.50
4.25
4.00
d
5d 5d
Isovels behind a row of spheres as measured by Schlichting. Secondary flow

in the boundary layer is marked behind (1), as calculated by K. Schultz-Grunow.
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 pseudo-wall
and does not affect the main flow.

s s
Isolated - roughness flow (k/s) - Form drag dominates

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

s s s
Wake interference flow (y/s)
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.
k j j j j
s s s
k is surface roughness height
s is the spacing of the elements
j is the groove width
y is the depth of flow
Quasi smooth flow - k/s or j/s becomes significant acts as Pseudo wall
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.
Concept of three basic types of rough surface flow
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.

v
Vr =CONSTANT
r
ri
rc
B r0
O
VELOCITY DISTRIBUTION IN POTENTIAL FLOW

IN A CURVED CHANNEL
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

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 cross-section 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 Sxy .
⎛ V 2 xy ⎞
⎜ ⎟
Sxy =
⎜ 2g ⎟
⎝ ⎠m
* 100 =
(
V 2 xy )
m * 100
⎛V ⎞ 2 2
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.

Following equations relate the deflection angle α1 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
⎡ P ⎤
⎢ r ⎥
c ⎥
(ii) tan α1 =17.4 ⎢ 0.25 for 2000 ≤ R e1 ≤ 45000
⎢ R e1 ⎥
⎢⎣ ⎥⎦
(iii) For a smooth triangular channel
⎡ P ⎤
⎢ r ⎥
c ⎥
tanα1 =13.4 ⎢ 0.25 for 2000 ≤ R e1 ≤ 15000
⎢ e1 ⎥
R
⎢⎣ ⎥⎦
In general,
⎡ P ⎤
⎢ r ⎥
c ⎥
tanα1 =K 3 ⎢ 0.25
⎢ R e1 ⎥
⎢⎣ ⎥⎦
If the channel is wide then

0. 5
⎡y⎤
⎢ ⎥
r ⎦
tanα1 =K 4 ⎣ c0.25
R e1
But Russian authors found that for a rectangular wide channel
⎛y⎞
tanα1 =11⎜ ⎟
⎝ rc ⎠
In general for a wide rectangular channel,
tanα1 =K 0
b
rc
( )
Φ R e1 for smooth flow
b ⎛ y ⎞
tanα1 =K 0 Φ⎜ ⎟ for rough flow
rc ⎝ K s ⎠
⎛b⎞
tanα1 =K 0f a ⎜ ⎟
⎝ rc ⎠

where f = friction coefficient and "a" is an exponent >1. The last equation can be
⎛ 8g ⎞
expressed in Chezy terms of coefficient ⎜⎜ C= ⎟ in the form
⎝ f ⎟⎠
a
⎛ 8g ⎞ ⎛b⎞
tanα1 = ⎜ 2 ⎟ K 0 ⎜ ⎟
⎝C ⎠ ⎝ rc ⎠
The value of tanα1 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.

6.1 Energy and Momentum Coefficients

Generally, in the energy and momentum equations the velocity is assumed to be steady
uniform and non-varying 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, υ 2 2g )m , (the subscript m indicating the mean value) need not necessarily be
2
equal to V 2g . 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 α and

proved that the selection of α and β (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 α is the proper
factor to use. To understand proper use of factors α and β and the energy principle or
momentum principle is used appropriately.
Box:
The weight of flow through an element of area dA is equal to ρ gυ dA ; the kinetic energy
2
per unit weight of this flow is V 2g ; The rate of transfer of kinetic energy through this
element is equal to
υ 2 ρ gυ 3
ρ gυ dA . = dA (1)
2g 2g
Hence, the kinetic energy transfer rate of the entire flow is equal to
A υ3
∫ ρ g dA (2)
0 2g

and the total weight rate of flow is equal to ρ g υ dA

ρ g Q = ρ g VA
mass
mass density ρ =
volume
kg
mass =ρ * volume = x m3 = kg
3
m
Force = N = mass * acceleration
m kg m
= kg * 2
=
s s2
kg m N
specific weight γ = ρ g = 3
* 2
=
m s m3
A
A Velocity v Velocity v
Real ideal
Velocity distribution along section AA
Velocity distribution in a Trapezoidal Section

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 υ3
⎛υ2 ⎞ ∫ dA ⎛ 2⎞
0 2g V ⎟
⎜ ⎟ = =α ⎜ (3)
⎜ 2g ⎟ VA ⎜ 2g ⎟
⎝ ⎠m ⎝ ⎠
in which α 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
A 3
∫ υ dA ∑ υi3dA
α= 0 3
i =1
≈ 3
i = 1.....N (4)
V A V A
Similar approach can be applied for computing the momentum term β VQ . The rate of
transfer of momentum through an element of area dA is equal to ρV 2dA ; Following
similar logic as above the momentum correction coefficient can be obtained as
A N
∫ υ dA
2
∑ υi2dA
i =1
β= 0
≈ i=1,2....N (5)
2 2
V A V A
β is also known as Boussinesq coefficient.

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 α and β for selected situations.

Table: Values of α and β for selected situations (after Chow, 1958)
Channel α β
Minimum Maximum Average Minimum Maximum Average
Regular
channels,
1.10 1.20 1.15 1.03 1.07 1.05
flumes,
spillways
Natural
streams
1.15 1.50 1.30 1.05 1.17 1.10
and
torrents
River
under ice 1.20 2.00 1.50 1.07 1.33 1.17
cover
River
valley,
1.50 2.00 1.75 1.17 1.33 1.25
over
flooded
The kinetic energy correction factor α and momentum correction factor β can be
expressed as (see box).
N
A 3
∫ υ dA ∑ υi3dA
0
α= 3 ≈
i =1
i = 1.....N (4)
3
V A V A
A N
∫ υ dA
2
∑ υi2dA
i =1
β= 0
≈ i=1,2....N (5)
2 2
V A V A

6.1.1 Determination of α and β
Many investigators have done extensive investigations on the computation of α and β .

Chow (1958) has summarised different equations for determination of α and β for
various velocity distributions.
Rehbock assumed a linear velocity distribution and obtained

α = 1+ ε 2
ε2
β = 1+
3
and for logarithmic velocity distribution.
α = 1 + 3ε 2 − 2ε 3
β = 1+ ε 2
⎧V ⎫
in which ε = ⎨ max − 1⎬ , Vmax 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 α
and β , as shown by Bakhmateff, is that β exceeds unity by about one-third of the
α +2
amount by which α exceeds unity. If β 1 + n and α 1 + 3n then β =
3
approximately. Generally, the coefficients α and β 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 α is more sensitive to velocity variation than β , so that for a
given channel section, α > β . Values of α and β can easily be calculated for
idealized two-dimensional velocity distributions.

n
υ ⎛ y⎞
Velocity Distribution =⎜ ⎟
υ0 ⎝ y 0 ⎠
υ0
v=
n +1
( n + 1)
3
α=
3n + 1
( n + 1)
2
β=
2n + 1
α − 1 ( n + 3)( 2n + 1)
=
β −1 ( 3n + 1)
1
If n =
7
α = 1.043, β = 1.015
The high value of α appropriate to laminar flow is of limited interest, since laminar flow
is rare in free surface flow problems. For turbulent flow in regular channels α seldom
exceeds 1.15. In view of the limited experimental data on values of α , the question
always arises whether the accuracy attainable with channel computations warrants its
inclusion!.
A practical method of arriving at the values of α and β 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 α and β may be classified as
1. Theoretical Methods
Based on experimental studies Strauss in 1967, has given empirical formulae for
computing α and β for general channel section based on the velocity distribution
given by the following equation.

V=ay1 n
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 ≤ ∞ .
α and β can easily be computed using following equations , if the velocity distribution
is known.
A 3
∫ υ dA
0
α= 3
V A
∫υ
2
dA
β= 0
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 α = 2 and β =1.33.
Strauss showed that
α = f ( n,∈1 ,B1 ,γ 1 )
β = f ( n,∈1 ,B1 ,γ 1 )
in which n is the exponent of the velocity distribution, and, ∈1 is normalized depths, B1 is
the normalized width of free surface to bed width, γ 1 is normalized bed width of berm
(including) to channel bed. The velocity distribution plays a dominant role in influencing
T
α and β and in trapezoidal channel in addition to B1 = . For rectangular channel the
b
exponent n of velocity distribution has a dominating effect. But Strauss’s method has
limited practical utility. It is not always true that the same velocity distribution prevails
along all the verticals of the cross-section, especially in non-rectangular channels. Also

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.
depth
of flow
y
Theoretical uniform flow Linear Power Logarithmic

velocity distribution velocity Law velocity
(Ideal) distribution distribution
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 vcell.

a b c d e
1
2
3
dy
4
dx
Grid for measuring Velocity
i+1
j j+1
Co ordinates of the nodes are (i, j), (i+1, j), (i+1, j+1), (i, j+1)
Corresponding velocities are v (i, j), v(i+1, j), v(i+1, j+1), v(i, j+1)
_ 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
by
1 ∑ vcell dA
v = ∫ ∫ v dy *db ≈
A00 A ( = by )
in which dA is 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 cross-section and the lines of equal velocities called
‘isovels’ (also called isotachs’) are drawn as shown in Figure.

Q = 17.95 l/s
y = 0.332 m
α = 1.041
0.3639
β = 1.01
0.3505
0.2987
0.2499
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 α and β and are
calculated using the following expressions.

A
a1 a2 a3 a4
v
v Vs elemental area
v2
v3
Graphical Method of determining α and β ( av , av2 , av3 )
∫ ν dA ∑ ν3dA
3
α= A
3
≈ A
3
(4) and
AV AV
∫A ν 2dA ∑ ν 2dA
β= ≈ A (5)
2 2
AV AV

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,v2, and v3 against the corresponding planimetered areas as shown in Figure. It is
evident that the areas under v2, and v3 curves are equal to ∑ν 3dA and ∑ν 2dA
respectively. V , α and β are computed as shown in the box.
12
8 Shaded area = A0
4
0
0 1 2 3
v, m/s
12
4
Shaded area = A1
0
0 2 4 6 8 10
v2, m2/s2
12
4 Shaded area = A2
0
0 4 8 12 16 20 24 28
v3, m3/s3
__
Graphical method of computing V, α and β

Shaded areas A 0 , A1, A 2 are planimetered.

The average velocity
y
shaded area A 0
V = ∫ v dy=
1
y
shaded area A1
Similarly, β = 2
V y
shaded area A 2
and α = 3
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 ∑ν da , ∑ν 2da , ∑ν 3 da are computed. In particular if the grids are so
chosen that their areas are equal, the computational work become relatively easier.
However, for greater accuracy the size of the grid should be chosen as small as
possible. Also near the boundaries, relatively smaller grids are to be chosen. The
advantage of this method is that it is less time-consuming than the graphical method as
the actual velocities need not be calculated and isovels need not be drawn. . For
purposes of comparison, α and β for rectangular channel shown in the above figure
are computed by this method and are given in the following Figure.

0.325 0.364 0.378
0.338 0.364 0.366
0.357 0.364 0.365
0.361 0.364 0.361
0.35 0.364 0.364
0.333 0.357 0.359
0.2188 0.262 0.252
Q = 17.95 l/s
y = 0.332 m
α = 1.041
β = 1.024
Grid Method

6.1.4 Methods based on the use of empirical formula
Assuming a linear velocity distribution law Rehbock has proposed the following
formulae for approximate values of α and β .
∈2
α = 1+ ∈2 ; β = 1 +
3
ν max
In which ∈= −1
V
Assuming a logarithmic velocity distribution law proposed the following expressions.
α = 1 + 3 ∈2 −2 ∈3 ; β = 1+ ∈2
In which ν 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 cross-section of flow.
6.1.5 Computation of α and β 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 cross-section of the flow,
α and β 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 α and β in Several Practical Cases
Actual values α and β in many practical cases (which are frequently met with in
Hydraulic Engineering) are presented in Table I. Some of these values are listed by
O’Brien and Hickox O’Brien and Johnson and King. They are reproduced here along
with several other cases for the sake of a comprehensive table of α and β values.

Sl. Channel Dimensions Hydraulic Coefficients Remarks

No. elements
width Max.de Hydraulic Area Critical Mean Grap

Graphi Rehb
(m) pth (m) Radius (m2) depth velocity hical
cal ock
(m) (m) (m/s)
1 0.60 0.862 0.222 0.519 0.198 0.320 1.20 1.10 1.07 Rectangular
channel 0.9144
m above weir
2 1.00 0.862 0.3250 0.895 0.216 0.53 1.22 1.20 1.08 and
obstructions
upstream
3 1.00 0.874 0.3249 0.893 0.219 0.365 1.41 1.37 1.12
Simson Tunnel
- centre of
4 1.01 0.429 0.2316 0.431 0.496 2.56 1.07 1.04 1.03
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
Horse shoe
6 1.987 1.50 0.6309 2.898 0.76 1.48 1.07 1.03 1.034 conduit straight
reach
Rhine 365.76
7 159.4 3.81 2.438 4.055 1.91 1.024 1.35 1.43 1.121 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 Sudbury
9 2.67 1.22 0.6492 3.009 0.658 0.874 1.04 1.04 1.014 Aqueduct with
10 2.74 0.914 0.548 2.19 0.600 0.792 1.04 1.03 1.014 a bottom slope
11 2.71 0.618 0.411 1.415 0.53 0.658 1.04 1.02 1.010 0.000189
12 2.65 0.460 0.326 1.014 0.499 0.569 1.04 1.03 1.012
Computed with
13 0.264 0.053 0.35 2.31 1.161
Bazin series 10
Computed from
14 0.244 0.0366 0.14 0.205 1.138
Nikurade's data
Series (E)
15 1.286 0.762 1.07 schoder and
Turner
Series (I)
Schoder and
16 1.286 1.524 1.08
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
Series D
Schoder and
19 1.286 3.07 1.80
Turner Runs
101 to 105
Series D, L, M.
20 1.286 2.743 2.00 Schoder and
Turner
Triangular
21 1.528 1.105
channel
Trapezoidal
22 1.665 1.225
channel
23 1.365 1.085 Pipe
24 1.460 1.164 Shallow ditch
25 1.422 1.136 Natural channel
Experiment
number 2C -
26 0.45 0.0911 1.222
Rajaratnam
Muralidhar

Diverging
27 0.45 0.350 3.72 2.14
channel
Rectangular
28 1.76 1.41 open channel
bend
Maximum α
and β in a
29 0.61 15.40 5.00 hydraulic jump
with an inflow
Froude number
of 7.4.
At the outlet
30 3.87 section of a
draft tube
Spiral flow
31 7.40 under a model
turbine wheel
(Serial No. 1 to 20 are from O'Brien and Johnson, Enr, Vol. 1113, page 214 - 216, 1934
August 16 th after Jagannadhar Rao and others).
From the table it may be seen that α values are larger in non-rectangular 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 cross-section, the values of α are
still larger. The value in the case of a diverging channel is 3.72. For spiral flows a value
of α as high as 7.4 has been quoted . All these examples show that there are several
practical cases in which the neglect of α and β 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 α and β along the Hydraulic jump
The variation of α and β along the length of hydraulic jump is given in figure below.
Jagannadha 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.

16
14 α
12
10
8
6
4
β
2
- 0.25 0 0.25 0.5 0.75 1.0

x
______
y2-y1
15
10
5 Jump Profile Roller Zone
0
- 0.25 0 0.25 0.5 0.75 1.0
x
______
y2-y1
Variation of α and β along the hydraulic jump

6.1.8 α , β for Flow in Natural Channels
The natural channels can be subdivided into distinct regions, each with a different mean
velocity.
Isovels in a single channel α is nearly 1.15
3 1
2
Berm Berm
2 Main channel (MC)

1 and 3 channel in the flood plains
natural channel: River α 2.0

Typical Cross Sections of natural channel
υ13 A1 + υ23 A2 + υ33 A3

α =
3
V ( A1 + A2 + A3 )
υ12 A1 + υ22 A2 + υ33 A3
β =
2
V ( A1 + A2 + A3 )
υ1 A1 + υ2 A2 + υ3 A3
V =
( A1 + A2 + A3 )
n = 0.035 n = 0.035 1
1
3 4m 3
1 1 2.5 m S0 = 0.001
2 2
n = 0.015
10 m 5m 10 m
COMPOUND CANAL CROSS-SECTION
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

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 α

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.
υ13 A1 + υ23 A2 + υ33 A3

α =
3
V ( A1 + A2 + A3 )
υ13 A1 + υ23 A2 + υ33 A3
α =
⎛ υ1 A1 + υ2 A2 + υ3 A3 ⎞
⎜⎜ ⎟⎟ ( A1 + A2 + A3 )
⎝ A + A + A
1 2 3 ⎠
α =
(υ13 A1 + υ23 A2 + υ33 A3 ) ( A1 + A2 + A3 )
2
(υ1 A1 + υ2 A2 + υ3 A3 )
3
⎛ N 3 ⎞⎛ N 2 ⎞
⎜⎜ ∑ υi Ai ⎟⎟ ⎜⎜ ∑ Ai ⎟⎟
= ⎝i = 1 N ⎠⎝i = 1 ⎠
∑ (υi Ai )
3
i=1
Similarly expression for β can be obtained.
υ12 A1 + υ22 A2 + υ32 A3

β= 2
V ( A1 + A2 + A3 )
υ A + υ2 A2 + υ3 A3
in which V = 1 1 .
A1 + A2 + A3
When flow resistance formula (Manning, Chezy, other formulae) is combined with the
above equations numerical values of α , may exceed much higher than 2 under certain
situations. Generally, the α value is taken as 1.0 when the information is lacking.

References:
1. Chow Van Te Open Channel Hydraulics, McGraw Hill Publications, 1958.
2. Henderson F.M. Open Channel Flow, MacMillan Publishing Company, 1966.
3. Jaganadha Rao, M.V., Lakshmana Rao, N.S., and Seetharamiah, K. "On the use of
Energy and Momentum coefficients in Hydraulic flow computations" - Journal - 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.314-323.
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 Johnson, J. W. : "Velocity Head Connections for Hydraulic Flows".
Engineering News Record. Vol. 113, No. 7, pp. 214 - 216, Aug. 16, 1934.

6.2 Energy, Momentum coefficients for different velocity

distributions
Rehbock obtained
1) For Linear Velocity Distribution
α = 1+ε2 Vmax
{ }
2
ε Vmax
β = 1+ , in which ε = −1 v
3 V yo
Given: α = 1 + ε 2 y
=> ε 2 = (α − 1 )
2
Linear velocity distribution
Substitution for " ε "in the expression for "β ", __
v y
= __
α - 1 3 +α − 1 α + 2 Vmax yo
β =1 + = =
3 3 3
α +2
β= (Linear relation)
3
α 1 1.6 2.2 2.8
β 1 1.2 1.4 1.6
The plot is shown below
2.1
2.0
Scale
1.9
X-axis 1 cm = 0.1α
1.8
Y-axis 1 cm = 0.1β
1.7
1.6
1.5
1.4
1.3
1.2
1.1
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
Kinetic energy correction factor α
(2) He obtained for Logarithmic Velocity Distribution the following equations.

Kinetic Energy correction factor, α = 1 + 3ε 2 − 2ε 3

Momentum correction factor, β = 1 +ε 2
2.5 v*
in which ε =
V
Given : β = 1 + ε 2
=> ε 2 = ( β − 1)
=> ε = ( β − 1)
Substituting for " ε " in the expression for " α ",
( )
3
α =1 + 3 ⎛⎜ ( β -1 ) ⎞⎟ -2
2
β -1 = 1 + 3 (β - 1 ) - 2 ( β - 1 )3 / 2
⎝ ⎠
=> α = 3β − 2( β − 1)3/2 − 2
β α
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

The plot is shown below
2.0
α=1+3ε2−2ε3
1.8
β=1+ε2
1.6
α
1.4
1.2
1.0 1.2 1.4

β 1.6 1.8 2.0
Relationship between α and β

6.2.1 Derivation of relationships

Assuming a wide channel with the two - dimensional velocity distribution given by
v πy
= sin and
V0 2yο
n
v ⎡ y ⎤
=⎢ ⎥ determine " α " and "β " ( as a function of exponent n in second case). Hence show
V0 ⎣ yο ⎦
α- 1
that (a) For laminar case = 2.76 and
β -1
α - 1 ( n+3 ) (2n+1)
(b) for turbulent case = .
β -1 (3n+1)
Solution:
v πy
Case ( a ) : = sin
V0 2yο
where v is the velocity at a depth of " y " from boundary, yο is the total depth of flow in wide channel.
Let B the width of wide channel.
πy
V = V0 sin
2yο
1
A∫
Mean velocity = V = v dA
1 y πy V yο πy
V=
Byο ∫0
Vο sin
2yο
B dy = 0
y0 ∫0 sin
2yο
dy
y
V ⎧ -2y πy ⎫ ο
= 0 ⎨ ο cos ⎬
y0 ⎩ π 2yο ⎭0
{ }
y
-2Vο ⎧ π y ⎫ ο -2Vο π
= ⎨ cos ⎬ = cos − cos(0)
π ⎩ 2yο ⎭0 π 2
- 2Vο
= {0 − 1}
π
2Vο
V=
π

6.2.2 Kinetic Energy Correction Factor

yο
1 1 π y
∫ v dA = ∫ V0
3 3
α= sin 3 B dy
{π }
3 3 2 yο
V A 2 Vο 0
By ο
yο
π3 π y
∫ sin
3
= dy
8 yο 0
2 yο
π y π y πy
sin 3 = sin sin 2
2 yο 2 yο 2yο
πy πy
cos = 1 − 2 sin 2 cos 2A = 1- 2 sin 2 A
yο 2yο
πy 1 1 πy
∴ sin 2 = − cos
2yο 2 2 yο
π y π y 1 π yπ y⎫ π y ⎧1
∴ sin 3 = sin sin 2
⎨ − cos = sin
⎬
2 yο 2 yο 2 yο 2 yο ⎩ 2 2 yο ⎭
1 π y 1 π y π y
= sin − sin cos
2 2 yο 2 2 yο 2 yο
sin ( A+B ) +sin (A -B )
sinA cosB =
2
⎧ π y π y⎫ ⎧ π y π y⎫
sin ⎨ + ⎬ + sin ⎨ − ⎬
π y π y ⎩ 2 yο yο ⎭ ⎩ 2 yο yο ⎭
∴ sin cos =
2 yο 2 yο 2
1 3π y 1 ⎧ −π y ⎫
= sin + sin ⎨ ⎬
2 2 yο 2 ⎩ 2 yο ⎭
1 3π y 1 π y
= sin − sin sin (- A) = -sinA
2 2 yο 2 2 yο
π y1 πy 1 πy π y
∴ sin 3 = sin − sin cos
2 yο 2 2yο 2 2yο 2 yο
1 πy 1 3π y 1 π y
= sin − sin + sin
2 2yο 4 2yο 4 2 yο
π3 y0 πy
8yο ∫0
∴α= sin 3 dy
2yο
:

π 3 ⎧1 y0 3 πy 1 y0 3π y 1 y0 πy ⎫
= ⎨ ∫0 sin dy - ∫ sin 3 dy+ ∫ sin 3 dy ⎬
8yο ⎩ 2 2yο 4 0 2yο 4 0 2yο ⎭
π 3 ⎧⎪ 1 ⎡ −2 yο ⎫⎪
yο yο yο
πy ⎤ 1 ⎡ −2 y ο 3 π y ⎤ 1 ⎡ − 2 y ο π y ⎤
= ⎨ ⎢ cos ⎥ − ⎢ cos ⎥ + ⎢ cos ⎥ ⎬
8yο ⎪ 2 ⎣ π 2yο ⎦ 0 4 ⎣ 3π 2yο ⎦ 0 4⎣ π 2yο ⎦ 0 ⎪
⎩ ⎭
π
{ }
3
y y y
= − ο ( 0 − 1) + ο ( 0 − 1) − ο ( 0 − 1)
8yο π 6π 2π
=
π 3 yο
8yο {π −
y o yο
+
6π 2π }
=
8yο {
π 3 6 yο -yο -3yο
6π }
π 3 8 yο π 2
= =
8yο 6π 6
π2
∴α =
6

6.2.3 Momentum correction factor

1 2 1 yo π y
β= ∫V dA= ∫0 V02 sin 2 B dy
{π }
2 2 2 yο
V A 2V0
B yο
π2 yo π y
4yο ∫0
= sin 2 dy
2 yο
cos 2 A = 1-2 sin 2 A
π y 1 1 πy
∴ sin 2 = − cos
2 yο 2 2 y0
π2 yo π y
4 yο ∫0
∴β = sin 2 dy
2 yο
π 2 ⎧1 yo 1 yo πy ⎫
= ⎨ ∫0 dy − ∫0 cos dy ⎬
4yο ⎩ 2 2 y0 ⎭
π 2 ⎧⎪ 1
y0 ⎫
1 ⎡y πy⎤ ⎪
⎨ [y]
y0
= − ⎢ ο sin ⎥ ⎬
4yο ⎪ 2 0 2⎣π y0 ⎦0
⎩ ⎭⎪
=
π2 1
4yο 2 { ( yο - 0 ) −
yο
2π
(0 − 0) }
π2 yο
=
4yο y
π2
∴β =
8
2
π
α −1 6 −1 π2 −6 8
∴ = 2 = = 2.76
β −1 π 6 2
π −8
−1
8

n
v ⎧y ⎫
case (b ): =⎨ ⎬
Vο ⎩ yο ⎭
where v is the velocity at a depth " y " from boundary, yο is the total depth of wide channel.
Let B the width of wide channel
n
⎧y ⎫
v = Vο ⎨ ⎬
⎩ yο ⎭
1
A∫
Mean velocity = V = v.dA
yο yο
1 yn Vο
∫ ∫y
n
∴ V= Vο B dy = dy
B yο 0
y 0n ( y0 ) n+1
0
y
Vο ⎡ ( y )n+1 ⎤ ο Vο ⎡ ( yο )n+1 ⎤
= ⎢ ⎥ = ⎢ − 0⎥
( yο )n+1 ⎢⎣ n+1 ⎥⎦
0
( ο )n+1
y ⎢⎣ n+1 ⎥⎦
Vο
=> V =
n+1
Kinetic energy correction factor :
y0
1 1 y 3n
∫ v dA = ∫ V0
3
α= 3
B dy
{ }
3 3
V A Vο 0
y o3n
B y0
n+1
yn
( n+1)3
∫y
3n
= dy
yο y 3n
0 0
y
( n+1)3 ⎡ y 3n+1 ⎤ ( n+1)3 ⎡ y 3n+1 ⎤
o
=> ⎢ ⎥ = 3n+1 ⎢ − 0⎥
y 3n+1
0 ⎣ 3n+1 ⎦ 0 y0 ⎣ 3n+1 ⎦
( n+1)3
=> α =
3n+1

Momentum correction factor

yο
1 1 y 2n
∫ V dA = ∫ V0
2 2
β = B dy
{ }
2 2
V A Vο 0
y 02n
B yο
n+1
y
( n+1 )2
∫y
n
= dy
yο yο2n 0
y0
( n+1 )2 ⎡ y 2n+1 ⎤
= ⎢ ⎥
yο2n+1 ⎣ 2n+1 ⎦ 0
( n+1)2 ⎡ y 02n+1 ⎤
= 2n+1 ⎢ − 0⎥
yο ⎣ 2n+1 ⎦
( n+1)2
=> β =
2n+1
( n+1)3 n 3 + 3n 2 + 3n+1-3n-1
−1
α −1 (3n+1)
∴ = 3n+12 =
β − 1 ( n+1 ) 2
n + 2 n+1-2n-1
−1
2n+1 (2n+1)
α − 1 ( n+ 3 )( 2n+1 )
=
β −1 (3n+1)
1
If n =
7
{ }
3
1
+1
( n+1)3 7
α= =
3n+1 1
3* + 1
7
1.4927
=> α = = 1.0449
1.4285
{ }
2
1
2 +1
( n+1) 7
β= =
2 n+1 1
2* +1
7
1.3061
=> β = = 1.0158
1.2857

Example:
Obtain α and β for the velocity distribution given below
y
u = 0.4 + 0.6 , h=1.0,
h
Solution:
1
1 1 ⎛ y⎞
u = ∫ ( udy ) = ∫ ⎜ 0.4 + 0.6 ⎟
h 10⎝ h⎠
1
⎡ ⎛ y2 ⎞⎤
= ⎢( 0.4 y ) + ⎜ 0.6 ⎟⎟ ⎥
⎜ 2
⎢⎣ ⎝ ⎠ ⎥⎦ 0
= 0.7 m/s
h 1
∫ ( u dy ) = 0.73*1 ∫ ( 0.4+0.6y ) dy
1 3 1 3
α= 3
u h 0 0
1
α=
1
∫
0.343 0
(
0.064+0.216y3 + 0.432 y 2 + 0.288 y dy)
1 ⎡ 1
= 0.064y +0.216y 4 + 0.432 y 3 + 0.288 y 2 ⎤
0.343 ⎣ ⎦0
α = 1.18
Problems:
1. The velocity distribution ( in m/s ) in an open channel 2m deep can be
represented by the equation,
y 1/2
v(y) = 0.6 +1.4 (
)
y0
Calculate the energy correction factor. Here in y is the height above bed and
yo = 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 α and β for a given
cross sections.
15 m
2.8 m/s
2:1 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 y = 10 m
2:1
105 m

⎛ 30y ⎞
3. For an assumed velocity distribution V = 5.75V* log ⎜ ⎟ Prove that
⎝ K ⎠
α = 1 + 3 ∈2 −2 ∈3 and β = 1+ ∈2 in which
Vmax
∈= − 1, Vmax is the maximum velocity, V is the mean velocity.
V

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
1
P1 = γ by12
2
1
and P2 = γ by2 2
2
If Force due to friction can be written as Pf = γ h′f by
in which h 'f is the friction head and y is the average depth, or ( y1 +y2 ) / 2. The
discharge through the reach is equal to
Q=
1
2
(
V1 + V 2 by )
Also, the weight of the body of water is
W = γ byL
z1 − z2
and sin θ =
L
Then the momentum equation, after substituting these expressions simplifies (see box) as
2 2
V1 V2 '
z1 + y1 + β1 = z 2 + y2 + β 2 +h f
2g 2g
2 2
V1 V2
z1 + y1 + α1 = z 2 + y2 + α 2 + hf
2g 2g 1-2
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 hf = h'f if
the difference between α and ρ is ignored. Similarity ends here. There are cases where
either momentum equation or energy equation can be used with the continuity equation.

Momentum Application
⎛ V1 +V 2 ⎞ ⎛ y1 +y 2 ⎞ ⎛ V1 +V 2 ⎞
Q= ⎜ ⎟b⎜ ⎟ =by ⎜ ⎟
⎝ 2 ⎠ ⎝ 2 ⎠ ⎝ 2 ⎠
W = γbyL= Specific weight * (Volume)
z1 -z 2
sinθ =
L
γ ⎛ V1 +V 2 ⎞ 1 1
⎟ ⎡ β 2 V2 -β1 V1 ⎤⎦ = γby1 - γby 2 + γby Lsinθ - γh f by
2 2 '
by ⎜
g ⎝ 2 ⎠⎣ 2 2
γby ⎡ 1 1 z -z
β 2 V1 V2 + β 2 V2 − β1 V1 − β1 V1 V2 ⎤ = γby12 - γby 22 + γby L 1 2 - γh 'f by
2 2
2g ⎣⎢ ⎦⎥ 2 2 L
divided by γb
y ⎡ 1 1 y +y y +y ⎛ y1 +y 2 ⎞
β 2 V2 − β1 V1 + β2 V1 V2 − β1 V1 V 2 ⎤ = y12 - y 22 + 1 2 z1 - 1 2 z 2 − h 'f
2 2
⎢ ⎥ ⎜ ⎟
2g ⎣ ⎦ 2 2 2 2 ⎝ 2 ⎠
y1 +y 2 ⎡ V2 V V2 ⎤ 1
2 2
V1 V V2 1 y y y y ' ⎛ y1 +y 2 ⎞
⎢β 2 − β1 + β2 1 − β1 1 ⎥ = y12 - y 22 + 1 z1 + 2 z1 − 1 z 2 − 2 z 2 + h f ⎜ ⎟
2 ⎢ 2g 2g 2g 2g ⎥ 2 2 2 2 2 2 ⎝ 2 ⎠
⎣ ⎦
Simplifying
2
V2 V2 β V V 2 − β1 V1 V 2 '
β2 − β1 1 + 2 1 = ( y1 -y 2 ) − z1 − z 2 − h f
2g 2g 2g
β 2 V1 V 2 − β1 V1 V 2
If β1 ≈ β2 we can neglect ≈0
2g
2 2
V1 V2 '
z1 + y1 + β1 = z 2 + y2 + β 2 + hf
2g 2g

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
∂
− ( p + γ Z ) = ρ as m direction
∂s binormal
∂
− ( p + γ Z ) = ρ an
∂s
rectifying
plane
normal
plane
tangent s direction
stream line
osculating
n direction plane.
(Principal normal)
The direction of the normal to s direction is towards the plane Centre of curvature is
considered as positive.
v2
Thus the acceleration an is given by a n =
r
in which v is the velocity of flow along the streamline, r is the radius of curvature of the
streamline.
(i) If an is zero then (a) v = 0, no flow and (b) r → ∞ , the streamlines are straight
lines.
∂
− (p + γ z) = 0
∂s
(a) v = 0, then
p
∴ + z = constant.
γ

h =hs
γh
Hydrostatic pressure distribution in parallel flows

p
At free surface = 0 , hence constant = z1
γ
Therefore, at any point x below the free surface, the pressure px
px
= the distance from the free surface say 'h'
γ
∴ px = γ h
h
Straight Gravity Dam
Ho
γh X
P=γΗο
Thus, the pressure varies linearly with depth from free surface and is known as
hydro static pressure distribution.
Hydrostatic
y
any
h
γ ___
g
γh

(ii) In general, when the flow is in the channel with small slope bed θ , 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 pointx can support the weight of the fluid.
Px ∆x = γ y∆x cos θ
or Px = γ y cos θ
A'
A
A'
A
h = y cos2 θ
A
y y
B B' c
Pressure distribution
on A'C
θ
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 θ
p
= h cos θ = d cos 2 θ
γ
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

for the hydrostatic pressure distribution is to be introduced and thus it can be written
as h = h s + c h = h s − c for convex.
r
β0
concave and convex profile on spillway

A c A
c
h
hs hs
h γc γh
γh
B
B' B B'
h = hs - c γc
h = hs + c
Convex surface: Centrifugal force Concave surface: Centrifugal force

opposing Gravity force in the same direction of
Example: Spillway Crest Gravity force
Example: Flip Bucket
Non Hydrostatic Pressure distribution Non Hydrostatic Pressure distribution
p an p ⎛ a ⎞
+z= r + c, for Concave = ⎜1 − n ⎟y for Convex section
γ g γ ⎝ g ⎠
For a Concave vertical sec tion
p ⎛ a ⎞
= y ⎜1 + n ⎟
γ ⎝ γ ⎠
an
thus h= h s ± c in which c =
g
v2 v2
In a curvilinear flow a n = ∴ c=
r 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.
(ii) v = c (free vortex)
r
(iii) v = rc (forced vortex)
2 2
v v
(iv) = , R 0.5 is the radius of curvature at the mid depth.
(r + d 2) R 0.5

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 the depth y it can be shown that the
pressure at any point r and θ is
2
p v ⎛ r ⎞
= ( r − R c + y ) cos β + ln ⎜ ⎟
γ g ⎝ Rc − y ⎠
(b) Effective piezometric head.
⎡ ⎤
2⎢ y 1 ⎥
v ⎢− + ln ⎥
⎢ Rc 1−
y ⎥
⎢ ⎥⎦
= Z2 + y cos β + ⎣
Rc
h cp
⎡ h ⎤
gR c ⎢ ⎥ cos β
⎣ Rc ⎦
__
V
β1
Rc
y
Flow in a bucket

Example:
Compute the overturning moment due to pressure on a retaining wall
soln: (i) Assume θ to be small
y
P
y
__
3
γy
Force acting on the retaining wall, P = Area of pressure triangle.
1 γ y2
= γy y=
2 2
Overturning moment = P * distance from the base at which P is acting
1 1 1
= γ y 2 * y = γ y3
2 2 6
(iii) If θ is not negligible, y = ycos2 θ
( )
2
γ y cos 2 θ γ y2
P= = cos 4 θ
2 2
y γ y3 γ y2
over turning moment = cos θ * = cos 4 θ 4
6 3 6

7.2 Pressure correction coefficients

A
β ′Az γ = ∫ γ hdA
0
A
1
Az ∫0
β′ = hdA
but h = ( h s ± C )
A
1
∴ β′ =
Az ∫ ( hs ± C ) dA
0
A A
1 1
= ∫
Az 0
h s dA +
Az ∫0
C dA
A
1
Az ∫0
β ′ = 1+ C dA
Show that
A A
1 1
α′ = ∫
Qy 0
hvdA = 1 +
Qy ∫0
cvdA
in which α ′ is the pressure distribution coefficient.

d v2
c= , d is the depth of flow in the section.
g r
Solution: h=h s +C
Head recorded in a curve = static pressure ± correction factor.
pressure = Mass of water * depth
A
= ρ g ∫ v dA h
0
Also pressure is α ′yQρ g

A
Thus α ′yQρ g = ρ g ∫ v dA
0
A A
1 1
∫ v dA ( h s + C )
Qy ∫0
∴α ′ = v dA h =
Qy 0
A A
1 1
α′ = ∫
Qy 0
h s v dA +
Qy ∫0
v dA C
A
1
Qy ∫0
α ′ = 1+ v dA C
∵ for uniform flow ∫ h v dA = Q y.

s

8.1 Specific energy equations for rectangular channels

Nm
Specific energy in Open channel is defined as the energy per of water at any
N
section of a channel measured with respect to the channel bottom. Thus it is
the total energy with z being zero.
E
d __
V d
θ b
Total Energy E r section A-A
Z Specific energy equation-

explanation of symbols.
A
See also Notation Table
Datum
Notations
⎛ 2⎞
V ⎟
Er = Total energy above datum = z + d cosθ + α ⎜
⎜ 2g ⎟
⎝ ⎠
⎛ 2
⎞ ⎛ q2 ⎞
⎜ V ⎟
E = Specific energy = d cos θ + α = d cosθ + α ⎜
⎜ 2g ⎟ ⎜ 2 gd 2 ⎟⎟
⎝ ⎠ ⎝ ⎠
Q = Discharge, b = channel width, d = flow depth,
q = Discharge per unit width = Q/b,
tan θ = Bed slope, α = Velocity coefficient, g = Acceleration due to gravity
Thus specific energy can be written as

2
V
E = d cos θ + α
2g

The concept of specific energy as it applies to open channels with small slopes is given
below.
Total energy equation is

2
p v
+ + z = cons tan t
γ 2g
In other words it can be rewritten as
2
v
y + + z = cons tan t
2g
If z = 0 then
2
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
Disch arg e Q = V A
Q
Therefore V =
A
2
2
⎛Q⎞ Q2
V =⎜ ⎟ = 2 2
⎝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
Q2
(E − y) =
2gy 2 b 2
Q
Defining q =
b
q2 Q2
Then ( E − y ) y = = 2
a constant
2gb 2 2g
( E − y ) y2 = constant

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 y1 and y2 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.
two imaginary roots two positive real roots

one negative real root one negative real root
d
depth
E<E c E=Ec E>E c
2 2
Specific energy where E=dcosθ + α (q /2gd )
When E > Ec (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 E = Ec the two positive roots become equal and this depth is
the critical depth. When E < Ec the two positive roots become imaginary and the third
one remains negative.

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 θ and the supercritical depth (and the negative root)
are equal to zero.
SubCritical flow
y2 q3 > q2 > q1
q3
q2
yc q1 A
y1
SuperCritical flow
o
45 C
Imaginary
Specific energy E
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.

For a given Specific energy E there are three possible depths: Two positive values and
one negative value. Two positive values are y1 and y2 respectively representing Super
critical and Sub critical depths. The minimum value of specific energy for the given
discharge represents the critical flow condition.
2
The locus of this represented by yc = E
3
For different values of discharges namely Q1, Q2, Q3 different specific energy curves
would be there.
The minimum specific energy represents the critical condition.

2
αV
E= d cos θ +
2g
dE α dV
= cos θ + 2V
dd 2g dd
2α Q 2 dA
=cos θ - =0
2gA3 dd
2α Q 2
cos θ = T
2gA3
2
α Q2 αV
D cosθ = =
gA 2 2g
2
αV
1=
gD cos θ
Making Froude Number, F = 1 for critical conditions, F can be defined as
V
F=
cos θ
gD
α
This is for non rectangular channel.
If α = 1, and θ is very small then it can be written as
V
F =
gD
In which D is the hydraulic mean depth .
Thus the specific energy is min imum when the flow is critical.
2
⎛ V ⎞
⎜⎜ ⎟⎟ = 1
⎝ gD ⎠
2 2
V V D
∴ =D or =
g 2g 2

8.2 Application of Specific energy

Transition Problem:
__ y1
V1
C1 y2
Sluice gate

Example 1:
Flow below a sluice gate

Pr oblem : If y1 = 2.5 m, y 2 = 0.60 m, b = 3.5 m, determine the discharge Q.
Solution :
Apply Bernoulli equation beteween sections 1 and 2, assuming losses are negligible.
2 2
V1 V2
y1 + + z1 = y 2 + + z2
2g 2g
z1 = z 2 and width b is constant
Q = A1 V1 = A 2 V 2
= b y1 V1 = b y 2 V 2
= 3.5 ( 2.5 ) V1 = 3.5 ( 0.6 ) V 2
8.75 V1 = 2.10 V 2
⎛ 8.75 ⎞
V2 = ⎜ ⎟ V1
⎝ 2.1 ⎠
V 2 = 4.16V1
2 2 2
V 2 V1 V1
= ( 4.16 ) = 17.36
2
2g 2g 2g
Substituting the values int o specific energy equation
2 2
V1 2 V1
2.5 + = 0.6 + ( 4.16 )
2g 2g
2 2
V1 V1
2.5 + = 0.6 + 17.36 ,
2g 2g
2
V1
16.36 = 2.5 - 0.6
2g
V1 =
( 2.5 - 0.6 ) * 2 * 9.81 V1 = 1.5095 ms -1
16.36
Q = 13.2081 m3 /s
V 2 = 6.2795 m/ s
V2
Froude number in the downstream F2 =
gy 2
6.28
= = 2.59
9.81* 0.6

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?
y
__ __
y1 V2
V1
y2
Positive step of z height
y
__ y1
V1 __
y2 V2
Z
Negative step of z height

2 2
V1 V2
y1 + = y2 + + ∆z
2g 2g
Q q2 q2
= q, y1 + 2
= y2 + + ∆z
b 2gy1 2gy 22
q2 q2
y2 + = y1 + − ∆z
2gy 22 2gy12
E 2 = E1 − ∆z
E1 and ∆z are known . E 2 is to be solved for y 2 by trial and error or
using solution of cubic equation.
Note :
Subcritical flow can change over to sup ercritical or subcritical flow depending on
the downstream conditions.
Similarly, sup ercritical flow can change over to sup ercritical or subcritical flow depending on
the downstream conditions.

two positive real roots
depth
m ax
E =E
_V
__c_
yc 2g
45
E1
For a given specific energy E1, the step height = E1 - E2

The maximum step height m ax
= E1 - Ec
Note, if the step height is more than the max
for the given q the choking occurs.
In other words the given discharge cannot flow over the step until the specific energy
increases.
y1'
y2
yc
y1 y'2
Flow over a transition

Given specific energy E1 has two depths namely y1 and y1'
(initial and alternate depth).
The flow correspondingly it would be super critical and sub critical flows
and could be vice versa.
In the downstream for the given specific energy E2 two possible depths are
y2' and y2 corresponding to super critical flow and sub critical flow respectively (alternate depths of E2).
If the critical depth occurs on the step then there are four possible situations
of water surface profiles.
Super critical to Sub critical 1. Super critical flow (y1) changing over to y2 subcritical through yc causing
classical hydraulical jump.
Super critical to Super critical 2. Super critical flow (y1) changing over to y'2 through yc.Thus it would be super critical to
super critical.
Sub critical to Super Critical 3. Sub critical approach flow y'1 changing over to y'2 via critical depth yc. Thus a
hydraulic drop occurs.
Sub critical to Sub critical 4. The sub critical approach flow y'1 changing over to y2 via critical depth yc.
The occurrance of one of the above type of profiles depends entirely on the downstream
condition.

For a given disch arg e

E1 - E 2 = ∆E
E1 = ∆E + E 2
2
If the flow is critical on the step then E c = E1
3
∆Z = E1 - E 2
2
If E 2 is to be equal to critical flow, then Ec = E1
3
2 1
∆Z = E1 - E1 = E1.
3 3
The step height is maximum.
Depths y1' and y'2 are known as alternate depths, and y 2 becomes yc .
The water surface profile downstream of this depends on downstream
controls.
If applied to a case of constriction of width, then critical width of contraction.
y3c =
( q2 * b)
or b
⎡ gy3 ⎤
= ⎢ 2c ⎥
min
g ⎢⎣ q ⎥⎦
Minimum specific energy line on vertical is C which is the critical depth. Therefore
Maximum constriction is obtained.
Flow positive pier - width reduces
Longitudinal water surface profile Looking from downstream
Water surface drops due to constriction in width. Example near the bridge piers.

Exercise problems:
db
1. Show that when, in a rectangular channel with a horizontal bed F = 1 and =0,
dx
the width must be a minimum and not a maximum. (Hint: Consider the variation
of v and with b for F > 1 and F < 1).
8.2.1 Normalisation of the specific energy curves

Consider the specific energy equation
q2
E = y+
2gy 2
Dividing by the critical depth y c , it can be written as
E y q2
= + but
y c yc 2gy 2 y c
⎛ q2 ⎞
⎜ 3 = 1⎟
⎝ gyc ⎠
q 2
yc2
=
2gy 2 yc 2y 2
y E
If = y ' and = E'
yc yc
yc2
∴ E ' = y '+
2y 2
1
E ' = y '+ 2
2y '
which is similar to general form of E ' Vs y '.
when critical depth yc is known from equation this specific energy curve
can be used for obtaining length scale for mod elling.
It can be shown for rectangular channel that
2 3
v q 2 gy3 ⎛ y ⎞
F2 = = 3 = 3c = ⎜ c ⎟
gy gy gy ⎝ y ⎠

Problem:
In case of Simple upward step ( ∆z being +ve),
y1 y 2 ∆z
For a geometrically similar model , , are same in model and prototype each
yc yc yc
y
case. Dynamic similarity condition should exist while the Froude similitude, and if
yc
are equal for two situations, then the ration of discharge is equal to q r = y3/
r .
2
Determine an expression for slope of the straight line to which the upper limb of the
specific energy curve is an asymptote for a channel having a bottom slope of θ .
Solution:
Let us consider Bernoulli's equation.

2
V
z+y+ =H
2g
Q
But V =
A
Let d → depth of flow normal to the channel bottom and θ → slope of the channel.
then y = dcos θ . Therefore the specific energy
α Q2
∴E = d cos θ +
2gA 2
Since d = y cos θ
α Q2
∴E = y cos θ + 2
2gA 2
Consider uniform
α Q2
E − ycos 2θ = constant (approximately)
2gA 2
Assume angle between slope of straight line and horizontal axis as φ
E − ycos 2θ = 0 is one asymptote

y
tan φ =
E
E = y cos 2θ
y
1= cos 2θ from figure
E
1=tan φ cos 2θ
1
tan φ =
cos 2θ
The expression for the slope of straight line to which upper limb of specific energy curve is
⎛ 1 ⎞
φ = tan-1 ⎜ ⎟
⎝ cos 2θ ⎠
The angle (φ ) depends upon the bed slope of the channel.
y d
θ
y
θ d
φ
E
Problem:
Plot the specific energy vs depth curves for Q = 400, 600 and 800 m3/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.

8.3 Problems
1. A rectangular channel, 9.15m wide carries 7.65 m3/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 m3/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 11.30m3/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 m3/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.

__2
V / 2g
1 Sluice Gate
3.0 m
__
0.6 m V2
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.

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
γQ
g
( )
β 2 V 2 -β1 V1 =P1 -P2 +Wsinθ-Pf
If θ = 0, and Pf = 0 and also if β1 = β 2 = 1, then the momentum equation simplifies can be written as
γQ
g
( )
V 2 − V1 = P1 − P2
The hydrostatic pressure forces P1 and P2 are respectively

P1 = γ z1 A1 and P2 = γ z2 A2
in which z1 and z2 are the distances to the centroids below the surface of flow
of the respective water flow areas (A1 and A 2 ).
_
z
centroid from free surface

Q Q
Also,V1 = and V 2 = .
A1 A2
Then, the momentum equation reduces to
Q2 Q2
+ z1A1 = + z2 A 2
gA1 gA 2
The two sides of the above equation are analogous and, hence, may be generally expressed
for any channel geometry as
Q2
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
M1 = M2. This means that the specific forces of sections 1 and 2 are equal, provided that

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 Pf. In Figure the direction of Pf 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
P2
P1
Pf
Definition Sketch - Momentum Equation
If there are any bluff body offering resistance force (Pf) to flow then
Pf
M1 - M 2 =
γ
The force Pf 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 (the sluice gate)

f
2.∆E ≠ 0, P = 0 (the simple hydraulic jump)
f
3.∆E ≠ 0, P ≠ 0 (the hydraulic jump with its formation assisted by some obstructions
f
in the flow such as dentated sill (Forced hydraulic jump)
Sequent depths of Normal Hydraulic jump

If Pf = 0 then the specific force equations can be simplified as
q 2 ⎛⎜ 1 1 ⎞⎟ 1 2
−
g ⎜y y ⎟ 2 1
⎝ 1 2⎠
(
= y −y 2
2 )
q2
i.e.,
gy y (1
= y +y
2 2 1 )
1 2
` The substitution q = v y leads to
1 1
v2 1y
1 =
g 2 y 2 1(2 y +y
)
1
v2 y ⎛y ⎞
or 1 = F2 = 1 2 ⎜ 2 + 1⎟
gy 1 2 y ⎜y ⎟
1 1⎝ 1 ⎠
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 y2 /y1, whose solution is
y y
given by 2 = 1 ⎡ 1 + 8F 2 − 1⎤ and 1 = 1 ⎡ 1 + 8F 2 − 1⎤
y 2 ⎢⎣ 1 ⎥⎦ y 2 ⎢⎣ 2 ⎥⎦
1 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 y1 and y2 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 y1 and y2. 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).

M
A
1
3
C
2 B
M
Specific-force
Specific-force
diagram
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.

Minimum value of specific force:
The specific force to be of a minimum value then the first derivation of M with
dM Q 2 dA d ( zA )
respect to y should be zero, i.e. =- 2
+ =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
dM Q 2 dA
=- 2 +A = 0
dy gA dy
Since, dA / dy = T, Q /A = V, and A / T = D. the above equation reduces to
2
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 E1, the specific - energy curve indicates two possible
depths, namely, a low stage y1 in the supercritical flow region and a high stage y2 in the
sub critical flow region. For a given value of M1, the specific-force curve also indicates
two possible depths, namely, an initial depth y1 in the supercritical region and a sequent
depth y2 in the sub critical flow region. If the low stage and the initial depth are both
equal to y1. Then the sequent depth y2 is always less than the high stage y'2.
Furthermore, the energy content E2 for the depth y2 is less than the energy content E1
for the depth y2. Hence, in order to maintain a constant value of M1, the depth of flow
may be changed from y1 to y2 which results in loss of specific energy is ∆E = E1 − E 2 .

y 0 2 y
M
Sluice 0
0 gate
2 2 2
E
y0 1 y2 y2'
c
c
yc
1
M y1 E1 E 1
Hydraulic
Specific-force jump Specific-energy
diagram diagram
Hydraulic jump at sluice gate outlet
y y
T M
E P dy B
dA
P' _ P
2 z P2
P"
2 y' y
2 Centroid C
C' y2
P'
1
yc P1
y1 yc y1 A
0 M1 M
E
0
45 for a channel E
of zero or small E2 E1
(a) (b) (c)
slope
Specific-force curve supplemented with specific-energy curve.
(a) Specific-energy curve; (b) channel section; (c) specific-force curve
Note:
Specfic energy diagram Specific force diagram
1. Given E1 as initial depth y1 (point P1' ). 1. Corresponding to initial depth y1 specific

force is M1 (Point P1). Initial depth y1 is
Initial depth y1 is super critical depth.
super critical depth.
2. Corresponding to E1 the alternate depth
2. Corresponding to alternate depth y'2 the
y'2 on sub critical limb P2'
specific force is point P.
3. The sequent depth due to hydraulic
3. The corresponding specific force for the
jump is y2 and the corresponding specific sequent depth is M1. In other words for
Normal Hydraulic Jump, the upstream and
downstream specific forces are the same.
4. The specific force corresponding to
sequent depth is indicated by the point P2.

In Hydraulic jump energy loss takes place. The depth corresponding to given E1 at high
stage is known as alternate depth to y1 and vice versa. Whereas the depths due to jump
are known as sequent depths.
y1, y2 are sequent depths.
y1, y'2 are alternate depths.

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, y1. Determine the water surface
profile.
Solution
b1 bx b2
x
0
L
Specific energy line
Critical depth line yx

Steps
Given Q and depth y1
Q
1. Q = A1 V1 , ∴ V1 =
b1y1
2
V1
∴ E1 = y1 + ,
2g
2. It is assume that no energy loss takes place along the transition
2
V2 Q12
3. E 2 = y2 + = y2 +
2g 2gy 22 b 22
4. E1 = E 2 , ∴ obtain subcritical depth y2 by trial and error or by direct solution
5. Let Sub critical depth at any sec tion x be is y x
Q2
E1 = y x +
2gy 2x b 2x
b x = b1 −
( b1 − b2 ) x
L
6. Solve for y x for var ious x.
a. Plot the profile :
In this case as the transition is a straight wall transition,
water surface can be joined between y1 and y 2 .
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 yc
Step1:
2
yc = 3 Q 2
gbx
Step2:
E1 = y1 + Q2
2gb12 y12
yc = 2 E1 ∴ bx can be det er min ed.
3
Thus maximum constriction at x is known. After determining the bx obtain super critical depths
(low stage depths) in the downstream of this constriction. If bx1 is the width at a distance x1
from the constriction then,
b x1 = b x +
( b2 − bx ) x
1
L
Q2
E2 = E x1 = y x1 +
2gb 2x1y 2x1
Solve for yx1 for super critical flow conditions. Figure shows the typical water surface profile.

b1 b2
bx
x
0
L
Plan

High stage
Sub critical flow
Critical depth line

yc Super critical flow
Low stage
Water surface profile with hydraulic drop

3. Super critical flow occurs in the approach channel of a transition in which b1 changes
to b2 as shown in figure. Locate the hydraulic jump if it occurs.
Solution
b1 bx b2
x1
Specific force line for high stage
Specific force line for low stage

for upstream
E Specific energy line

for downstream
High stage
Low stage
a) When jump occurs energy loss takes place
E1 − E 2 = ∆E
2
V1 Q2
E1 = y1 + = y1 +
2g 2gy12 b12
V1
b) y1 is given Q is given ∴ F1 =
gy1
c) Given Q, b 2 , y 2 at the downstream section, y 2 should be sub critical depth if the jump
has to occur.
d) Jump occurs but (i) can occur in the transition reach, (ii) not in the transition reach.

Step1: Compute E1 and plot the line.

Compute E 2 and plot the line.
Q2
Step 2 : Compute specific force M1 = zA1 +
gA1
Q2
Similarly compute M 2 = z +
gA 2
Step 3:
When the specific force M1=M2, 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.
low stage
Specific Specific
depth for Remarks
x bx force for force for
specific
low stage high stage
energy E1
0 b1 y1 M1 M2
2
5
-
-
x bx
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 Hx of the water surface profile is known.
2. Assume that E1 to E2 loss is distributed linearly over transition.

3. It can be written that

H x = yx + zx ,
∴ zx = Hx − yx
At any point x the specific energy is given by
2
⎡ ( E1 − E 2 ) x ⎤ = V x + H
⎢ E1 - ⎥ x
⎣ L ⎦ 2g
I n the above equation only unknown is velocity. Hence calculate the velocity .
But V x =
Q2
=
Q2
. Calculate y x knowing bx = b1 +
( b2 − b1 ) x
A 2x b 2x y 2x 2g L
∴ Obtain z x from the known H x by subtracting the depth y x
and plot the z as a function of x to obtain the hump profile.
Alternative solution for eliminating the jump is to increase the roughness
In other words increasing the friction . Also by changing the width .
b1 bx b2
x1
Specific force line for high stage

for upstream
Specific force line for low stage

for downstream
__2
Vx
____
2g
yx
zx

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.
7.63 cm
45.7
Z = 4.57 cm
Flow over a step

Solution
y 2 = 45.7 − 4.57 − 7.63 = 45.7 − 12.2 = 33.5 cm
2 2
V1 V2
E1 = y1 + E 2 = y2 +
2g 2g
V1 y 2 33.5
by1 V1 = by 2 V 2 ∴ = = = 0.733
V 2 y1 45.7
So V1 = 0.733V 2 or V 2 = 1.364 V1
⎛ 2
⎞ ⎛ 2
⎞
∆z = ⎜ y1 + ⎟−⎜y + ⎟
V1 V 2
⎜ 2g ⎟ ⎜ 2 2g ⎟
⎝ ⎠ ⎝ ⎠
2
⎡ 2
⎤
V1 ⎢ V 2 ⎥
4.57 = 45.7 − 33.5 + 1− 2
2g ⎢ ⎥
⎣ V1 ⎦
2
V1 ⎡ ⎛ 1 ⎞2 ⎤
⎢1 − ⎜ ⎟ ⎥ = 4.57 − 45.7 + 33.5
2g ⎢⎣ ⎝ 0.733 ⎠ ⎥⎦
∴
V1 = 131.866 cm / s
Q = 6.026 l / s / cm

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
3m
3m
3 m/s 0.61 m
3m
3 m/s
3m b=?
Q = 3*3*3 = 27 m3 / s
Q 27
q= = = 9.0 m3 / s / m
b 3
1/ 3 1/ 3
⎛ q2 ⎞ ⎛ 9.02 ⎞
yc = ⎜ c ⎟ =⎜ = 2.021 m
⎜ g ⎟ ⎜ 9.81 ⎟⎟
⎝ ⎠ ⎝ ⎠
V12 32 Nm
E1 = y1 + = 3+ = 3.46875
2g 19.2 N
32
3+
E 19.2 = 1.72
E1' = 1 =
yc 2.021
∆Z 3.46875 0.61
Downstream specific energy E'2 = E1' − = − = 1.4145
yc yc 2.021
If the flow has to be critical
E2
E 2 = 3.46875 − 0.61 = 2.85875 E '2 = = 1.5
yc2

E 2 2.85875
yc2 = = = 1.9058
1.5 1.5
Q = b 2 yc2 Vc2 = b 2 yc2 gyc2
27
∴ b2 = = 3.2765 m
1.9058 9.81*1.9058
∴ For critical flow to occur downstream width must be 3.2765 m.
Minimum exp ansion permitted is 0.2765 m in 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.
70 cm
23 cm 481 cm
Plan
400 cm
Q = 11 l/s yn
S0 = 0.0016
25 cm 15 cm
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

downstream after the transition and if the jump has to occur downstream of the
transition, what necessary modifications are required.

Y
B B
__ B/2
2
B
1.5 B Y B Section along "YY"
Plan
75 50
50
Longitudinal Section along a Transition
(All dimensions are in cm)
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
m3/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.

9.3 Application of Specific Force and Specific Energy

1. Determine the energy Loss in a NHJ
Solution:
Applying Momentum equation

γQ
( V2 − V1 ) = P1 − P2
g
γ Q 2 ( y1 − y 2 )
gb 2 y1y 2
(
= y12 − y 22 )
Q2
=
( y1 + y2 ) y1y2
2 4
2gb
Q2
2
= q2
b
q 2 ( y1 + y 2 ) y1y 2
= (1)
2g 4
Specific energy equation
V12 V2
y1 + = y 2 + 2 + ∆E
2g 2g
Q2 Q2
∆E = y1 + − y2 −
2gy12 b 2 2gy 22 b 2
q2 ⎡ 1 1 ⎤
∆E = ( y1 − y 2 ) + ⎢ 2 − 2⎥
2g ⎣⎢ y1 y 2 ⎥⎦
⎡ q2 1 ⎤
= ( y1 − y 2 ) ⎢1 + 2 2
( y2 + y1 )( −1)⎥
⎢⎣ 2g y1 y 2 ⎥⎦
Substituting from momentum equation
⎡ ( y + y ) y y ( y + y )( −1) ⎤
= ( y1 − y 2 ) ⎢1 + 2 1 1 2 2 2 1 2 ⎥
⎣⎢ 4 y1 y 2 ⎦⎥
=
( y1 − y2 ) ⎡ 4y y − ( y1 + y 2 ) ⎤
2
1 2
4y1y 2 ⎣ ⎦
=
( y1 − y2 ) ⎡ 4y y − y12 − y 22 − 2y1y 2 ⎤
4y1y 2 ⎣ ⎦
1 2
∆E =
( y2 - y1 )3
4y1y 2

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
0.5
1
3
A= (b+my)y
(b+my)y
______________ =R
b+2 m2+1 y
Q = A.ν =
= (3+0.5) 0.5*4
3.5
= *4 = 7 m3 /s
2
Q2
M1 = + zA
gA
72 ⎛ 3*0.52 1*0.53 ⎞
= +⎜ + ⎟
3.5 ⎜ 2 3 ⎟⎠
9.81* ⎝
2
= 2.86 + 1.5*0.25 + 0.041666 = 3.2767m3
M1 =M 2
72 ⎛ 3*y 2 2 2 3 ⎞
M1 = +⎜ + y2 ⎟
⎜ 2
9.81* ( 3+y 2 ) y 2 3 ⎟
⎝ ⎠
Solve by trial and error y 2 = 1.05m.
Alternative approach is:
z M1 1*3.2767
3
= = 0.12135
b 33
m2Q2 1*7 2
5
= 5
= 2.05×10-2
gb 9.81*3
From graph y 2 = 1.05m.

9.3.1 Hydraulic Jump
1. For the case of hydraulic jump in a rectangular channel, complete the following table.
y1 ( Head
V 1 (m /s) q (m3 /s) y2 (m) V 2 (m / s)
m) 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 m3/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.
0.6 m 0.5 m
Section
8m
Plan
Stilling basin

4. A rectangular channel 6m wide carries 11.5 m3/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 m3/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 m3/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.
⎡ b y 2 my 2 ⎤ Q2ρ
ρg ⎢ + ⎥+ =Constant
⎣ 2 3 ⎦ ( b+my )y
Draw the force-momentum 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 m3/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 m3/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 m3/s at a flow depth of 5 m.

(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 m3/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.

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
1 1
1 1
6.096 m

3.6576 0.957
4.980 m
6.096 m 15.24 m
0.957
3.6576
6.096
x2
______ y2
______
+ =1
6.0962 2.492
Plan of transition - Trapezoidal to Horse shoe Tunnel
x2 y2
+ =1
37.1612 6.2001
d o = 6.0198 m
do
= 3.099 m
2
3.6576 m 6.0198 m
4.93776 m
15.24 m
Design Discharge = 62.297 m3/s

n = 0.018

d
__0
0.2153 2 0.2153
d0 d0
1 1
d0
4.9804
Exit Tunnel
4.2672 m
u/s
33.528 m 6.096 m 1 4.2672 m 1
1 1
6.0198 m 4.2672 m 6.096 m

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".
yc = 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 Sc". 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
E-y 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.

10.1.2 Analytical Properties of Critical Flow

v2 q2
Consider the Specific energy equation E = y + = y+
2g 2gy 2
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
dE q2
=1- 3 =0
dy gy
q2
∴ q 2 = gy3c or yc = 3
g
and Vc 2 = gyc
The subscript c indicates critical flow conditions.
Thus the critical depth yc is a function of discharge per unit width alone.
Further, the above equation it can be written as
Vc 2 1
= yc .
2g 2
Thus the specific energy for critical flow can be expressed as
Vc 2 3
E c = yc + = yc
2g 2
2
or yc = E c
3
d 2E
The second derivative should be negative i.e., = - ve
dy 2
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.

q max Maximum point
E0
2
yc= __ E0
3
q
0
Variation of the Discharge with depth for a given specific energy value
How q varies with y for a given E= Eo?
When y → Eο and then q → 0. Similarly, when y → 0, q → 0, and there will clearly be a
maximum value of q for some value of y between 0 and E0 (y cannot be greater than
E0). The relationship can be written as q 2 = 2 gy 2 ( Eο − y ) and differentiating the above
equation with respect to y,
dq
2q = 4gyE ο - 6gy 2 = 0
dy
∴ 6gyc2 = 4gyE 0
2
i.e., yc = E0
3
d2E
Differentiating again it can be established that = + ve
dy 2
Alternative approach:
Show that the flow is maximum when it is critical flow for a given specific energy plot the
graph " E0” verses” q”
Solution:

V2 Q2
Specific energy = E = y + =y+
2g 2gA 2
when " α = 1.0 "
∴ Q = 2gA 2 ( E-y ) = A . 2g ( E - y )
dQ
For the flow to be maximum, " = 0"
dy
dQ d
= ⎡A 2g (E - y ) ⎤ = 0
dy dy ⎣ ⎦
1
dA d
= 2g (E - y ) + A 2g . (E-y ) 2 = 0
dy dy
dA A 2g
= 2g (E - y ) − =0
dy 2 E- y
dA A 2g
2 g (E-y ) =
dy 2 E- y
dA
2 ( E- y ) =A (1)
dy
But Q = A 2 g (E - y)
Q2
=> 2(E - y ) =
gA 2
dA
Substituting in eqn: (1 ) and taking = T,
dy
Q2 T
=A
gA 2
Q2T
=> =1
gA 3
Q2 T V2 T V2
But = =
gA 3 gA gD
V2
∴ =1
gD
V2
But = F2
gD
=> F2 =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.

Any one of the above three equations may be used to define critical flow. For example:
(1) The crests of E-y curves drawn for all values of q can be joined by a straight line
having the equation y = 2E / 3, as shown in Figure.
q1 SubCritical flow limb

q2 q3 q3 > q2 > q1
o SuperCritical flow limb

45
E
(2) yc 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 θ is small than y → 0, E → ∞, an asymptote.
Similarly y = E is another asymptote. The specific energy equation can be written as
q2
( E − y ) y2 = = 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.

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).

Transition (flow basis):
1. Sub critical to Sub critical
2. Sub critical to Super critical (Hydraulic drop)
3. Super critical to Sub critical (Hydraulic Jump)
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.

Following assumptions are made
1. Constant rectangular channel
2. Short reach.
3. No frictional loss.
4. Hydrostatic pressure distribution is assumed.

__ 2 __ 2
_V_ V
__
2g 2g
T.E
y
y
z
x z = f(x)
Longitudinal section
q
b
Plan

⎛ Q ⎞
The total energy H and q ⎜ = = discharge per unit width ⎟ are constant,
⎝ b ⎠
q2
H = y+z+ = E+z = constant
2gy 2
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
Substituting and simplifying
dy
dx
(
1-F2 +)dz
dx
=0 (∵
dE
dy
=1-F2 ; F =
V
gy
)
2
V
E=y+
2g
dE d ⎛ Q2 ⎞ Q2 ⎛ -3 dA ⎞
= 1+ ⎜ ⎟ = 1+ ⎜ −2A ⎟
dy dy ⎜⎝ 2gA 2 ⎟⎠ 2g ⎝ dy ⎠
dE Q2
=1- 3
T = 1- F2
dy gA
Q 2T
(i.e) F2 =
gA3
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 E-y curve.
If there is an upward step in the channel bed, i.e., if dz/dx is positive, then the product
(1-F2 ) dx
dy
must be negative and vice versa (see box).

dz
If is positive
dx
z
Bed
dy
dx
( )
1 − F2 = negative
dy
F < 1 ( Subcritical ) -ve ( depth decreases along x )
dx
dy
F > 1 ( Supercritical ) +ve ( depth increases along x )
dx
dz
If is negative
dx
Bed
z
dy
dx
( )
1 − F2 = positive
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
dx
(
= 0 ,. Then the product 1-F2 )
dy
dx
is then
dy
equal to zero. Hence, either = 0 or F = 1 (critical flow) .
dx
dz dy
The first situation occurs in the step-transition problem when =0, = 0 both
dx dx
upstream of the step and over the step, and in both cases F ≠ 1 .
dz
For the second situation, the question is " Can a situation be visualized in which =0
dx
dy
and ≠ 0?
dx

The answer is yes.
Consider the Free outflow from a Lake as an example of critical flow.
Flow Eo 2
y = E
c 3 0
P
dz dy
= 0, 0, F = 1
dx dx
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.
dz dy
At the crest P, = 0 the flow is accelerating at this point, resulting in ≠ 0 . Then the
dx dx
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
5
yc , (i.e 0.715 yc ) and that y = yc at a distance upstream from the overall edge of weir of
7
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

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 over a sharp crested weir Free overfall (drop)
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.

When the canal drops suddenly, a free overfall is formed, since flow changes to
supercritical flow can be used as a measuring device.
__ 2
v
__ y
__
2g yc
H1
yc
yb
-3 -2 Level -1 0 +1 X
__
yc
- 0.5
- 0.6
Minimum drop distance
Free overfall profile
The drop distance should be more than 0.6yc. Brink depth yb 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.
q2
Ho = y + α
2gy 2
Differentiating w.r.t 'y' assuming Q to be constant.
dH o q2
= 1− α 3
dy gy
dH o αq 2
= 0 if the flow is critical, hence yc = 3
dy g
If α =1, then Q = b g y3/2
c
Rouse showed y b = 0.715yc
3/ 2
⎡ y ⎤
Thus Q = b g ⎢ b ⎥
⎣ 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 yc. This is applicable to canals with
slopes upto 0.0025.

Thin weir plate - Free over fall

x Brink depth
yc or
End depth (yb)
L
_y_c 1.4 , x = 3 to 4 yc
yb
Brink depth

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
V2
Total energy = TE = H = z + y + ( α = 1.0 )
2g
Q2 q 2 .b2
=>H=z+y+ = z + y +
2gA 2 2g .b 2 y 2
2
⎡q ( x )⎤⎦
H=y+z + ⎣
2gy 2
Differentiating both sides with respect to " x ",
d ⎧⎪ ⎡⎣q ( x )⎤⎦ ⎫⎪
2
dH dy dz
= + + ⎨ ⎬=0
dx dx dx dx ⎪ 2gy 2 ⎪
⎩ ⎭
dH dz
If = 0 and = 0 ( No energy loss, Horizontal channel )
dx dx
dy q 2 dy q dq
- + =0
dx gy 3 dx gy 2 dx
and by continuity equation q b = a constant, Q.
Then
dQ dq db
=0=b +q =0
dx dx dx
dq db
b =−q
dx dx
dq
Eliminating , between above two equations then it may be written as
dx
dy
dx
(
1-F 2 - )
q q db
gy 2 b dx
=0
i.e.,
dy
dx
( )
1-F 2 - F 2
y db
b dx
=0
db
It can be concluded that critical flow occurs when , i.e., at a section of maximum
dx
horizontal constriction. The critical flow will not occur at a section of maximum width, but
only at a section of minimum width.

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 depth increases as x increases
dx dx
dy
F>1 supercritical then < 0 depth increases as x increases
dx
Converging channel
F<1 F>1
dy
__ < 0 dy
__ > 0
dx dx
Sub critical db
__ < 0 Super critical
dx
Diverging channel
F<1 db F>1
Sub critical __ > 0 Super critical
dx
Horizontal constriction

Derive the following equation for a non prismatic channel, assuming no energy loss.
y 3c db
Sο + 2 .
dy by dx
= 3
dx ⎛y ⎞
1−⎜ c ⎟
⎝ y ⎠
Solution:
Total energy at any section is given by
V2
H= z+y+ ( α = 1.0 )
2g
Differentiating wrt "x",
dH dz dy d ⎪⎧ V 2 ⎪⎫
= + + ⎨ ⎬ − − − − − − − − − −− > (1)
dx dx dx dx ⎩⎪ 2g ⎭⎪
dH
But = −S f
dx
dz dH
Similarly = −Sο But Sf = 0 => =0
dx dx
Substituting in eq: ( 1 ),
dy d ⎪⎧ V 2 ⎪⎫
0 = - Sο + + ⎨ ⎬ − − − − − − − −− > (2)
dx dx ⎪⎩ 2g ⎪⎭
dy d ⎧⎪ Q 2 ⎫⎪
0 = - Sο + + ⎨ ⎬
dx dx ⎪⎩ 2gA 2 ⎭⎪
Consider a rectangular channel with varying width
d ⎪⎧ Q 2 ⎫⎪ d ⎪⎧ Q 2 ⎪⎫ Q 2 d ⎪⎧ 1 ⎪⎫
⎨ ⎬= ⎨ ⎬= ⎨ ⎬
dx ⎩⎪ 2gA 2 ⎭⎪ dx ⎩⎪ 2g b2 y 2 ⎭⎪ 2g dx ⎩⎪ b2 y 2 ⎭⎪
Q2 ⎪⎧ −2 db −2 dy ⎪⎫
= ⎨ 3 2 − 2 3 ⎬
2g ⎪⎩ b y dx b y dx ⎪⎭
Q 2 db Q 2 dy
= −
gb3 y 2 dx gb2 y 3 dx

Substituting this expresstion in eq: ( 2 ),

dy Q 2 db Q 2 dy
- Sο + − 3 2 − 2 3 =0
dx gb y dx gb y dx
dy ⎛ q2 ⎞ q 2 db
=> - Sο + ⎜ 1 − ⎟ − =0
dx ⎜⎝ gy 3 ⎟⎠ gb y 2 dx
q2
But = y 3c
g
y 3c db
Sο + .
dy by 2 dx
∴ = 3
dx ⎛y ⎞
1−⎜ c ⎟
: ⎝ y ⎠

11.1 Critical depth in Trapezoidal and Circular channels

Problem:
( y'c + 1)
3 3
Q m 2 3 y'c myc
For trapezoidal channel show that = , where y'c =
gb 5
2yc' +1 b
Solution:
The most important basic problem is to determine the critical depth.
From the dimensional analysis
yc = f ( Q, b, g, m )
yc ⎛ Q2 ⎞
Combining it can be rewritten as = f ⎜ 5 ,m⎟
b ⎜ gb ⎟
⎝ ⎠
V
For critical flow F = = 1.
gD
2
V
∴ =D
g
2
Q2
From continuity equation V =
A2
Q2
A 2D = .
g
Q
∴ Section factor Z =
g
Consider A= ( b+myc ) yc
⎛ my ⎞
= ⎜ 1+ c ⎟ byc
⎝ b ⎠
⎛ 2myc ⎞
T = ⎜1 +
b ⎟⎠
b
⎝
⎛ myc ⎞
A ⎜⎝ b ⎟⎠
1+ yc
=
T ⎛ 2myc ⎞
⎜1 + b ⎟
⎝ ⎠
⎡ ⎛ myc ⎞ ⎤
⎤ ⎢ ⎜⎝ b ⎟⎠ ⎥
2 1+ yc
⎡⎛ my ⎞
Z = A D = ⎢⎜ 1+ c
2 2
⎟ c⎥
by ⎢ ⎥
⎣⎝ b ⎠ ⎦ ⎢ ⎛1 + 2myc ⎞ ⎥
⎢⎣ ⎜⎝ b ⎟⎠ ⎥⎦
myc
Defining y'c =
b
( 1+y'c ) ( 1+y'c )( yc3 b 2 ) Q 2

2
2
=
(1+2y'c )
Z =
g
m3
Multiplying on both sides by we get
b5
() ( )( yc3b2 ) ⎛⎜ m3 ⎞⎟ = (1+y'c ) (1+y'c ) y'

2 2
Q2 m3 1+y'c 1+y'c 3
=
g b5 ( )
1+2y'c ⎜ b5 ⎟
⎝ ⎠ (1+2y'c ) c
Q 2 m3 ( 1+yc ) yc
' 3 ' 3
=
g b5 (1+2y'c )

Problem:
Show that for circular channel

2
⎧⎪ Q ⎫⎪ β − sinβ cosβ
⎨ 2 ⎬ =
⎪⎩ do gdο ⎪⎭ 64 sinβ
In which “y” is the depth of flow " d0" is the diameter of the circular channel.
Solution:
When flow is critical,

V
Froude number F = =1
gD
Q
=> V = gD =
A
Q2
=> = gD − − − − − − − −− > (1)
A2
r2 d 02
Area of flow A = (θ − sinθ ) = (θ − sin θ )
2 8
substituting θ = 2 β it can be written as
d 02
A= ( 2 β − sin2β )
8
⎡d ⎤
Top width = T = 2 ⎢ ο sin (π − β ) ⎥ = dο sin (π -β ) = dο sin β
⎣ 2 ⎦
A dο ⎡ ( 2 β − sin2β ) ⎤ dο ⎡ ( 2 β − 2sinβ cos β ) ⎤
Hydraulic depth D = = ⎥= 8 ⎢
T 8 ⎢⎣ sinβ ⎦ ⎣ sinβ ⎥
⎦
d ⎡ ( β − sinβ cos β ) ⎤
D= ο ⎢ ⎥
4 ⎣ sinβ ⎦
Q2 Q2
From eq: ( 1 ) gD = =
A 2 dο4 ⎡
⎣
( 2 β − 2sinβ cos β )2 ⎤⎦
64
gdο ⎡ β − sin β cos β ⎤ Q2
gD = =
4 ⎢⎣ sin β ⎥ 4
⎦ dο ⎡ β − sinβ cos β 2 ⎤
( ) ⎦
16 ⎣
Q2 g dο ⎡ β − sin β cos β ⎤
∴ 4 =
dο ⎡ 2⎤ 4 ⎢⎣ sin β ⎥
⎦
⎣
( β − sinβ cos β ) ⎦
16
1 ⎡ ( β − sin β .cos β ) ⎤
3
Q2
=> = ⎢ ⎥
gdο5 64 ⎢⎣ sinβ ⎦⎥
Q2 1
∴ = ( β − sin β cos β )3
dο gdο 64 sinβ
2

11.2 Hydraulic exponent for critical flow: M

The section factor Z for critical flow in general can be expressed as
Z2 = Co ycM
in which M is an exponent to be determined and C0 is a constant proportionality.
Taking logarithm on both sides
2 ln Z = ln Co + M ln yc
Differentiating with respect to y
d d d
2 ( ln Z ) = ( ln Co ) + M ( ln yc )
dy dy dy
d M
( ln Z ) = (1)
dy 2 yc
A A
But from definition Z =
T
Taking logarithm it may be expressed as
( ) (
ln Z = ln A3/ 2 + ln T −1/ 2 )
d 3 d 1 d
( ln Z ) = ( ln A ) − ( ln T )
dy 2 dy 2 dy
d 3 1 dA 1 dT
( ln Z ) = − ( 2)
dy 2 A dy 2 T dy
Comparing equations (1) and (2) it may be written as
⎡ 3 1 dA 1 dT ⎤
M = 2 yc ⎢ − ⎥
⎣ 2 A dy 2 T dy ⎦
dA
But = T, then
dy
⎡ T 1 dT ⎤
M = y c ⎢3 - ⎥
⎣ A T dy ⎦
y ⎡ A dT ⎤
M = c ⎢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.

(a) If channels of rectangular cross section,
dT
=0
dy
3y T
∴M = c
byc
M = 3.0 (∵ T = b )
( b) For trapezoidal channel obtain the following expression
M=3
(1 + 2y'c )
−
2y'c
(1 + y'c ) (1 + 2y'c )
yc
i n which y'c = m .
b
Solution :
For trapezoidal channel ( for critical flow )
dT
A = ( b+ my ) yc , T = b+ 2 myc , = 2 m.
dy
yc ⎡ A dT ⎤
Substituting the above in the standard expression for M =
A⎣ ⎢ 3T-
T dy ⎥⎦
yc ⎡ ( b+m yc ) yc 2m ⎤
M= ⎢3 ( b+ 2 myc ) − ⎥
( b+m yc ) yc ⎣⎢ ( b+2m yc ) ⎥⎦
⎡ ⎛ myc ⎞ ⎤
⎢ b ⎜1+ ⎟ 2 myc ⎥
1 ⎛ 2 myc ⎞ ⎝ b ⎠
M= ⎢3b ⎜ 1+ ⎟ - ⎥
⎛ myc ⎞ ⎢ ⎝ b ⎠ ⎛ 2 myc ⎞ ⎥
b ⎜ 1+
b ⎟⎠ ⎢⎣ ⎜ b ⎟⎠ ⎥⎦
b 1+
⎝ ⎝
⎡ ⎛ myc ⎞ 2 myc ⎤
b ⎜1+
1 ⎢ ⎛ 2 myc ⎞ ⎝ b ⎠⎟ b ⎥
M= ⎢3b ⎜ 1+ ⎟ - ⎥
⎛ myc ⎞ ⎢ ⎝ b ⎠ ⎛ 2 myc ⎞ ⎥
b ⎜ 1+
⎝ b ⎟⎠ ⎢⎣ ⎜1+ b ⎟ ⎥
⎝ ⎠ ⎦
⎡ ⎛ 2 myc ⎞⎛ 2 myc ⎞ ⎛ myc ⎞ 2 myc ⎤
⎢ 3 ⎜ 1+ b ⎟⎜ 1+ b ⎟ - ⎜ 1+ b ⎟ b ⎥
⎢ ⎝ ⎠⎝ ⎠ ⎝ ⎠
1
M= ⎥
⎛ myc ⎞ ⎢ ⎛ 2 myc ⎞ ⎥
⎜1+ b ⎟ ⎢ ⎜ 1+ b ⎟ ⎥⎦
⎝ ⎠⎣ ⎝ ⎠
myc
If y'c =
b
1 ⎢ ( )( c )( )
⎡ 3 1+ 2 y' 1+ 2 y' - 1+ y' 2 y' ⎤
c c c
⎥
(1+ ) ( )
M=
y'c ⎢ 1+ 2 yc' ⎥
⎣ ⎦
M=⎢
⎡
⎢ 3 (1 + 2y c ) − 2 (1 + yc ) y c ⎥
' 2 ' ' ⎤
⎢
⎣
(1 + y'c )(1 + 2y'c )
⎥
⎥
⎦
M=3
(1 + 2y'c ) − 2y'c
(1 + y'c ) (1 + 2y'c )

Show for Triangular channel M = 5.0

Solution :
⎛ y⎞
Section factor Z = A D = my 2 ⎜⎜ ⎟⎟
⎝ 2⎠
2
⎡ ⎛ y ⎞⎤ 2 4⎛y⎞ m
2
Z = ⎢ my 2 ⎜⎜
2
⎟⎟ ⎥ = m y ⎜ ⎟ = y5
⎢⎣ ⎝ 2 ⎠ ⎥⎦ ⎝2⎠ 2
Z2 = Cy M
Comparing these two equations M = 5
Critical flow exponent for non prismatic channel ( Nature channel ) :
log Z2 - log Z1
tan θ =
log y 2 - log y1
⎛Z ⎞
log ⎜ 2 ⎟
tan θ = ⎝ Z1 ⎠
⎛y ⎞
log ⎜ 2 ⎟
⎝ y1 ⎠
Z2 = CycM
2 lnZ = ln C+ M lnyc
ln Z
∴M = 2 = 2 tan θ
ln yc
M = 2 tan θ
2 lnZ1 = lnC+ M lnyc1
2 lnZ2 = lnC+ M lnyc2
Subtracting
2 lnZ1 - 2 lnZ2 = M ⎡⎣lnyc1 - lnyc2 ⎤⎦
⎛Z ⎞ ⎡ yc ⎤
2 ln ⎜ 1 ⎟ = M ln ⎢ 1 ⎥
⎝ Z2 ⎠ ⎣⎢ yc2 ⎥⎦
⎛Z ⎞
2 ln ⎜ 1 ⎟
M= ⎝ Z2 ⎠ = 2 tan θ
⎡ yc ⎤
ln ⎢ 1 ⎥
⎣⎢ yc2 ⎦⎥
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
yc Q
Similarly the graph for Vs 2 can be constructed.
d0 D gD

11.3 Problem
Derive the value of M, N for rectangular (narrow, wide), Trapezoidal, Triangular channel
by using the following expression.
y⎛ A dT ⎞ 2y ⎛ dP ⎞
M= ⎜ 3T − ⎟ and N = ⎜ 5T − 2R ⎟
A⎝ T dy ⎠ 3A ⎝ 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 ⎥
3 by ⎣ 2 ⎦
2 y
= [5b − 2b] = 2.0
3 by
For Chezy relationship
y⎛ dp ⎞
N= ⎜ 3T − R ⎟
A⎝ dy ⎠
y ⎛ by ⎞
N= ⎜ 3b − ( 0) ⎟
by ⎝ b + 2y ⎠
N = 3.0

(b) Triangular Channel:

1) Area of triangle
A = my 2
2) P = 2y 1+m 2
my
3) R =
2 1+m 2
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 + m2 ⎥
3A ⎢⎣ 2 1 + m2 ⎥⎦
2y ⎡ my ⎤
= ⎢10my − 2 1 + m 2
⎥
3my 2 ⎣⎢ 1 + m2 ⎦⎥
2 y2
= [10m − 2m]
3 my 2
2 16
=8m* =
3m 3

(c ) Trapezoidal Channel
y⎛ A dT ⎞
M= ⎜ 3T − ⎟
A⎝ T dy ⎠
y ⎡ ( b+my ) y d ( b+2my ) ⎤
= ⎢3 ( b+2my ) − ⎥
( b+my ) y ⎣ ( b+2my ) dy ⎦
y ⎡ ( b+my ) y 2m ⎤
= ⎢3 ( b+2my ) − ⎥
( b+my ) y ⎣ ( b+2my ) ⎦
y ⎡ 3 ( b+2my )2 − ( b+my ) 2my ⎤
= ⎢ ⎥
( b+my ) y ⎢⎣ ( b+2my ) ⎥⎦
3 ( b+2my ) − ( b+my ) 2my

2
=
( b+my )( b+2my )2
2
⎡ my ⎤ ⎛ my ⎞ my
3b ⎢1 + 2
2
− ⎜ 1+ ⎟2 b
⎣ b ⎥⎦ ⎝ b ⎠ b
=
⎛ my ⎞ ⎡ my ⎤
⎜ 1+ ⎟ by ⎢1 + 2 b
⎝ b ⎠ ⎣ b ⎦⎥
2
⎡ my ⎤ ⎛ my ⎞ my
3 ⎢1 + 2 − ⎜ 1+ ⎟2
⎣ b ⎥⎦ ⎝ b ⎠ b
M=
⎛ my ⎞ ⎡ my ⎤
⎜ 1+ ⎟ ⎢1 + 2
⎝ b ⎠⎣ b ⎦⎥
my c
if y 'c =
b
( )
2
3 ⎡⎣1 + 2 y 'c ⎤⎦ − 1+y 'c 2y 'c
M=
( 1+y'c ) ⎡⎣1 + 2y'c ⎤⎦
_______________________________________________________
2y ⎛ dp ⎞
N= ⎜ 5T − 2R ⎟
3A ⎝ dy ⎠
2y ⎛ ( b+my ) y ⎞⎟ − 2 1+m2
= ⎜ 5 * ( b+2my ) − 2
3 ( b+my ) y ⎜ b+my 1+m 2 ⎟⎠
⎝
⎛ y⎞ y
10 ⎜ 1+2m ⎟ 8 1+m 2
⎝ b⎠ b
N= −
⎛ y⎞ ⎛ y ⎞
3 ⎜ 1+m ⎟ 3 ⎜1 + 2 1+m 2 ⎟
⎝ b⎠ ⎝ b ⎠
y
N=
10 1+2y ' ( −
8
b) 1+m 2
in which y ' =
my n
3 1+y ' ( ⎛ )
3 ⎜1 + 2
y
1+m 2 ⎟
⎞ b
⎝ b ⎠

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.
A = ( b + myc ) yc
D=
( b + myc ) yc
( b + 2myc )
Q
Zc = = constant = C1 = known
g
⎧⎪ ( b + myc ) yc ⎫⎪
1/ 2
C1 = ( b + myc ) yc ⎨ ⎬ (1)
⎩⎪ ( b + 2myc ) ⎭⎪
C12 ( b + 2myc ) = ( b + myc ) y3c
3
leads to
y6c + py5c + qyc4 + ry3c + syc + t = 0
in which the cons tan ts p, q, r,s and t are known.
Solve this by polynomial or by trial and error method.
It would be easier to solve the equation (1) by trial and error procedure.
After obtaining the answer check for the Froude number which should be equal to 1.

Example:
Consider a Rectangular channel and obtain the critical depth for a given discharge.
Solution:
A by
Area = b y D = = =y
T b
Q
∴ Z = = b y y1/ 2
g
Q
c =
y3/ 2
b g
2/3 2
⎛ Q ⎞ ⎛ q ⎞ q2
yc = ⎜ = ⎜ =
⎜ b g ⎟⎟ ⎜ g ⎟⎟
3
⎝ ⎠ ⎝ ⎠ g
12.1.2 Trial and error method

For a given trapezoidal channel obtain the critical depth by trial and error method.
Solution:
For trapezoidal channel
⎡( b + myc ) yc ⎤⎦
3/ 2
A D=⎣
( b + 2myc )1/ 2
⎛ ( b + my ) y ⎞ 3 Q 2
3
Squaring ⎜⎜ ⎟⎟ y = = constant
⎝ ( b + 2my ) ⎠ g
For a given b, m, Q, select a value of yc
Assume b = 6 m, m = 2m, Q = 12 m3 / s Solve for yc
( )
3
6 + 2yc y3c
=
144
= 14.679
6 + 4yc 9.81
( 3 + yc )3 y3c = 36
= 3.6697
3 + 2yc 9.81
Q
Assume a value of yc and compute A D and compare with the value obtained by .
g
yc 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

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 yc 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
Q
curve is generated assuming different values of yc and Z. The value of is computed
g
and yc is obtained from the chart. A one meter diameter culvert carries a discharge of
0.7 m3/s. Determine the critical depth.
d0 y
θ
⎡ ⎤
1 ⎢ θ − sin θ ⎥
D= ⎢ ⎥ do
g ⎢ sin θ ⎥
⎣ 2 ⎦
2 (θ − sin θ )
1.5
Z= 0.5
d 0.5
0
⎡ θ⎤
32 ⎢sin ⎥
⎣ 2⎦
Knowing the value of d0 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 m3/s. Determine the critical depth.
Q 0.7
Zc = = = 0.2235
g 3.132
Construct a graph of yc Vs Z and obtain the value of yc
From the graph yc = 0.4756
From the design chart determine the critical depth for a circular channel of 0.9 m
diameter. Discharge 0.71 m3/s.

Solution:
0.71
Z= = 0.22669
9.81
Z
= 0.29499 ( from table )
d 02.5
yc
= 0.56, yc = 0.49527 m
d0
1.0 (17 )
2
α Q2
ψ= = = 29.5
g 9.81
0.27
⎡ 29.5 ⎤ 6
yc = 0.81 ⎢ 0.75 1.25 ⎥ − = 0.86 m
⎣2 6 ⎦ 30 ( 2 )

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.
yc
Z=A D
Graph showing variation of section factor
with critical depth for a given pipe of
diameter do
yc
__ y
or __c
d0 b
A D A D
_____ or _____
2.5 2.5
d0 b

Reference:
Straub W.O, Civil Engineering, ASCE, 1978 Dec, pp 70 - 71 and Straub 1982.
Table: Semi empirical equations for the estimation of yc (Straub, 1982) MKS units
Channel type Equation for yc in terms of

ψ = α Q2 / g
1/ 3
⎛ψ ⎞
⎜ 2⎟
⎝b ⎠
b
Rectanglar
Range of applicability
1 0.27
Q
0.1 < 2.5 < 4.0
m ⎛ Ψ ⎞ b b
0.81⎜ 0.75 1.25 ⎟ − Q
For 2.5 < 0.1
⎝m b ⎠ 30m b
b use equation for rectangular channel
Trapezoidal
l 0.20
⎛ 2Ψ ⎞
m ⎜ 2⎟
⎝m ⎠
TRIANGULAR
y
( 0.84cΨ )0.25 y = cx 2
y = cx2
x
Parabolic

⎛ 1.01 ⎞ 0.25
⎜⎜ 0.26 ⎟⎟ Ψ
⎝ d0 ⎠
d0 Q0.52 Range of applicability
yc = 0.053 yc
d 0.3
0 0.02 ≤ ≤ 0.85
do
yc = [ m ]
Circular Q = m3s −1 , d 0 = [ m ]
y
Range of applicability
y
0.25 0.05 ≤ c ≤ 0.85
⎛ψ ⎞ 2b
b x 0.84b0.22 ⎜ 2 ⎟
⎝a ⎠ a = major axis
b = minor axis
a
Elliptical
y
1/ ( 2m +1)
y = cx ( )
1/ m −1
1
____ ⎛ m3ψ c 2m −2 ⎞
m-1 ⎜⎜ ⎟⎟
y = cx ⎝ 4 ⎠
x
Exponential
Example:
b = 6.0 m, m = 2, Q = 17m3 / s determine yc
Solution:
From table
0.27
⎛ ψ ⎞ b Q
yc = 0.81⎜ 0.75 1.25 ⎟ − for 0.1 < < 4.0
⎝m b ⎠ 30m b 2.5
α Q2
where ψ =
g
Q 17
The value of 2.5
== 0.19,
b 6 2.5
It is in the range of the equation. Substituting the appropriate values,
1(17 )
2
ψ= = 29.5
9.8
0.27
⎛ 29.5 ⎞ 6
yc = 0.81⎜ 0.75 1.25 ⎟ − = 0.86 m
⎝2 6 ⎠ 30 ( 2 )

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 m3/s.
1 1
m m yc
6m
Solution:
Trial and error procedure
A= ( b+my ) y = ( 6 + 2.5 yc ) yc
T = b+2my=6 + 5yc
A ( 6 + 2.5yc ) yc
D= =
T 6 + 5yc
Q
Zc = =A D
g
⎧ ( 6 + 2.5yc ) yc ⎫
0.5
V 2 Q2 / A 2 20* 20
= = = ( 6 + 2.5yc ) yc ⎨ ⎬
2g 2g [6 + 2.5 yc ]2 19.62 ⎩ 6 + 5yc ⎭
yc = ?
vc = gyc = ?
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.
Q 2T
1− =0
gA3
The formula must be satisfied by some positive depth yc greater than 0 (a lower bound)
and less than, an arbitrarily selected upper bound say, 10 m.

T is the free surface width b + 2myc . The interval is bisected and this value of yc 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.
T F(x)
1y 0
1
m m 100
b
Trapezoidal Bisection
Newton Raphson Method is discussed elsewhere.

12.2 Problems
There are three types of problems in critical flows as shown in table.
Type m Q yc b or d
I ?
II ?
III ?
Types I and II are easy to solve.
Type III problem requires a different approach.
Type I problem
myc
is known.
b
Qm3/ 2
∴ From the graph yc Vs Z =
b 2 gb
Q can be determined.
Type II problem
Here the solution is for obtaining critical depth. There are different methods. Graphically
Qm3/ 2 myc
Z= 2
can be computed and value of can be obtained from which
b gb b
yc can be computed.
Type III problem

This problem can be solved using simultaneous solution of two algebraic equations
which is illustrated below.
Defining
myc yc
Y1 = ( for trapezoidal channel ) or ( for circular channel ) .
b d0
Qm3/ 2 Q
and X1 = 2
or
b gb d 02 gd o
Then
myc
b =
Y1

Q m3/ 2 Q m3/ 2 Y15 / 2

X1 = =
m 2 yc2 g ( myc ) m 2 yc2 m1/ 2 y1/
c
2
g
Y12 Y11/ 2
Q Y15 / 2
= = M1Y15 / 2
m yc2 gyc
X1 = M1Y15 / 2
In which
Q
M1 = 2
my c gyc
Given Q, yc , m,
Qm3 / 2 myc
X1 = , Y1 =
b 2 gb b
y'c (1 + y'c )
3 3
2 3
Qm
=
gb5 1 + 2y'c
Q
M1 = and is known.
myc2 gyc
X1 = M1Y15 / 2
Substitutingin the above equation
5/ 2
Qm3 / 2 ⎛ myc ⎞
= M1 ⎜ ⎟
2
b gb ⎝ b ⎠
Qm3 / 2 ( myc2 ) gyc
= b 2.5 y15 / 2
Q g
Solve the equation and obtain the solution for bed width b for trapezoidal channel.
Similarly solve for diameter for the pipe line.

Problem
Compute the critical depth in a trapezoidal channel for flow of 30 m3 s-1. The channel
bottom width is 10.0 m, side slope m =2. The bottom slope is negligible and α = 1
T
1 1
m m y
.
Trapezoidal
Solution
Given
Bottom width b = 10m

Sideslope m =2
Flow Q = 30 m3s -1
α =1
Critical Depth y c = ?
For finding the critical depth,
Cross sectional area of the channel A= ( b + 2 y C ) * y C
= ( 10 + 2 y C ) * y C
Section factor Z=A D
in which D = A / T
for trapezoidal channel the top width T = ( b + 2m y C )
D = ( 10 + 2y C ) * yC /(10 + 2 * 2 * yC )
then the section factor Z=A ( 10 + 2 yC ) * yC / (10 + 2 * 2 * yC )
30
by u sin g the equation A D =Q / g, =
= 9.578
9.81
Substituting all the parameters A, P, T, D, and Q in the above equation and solving for y c one gets
3
⎡⎣(10 + 2 y c ) y c ⎤⎦ 2
1
= 9.578
(10 + 4y c ) 2
3 1
⎡⎣(10 + 2 y c ) y c ⎤⎦ 2 − 9.578 (10 + 4 y c ) 2 =0
⎡⎣(10 + 2 y c ) y c ⎤⎦ − 91.743 (10 + 4 y c ) = 0

3
by trial and error, y c = 0.91 m

Problems
1. A trapezoidal channel with side slopes of 2 horizontal to 1 vertical is to carry a flow of
16.7 m3/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 m3/s. Find the critical depth yc and critical velocity
Vc 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.
___
V1
Broad Crested Weir
H
y1
yc

P is the height of weir, B is the breadth of the weir. Assume the approach velocity V1 to
be very small.
⎛ 3 yc 2 ⎞
⎜ Answer: H = ⎛Q⎞ 1
, yc = ⎜ ⎟ , Q = 0.544L gH3/2 ⎟
⎜ 2 ⎝L⎠ g ⎟
⎝ ⎠

13.1 Measuring Flumes

Measuring flumes, on the contrary, introduce a width-wise 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 measurements-one upstream and one at the throat
(narrowest cross-section), 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 cut-throat flume (Fig 1.9 d)
are the results of such studies.

X X
Throat
FLUME LENGTH
PLAN
yc
SECTION along XX
Note: Sub-Critical to Super-Critical - Single Measurement
Sub-Critical to Sub-Critical - Double Gauging
Venturi Flume with an hump in the bed
R
GAUGE WELL
Z GAUGE WELL Z
D
P W
C
THROAT
PLAN
R
M B F G
E WATER SURFACE
LEVEL FLOOR
K
SECTION ON ZZ
PARSHALL FLUME
Standard dimensions are indicated
in the figure by letters

Y Y
PLAN
yc
Throat
SECTION ON YY
STANDING WAVE FLUME
Long throated flumes
Cut throat flumes
Parshall Flumes
H- Flumes
Throatless flumes with rounded transition

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 stage-discharge relations of several critical depth flumes in general
can be expressed as Q = C0 h n
where 'C0' is a coefficient depending on the breadth (b) of the throat, on the velocity
2
head 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 head-discharge relationship are the
Venturi flume, Long throated flume, Parshall flume, cut-throat flume, and H-flume.
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.

14.1 Weirs-Introduction
A control structure can be defined as a change in the cross-section of the flow whereby
the regime of the flow is modified. At such a section, a definite stage-discharge
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 tail-water conditions. In such a case, the flow can be measured by a
single depth measurement upstream of the critical section.


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

14.3 Proportional Weirs

'Proportional Weirs' are weir shapes designed to achieve a particular head-discharge or
head-velocity 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
monotonically-increasing 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 X AXIS
Y=f(x)
Y AXIS Y AXIS
O 2 O
__s
DATUM 3 s s
W
W
LINEAR PROPORTIONAL WEIR (SUTRO WEIR) LOGARITHMIC WEIR
X AXIS X AXIS
Y AXIS
Y AXIS d
O 1s
__
DATUM 3 s s
W W
QUADRATIC WEIR LINEAR PROPORTIONAL
(ORIFICE NOTCH) (ORIFICE NOTCH)
Some examples of Proportional Weirs

30 cm
15 cm
T = 50.88 mm
T = 25.4 mm
W' = 76.2 mm
W = 50.8 mm
0 X 150 mm 0 150 mm
Profile of a typical baseless weir (NBW-1) Profile of a typical baseless weir (NBW-2)
Y
Y
30 cm
15 cm
Proportional Proportional
portion Portion
0
X
d = 127 mm 0 X
d = 228.6 mm
Rectangular
a = 101.6 mm a = 76.2 mm
Rectangular orifice
orifice
2b = 1219.2 mm 2b = 1219.2 mm
p = 215.9 mm p = 228.6 mm
1
1 Bed of chamber
Trapezoidal chamber
Profile of a typical Sutro-parabolic weir Profile of a typical linear proportional orifice notch
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", Journal of the Franklin Institute,
Volume 285, Number 5, May 1968, Page 347 - 363.

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 semi-empirical in nature.
Discharge Coefficient:
For assessing the discharge passing over a weir, many semi-empirical 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.

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.
Approach flow
Plan
Skew Weir
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 non-modular. "The ratio of the
downstream and the upstream water depths above the weir crest is defined as the
submergence ratio, σ “. The limiting value of σ 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.

CLINGING NAPPE
Adhering to the crest and
downstream face of the weir
DEPRESSED NAPPE NAPPE WETTED

UNDERNEATH
The coefficient of discharge is generally
above the free discharge coefficient upto
H/L < 0.8
Ventilated
VENTILATED
(FREE) NAPPE
The ventilated nappe springs
clear of the crest
DIFFERENT NAPPE SHAPES
The weir geometry influences the coefficient of discharge. It depends upon the pressure
distribution along the geometric profile, boundary layer growth and separation zones.

14.5 Polygonal weirs-Introduction

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, duck-bill overfalls.
97.5 m
30.0 m
Stilling basin
Duck-bill overfall
corner angle: 90
148.0 m
14.4 m
Labyrinth weir
corner angle: 45.6
314.4

Labyrinth weir
corner angle: 117
242
20.4 m
11.7 m
Polygonal weir
corner angle: 133 , 152 , 255

5.0 m
to 8.0
5.0 m stilling 3.0
basin
Square intake Rectangular intake tower

corner angle: 90 corner angle: 90
14.9 m 15.0 m
3.5 m
3.0 m
Polygonal intake tower Rectangular spillway

corner angle: 84.3 corner angle: 90
Layouts of Overfall structures with
Polygonal Center Line of Weir Crest
These weirs consist mainly of straight parts with corners in-between. 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 [≤ 30.0 m ] and small floods ⎡⎣ ≤ 100 m3 /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.

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, α , is formed by both the straight sides of the weir and is
measured in the downstream.
side wall
hc
α
Wc
l /2
d
lc /2
"corner"-weir
Isometric view of corner-weir looking downstream

side wall
weir crest
lc/2 lc/2
α ld/2
ld/2
Disturbed area
(overlapping flow region)
Plan of Corner-weir
T
b
Rectanglar
h
R X
2W
POEBING WEIR
Angle ' α ' varies within the limits, Convex angle 0D < α ≤ 180D
Concave angle 180D < α ≤ 360D

In the physical system ("corner weir ") in the range of 180D to 360D does not perform
satisfactorily. Therefore α1 + α 2 = 360D in physical system.
α /2
1
α2
α /2
1
Concave angle Convex angle
o o o o
0 < α1 < 180 180 < α2 < 360
Layout of Corner Weir
The flow over the corner weir can be apportioned as (1) disturbed area near the corner
are shown in green color and
(2) With two-dimensional flow.
The length of the area of disturbed flow = ld / 2 + ld / 2

The length of the corner weir is lc . Hence, 0 ≤ ld ≤ lc
The local disturbance factor 'DF' with a distance, l, from the corner is defined as
C (l )
DF ( l ) =
Cn
in which C ( l ) and Cn are the coefficient of discharges for the corner weir of length 'l'
and for the normal flow condition.
3Qc
The discharge over the weir is written as Cm = * 3/ 2
2Cnlc 2 g hc,n
in which Q is the discharge, in m3s-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.

⎛l ⎞
DF ⎜ d ⎟ = 1
⎝2⎠
For values l ≥ ld / 2 , DF = l . Accordingly, the mean distribution coefficient, Cm , of the
overlapping flow zone can be written as
l /2
2 d
cm = ∫ DF ( l ) dl
ld 0
The discharge, Qc , of the “corner " weir is
2
Qc = Cn 2 g hc,n
3/ 2
⎡⎣lc − (1 − Cm ) ld ⎤⎦
3
The overall head, hc,n , belonging to the discharge coefficient, Cn , under normal flow
conditions (two-dimensional flow) can be estimated from
ν c2 −ν c,n
2
hc,n = hc +
2g
in which hc = overfall head, assuming three dimensional flow, in meter, at the " corner "
weir ; ν c = flow velocity, assuming three-dimensional flow, in meter per second, at the "
corner " weir ; and ν c,n = flow velocity under normal flow conditions (two-dimensional
flow) at the " corner " weir. For the hydraulic calculation the length, ld , of the disturbed
area and the value of Cm must be known.
14.5.2 LENGTH ld OF OVERLAPPING ZONE
With increasing overfall heads hc,n the length of the overlapping zone, ld , grows
symmetrically to the corner, as far as ld = lc .
3Qc
In this case the mean disturbance coefficient is Cm = * 3/ 2
2Cnlc 2 g hc,n
If the corresponding limiting value for hc,n = hc,n
*
.
Using the length, ld , of the overlapping zone of flow, depending on the strength of
disturbance, from Eq. 7 one may obtain
⎛ 3Qc ⎞ 1
ld = ⎜ lc − ⎟
⎜ 2C 2 g h 3/ 2 ⎟1− C
⎝ n c,n ⎠ m

3Qc
l d = lc − 3/ 2
2Cn 2 g hc,n
1
Thus ld = l d
1 − Cm
in which l d represents the length of disturbance. The independent variables
lc ,Qc ,Cn, and hc,n , are determined from experiments.
Indlekofer and Rouve have conducted investigations for sharp-crested “corner “weirs
with corner angles α = 46.81D , 62.08D , 89.64D , and 123.45D . The crest thickness was 2mm.
The discharge was determined by the Rehbock formula
2⎛ h + 0.0011 ⎞
⎟ b 2 g ( h + 0.0011)
Q = ⎜ 0.6035 + 0.0813 3/ 2
3⎝ P ⎠
in which h is the overfall head, in meters; P is the weir height, in meters; and b is the
width, in meters and Cn is the coefficient of discharge.
Length of Overlapping Zone Area l d
The length of overlapping zone area, l d , can be calculated using Eq. 12.
The length of disturbance, l d , is related to the overall head, hc,n , by a simple linear
function, l d = A + Bhc,n
in which A is a constant, in meter; and B slope for l d . It must be mentioned that the
constant, A is very small, and either positive or negative.
14.5.3 Length ld of Overlapping Zone
Based on the laws of similitude, one can assume a linear relation between the length,
ld , of the zone of disturbance and the overfall head. The length, ld , will be
B
ld = hc,n
1 − Cm
or using the slope B, for the length, ld , of the overlapping zone
ld = Bhc,n
lc
Because of the linearity B = *
hc,n

Mean Disturbance Coefficient Cm in overlapping zone - The mean disturbance
coefficient, Cm , which considers the influence of the disturbance with a length, ld ,
compared with the flow normal to a straight weir, can be calculated by,
B
Cm = 1 −
B
using the slope B and B .
Figures show the typical Polygonal Plan with angles 0D < α < 180D .
lc (2,3)
2
α n-1 n-1
lc (1,2) αn lc (n-1,n)
1 n
lc (n,1)
Plan of the typical polygonal corner weir , Corner Angles 0 < α < 180 ;

Overlapping zones
Undisturbed zones
Length of Overlapping zone for Constant Overfall Head
3/2 ⎡ ⎤
n
2
Qc = Cn 2 g h c, n ⎢lc( n ) − ∑ l d ,i ⎥
3 ⎢⎣ i =1 ⎥⎦
Example:
Discharge of Sharp-Crested Shaft Spillways with Equilateral Polygonal Plan
In case of shaft spillways, with equilateral polygonal in plan above Equation can be
simplified as
2 3/ 2 ⎛ ld ⎞
Qc = Cn 2 g n lc hc,n ⎜1 − ⎟
3 ⎝ lc ⎠
in which lc 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," Journal of
the Hydraulics Division, ASCE, Volume 101, Number HY#, Proceeding paper 11178,
March 1975, pp. 385 - 401.

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. Flat-vee 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
V-notch weir sill
Triangular profile two dimensional weir
Triangular broad-crested weir

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
0.05 ≤ H1 L ≤ 0.08 . The upper limit as H1 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 H1 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 E1 = E2
2 2
V1 V2
h1 + α1 =E 2 = y 2 + α 2
2g 2g
In other words
V= {2g ( E1 -y 2 )}
0.5
α −0.50
In which E1 equals the upstream specific energy over the weir crest.
α1 V12
______
α 2 V22
2g ______
2g
E1 h1
E2 yc h2
y2
y1
P weir
Flow over a broad crested weir under submerged condition
substituting Q = VA and putting α 2 = 1.0 gives

0.5
Q = A { 2g ( E1-y2 )}

If the critical flow occurs at the control section (y = yc), a head -discharge relationship for
various throat geometries can may be derived from
0.50
{
Q =Ac 2 g ( E1 − yc ) }

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

14.9 Bear-trap weir

Hydro-electric 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 bear-trap 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 bear-trap weir for flood disposal
Raised position of bear-trap weir for flood disposal

15.1 Flow below a Sluice Gate

___2
V1
___
2g
Sluice
Gate
Hydraulic
H1 y 1
jump
1
2 3 y
W y = δW 2
l
Rapidly varied flow with Hydraulic jump (1 and 3 subcritical flows,
2 Super critical flow)
Flow below a Sluice gate
v12
H1 = y1 +
2g
opening w
y y
depth at vena contracta y = δw or δ= ,n= 1
w w
Q = Cd C v bw 2g ( y1 -y )
= Cd C v bw1.5 2g ( n-δ )
= K bw1.5 2g
y1
The value of n = , range is 1.50 to 5.00.
w
δ ranges between 0.648 to 0.624.
Cd ranges between 0.607 to 0.596.
As 'n' increases from 1.5, Cd decreases upto 2.40 with a value of 0.600 to 0.596. Then
further increase in n (> 2.40) the Cd value increases from 0.596 to 0.624. For the same
range, 'K' increases from 0.614 to 1.279.
For
n = 2 δ = 0.630
n = 3 δ = 0.625

n = 10 δ = 0.620
y1
Vena contra is located at a distance of = 1.
nδ
The sequent depth of jump should not exceed the value given by
y2 δ ⎡ ⎛H ⎞ ⎤
= ⎢ 1 + 16 ⎜ 1 − 1⎟ − 1⎥
w 2 ⎢⎣ ⎝ δw ⎠ ⎥⎦
Figure shows the limiting tail water level for modular flow below a sluice gate.
5
δ = 0.611
4
Submerged
Flow
3
Modular Flow
2
assuming y1 __ H1
1
0 1 1.5 2 3 4 5 6 7 8 9 10
Ratio n = y1/w
Limiting tail-water level for modular flow below a sluice gate
Henderson proposed an equation for the contraction coefficient δ for the radial (Tainter)
gate which depends on inclination angle θ .

2
⎛ θ ⎞ ⎛ θ ⎞
δ =1- 0.75 ⎜ D ⎟ + 0.36 ⎜ D ⎟
⎝ 90 ⎠ ⎝ 90 ⎠
The expected error is less than 5% provided that θ < 90o . Thus the discharge coefficient
δ
for radial gate is given by Cd = 0. 5
.
⎛ δw ⎞
⎜1+ ⎟
⎝ y1 ⎠

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.
x Brink depth
yc or
End depth (yb)
L
_y_c 1.4 , x = 3 to 4 yc
yb
Brink depth
The drop distance should be more than 0.6yc. 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.
q2
Ho = y + α
2gy2
Differentiating w.r.t 'y' assuming Q to be constant.
dH o q2
= 1− α
dy gy3
y1 αq 2
n= if the flow is critical, hence yc = 3
w g
If α =1, then Q = b g yb
Rouse showed yb = 0.715y
⎡ y ⎤
Thus Q=b g ⎢ b ⎥
⎢⎣ 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 yc. This is applicable to canals with
slopes upto 0.0025.

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.

STAGE (m)
H
DISCHARGE m3/s
Typical Stage Discharge Relationship
Measuring stage:
Water surface EL 210.00 m
Bed EL 205.00 m
H (M.S.L)
(Above Mean Sea Level)
EL 200.00 m Datum
Definition of stage
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.

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.

(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 zero-displacement 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

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 C1 is injected at
the constant injection rate QT
X2
X1
INJECTION
QT
SKETCH OF THE REACH C2

X1 X2
C1 Concentration
C2
TIME
CONCENTRATION-TIME CURVE MEASURED AT X2
The tracer disperses laterally into the flow and tracer concentration distribution is similar
to as shown in figure. At some point X2 downstream, where the tracer material is
approximately uniformly mixed, the flow is sampled continuously.
By continuity QT C1 = ( Q+QT ) C2
in which
Q is the stream discharge, C2 is the concentration at X2
If QT << Q and if there is no tracer loss

QT C1 = QC2
then
QT C1
Q=
C2
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 t0. The cloud of tracer disperses laterally and longitudinally as it
moves downstream.
Q W
X2, t2
X0, t0 X1, t1
DYE CLOUD DISPERSING DOWN STREAM
Concentration at x1
Concentration at x2
TIME
CONCENTRATION Distribution at X1 and X2
At the section X2, where the tracer is completely mixed literally, the flow is sampled
∞
continuously. From the conservation of mass W= ∫ Q C dt
0
W
in which Q is nearly constant through sampling period Q = ∞
∫ C dt
0
The common tracers used are
(a) Salt solutions
(b) Radio active tracer are detected by its scintillation detectors

(c) Fluorescent dyes with flourometers.
Advantage of dilution method, they condensed in closed conduits, such as penstocks,
sewers pipe lines, where current-meter 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.
40
WATER SURFACE
30
20
10
0
SEDIMENT CONCENTRATION
15.3.3 MODERN MEASURING TECHNIQUES

The two principle techniques that are used in the modern measuring instruments are
'Electro-acoustics' (ultrasonic frequencies) and
'Electro-optics'.
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

C =141000 + 421 T - 3.7 T 2 + 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.
A
θ
FLOW
v
rp

Notation:
L Path length θ path angle

(usually 30° to
60°)
Average velocity of flow of river rp path velocity at
V
at depth y. depth y
V Average velocity of flow of river C Velocity of sound
in water
d actual depth of flow a area of flow
t1 time taken for a pulse to travel t2 time taken for a
from A to B. pulse to travel
from B to A.
F1 output frequency corresponding F2 output frequency
to 1 / t1 corresponding to
1 / t2
Fc output frequency corresponding M Multiplication fig.
to 1 / t1 minus 1 / t 2 of variable
frequency
oscillator
T measuring period N no. of
coincidences in
dif.fre.store = FcT
L
The time taken for a pulse to travel from A to B t1 =
C + VP
Similarly time taken for a wave front to travel in the opposite direction is
L
t2 =
C − VP
1 1 2Vp
− =
t1 t 2 L
L ⎛1 1⎞
Vp = ⎜ − ⎟
2 ⎝ t1 t 2 ⎠
Average velocity of river flow at depth y is given by
Vp
V=
cos θ
L ⎡1 1 ⎤
V= ⎢ − ⎥
2cos θ ⎣ t1 t 2 ⎦
Transducers may be so positioned in the vertical plane so as to make average velocity

L ⎡1 1 ⎤
V at depth of equal to the average velocity of flow V then V = ⎢ − ⎥
2cos θ ⎣ t1 t 2 ⎦

If 'Q' is discharge
Q = aV
= V- dsinθ
L2 ⎡ 1 1 ⎤
Q= ⎢ − ⎥ d tanθ
2 ⎣ t1 t 2 ⎦
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 co-efficient of discharge as in first method.

Detector
Projector
Flow V L
θ
Receiver
VP
Projector
Pulser
Receiver Receiver
Clock
A B
Computer Recorder
Schematic illustration for the Arrangement of

Sound Projector and Receiver
2. ELECTRO-ACOUSTIC 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.

CLOCK
PULSER RECORDER
RECEIVER COMPUTER
TRANSDUCER
d
ACCOUSTIC AXIS
TARGET AREA
Schematic arrangement for a Depth Sounder

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

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 ELECTRO-OPTICAL 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 wide-frequency band, visible light source which are the second
utilizes a narrow band coherent laser beam.
ELECTRO-OPTICAL 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.

To
From Photomultiplier
Light Tube
Source
Lens 1 1 cm. Gap
Focal
Volume
Lens 2
Mirror Mirror
Dark Opaque
Zone Coating
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 ELECTRO-OPTICAL 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. Electro-optical velocity instrument is also called, Laser Velocimeter, Laser
Doppler Velocimeter and laser anemometer.

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 1953-54 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.

OUTPUT E α v
FIELD N
Supply Velcoity of
flow = V
Basic principle of fluid flow

measured in pipes by electromagnetic
induction
Measuring System
Velocity
of Water
Electrical
Potential
Generated
Induced Magnetic Field
Probes
Principle of electromagenetic
river gauging

Noise cancellation probes

Bed Conductivity probe
Signal cable duct
Signal
probes
Bed
Conductivity
probe
Noise cancellation Cell for producing Shelter for

probes Instrumentation
magnetic field
Diagrammatic view of an electromagnetic river gauging station
Buried
Coil
Coil Signal Probes Signal

Drive Recovery
Timing
Signals Timing
Signals
Flow of water
Stage
Telemetry
system Data Processor
Water
Conductivity
Bed
Conductivity
Data Recording
Devices
Typical block diagram of an electromagnetic river gauging station

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=Hvb
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.

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.

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.
Liquid flow measurement in open channels -
ISO: 555/ II 1974
dilution methods, Part II Sudden injection method.
Liquid flow measurement in open channels by
ISO: 748 1973
velocity-area methods.
ISO: 772 1973 Vocabulary and symbols.
ISO: 1100 1973 establishment and operation of a gauging station
and determination of the stage-discharge relation.
Liquid flow measurement in open channels - slope
ISO: 1070 1973
- area method.
Collection of data for determination of errors in
ISO: 1088 1973
measurement by velocity area methods.
ISO: 2425 1974 Measurement of flow in tidal channels
Liquid flow measurement in open channels - cup -
ISO: 2537 1974
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).
Liquid flow measurements in open channels -
ISO: 3454 1975
sounding and suspension equipment.
ISO: 3455 1976
calibration of current meters in straight open tanks.
Methods of measurement of suspended sediment
ISO: 4363 1977
in open channels.
ISO: 4364 1977 Bed material sampling.
Functional requirements and characteristics of
ISO: 3716 1977
suspended sediment load samplers.
ISO: 3846 1977
weirs and flumes - rectangular broad crested weirs.
Liquid flow measurements in open channels by
ISO: 3847 1977
weirs and flumes - end depth method.
ISO: 4359 1978
flumes.
ISO: 4360 1978
weirs and flumes - triangular profile weirs.
ISO: 4361 1978 weirs and flumes - round nosed broad crested
weirs.
ISO: 4373 1978 Water level measuring devices.
ISO: 4369 1978 The moving boat method
Calculation of the uncertainty of a measurement of
ISO: 5168 1978
flow rate.
ISO: 4377 1978 Flat V- weirs.
ISO: 4375 1978 Cableway system.
ISO: 6418 1978 Ultrasonic (acoustic) velocity meters.

ISO: 4366 1978 Echo Sounders.

Reports
Investigation of the total error in measurement of
ISO: Data 1978
velocity-area methods.
Note: In 1973 all existing recommendations become standards.

16.1 Concept of Uniform Flow

1. The depth average flow velocity (integrated over depth), area of flow cross-sections
are every where constant along the channel.
2. The energy grade line Sf, water surface slope Sw and channel bed slope S0 are all
parallel, i.e. Sf = Sw = So
Figure shows Boundary layer growth in open channel with an ideal entry condition.
yc
Development of uniform flow in a long channel
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.

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 (yn>yc).
Steady Turbulent
Uniform flow
Unsteady 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.

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 V = CR x Sy in which
x and y are components, and vary depending on uniform formula.

yb
1
V = ∫ ∫ v dx dy
A00
Momentum Equation:
(1) (2)
_
v1
y1 W sin θ
A
_ y2
v2
τ0
P
l Datum

γ
Q
g
( )
β 2 V2 − β1 V1 = P1 -P2 + W sinθ - Pf
If V2 = V1 , β1 = β 2 , P1 =P2 then
W sinθ = Pf (1)
Pf = shear force acting on boundary = Shear stress * Area
= τo * Area
= τo PL
P is the wetted perimeter, Sin θ = So
Weight W = ρ g AL
W Sinθ = ρ g AL Sinθ
Substituting in equation (1)
ρ g AL So = τo PL
ρ g AL So
τo = = γRSo (2)
PL
τo
Note υ* = = gRSo Critical shear velocity
ρ
2
V
But τo = cf ρ (3)
2
2
V
γRSo = cf ρ
2
1/ 2
⎡ 2γ ⎤
or V = ⎢ RSo ⎥
⎣ cf ρ ⎦
2g
V= RSo
cf
2g
If = C then
cf
V = C RSo .
This is known as Chezy equation. The coefficient C is either estimated or
determined experimentally. C has dimension of ⎡ L1 / 2T −1 / 2 ⎤
⎣ ⎦

2. Consider Darcy Weisbach equation for loss in pipe due to friction

L V2
hf = f
d o 2g
2 1 hf ⎡ hf ⎤
∴ V = 2g d o , ⎢ L = So = Sf ⎥
f L ⎣ ⎦
⎡ πd 02 ⎤
2 1 ⎢R d ⎥
V = 4R*2g *Sf , ⎢ = 4 = 0⎥
f ⎢P πD 4⎥
⎢⎣ ⎥⎦
8gRSf
V=
f
Comparing with Chezy equation:
8g
C=
f
C 1
=
8g f
Manning formula is an emprical relation based on field observations and is given by
1 2/3 1/2
V= R So
n
⎡ -1 ⎤
in which V in m/s, R in m. Thus 'n' has dimensions of ⎢ L 3 T ⎥
⎢ ⎥
⎣ ⎦
[ If R=15 cm, n = 0.015, So = 0.0004 , then V = 0.376 m/s ]
The hydraulic engineers 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.

16.3.1 Resistance in Open Channel Hydraulics

If Manning and Chezy equations are compared
2 1 1 1
1 3 2
R S0 = CR 2 S02
n
2 1 1
-
R3 2 R6
C= =
n n
1
R6
C=
n
For laminar flow:
K
f=
R e1
VR
R e1 =
υ
VR
K= f
υ
8gSR 8gVR 2S 8gR 2S 8g
But f = 2
∴ K= = = 2 R e1
V υV 2 υV C
⎡ 8g ⎤
C2 = ⎢ ⎥ R e1
⎣K⎦
VR 8g
If R e1 = f=
υ C2
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)

16.3.2 Laminar Flow with Roughness

60
f= ← for a 90D V shape channel. Roughness 0.3023 mm
R e1
33
f=
R e1
1.0
0.8 Reference:
0.6 "Chow Ven Te- Open Channel
Hydraulics", Mc Graw Hill Company,
0.4 International student edition, 1959, page - 10
24
0.2 f = ___
Re1
14
f = ___
Re1
0.1
0.08
f 0.06
0.04
0.02
0.01
0.008
0.006
Laminar Transitional Turbulent
0.004
10 2 4 6 8 102 2 4 6 8 103 2 4 6 8 104 2 4 6 8 105 2 4 6 8 106
Re1
Variation of friction coefficient f with Reynolds number Re1(= vR
__ )
υ
in smooth channels

Reference:
"Chow Ven Te- Open Channel Hydraulics",
Mc Graw Hill Company, International
student edition, 1959, page - 11
1.0 0.1 37.5 cm

0.8 0.08 Varwick 25 cm
0.6 0.06
0.4 0.04
60
f = ___
Re1
14
f = ___
0.2 0.02
Re1
103 2 4 6 8 104 2 4 6 8 105 2 4 6 8 106 2 4 6 8 107
33
f = ___
Varwick
f
0.1 1
0.08
Re1 1
20 cm
0.06
0.04
0.02
Laminar Transitional Turbulent
0.01
0.008
0.006
10 2 4 6 8 102 2 4 6 8 103 2 4 6 8 104 2 4 6 8 105 2 4 6 8 106 2 4 6 8 107
Re1
Variation of friction coefficient f with Reynolds number Re1 (= vR
__ )
υ
in rough channels
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

(i) Hydro dynamically smooth turbulent flow
(ii) Hydro dynamically Rough turbulent flow
(iii) Hydro dynamically transition turbulent flow.

The boundary layer δ for flow past a flat plate is given by
−1 / 2
δ ⎛V x⎞
= 5⎜ o ⎟ Laminar
x ⎝ υ ⎠
−1 / 5
δ ⎛V x⎞
= 0.38 ⎜ o ⎟ turbulent R e > 2 * 107 logarthmic velocity law holds
x ⎝ υ ⎠
99% V
Turbulent
Pseudo boundary
δ
δ∗ Transitional y
region
δ0 Viscous sub layer
Velocity
Velocity distribution

Smooth kc =
__
5υ
v*
kc is critical roughness height
k is roughness height
kc = 100 __ for average condition
υ
v*
δ0 k δ0 k δ0
k
kc (a) Smooth kc kc
(b) wavy (c) rough
Different surface roughness

Viscous sublayer
ks
(i) Hydrodynamically smooth

turbulent flow f = f(Re)
ks
Viscous sublayer
(ii) Hydrodynamically transition
flow f = f (Re, ks/y)
ks
Viscous sublayer
(iii) Hydrodynamically rough
turbulent flow f = f (ks/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.
v*K s τo
If we define R e* = shear Reynolds number . ; and v* = gRSf = .
υ ρ

The flow is classified as follows:
v*K s
<4 Hydrodynamically smooth
υ
vK
4 < * s < 100 Hydrodynamically transition
υ
v*K s
> 100 Hydrodynamically fully developed turbulent flow
υ
Summary of Velocity-Profile Equations for Boundary layers with dp =0

dx
Zone Smooth Walls Rough Walls

Law of the Universal equations
wall
Laminar v* y v v* y
<4 =
sub layer υ v* υ -
(y≤δ )
Buffer v* y
4< < 30 to 70
zone υ - -
Logarithmi v* y v v y v k
> 30 to 70 = A log * + B = A log + B
c zone υ v* υ v* y
(also y
called turb < 0.15 v v y v k
δ = 5.6 log * + 4.9 = −5.6 log + B
ulent layer) v* υ v* y
B=f
(roughness size, shape and
distribution)
Velocity-defect law
Inner y V−v y
< 0.15 = − A log + B
region δ v* δ
(overlaps
with V−v y
= −5.6 log + 2.5
logarithmi v* δ
c wall law)
V−v y
Outer y = −A log
< 0.15 v* δ
region δ
(approxim V−v y
ate = −8.6 log
v* δ
formula)
(3000 Power Law
< Re < v ⎛v y⎞
1
7
70,000) - = 8.74 ⎜ * ⎟ -
v* ⎝ υ ⎠
outer
region
A and B are constants.

Table shows velocity distributions for different conditions
VR VR
Pipe flow equation R e = Open channel flow R e = 4
υ υ
Blasius 0.3164 C=18.755 R1/8 ⎡ mks units for g = 9.806 m/sec 2 ⎤
equation f= upto R e <105 e ⎣ ⎦
R 0.25
e
for 0.223
1 R f f = 0.25
smooth Re
=2log e
flow f 2.51
R e 8g
R e > 105 C = 4 2g log
2.51 C
R 8g
C = 17.72 log e
2.51 C
⎡ 3.5294R e ⎤
C = 17.72 log ⎢ ⎥
⎣ C ⎦
Smooth
pipe flow
1
f
( )
= 0.86 ln Re f - 0.8 C R 8g
= 2 log e
C 2.51
Nikurads 8g
1 ∈
e Rough = 1.14 - 0.86 ln C ⎡12R ⎤
*
f do = 2 log ⎢
pipe ⎥
Nikurads 8g ⎣ ks ⎦
e
White 1 ⎡∈ /d o 2.51 ⎤ C ⎡ ks 2.52 8g ⎤
and = 0.86 ln ⎢ + ⎥ = -2 log ⎢ + ⎥
f ⎣ 3.7 Re f ⎦ 8g ⎢⎣14.83R R e f ⎥⎦
Colebroo
k formula
Suggested modification to equation is
C ⎡ k 2.5 ⎤
= -2 log ⎢ s + ⎥
8g ⎣12R R e f ⎦
[ASCE Task Force Committee 1963]. R is hydraulic mean radius, 4R = Diameter of
pipe.
In open channel flow following aspects come into picture
f = f ( R e , K, C,N, F,U )
(1) (2) (3)
In which Re is the Reynolds number, K is the Relative Roughness, C Shape factor of the
cross-section, 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.

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 cross-section.
C 1 R
= =A log +B
2g f y'
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 ks. 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 s
Isolated - roughness flow (k/s) - Form drag dominates

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

s s s
Wake interference flow (y/s)
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.
k j j j j
s s s
k is surface roughness height
s is the spacing of the elements
j is the groove width
y is the depth of flow
Quasi smooth flow - k/s or j/s becomes significant acts as Pseudo wall
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.
Concept of three basic types of rough surface flow
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).

16.3.3 Areal concentration or Density Distribution Roughness
Elements
Spheres 1 2
Schlichting, 1936 Spatial distribution of roughness
O'Loughlin and Mcdonald (1964) Cubes arranged
as in (1) abd (2) also sand grains (2.5 mm dia)cemented
to the bed .
Koloseus (1958) and Koloseus and Davidian (1965)
conducted experiments using Cubical Roughness
Symmetrical diamond shaped pattern.

4
Spheres
3
Cubes Sand
ks
___ 2
y
1
Nikuradse
0
0 0.2 0.4 0.6 0.8 1.0
λ − Areal concentration
Effective roughness as a function of form pattern, and
concentration of roughness elements. (Assuming high
Reynolds number)
Schlichting (1936) - Sphere spacing
Koloseus (1958)
Koloseus and Davidian (1965)
Cubical Roughness
Symmetrical diamond shaped pattern
O'Loughlin and Mcdonald (1964)

Cubes arranged as in 1 and in 2.
Also sand grains centered to the sand grains
(2.5 m diameter)
10
ks
___ 1
y Sand
0.1
0.001 0.01 0.1 1.0
λ − Areal concentration
Logarithmic plot of data from figure at low
concentration
Open channel resistance (after H. Rouse, 1965)

1.5
F = 1.5
1.0
1.0 0.5
V
d = 3b
3b
0.5
0
0 0.5 1.0 1.5 2.0 2.5
Froude number, F
Resistance of a bridge pier in a wide channel, after Kobus and Newsham
1.5 S
__ =5
D 7.5
30
1.0
CD
0.5
D
V d D S
d = 30
0
0 0.5 1.0 1.5 2.0
Froude number, F
Variation of pier resistance with lateral spacing "S"

0.6
y/b = 1/16
0.4
y/b = 1/8
ζ 0.2
y
2b
4b
y/b = 1/4 b
0
0.1 90
0.2 0.4 0.6 0.8 1.0 2.0 4.0

Froude number, F
Loss at one of a series of channel bends after Hayet
Some of the important References:
(i) Task force on friction factors in open channels Proc. ASCE JI. of Hyd. Dn. Vol. 89.,
No. Hy2, March 1963, pp 97 - 143.
(ii) Rouse Hunter, "Critical analysis of open channel resistance" , Proceedings of ASCE
Journal of Hydraulic division, Vol.91, Hyd 4, pp 1 - 25, July 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
f DhS
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 cross-sectional shapes. The

lograthmic velocity distribution can be integrated only for the wide rectangular and
circular sections.
Effect of boundary non-uniformity 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. Bazin’s Formula 1897 (MKS)
3. Powell Formula (1950) FPS while using Powell formula C must be multiplied by
0.5521 to get C in m1/2 s-1
4. Pavlovskii Formula (1925)
Manning equation is applicable to fully developed turbulent rough flow.
Slope of the straight line is 1:3

1/ 3 1/ 6
⎛ ks ⎞ g ⎛ g ⎞
f ⎜R⎟ C ⎜ ⎟
⎝ ⎠ C2 ⎝ ks ⎠
n k s1/ 6
1/ 3
⎛k ⎞
f = 0.113 ⎜ s ⎟
⎝R⎠
If we replace k s by diameter of the grain size (d)
1/ 3
⎛d⎞
f = 0.113 ⎜ ⎟
⎝R⎠
1/ 6
8g 8g ⎛ R ⎞
C= = ⎜ ⎟
f 0.113 ⎝ d ⎠
for MKS units g = 9.806 m/s 2
1/ 6 1/ 6
8 * 9.806 ⎛ R ⎞ ⎛R⎞
C= ⎜ ⎟ = 26.3482 ⎜ ⎟
0.113 ⎝ d ⎠ ⎝d⎠
1/ 6
⎛R⎞
or C = 26.34 ⎜ ⎟
⎝d⎠
1/ 6
⎛R⎞
n= ⎜ ⎟
⎝C⎠
1
n= *d1/ 6 = 0.0379 d1/ 6
26.34
A number of empirical methods to relate n diameter of the particle are advanced.
1 Strickler n = 0.02789 d1/6 [ d in m ] This is not applicable to mobile

(1923) bed
50 [ d in feet ]
n = 0.034 d1/6
2 Henderson's
interpretation
of Strickler's
formula
3a Raudkivi n = 0.047 d1/6 [ d in m ]
(1976)
65 [ d in mm ]
n = 0.013 d1/6
3b Raudkivi d65 = 65 % of the material by
(1976) weight smaller.
65 [ d in feet ]
n = 0.034 d1/6
3c Raudkivi
(1976)
50 [ d in feet ] 0.039 * ( 0.3048 )
1/ 6
4 Garde and n = 0.039 d1/6
Ranga Raju
= 0.039 ( 0.82036 ) = 0.03199
( ) 1/6
n = 0.03199 d50 , d is in 'm'
50 [ d in m ]
n = 0.0475 d1/6
5 Subramanya
90 [ d in m ]
n = 0.038 d1/6
6 Meyer and (Significant proportion of coarse
Peter and grained material)
Muller

7 Simons and n = 0.047 d1/6 [ d in mm ]

Sentrvrk
(1976)
8 Lane and n= 0.026 d1/6
75
(d in inches and d75 = 75% of the
Carbon material by weight is smaller)
(1953)
8) Consider
v* = g R Sf
υ* k s
4< < 100 Transition flow
v
R1/6 R1/6
n= but C = 26.35
C d
R1/6d1/6 1
∴ n= = d1/6 = 0.03795 d1/6 (d in m)
R1/6 ( 26.35 ) 26.35
Conditon for fully developed rough flow
v*k s n6
= 100 d= = 3.3458 * 108 n 6
υ ( 0.03795) 6
6
⎡ n ⎤ 1
g R Sf ⎢
⎣ 0.03795 ⎥⎦ υ
Assuming
υ = 1.01 * 10-6 m 2 /s g = 9.806 m/s 2
9.806 1
n 6 R Sf −6
= 100
1.01* 10 ( 0.03795)6
n 6 RSf ≥ 9.635 * 10−14

Blasius equation (Re <105)

1
__ Re f
= 2.0 log ( _____ )
f 2.51
C = 4 2g log ( Re 8g
_____ )
2.51C
180
Smooth surfaces
150 0.316
Laminar flow f = _____
140 Re
0.25
do
_____ 2R
= _____
130 1
__ 2ks ks
8
(C = 15.746 Re , mks)
120 507
110 252
100 126
90 60
80
30.6
70
15
60 10
v* ks
______
υ = 100
50 1 __ C
Transition zone Fully rough zone __ = = 2.0 log ( 12R
___ ) or Manning
f 8g ks
Commercial surfaces
40
Sand coated surface (Nikuradse)
30
103 2 4 6 104 105 106 107 108
Reynolds number Re = 4 V R/v
Modified Moody Diagram showing the Behavior of the Chezy C after Henderson

16.4 History of Uniform Flow Velocity and Resistance Factor

The design of the cross-section 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 cross-sectional area on the other.
LEONARDO DA VINCI (1452-1519):
“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.

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 cross-section 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 cross-section of a canal to supply water to
the city of Paris from the river Yvette.

2
Chezy put forth that V / RS 0 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
2
channel. Chezy, however, did not assume that the value V / RS 0 was a constant for all
streams, as he found this value to vary from one stream to another.
The present–day- Chezy-formula is written as V = C RS where C is known as ‘Chezy

0
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
[ L] 2 [T ]−1 , where [ L ] and [T ] are units of length and time of any measuring system.
PIERRE LOUIS GEORGES DU BUAT (1734-1809): He proposed a formula for average
velocity.
48.85 R -0.8
V= - 0.05 R [in metric units].
1/S0 - ln (1/S0 ) + 1.6
The surface roughness of the boundaries was ignored in the formulation of the above
equation.

JOHANN ALBERT EYTELWEIN (1764-1848), published, at Berlin in 1801, a formula for
open channel flow, namely V = 50.9 RS0 (in metric units)
A firm proponent of non dimenisonal quantities in the analysis of any problem, JULIUS
WEISBACH (1806-1871), was the first to write a formula for resistance to flow through
2
L V
closed pipes as h L = f
d 0 2g
in which f is a non dimensional friction coefficient, which is commonly known at present
as the Darcy-Weisbach friction factor, h L the head lost due to the frictional resistance, L
the length of pipe in which the head loss h L has occurred and d0 the diameter of the
pipe. Weisbach reported that f is a function of the Reynolds number R e and the relative
roughness, for a given shape of cross-section.
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 triple-factor formula V = CR x S0y
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 D’ARCY (1803-1858)
in 1855 in France, and were continued after his death by his worthy assistant HENRI
EMILE BAZIN (1829-1917).
D’ARCY 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 cross-section could be changed. Rectangular, trapezoidal,
triangular and semi-circular cross-sections were tested. The different surfaces tested

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 (1818-1894) and WILHELM
RUDOLPH KUTTER (1818-1888) 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.
l m
a+ +
n S0
C=
⎛ m⎞ n
1 + ⎜ a+ ⎟
⎝ S0 ⎠ R
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
PHILIPPE-GASPARD GAUCKLER (1826-1905) made a proposal of two formulae for
use in different slope ranges, as follows:
V = λ1R 4/3S0 or C = λ1 R 5/6 S1/2

0 for S0 less than 0.0007
1
2/3
and V = λ 2 R S1/2
0 or C = λ 2 R6 for S0 greater than 0.0007
in which λ1 and λ2 are coefficients to be determined experimentally.
In 1889, ROBERT MANNING (1816-1897), 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

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 D’ARCY and Bazin. He
adapted an approximate value of 2 / 3 for this exponent. MANNING finally proposed, for
earth channels in good condition, the formula.
⎛ 1 R ⎞
In metric units, V = 34 S1/2
0 ⎜ R 2 + - 0.03 ⎟
⎜ 4 ⎟
⎝ ⎠
Or
⎛ R 0.03 ⎞
C = 34 ⎜⎜1 + − ⎟
⎝ 4 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 λ , expressed in metric units,
corresponded very closely to the roughness coefficient n associated with Ganguillet-
1 2/3 1/2
Kutter formula. Thus, in 1891, FLAMANT gave the formula V = R S0 (in metric
n
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.
V = MR 2/3 S1/2
0
1 2/3 1/2
Manning formula reads as V = R S0
n
1 1/ 6
and the coefficient C turns out to be 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
RS ⎡ ⎛ ϒ ⎞⎤
2
= ⎢0.0115 ⎜ 1 + B ⎟ ⎥ (in metric units)
V ⎣ ⎝ R ⎠⎦
which can be reduced to the form

86.96
C= (in metric units)
ϒB
1+
R
The term ϒ B in the above equation is a roughness factor. However, Bazin’s ϒ B exhibits
a thirty fold variation for a threefold variation in Kutter’s n.
As the slope of the channel is, once again, not considered in the above equation,
Bazin’s C is considered to be a function of R alone and not S0.
Another empirical formula for the Chezy coefficient C was given by PAVLOVSKII , in
1925.
Ri
The formula is C = (in metric units)
n
(
in which i = 2.5 n − 0.13 − 0.75 R n − 0.10 )
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.,
i = 1.5 n for R less than 1 meter and
i = 1.3 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):

48.85 R − 0.8
V= − 0.05 R
1 1
− ln + 1.6
s0 s0
3. GIRARD FORMULA (1803):
4. DE PRONY FORMULA (1804):

2
RS0 = 0.00004445 V + 0.00030931 V (in metric units)
5. EYTELWEIN FORMULA (1814-1815):

2
RS0 = 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.
RS0 W
= 0.0004021 + 0.0002881 (in metric units)
V V rc
in which W is the width of the river and Rc the radius of curvature of the river. For a
straight reach of the river, the term containing Rc should be dropped out. It is to be
noted that the term W / rc is reported under the root sign by LELIAVSKY.
7. ST. VENANT FORMULA (1851):
V = 60 ( RS0 )
11/21
(in metric units)
8. TADINI FORMULA (1850): (in metric units)
V = 50 RS0
The same formula is also attributed to COURTOIS.
9. HUMPHREYS and ABBOT FORMULA (1861)

2
⎡ 0.0025 * 0.933 0.933 ⎤
V= ⎢ + 68.72 R ' S0 - 0.05 ⎥ (in metric units)
⎣ R + 0.457 R + 0.457 ⎦
10. GANGUILLET AND KUTTER FORMULA (1869):
In metric units, C is given by
1 0.00155
23.0+ +
n S0
C=
⎛ 0.00155 ⎞ n
1+ ⎜ 23.0+ ⎟
⎝ S0 ⎠ R
11. ‘REDUCED’ FORM OF GANGUILLET – KUTTER FORMULA:
100 R
n+ R
12. GIBSON FORMULA:

1
24.55 +
C= n (in metric units)
n
1 + 24.55
R
13. MANNING FORMULA (1889):
1 2/3 1/2
V= R S0 (in metric units)
n
14. BAZIN FORMULA (1897):
86.96
ϒB
1+
R
15. SIEDEK FORMULA (1901):
This formula was given, in metric units, for the case of natural streams and rivers.
y mean 1000 S0
V=
( W )1/20
where ymean 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 +
ϒV
25.0ϒ V
1+
R
Where ϒ 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.
( i ) V =30.7 y S0.5 for y mean < 1.5 m

( ii ) V = 34.0 y0.75 S0.5 for 1.5 < y mean < 6 m
( iii ) V = 44.5 y0.60 S0.5 for y mean > 6 m
18. MATAKIEWICZ FORMULA (1911): (in metric units) reported by STRICKLER

(i) V = 35.4 y0.7 S0.493 + 10S

(ii) V = 35.4 R 0.7 Sβ
in which β is a variable exponent dependant on boundary roughness
19. KOCHLIN FORMULA (1913): (in metric units)
(
V = CK 1 + 0.6 R ) RS
where CK is the roughness parameter.
20. BARNES FORMULA (1916):
V = C R α Sβ
in which C, α and β vary depending on the type of the channel boundaries
21. STRICKLER FORMULA (1923):
V = M R 2/3 S1/2
22. FORCHHEIMER FORMULA (1923):
V = C R 0.7 S0.5
where the value of the coefficient C varied from 143 to 43 (in English units)
23. PAVLOVSKII FORMULA (1925): (in metric units)
1 i
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.

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.
V Lρ VL
Re = = he showed that a corresponding change in the law of resistance
µ υ
occurred with the change in the type of motion.
By this time, the Darcy-Weisbach equation for head loss through circular pipes
2
L V
hf = f was well established.
d 2g
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 Re and the relative
k
roughness of the pipe , where k is the average depth of pipe wall roughness and d 0
d0
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.

From dimensional analysis also, the same result is obtained, f = f ( R e , k / d o )
In 1932 - 33, NIKURADSE conducted a series of well-planned 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, Nikuradse’s
experimental findings have led to the establishment of semi rational formulae for
velocity distribution and hydraulic resistance for turbulent flows in circular pipes.
The Hagen-Poiseuille equation can be written as
32γ µ V L
hf =
d 02
Where ϒ 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
0.3164 0.3164
f= = .
( )
0.25
V d0 / υ Re0.25
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 ≤ Re ≤ 100 ,000 , 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.
1 ⎡⎛ k ⎞ 18.7 ⎤
= 1.74 − 2.0 log ⎢⎜ s ⎟+ ⎥
f ⎣⎢⎝ r0 ⎠ Re f ⎥⎦

MOODY has plotted the above equation to appear in the form of a family of
1 ⎡⎛ k ⎞ 18.7 ⎤
log f vs log Re curves for various = 1.74 − 2.0 log ⎢⎜ s ⎟+ ⎥ values.
f ⎢⎣⎝ r0 ⎠ Re f ⎥⎦
Application of the semi-rational formulae to open channel flows:
Analysing Bazins experimental data in this connection, KEULEGAN arrived at the
equation 1 / f = 2.034 log ( R / k s ) + 2.211 in the case of turbulent flow in rough-walled
channels.
Thus, the logarithmic formulae for rough walled channels were expressed as follows:
V ⎛ Rv ⎞ V ⎛R⎞
= A S + 5.75 log ⎜ * ⎟ for smooth channels and = A r + 5.75 log ⎜ ⎟ for rough
v* ⎝ υ ⎠ v* ⎝ kS ⎠
channels.
in which the characteristics AS and A r are functions of the Froude number.
16.4.2 Exponential Formulae
STRICKLER expressed the Manning 'n' in terms of roughness k s as n = 0.00106 k1/6

s
( k s in cm)
But he started with the numerical value of 1.476 instead of 1.486 in the Manning
formula. Strickler’s formula for n is given by n = 0.0342 d1/6

m
in which d m is the median sieve size of the sand grains and in feet.
n = 0.02789 d1/6
m
in which d m is in "m".
WILLIAMSON from his experimental data and also with some suggested corrections to
Nikuradse’s data, gave the formula
n = 0.031 k1/6
s (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 4.32 < R / k s < 276 requires Manning formula to be as given below.

n = 0.0387 k1/6
s ( k s in meters)
Manning formula is an exponential equation applicable to a particular range. In the first
place, Manning’s 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.
C = 8g / f = R1/6 / n or n = R1/6 f / 8g
For fully developed flow at high Reynolds number, f is found to be independent of
Reynolds number, and nearly proportional to 1 / R1/3 . 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
( n6 ) RSf ≥ 3.0755 * 10-14 with the assumption n = 0.03795 d 1/6
υ = 1.01* 10−6 m 2 /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 silt-laden 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

radius, sharp curvatures of the channels resulted in increased values of n. The effect of
channel irregularity, non-linear 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 n'0 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) n1' due to the surface irregularities;
(ii) n'2 due to the variation in the shape and size of the channel cross-sections;
(iii) n'3 due to the presence of obstructions in the flow;
(iv) n'4 because of growth of vegetation, algae or weeds; and,
(v) n'5 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, n = n'0 + n1' + n'2 + n3' + n'4 n5'
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.

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 non-uniform as well for unsteady flows. Only a few
attempts have made to investigate the validity of these assumptions s for non-uniform
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.
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0 1 X 106 2 X 106 3 X 106 4 X 106 5 X 106 6 X 106
UH
Reynolds number, n = __
v
Manning n vs. Reynolds number for three-mile Slough near
Rio Vista, California. (After Baltzer and Lai)

CORRECTION FUNCTION
FOR MANNING'S COEFFICIENT
1.2
1.1
1.0
0.9
0.8
0.7
Qo=Calculated flow for no = 0.025
0.6
0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
n/n0
Variation of flow rate with n for the Detroit River
Reference Mahmood. K and Yevjevich. V (Ed) Unsteady flow in Open Channels,
Volume - I Water Resources publications, Fort Collins, Colorado 1975,
Page No. 47
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 under-estimating 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 Q0. 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.

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).
0.00155 1
23 + +
Sο n
C=
n ⎛ 0.00155 ⎞
1+ ⎜ 23 + ⎟
R⎝ Sο ⎠
Manning's n ( Irish Engineer,1889)
1
V= R 2 / 3 Sο1 / 2
n
Chezy's C ( French Engineer,1768 )
R1 / 6
C=
n
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 &
0.045
some vegetation along banks

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 = C0 A Rx Sy = K Sy
in which K = C0 A Rx .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
2
1 3
conveyance can be written as K=CAR 1/2
and K = A R respectively and in which C
n
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
2
1 3
the following discussions and computations will be based on K = A R
n

17.4 The Section Factor for Uniform-Flow Computation

2/3
The term AR is known as the section factor for uniform - flow computation; in case of
1/2
Manning formula. This would be AR 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
For Manning formula For Chezy formula

2 1
nQ 2 Q
AR 3 = AR =
S0 C S0
2 1
2 K
3
AR = n K AR =
C
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 S0, there is only one possible depth for maintaining a uniform flow, provided
2/3 1/2
that the value of A R (or AR ) always increases with the increase in depth, which
is true in most cases. This depth is the normal depth yn. 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 (
1 / 2
or A R ) 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=kyi
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 yn can be computed. On the other hand,

when n or C, S0, and the depth (hence the section factor), are given, the normal
discharge Qn 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
2/3
given section factor, and vice versa. The A R 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.

10
8
6
4
y d0
1 m = 1.5
0.8 m = 2.0
0.6 m = 2.5
m = 3.0
0.4 m = 4.0
0.2
0.1
0.08
0.06
0.04 1 y
m
0.02
b
0.01
0.0001 0.001 0.01 0.1 1 10
2/3 2/3
AR AR
Values of _____ and _____
8/3 8/3
b do
Curves for determining the normal depth

Problem: Calculate conveyance factor K using Manning equation for a trapezoidal
channel.
Solution:
AR 2 / 3
∴K =
n
=
( b + my ) y ( b + my )2 / 3 y2 / 3
( )
2/3
n b + 2 1 + m2 y
2/3
⎡ my ⎤ ⎡ my ⎤
b ⎢1 + y 1+ b2 / 3 y2 / 3
⎣ b ⎥⎦ ⎢⎣ b ⎥⎦
= 2/3
⎡ y⎤
n ⎢1 + 2 1 + m 2 ⎥ b 2 / 3
⎣ b⎦
5/3
⎡ my ⎤ ⎡ 5 / 3 ⎤ 5 / 3
⎢1 + b ⎥⎦ ⎣ y ⎦ b
= ⎣ 2/3
⎡ y⎤
n ⎢1 + 2 1 + m 2 ⎥ b 2 / 3
⎣ b⎦
5/3
⎡ my ⎤
1 ⎢⎣1 + b ⎥⎦ [ by]5 / 3
K= 2/3
n⎡ 2 y⎤
⎢⎣1 + 2 1 + m b2 / 3
b ⎥⎦
5/3
⎡ my ⎤
⎢⎣1 + b ⎥⎦
1/ 3
1 ⎡ y5 b 5 b 3 ⎤
K= 2/3 ⎢ 2 3 ⎥
n⎡ 2 y⎤ ⎢⎣ b b ⎥⎦
⎢⎣1 + 2 1 + m b ⎥⎦
5/3
⎡ my ⎤
b8 / 3 ⎢1 +
b ⎥⎦
5/3
⎣ ⎡y⎤
K= 2/3 ⎢⎣ b ⎥⎦
⎡ y⎤
n ⎢1 + 2 1 + m 2 ⎥
⎣ b⎦

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 y y
= 0.938 or 302 22' = 0.95 or 308
d0 d0
Maximum velocity y y
0.81 = 0.81
d0 d0
256 27' 56'' θ = 257 27'
Solution
Chezy equations
(a) Circular section (Maximum discharge)
r2 T
A= (θ − sin θ )
2
p = rθ , Q = AC RS
r2 52
Q= (θ − sin θ ) C RS0 d0 y
2
θ
r2 r (θ − sin θ )
Q= (θ − sin θ ) C S0
2 2 θ
⎡ r2 ⎤
⎢ A (θ − sin θ )
r ⎥
⎢∵ R = = 2 = (θ − sin θ ) ⎥
⎢ P rθ 2θ ⎥
⎢⎣ ⎥⎦
Q=
(θ − sin θ )
3/ 2 5/ 2
r
S0 C
θ1/ 2 2 2
1/ 2
d ⎡ A3 ⎤ dA dP
⎢ ⎥ = 3P −A =0
dθ ⎣⎢ P ⎦⎥ dθ dθ
Let x = (θ − sin θ ) θ −0.5 ;

3/ 2
xr 5 / 2
Q= S0 C
2 2
Differentiating x w.r.t to θ and equating to zero.

dx ⎡ 3 −0.5 ⎤ ⎡ 1 3/ 2 ⎤
= ⎢ (θ − sin θ ) (1 − cos θ )(θ ) ⎥ − ⎢ θ −3 / 2 (θ − sin θ ) ⎥ = 0
1/ 2
dθ ⎣ 2 ⎦ ⎣2 ⎦
3 (θ − sin θ )
1/ 2 3/ 2
1 θ − sin θ ⎞
= (1 − cos θ ) = ⎛⎜ ⎟
2 θ 2⎝ θ ⎠
θ − sin θ
3 (1 − cos θ ) =
θ
sin θ = θ ( 3 cos θ − 2 ) ;
θ = 308 Radians.
Then the depth for maximum discharge.
θ⎞
⎛
(
y = r + r cos ⎜180 − ⎟ = r 1 + cos 26 = 1.899r
⎝ 2⎠
)
128
308 − 180 = = 64 ,
2
90 - 64 = 26
⎡ 1.899 ⎤
y = 0.95do ⎢∵ = 0.95⎥
⎣ 2 ⎦
(a) Manning Equation - Maximum Discharge
d ⎡ 2/ 3⎤
AR =0
dθ ⎣ ⎦
1/ 3
d ⎡ A5 ⎤
⎢ 2⎥ =0
dθ ⎢⎣ P ⎥⎦
dA dP
5P − 2A =0
dθ dθ
dA r 2 ⎡ r2 ⎤
= (1 − cos θ ) ⎢∵ A = (θ − sin θ ) ⎥
dθ 2 ⎣⎢ 2 ⎦⎥
dP
=r [∵ P = rθ ]
dθ

⎛ r2 ⎞ r2
∴ 5 ( rθ ) ⎜ ⎟ (1 − cos θ ) − 2 (θ − sin θ ) r = 0
⎜ 2 ⎟ 2
⎝ ⎠
r3
5 θ (1 − cos θ ) = r 3 (θ − sin θ )
2
1
∴ θ (1 − cos θ ) = (θ − sin θ )
2 .5
5θ (1 − cos θ ) = 2 (θ − sin θ )
5θ − 5θ cos θ = 2θ − 2 sin θ
3θ = 5θ cos θ − 2 sin θ
θ = 302 22'
θ d0
y = r − r cos = 1.876r = 1.876
2 2
∴ y = 0.938 d 0
(b) Circular section (Maximum velocity)

Using Manning equation
r2 r ⎛ sin θ ⎞
A = (θ − sin θ ) ; R = ⎜1 −
2 2⎝ θ ⎟⎠
V ∝R
d ⎡ 2/ 3⎤
R =0
dθ ⎣ ⎦
d ⎡ A2 / 3 ⎤
⎢ 2/ 3 ⎥ = 0
dθ ⎢⎣ P ⎥⎦
dA dP
P −A =0
dθ dθ
⎡ r2 ⎤ r2
( rθ ) ⎢ (1 − cos θ )⎥ − (θ − sin θ ) r = 0
⎣⎢ 2 ⎦⎥ 2
r3 r3
∴ θ (1 − cos θ ) = (θ − sin θ )
2 2

−θ cos θ + sin θ = 0
θ = tan θ
θ = 257 27' 56'' ≈ 257 30'
⎛ 257.5 ⎞
The depth of water for maximum velocity is y = r + r cos ⎜180 − ⎟ = r + r cos 51.25 =
⎝ 2 ⎠
0.81 diameter = 0.81d0
Problem
What would be the difference in discharge when it is running full and when it is
yn
under = 0.938
d0
Solution
yn
= 0.938 AR 2 / 3
= 0.3353
d0 d08 / 3
yn
= 1 .0 AR 2 / 3
= 0.3117
d0 d08 / 3
Qmax 0.3353
= = 1.0757
Q full 0.3117
i.e. Maximum discharge is 7.6% higher than discharge in pipe when flowing full.
If Manning's equation is used.
yn
If Chezy's equation is used, = 0.95
d0
A
= 0.77072
d02
P
= 2.69057
d0

A
R d02
= = 0.28645
d0 P
d0
AR1 / 2
∴ = 0.41249
d 05 / 2
AR1 / 2 π d02 d0 π 1
When full = = = 0.39269
d 05 / 2 4 4 4 4
Qmax 0.41249
= = 1.0504
Q full 0.39269
∴ 5.04% excess.

18.3 The channel- bed slopes may be classified into the
following five categories

Mild Slope (M) Horizontal H
Sustainable Slopes Critical Slope (C) Non Sustainable Slopes
Steep Slope (S) Adverse Slope A
The slope that can uniform flow is called sustainable slopes. The mild slope sustains
sub critical ( Fr < 1) uniform flow, denoted as M. The critical slope sustains uniform flow
at critical depth ( Fr = 1) denoted as C. steep slope sustain the supercritical uniform flow
( Fr > 1) denoted as S.
When the slope is zero (Horizontal) then
1 23
V= R So = zero
n
Q = V yn ∴ yn → ∞ indicated as yn*
1 23
V= R So
n
Thus yn is imaginary or negative (from chezy's equation).
Slopes are also classified as

(i) Mild always subcritical uniform flow is sustained F<1
Sustaining slope (ii) Critical always sustains critical uniform flow F=1
(iii) Steep always sustains super critical uniform flow F>1
(i) Horizontal Normal depth y*

Non-Sustaining slope n
(ii) Adverse
Positive slope So = - dz
__
dx
Slope
Negative slope So = dz
__
dx
dH
Energy slope Se = __
dx
Friction slope dhf
Sf = __
dx
Bed slope dz
__
S0 = dx
Water surface slope Sw = __dy

dx

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 Qn, The mean velocity of flow V , The normal depth yn, The
coefficient of roughness n, The bed slope of channel S O 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 Qn: 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 yn: 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 So: 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 *.

TABLE 1: TYPE OF PROBLEMS OF UNIFORM - FLOW COMPUTATIONS
Type of Discharge Velocity Depth Roughness Slope Geometric

problem elements
Q V y n So
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, So, and n carries a discharge of Q
m3 / 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 yn as

Substituting the given quantities in the above expressions in the Manning formula and
simplifying, one gets an algebraic expression as follows
ao+ a1yn = [y (b+y)] 2.5
in which ao and a1 are constants, b is the bed width of the channel in meter. This
equation is to be solved by trial and error for yn in meter. Then area of the flow an
square meter and velocity Vn 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/S0. Then, assume a value of y and compute the
2/3 2/3
section factor A R . Make several such trails until the computed value of AR 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 R2/3 A R2/3 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
2/3 2/3
found from the y - A R curve where the coordinate of A R 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.

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 ( )
y
600 600
b
Triangular
Solution:
(
Area A = b - y / 3 * y )
= by -0.58y 2
(
Perimeter P = b + 4 y / 3 )
= b + 2.31 y
For a given depth, the discharge is to be maximum.
According to Manning formula,
⎛1⎞
Q = ⎜ ⎟ * AR 2 / 3S1ο / 2
⎝n⎠
Where n and Sο are cons tan ts
Now, R=A/P =
( by - 0.58 y 2 )
( b + 2.31 y )
Substituting the values in the above expression, then,
{( }
1/3
)
5
/ ( b + 2.31 y )
2
Q= by - 0.58 y 2
differentiating the right hand side of the equation, then,
( b + 2.31y )2 * 5 * ( by − 0.58y 2 ) ( )
4 dA 5 dP
* − ( b + 2.31y ) * 2 * by − 0.58y 2
dy dy
=0
( b + 2.31y )4

But dA / dy = ( b - 1.16 y ) and dP / dy = 2.31.

Then the above equation reduces to
5b2 -10.72 y 2 - 10.37 =0
5b2 -10.37
y2 =
10.72
( 0.466b2 − 0.97 )
1/2
y =
is the depth at which the discharge is maximum.
Computation of uniform flow depth in CBI&P channel
1
Q = 30m3s-1 , S0 = , m = 1.25, n = 0.015
1600
⎛1 ⎞ 1
A= 2 ⎜ y 2 Cotθ ⎟ + y 2 2θ
⎝2 ⎠ 2
= y 2 ( θ+Cotθ )
y 1
P = 2y ( θ+Cotθ )
R=
A y
=
m
P 2
Cotθ =1.25
θ = 38.6D = 0.644 radians
∴ A = 1.894y 2
1
Q = AR 2/3S1/2
n 0
2/3
1.894 2 ⎛ y ⎞
30 = y ⎜ ⎟ 1/1600
0.015 ⎝ 2 ⎠
2
8/3 0.015×30 ⎛ 1 ⎞ 3
∴y = ⎜ ⎟
1.894 1/1600 ⎝ 2 ⎠
y = (14.1)
3/8
= 2.70m
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 m3/s, calculate
(a) the normal slope,
(b) the critical slope and critical depth for 10.20 m3/s, and
(c) the critical slope at the normal depth of 915mm.

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 m3 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.
(τ ο = γ R Sf )
4. A discharge of 40.0 m3/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 m3/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 m3/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, Re.
Classify the flow according to Froude Number.
8. In the previous problem if the discharge is changed to 30 m3/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

1 1
3 3
10 m 2.5 m 10 m
1 1
2 2
5m
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 m3/s.
(a) Calculate the discharge for a proportional depth of 0.938 do 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 m3/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
y1 1
y
m
Manning n, depth of flow y, and longitudinal slope So are given. Express discharge as a
function of side slope, depth, n and bed slope.
Answer
Q = ( l / n ) f ( m ) y 8/3
0.3142 m 5 /3
f(m)=
(1 + 1 + m 2 )2 /3
(a) Compute the discharge when n = 0.017, y = 6.5 cm and So = 0.03, m = 24.
(b) Compute the discharge when m = 24, n = 0.015, y = 8.00 cm, So = 0.04.

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 m3/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 m3/s. Assume Chezy
coefficient as 65 m1/2s −1 .
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 m1/2s −1 .
18. A trapezoidal channel has slope 1.5 horizontal to 1 vertical. It is to discharge 20
m3/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
1/2 −1
discharge that can be secured at atmospheric pressure, if the value of C is 50 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 m3/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 S0 , 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?

22. What flow can be expected in a 1.20 m wide rectangular cement-lined 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 m3/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
m3/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
m3/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 m3/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 m3/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 m3/s. If slope S0 =
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 yn
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 m3/s. The slope of the channel
bottom is 0.001 and n = 0.013.

19.2 Problem: Irregular Channel

Compute the uniform flow depth for the C.S. when flow is 283 m3/s. n = 0.024,
S0 = 0.0001
River stage (in m) above Distance to first perimeter Distance to second perimeter
an arbitrary Datum intersection from south intersection from south Bank
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
20 (x, y) (xi, yi)
15
River bed
elevation
0 60 120 365
Distance from South Bank
River bed elevation has a function of the distance (after French)

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
Q 2T
the dimensions of the channel. The corresponding equation is 1- =0 must be
gA3
satisfied by some positive depth yc greater than 0 and less than an upper bound say
100 m. T is the top width given by (b + 2 myc). The interval is bisected and this value of
yc 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.
T F(x)
1y 0
1
m m 100
b
Trapezoidal Bisection
Similarly it could be used for obtaining uniform flow depth using the following equation.
2
1
Q= AR 3 S0
n
2
1
Q - AR 3 S0 = 0
n
2
nQ
- AR 3 = 0
S0

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 yk is selected and
the flow rate Qn, is computed using manning formula using the area and hydraulic
radius corresponding to yk. This Qk is compared with actual flow Qn; then the objective is
to chose y such that the error.
f (yk) = Qk - Qn is within the tolerance limit. The gradient of f with respect to y is
df(yk ) dQk
=
dyk dyk
because Qn is constant. Hence, assuming manning roughness is constant,
⎛ df ⎞ ⎛1 1 2 ⎞
3
⎜ ⎟ = ⎜ So 2 A k R k ⎟
⎝ dy ⎠k ⎝ n ⎠
⎛ -1 ⎞
⎜ 2
1 1 2A R 3 dR dA ⎟⎟
= So 2 ⎜ +R 3
n ⎜ 3 dy dy ⎟
⎜ ⎟
⎝ ⎠k
1 1 2 ⎛ 2 dR 1 dA ⎞
= So 2 A k R k 3 ⎜ + ⎟
n ⎝ 3R dy A dy ⎠k
⎛ df ⎞ ⎛ 2 dR 1 dA ⎞
⎜ ⎟ = Qk ⎜ +
⎝ dy ⎠k ⎝ 3R dy A dy ⎟⎠k
in which the subscript k out side the bracket indicates that the
quantities in the bracket computed for y = yk.
In Newton's method,
given a choice of yk , yk+1 is chosen to satisfy
⎛ df ⎞ 0- f (y)k
⎜ ⎟ =
⎝ dy ⎠k yk + yk+1
This yk+1 is the value of yk ,
f (yk )
yk+1 = yk -
( df / dy ) k

Which is the fundamental equation of the Newton's method. Iterations are continued
until there is no significant change in yn; this will happen when the error is nearly zero or
an acceptable prescribed tolerance.
Thus for manning equation it may be written as

1 - Q / Qk
y k+1 = y k -
⎛ 2 dR 1 dA ⎞
⎜ + ⎟
⎝ 3R dy A dy ⎠ k
For rectangular channel A = bo y and R = bo y / ( bo + 2y ) where bo is the
channel width; The quantity in denominator can be for rectangular channel
d
dy
d
(R ) = A P
dy
( )
1 dA A dP
= −
P dy P 2 dy
⎡ T R dP ⎤
=⎢ − ⎥
⎣ P P dy ⎦
consider
2 dR 1 dA
+
3R dy A dy
2 P ⎡ T R dP ⎤ T
− +
3 A ⎢⎣ P P dy ⎥⎦ A
21⎡ dP ⎤ T
⎢T −R ⎥+
3 A⎣ dy ⎦ A
2 T 2 R dP T
− +
3 A 3 A dy A
⎡ 5 T 2 1 dP ⎤
⎢ 3 A − 3 P dy ⎥
⎣ ⎦
For rectangular channel
5 bo 2 1
− 2
3 bo y 3 ( bo + 2 y )
51 4 1
−
3 y 3 ( bo + 2 y )
5 ( bo + 2 y ) − 12 y 5bo + 10 y − 4 y
=
3 y ( bo + 2 y ) 3 y ( bo + 2 y )
5bo + 6 y
=
3 y ( bo + 2 y )
1 - Q /Qk
y k+1 = y k -
⎛ ⎞ 5 bo + 6 y k
⎜ ⎟
⎜ 3y b + 2 y ⎟
⎝ k o k ⎠ ( )
Similarly the channel shape function ⎡⎣( 2/3R )( dR/dy ) + ( 1/A ) (dA/dy)⎤⎦
for other cross sections can be derived.

Trapezoided Channel
( bo + 2my ) + 6 y (1 + m2 ) + 4my 2 (1 + m2 )
3y ( bo + my ) ⎛⎜ bo + 2 y (1 + m 2 ) ⎞⎟
⎝ ⎠
8
Triangular Channel
3y
4 ( 2sinθ + 3θ − 5θ cos θ )
Circular Conduit
⎛θ ⎞
3do (θ )(θ − sin θ ) sin ⎜ ⎟
⎝2⎠
in which
⎛ 2y ⎞
θ = 2 cos −1 ⎜1 − ⎟
⎝ do ⎠
Example:
Compute the flow depth in a 0.6 m wide rectangular channel having n= 0.015, S0 =
0.025, and Q = 0.25 m3s-1.
B
Solution:
Let wide bo = 0.6m

Manning coefficient n = 0.015
bed slope Sο = 0.025
discharge Q = 0.25m3 s −1
normal depth y = ?
A bo
Hyraulic mean radius R = =
p bo + 2 y
2 1
1
Q = AR 3 Sο 2
n
2
1
1 ⎛ byk ⎞3
Q = bo y ⎜ ⎟ Sο 2
n ⎝ b + 2 y k ⎠

⎡ 5 ⎤
1⎢
Q= ⎢
( by )
k 3 ⎥
1
S 2
n 2⎥ ο
⎢⎣ ( b + 2 yk ) 3 ⎥⎦
⎡ 5 ⎤
1 ⎢ ( 0.6 × yk ) 3 ⎥ 1
Qk = ×⎢ 2⎥
× ( 0.025 ) 2
0.015
⎢⎣ ( 0.6 + 2 yk ) 3 ⎥⎦
5
53 53
0.6 3 y 4.4993 y
Qk = 10.5409* k = k (1)
(
0.6 + 2 yk
23
) (
0.6 + 2 yk
23
)
5bo + 6 yk
Shape function =
3 yk ( b + 2 yk )
5 ( 0.6 ) + 6 yk 3 + 6 yk 1 + 2 yk
= = =
3 yk ( 0.6 + 2 yk ) 3 yk ( 0.6 + 2 yk ) yk ( 0.6 + 2 yk )
⎛ 0.25 ⎞
⎜1 − ⎟ yk ( 0.6 + 2 yk )
y k+1 = yk − ⎝ Qk ⎠
(2)
(1 + 2 yk )
Iteration (k) 1 2 3
yk ( m ) 0.100 0.1815 0.1727
Q(m3s -1 ) 0.1125 0.2684 0.2488
V Q/ A
Froude number F = =
gy gy
0.2488 / ( 0.6* 0.1727 )
F= = 1.844
( 9.81*0.1727 )
∴ super critical flow

Hydraulics Prof. B.S. Thandaveswara Rao
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 slope-area 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
V2 V2
velocity = 1 m/s, = 0.05 m and if the velocity = 2 m/s, = 0.20 m ) and,
2g 2g
(c) The fall should be equal to or greater than 0.15 m.
( )
When the reach is contracting V u < Vd , k=1.0 . When the reach is expanding
( Vu > Vd ) , k=0.5 . 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.

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, cross-sectional 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 cross-sectional 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, Regula-Falsi (false position), secant, and the Van Wijngaarden-Dekker-Brent
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.

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

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

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 1.0 to 1.0 2.0 to 1.0
soils
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 S0 = 0.01

Moderate 10 to 20 m / S0 = 0.01 to
slope km 0.02
Steep slope ≥ 20 m / km S0 ≥ 0.02

0.1
0.05
Median (d50)
Grain Size in mm
0.02
F = 1.0
0.01
F = 0.85
0.005
0.002
0.001
0.0005
0.0002
0.0001
Bank Full Discharge, m3/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

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

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 non-flushing 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
Lining Lining Type n-value different depth ranges

Category Depth ranges
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
Rigid Soil Cement 0.025 0.022 0.020
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
Jute Net 0.028 0.022 0.019
Fiberglass Roving 0.028 0.021 0.019
Straw with Net 0.065 0.033 0.025
Temporary Cured Wood Mat 0.066 0.035 0.028
Synthetic Mat 0.036 0.025 0.021
2.5-cm (d50) 0.044 0.033 0.030
Gravel
Riprap 5 -cm (d50) 0.066 0.041 0.034
15-cm (d50) 0.104 0.069 0.035

Rock
Riprap
30-cm (d50) - 0.078 0.040

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 mis-operation 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 semi-circular
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
m3 /s at relatively large depths of flow. A prelimimary estimate of freeboard for an
unlined channel can be obtained from USBR formula.
FB = Cy
in which FB is the freeboard in feet, y is the design depth of flow in feet,
C is a coefficient. However, it may be noted that C has dim ensions of L1/ 2 .
C var ies from 1.5 at Q = 0.57 m3 / s to 2.5 for canal
capacity equal to and more than 85 m3 / s.

The free board recommended by USBR for channels are given below
Q m3/s Free board FB 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 m3/sec.
Free board as per Indian Standards (IS 4745 - 1968), (IS 7112 - 1973)
Discharge Q (m3/s) Free board (m)

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
1 MEMBRANE LINING 1
ABOVE W.S.
HEIGHT OF EARTH
LINING ABOVE W.S
0
0.1 .2 .4 .6 .8 1 2 4 6 8 10 2 4 6 8 100 2 4 6 8
1000
DISCHARGE CAPACITY IN m3/s
Bank height for canals and free board for hard
surface or buried membrane and, earth lining

Free boards provided in some of the major lined canals in India are given below
Sl.No. Name of Canal Free Board FB 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 (m3/s) < 0.15 0.15 - 0.75 0.75 - 1.50 1.50 - 9.00 > 9.00
Free board 0.30 0.45 0.60 0.75 0.90
(m)
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
Vmb ⎛ 2T ⎞
∆y max = ⎜ ⎟
2g ⎝ rc ⎠
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 m3/s ( Houk, 1956 ). Suggested radii for different
discharges are given in table below.

Radius of curves for lined canals
Discharge (m3/s) Radius (minimum) in m

280 and above 900
Less than 280 to 200 760
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

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 A P R T D Z=A D
Section
Rectangular 2y2 4y 0.5 y 2y y 2y2.5
3y 2 2 3y 0.5 y 4 3 3 3 2 .5
y y y ( 1.5 y 2.5 )
(1.732 y2 )
Trapezoidal 4 2
(3.464y) 3
(2.3094y) (0.75y)
y2 2 2y 2 2y y 2 2. 5
Triangular y y 0.707 y 2.5
2.828y 4 2 2
0.3535y 0.5y
Semi π πy 0.5 y 2y π π 2.5
y2 y y 0.25π y 2.5
Circular 2 4 4
Parabola 4 8 y/2 2 2y 2 8
2 y2 2y y 3y 2.5
4 3 3 0.5y 3 9
2 y2 2.83y
3 1.89 y 2 3.77y 0.667y 1.5396 y 2.5
Hydrostatic 1.40 y2 2.98 y 0.468 y 1.917 y 0.728y 1.91 y2.5
Catenary
** 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:

(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 water-table.
(c) Reduction in area of cross-section (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. Rigid-impermeable Lining,
2. Flexible and Permeable Permanent Linings and
3. Flexible Temporary Linings.
Example for the same are indicated in the box.
Canal Lining
Rigid Impermeable Linings Flexible and Permeable Permanent Linings Flexible Temporary
Rubble Masonry Rip-rap or Stone Blocks Bare Soil

Cast-in-place Concrete Gabions Straw with Netting
Grouted Rip-rap or Grouted Interlocking Pre-cast Concrete Hemp or Jute Mats
Pre-cast Concrete Interlocking Synthetic Units Synthetic Matting
Soil Cement Vegetation and Grasses
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.

Allows infiltration into channel bed, hence loss of water. Partial failure can occur and still
can resist erosion.



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.

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 S0, 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 S0
2 Compute the value of section factor AR 2/3 = nQ/ So or AR1/2 = Q/ C So ( )
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
3 b
shown as = 4 − m . In India, y for the trapezoidal channel can be taken as
y
b
0.577 A which corresponds to = 3 − m for earth canals).
y
4 If hydraulically efficient section is required, then the standard geometric
characteristics (click) are used and yn is to be computed.
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
5 (Check Froude number and other velocity constraints such as ( for non-
reinforced concrete linings V ≤ 2.1 m/s and Froude number ≤ 0.8 . For
reinforced linings 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.

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 m3/s on a slope of 0.0015. Consider the hydraulically efficient
trapezoidal shape.
Solution
For hydraulically efficient trapezoidal channel
y
A = 1.73 y 2 , P = 3.46 y, R =
2
n = 0.015,
1
Q = AR 2/3 S10/ 2
n
⎛ 2
1⎞
1 ⎜ 2 ⎛ y ⎞3
20 = ⎜ 1.73 y ⎜ ⎟ ( 0.0015 ) 2 ⎟⎟
0.015 ⎜ ⎝2⎠ ⎟
⎝ ⎠
8
y3 = 7.107
y = 2.086 m
For Trapezoidal channel width is given by
2
b= y,
3
b =1.15y = 2.409 m
m=
3
3
(
= 0.5773 i.e., = 60o )

Q 20
Velocity = = = 2.656 m/s
A 1.73y 2
A 1.73y 2
Hydraulic mean depth D = = = 0.749y = 1.563 m
T 4
y
3
V
Froude Number = = 0.678
gD
Freeboard for discharge Q = 20m3 /s is 0.75 m to 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.
Free board = 0.764 m
2.85 m
1.0
2.086
0.58
o
60
b = 2.4 m
Design a trapezoidal channel to carry Q = 20.25 m3/s, V = 1.5 m3/s, n = 0.025, S0 =
0.0016, side slope m = 2. Assume a bed width of 6 m.
Solution
Step 1: Q, n, S0 and m have been given
A = ( b + my )
P = b + 2 y 1 + m2
R=
( b + my )
b + 2 y 1 + m2

nQ 20.25
AR 2 / 3 = = 0.025* = 12.656
S0 0.0016
Discharge 20.25
Area = = = 13.5 m 2
Velocity 1.5
13.5 = ( 6 + 2y ) y
Solving for y, we get y = 1.5 m
b
=4
y
Add a free board of 0.75 m.
Designed channel is shown in figure.
Fb=0.75
y = 1.5 1
2
b=6m
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.

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.

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.
100
80 USSR Data
Sand Gravel Pebbles
60 Silt F M C F M C F M C L
50 20 100 200
0.005 0.01 0.02 0.05 0.1 0.2 0.5 1 2 5 10 50
40
Average particle size, mm
30
20
10
8
6
5
4
3
2
1
0.8
0.6
0.5
0.4
0.3
0.2 U.S. standard mesh sieve sizes
Clay Silt V.F. F M C F M L
Sand Gravel
U.S. Dept. Agriculture, Bureau of Soils Classification
0.1
Legend: V.F. - very fine; F- fine; M-medium; C-coarse; L-large
U.S. and U.S.S.R. data on permissible velocities for noncohesive soils.
Following Steps are used for Designing
Given a particular soil type, the channel is designed so that the design velocity does not
exceed Vmax 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.

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
m3s-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
1
3
b=6m
A= ( b+my ) y = ( 6+ 3*1.3) *1.3 = 12.87 Sq.m

Q 20
V= = =1.554 m s-1
A 12.87
which is higher than the permissible velocity (of V = 1.25 ms-1)
From graph
So = 0.0015 < 0.0065 ( 0.65% )

∴ Side slope adopted 3:1 which is < (1 : 1)
Suggestion : Increase width, b, to reduce velocity:
For b = 8.4 m, y n = 1.3 m Corresponding area of flow A = ( 8.4+3) *1.3 = 15.99 m2

Q 20
V= = = 1.251 m/s which is equal to the permissible velocity
A 15.99

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.
Side Slope, m: 1 = 1.5 : 1
y
4y
0.750 γ ySo γ ySo

0.750
γ ySo
0.970
Tractive force distribution obtained using membrane analogy

This distribution varies depending on the cross section and material

0.3 m
0 0.9 m
Oslon - Florey
Cruff
Simon
Normal
Preston Tube
Boundary shear distribution,
Central Water Power Research Station
(August, 1968)
Discharge: 0.06 and 0.11 m3/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
⎡ ⎤
⎢ u −u ⎥
τ0 = ρ ⎢ 2 1 ⎥
⎢ 2.3 log y 2 ⎥
⎢⎣ k y1 ⎥⎦
Cruff's Method: Uses the above equation but k value is obtained from velocity profiles.
Oslon and Florey Method: Membrane analogy.

2.0
Bed Prototype Rough
Trapezoidal Boundary
1.5
τmax
______
γRS
1.0
0.5
0 5 10 15 20 25 30
b/y or b/R
2.0
Sides Prototype Rough Trapezoidal
Boundary (Left)
Prototype Rough Trapezoidal
τmax 1.5
______ Boundary (Right)
γRS
1.0
0.5
0 5 10 15 20 25 30
b/y or b/R
Maximum shear on bed and sides for alluvial channel
based on Normal's Method.
(U.P. Irrigation Research Institute Roorkee,
Annual Review, 1971)
τmax
______ b
__
γRS = 1.2 for bed and 0.6 for the sides when y exceeds 10
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 γ ys0 and on the sides 0.76 γ ys0 .

b
The figures show the maximum unit tractive forces in terms of γ ys0 for different
y
ratios.
1.0 1.0
0.9 Trapezoidal,
0.9 m = 2 and 1.5
Trapezoidal, m = 2
0.8 0.8
0.7 Trapezoidal, m = 1.5
0.7
0.6 0.6
Rectangle Rectangle
0.5 0.5
Trapezoidal, m = 1
0.4 0.4
0.3 0.3
0.2 0.2
0.1 0.1
0 0
0 1 2 3 4 5 6 7 8 9 10 0 1 2 3 4 5 6 7 8 9 10
b/y b/y
On sides of channels On bottoms of channels
Maximum unit tractive forces in terms of γyS 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.

42
40
38
36
34
32
30
28
26
24
22
20
0.1 0.2 0.3 0.4 0.6 0.8 1.0 2.0 3.0 4.0 inches
mm
Particles size in inches (1 inch = 25.4 mm)

Angles repose for non cohesive material
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.

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.
20000 DESIGN OF SMALL DAMS USBR' 87

4
3
10000
9000 2
8000
7000
6000 Fortier & Scobey - Recommended for canals in fine sand 1
5000 NK 0.9
4000 with water containing colloids Line representing relations of tractive forces b/ft2 = 0.5 0.8
3000 Diameter in inches 0.7
0.6
Tractive force kg/m2 = diameter in centimeters (approx) 0.5
2000
U.S.S.R. 0.4
Recommended value for NK 0.3
U.S.S.R. Canals with 0.2
1000 canals with high content of Canals with
2.5 % colloids in water
fine sediment in the water 0.1% colloids
900
800 in water
700
0.1
600 Schoklltach - Recommended Recommended values for canals0.09
500 for canals in sand with low content of fine sediment
0.08
400 0.07
in the water 0.06
300 NUERNBURG KULTURAMPT (NK) Recommended values for canals in 0.05
coarse, non-cohesive material size
200 25% or larger 0.04
0.03
0.02
100 Recommended values for canals with clear water
90 Stroub values of critical force
80
70 U.S.S.R. Canals with clear water
60 0.01
50
0.009
40
0.008
30 0.007
20 0.006
0.005
0.004
10 Fortier & Scobey - Recommended for canals, in fine sand and clear water 0.003
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
MEAN DIAMETER, MILLIMETERS
Conversion Factor 1 lb / ft 2 = 47.87 N / m2
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.

P = 4.75 Q
f s = 1.76 d1/ 2
1/ 3
⎛Q⎞
R = 0.47 ⎜ ⎟
⎝ fs ⎠
S0 = 3*10−4 f s5 / 3 Q1/ 6
In which P is the wetted perimeter (m), R is the hydraulic mean radius (m), Q is the flow
in m3/s, d is the diameter of the sediment in mm, fs is the silt factor, S0 is the bed slope.
Table: particle size and silt factors for various materials
Material Size (mm) Silt factor

Small boulders, cobbles,
64 - 256 6.12 to 9.75
shingles
Coarse gravel 8-64 4.68
Fine gravel 4-8 2.0
Coarse sand 0.5-2.0 1.44 - 1.56
Medium sand 0.25-0.5 1.31
Fine sand 0.06-0.25 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.
V = 10.8 R 2 / 3 S1/ 3
0 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;

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 γ ALS0 in
which, γ is the unit weight of water, A is the wetted area, L is the length of the channel
reach, and S0 is the slope. Thus, the average value of the tractive force per unit wetted
γ ALS0
area, is equal to τ 0 = = γ RS0 , in which P is the wetted perimeter and R is the
PL
hydraulic mean radius; For wide rectangular channel, it can be written as τ 0 = γ yS0
The tractive force is also called Drag Force.
Consider a sediment particle submerged in water and resting on the side of a
trapezoidal channel. For this case the tractive force A p τs must be equal to gravity force
component w s sin α
Let
τb be the critical shear stress on bed, τs be the critical shear stress on side-walls
Ap be the effective surface area of typical particle on bed or side wall
θ0 be the angle of the Side slope and
α be the angle of repose (angle of internal friction) of bank material.

L
C
b/2 my
Flow
τbAp
τsAp Ws
θ0
Plan View
Ws = submerged weight of the particle
On the surface of the side slope
θ0 θ0

From Force diagram, resultant Force, R:
( )
2
R= ( Ws sinθ0 ) + A pτ s
2
Resisting Force, Fs:
Ws cosθ0 is the weight component normal to side slope

tanα is the coefficient of friction ( due to angle of internal friction )
Fs = Ws cos θ0 tanα
Therefore
R = Fs at incipient motion.
Ws cos θ0 tan α = Ws sin 2 θ0 + A p2 τ s2

Solving for the unit tractive force τ s that causes impending motion on a sloping surface
Ws tan 2θ0
∴ τs = tan α cos θ0 1-
Ap tan 2α
On the channel bed, with θ 0 being zero it reduces to
Ws tanα
A p τ b = Ws tanα → τb =
Ap
Tractive Force Ratio
τs tan 2 α sin 2α
K= = cos α 1 − = 1-
τb tan 2 θ0 sin 2θ0
K is the reduction factor of critical stress on the channel side.
Thus the ratio is a function of only side slope angle θ 0 and angle of repose of the
material α .

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
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 m3s-1
Gujarath - Rajasthan border 71m3s-1
No. of branches 42
Length of distribution Network 66000 km
concrete lining of 100 mm to 125 mm concrete
Total Lining
Phase I 150.58 + 93.93 + 39.26 = 283.77
Phase II 126.14 + 1.08 +22.60 = 149.82
Total = 435.59 x 105 Sq.m
2) Sardar Sarovar Project

design disharge 86937.2 m3s-1 will be the 3rd largest in the world.
Gazenba, china 1.13 x 105 m3s-1
Tucurri Brazil 1.0 x 105 m3s-1
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
Height of dam from foundary 163.00 m
Gross storage 9497.07 m3
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 rip-rap, 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 S0 ≈ 0.01 .

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 τ max vs. τ permissible or Vmax vs. Vpermissible , and if they satisfy their
add the freeboard and the design is complete. Table below lists the values for various
lining types.
Table : Permissible shear stresses for lining materials
Lining Category Lining Type Permissible Unit Shear Stress (kg/m2)

Woven Paper Net 0.73
Jute Net 2.20
Fiberglass Roving
Single 2.93
Temporary
Double 4.15
Stream with Net 7.08
Cured wood Mat 7.57
Synthetic Mat 9.76
Class A 18.06
Class B 10.25
Vegetative Class C 4.88
Class D 2.93
Class E 1.71
Gravel Riprap 2.5 cm 1.61
5 cm 3.22
Rock Riprap 15 cm 9.76
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.

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 m3/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:
n for gravel ranges: 0.020 - 0.030

Assume n = 0.025
n = 0.038 ( d 75 )
1/ 6
= 0.020
2. Angle of repose:
d 75 = 2cm = 0.8 in → α = 32o
3. Slightly sinuous channel: Cs = 0.90
4. Side Channel slope: Try 2H:1V
⎛1⎞
θ = tan -1 ⎜ ⎟ = 26.6D
⎝2⎠
5. Tractive force ratio:
τs sin 2θ sin 2 26.6D

K= = 1- = 1- = 0.53
τb sin 2α sin 2 32D
6. Permissible Tractive Force:
( )
Bed: τ b = Cs 16 N/m 2 = 14.4 N/m 2
Side: τs = Kτ b = 0.53 (14.4 ) = 7.6 N/m 2
7. Assume incipient motion on side wall:

τs = 0.76γ yoSo = 7.6 N/m 2

τs 7.6
yn = = = 0.68 m
0.76γ So 0.76 ( 9790 )( 0.0015 )
8. Solve for bottom width b:
1 2/3 1 / 2 1 A5/3 1 / 2
Q = AR S0 = S0
n n P 2/3
where A=by+my 2 , P= b+2y 1+m 2
b = 2.42m ( smallest positive real solution )
9. Tractive force on bed:
τ b = 0.97 γ y0S0 = 0.97 ( γ )( 0.68 )( 0.0015 ) = 9.7 N/m 2

1.7 N/m 2 < 14.4 N/m 2
10. Check velocities:
Area = by + my2 = ( 24.2 )( 0.68 ) + 2 ( 0.68 ) = 17.4 m 2

2
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:
For Q = 20 m3 /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

Side slope (assume) = 2.5 : 1 = 21.80° = 21°48'
1
θ = tan -1
2.5
From fig for 8 mm diameter moderately rounded angle of repose
sin 2θ0 sin 2 (21.80) 0.1379

K= 1- 2
= 1- 2
= 1-
sin α sin 24 0.1654
0.1663 = 0.4077
∴ Critical Shear Stress = 0.13 * 47.87 = 6.2231 N/m2

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 m3/s-1
2
⎧ ⎫3
1 ⎪ ( B+my ) y ⎪
( b+my ) y ⎨ ⎬ S0 = Q
n
⎩ ( 2
⎪ B+2 1+m y ⎪
⎭ )
5
1 { ( b0 +2.5 ( 3.385) ) 3.385 } 3
0.00025 = 20
0.025 2
{( b +2 5 ) 3.385}
0
3
5
{( b0 +8.4625) 3.385} 3
= 31.6227
2
{( b0 +15.138)} 3
Solve by trial and error for b.

21.1.7 Economic Aspects of Canal Design
( by + my2 )
5/3
Qn
(i) AR 2 / 3 = 2/3
−
⎡ b + 2y 1 + m 2 ⎤ S0
⎢⎣ ⎥⎦
solve the above equation for y
1/ 4
⎡b / y + 2 1 + m2 ⎤
⎣⎢ ⎦⎥
( Qn / )
3/ 8
(ii) y = S0
( b / y + m )5/ 8
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
( iii ) Cost of bed material Cb = µB t b ( b + 2b ') = Bb + k per unit length
(iv) Cost of side material Cs = µs t s ( 2E + E ') = 2 ⎡( y + FB ) 1 + m 2 ⎤

⎣⎢ ⎦⎥
Therefore C=Cb + Cs = bB + k + 2 ⎡( y + FB ) 1 + m 2 ⎤
⎣⎢ ⎦⎥
Notation :
C = total material cost per unit length,
Cb = material cost per channel base per unit length,
Cs = material cost of sides per unit channel length,
b' = bottom corner width,
tb = thickness of the base material,
t = channel side lining thickness,
µB = cost of base lining material per unit volume,
µs = cost of side lining material per unit volume,
B = cost of base material for specified thickness per unit area,
= cost of side lining material for specified thickness per unit area,
FB = vertical free board requirement,
E = wetted length of the side,
E' = side length of the free board.

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.
(
∂ AR 2 / 3 ) ∂C
∂b = ∂b
(
∂ AR 2 / 3 ) ∂C
∂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
⎛ y⎞ ⎛ y⎞
K1 ⎜ ⎟ + K 2 ⎜ ⎟ + K 3 = 0
⎝b⎠ ⎝b⎠
( )⎡
⎣
⎛ B ⎞⎤
K1 = 20 z 2 + 1 − ⎢1 + 4 ⎜ ⎟ ⎥ 4z z 2 + 1
⎝ ⎠⎦
⎡ ⎛ B ⎞⎤ ⎛ B⎞
K 2 = ⎢1 − ⎜ ⎟ ⎥ 6z z 2 + 1 − 10z ⎜ ⎟
⎣ ⎝ ⎠⎦ ⎝ ⎠
B
K 3 = −5 then,
b 2K1
= 1/ 2
y ⎡ B ⎤
− K 2 + ⎢ K 22 + 20 K1 ⎥
⎣ ⎦
which is a function of z and the ratio of the unit costs of the base to side slope material viz;
B Unit Cost of Base Material

=
Unit Cost of Side Material
Solution Steps
1. S0 , B, , n, z and Q are given. Determine K1, K2, and K3.
2. Estimate b/y for minimum cost using equation.
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.
n/Q
5. Generate the graphs for y Vs b for different values of B/ and for a given value of z.
S0
6. Also study the sensitivity of lining cost to variations of side slope (or side slope ratios).

Sample Run
Data
Q in cumecs, B, and k in R s , FB 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 flow = 0.366 m
⎛ 105 ⎞
K1 = 20 ( 0.25 + 1) − ⎜1 + 4 ⎟ 4 ( 0.50 ) 1 + 0.25 = 8.3189
⎝ 65 ⎠
⎡ ⎛ 105 ⎞ ⎤
K 2 = ⎢1 − ⎜ ⎟ ⎥ 6 (1 + 0.25 ) − 10 ( 0.50 )(1.615 ) = −12.2005
⎣ ⎝ 65 ⎠ ⎦
K 3 = −5 (1.615 ) = −8.075
b 2 ( 8.315 )
=
y 12.20 + ⎡12.252 + 20 (1.615 ) 8.315⎤
⎣ ⎦
References
1. Hager, W.H. 1985, Modified venturi channel. Journal of the Irrigation and Drainage
Engineering, ASCE, 3(1): 19-35.
2. Hager, W.H. and P.U. Volkart, 1986, Distribution channels, Journal of Hydraulic
Engineering, ASCE, 112(10): 935-952.
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,

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.

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

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, ultra-sonic, and electro-magnetic, 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 well-developed 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

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, well-trained team.
The inflow/ outflow method is very sensitive to canal size.

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.
Canal in cutting
Canal in filling Original ground level
Embankment
Original grand slope

Retaining wall
Original ground level

Part in cutting and in filling

braced type

Typical cross sections of Power canals
Covered duct R.C.C

on piles when there is
land slide problem

Side slopes and other salient features
Width Side
Average
'b' in Length slope Depth y Discharge
Velocity S0
in km (m) Q (m3/s)
(m/s)
m m:1
Nangal - - 1.25 : 1 - 354 - -
Sutlej -
9.45 11.8 1.5 : 1 6.26 255.0 2.1 1 / 6666
Beas link
Lower
11.9 15.6 1.5 :1 3.97 127.4 0.665 -
Sileru
Yamuna
Hydel 11.0 - 1.5 : 1 - 200.0 - -
Stage I
Hirakud 51.0 - 1:1 6.3 707 1.97 -
Maximum permissible velocity (safe against erosion)
(1) Stenberg’s formula
Vb = 4.43 2d
in which Vb in m/s, d is the diameter of the particle in m
(2) Bogardi and Yen’s formula
V = 22.9 d 4/9
m γ-1
In which V is the velocity in cms-1, dm is the effective size of particle in cm, γ is the
specific gravity of the particle.
Minimum permissible velocity
1. Ludin suggested for preventing sedimentation (Minimum permissible velocity)
a. V = 0.3 in m s-1 in case of water containing silt.
b. V > (0.3 to 0.5) in m s-1 in case of water carrying fine sand.
2. Kennedy's formula for Minimum permissible velocity
V = C y0.64
In which V is in m/s, y is the depth of water in m, C is a coefficient between 0.54 and
0.70 depending on the size of the silt.
In general, a guideline is 0.6 m/s and a minimum water depth of 1.5 m.

4.572 m Jeepable road
1:3:6 lining
1.524
Stones of medium size 1.524 m
3.96 m
Rock fill
1.524 m 38.4 m
Impervious material
Impervious core
300 mm thick small Stones of small size
300 mm sand layer stone layer
Channel in the hill at Hirakud Project
Recommended Manning N for different conditions
The permissible rugosity coefficient N depending on the discharge capacity in soils
other than rock.
Q m3/s N
0.15 0.03
0.15 to 1.40 0.025
1.40 to 14.10 0.0225
> 14.10 0.02
Gangguillet and Kutter

NR
V= (Chezy coefficient)
R + D0
⎧⎪ 1 ⎛ 0.00155 ⎞ ⎫⎪
N = ⎨ + ⎜ 23 + ⎟ ⎬ s0
⎪⎩ n ⎝ s0 ⎠ ⎪⎭
⎛ 0.00155 ⎞
if D0 = ⎜ 23 + ⎟n
⎝ s 0 ⎠
1 ⎛ 0.00155 ⎞
+ ⎜ 23 + ⎟
n ⎝ s0 ⎠
C=
⎛ 0.00155 ⎞ n
1 + ⎜ 23 + ⎟
⎝ s0 ⎠ R
in which C is Chezy coefficient (MKS) in terms of Manning n.

Recommended Manning n for different conditions
Ashlar and well laid brick work 0.013

Rough brick work, good stone work in fair order 0.015
Rough brick work, good stone work in inferior condition 0.017
Rubble masonry, coarse brick work and masonry 0.020
Canals in earth above the average in order and regime 0.0225
Canals and rivers in earth in tolerably good order and regime 0.025
Canals and rivers in earth below the average in order and regime 0.0275
Canals and rivers in bad order and regime 0.030
Torrents encumbered with detruits 0.050
Safe velocities for different soils
Ordinary earth 0.5 to 1.0 m/s

Firm gravel or clay 1.0 to 1.5 m/s
Broken stone and light pitching 1.0 m/s
Firm conglomerate 1.75 to 2.25 m/s
Sound rock 3.5 m/s
Proportion of bed width to depth
b
= 2 1 + m 2 − 2m
y
1 b
m is the side slope, equal to to 1, the = 1.25
2 y

22.1 Lining the Canals

It is always assumed that seepage losses would be reduced from the theoretical
calculations as between 3.7-1.8 cumec per million square metres (cumec/Mm2) for an
unlined canal in the sandy or clayey loams to 0.6 cumec/Mm2 for a lined canal. In 1988
a field study of the performance of lined distributaries were losing 3.5 cumec/Mm2 and
watecourses were losing 3.7 cumec / Mm2.
Why is there such a gap between assumptions at the feasibility stage and what
happens in practice?
Old earthen irrigation channels in permeable soils can lose a lot of water through
seepage. Large losses through the bed and sides of canal lead to low conveyance
efficiency; that is, (the ratio of water reaching farm turnouts to that released at the
source of supply from a river or reservoir). Earthen canals also get clogged up with
weeds which reduce the water-carrying capacity.
These two factors combine to disadvantage of the tail end farmers. Therefore Unlined
canals are inefficient, inadequate from the point of view of equitable performance.
In Punjab, the expected saving by brick lining is of the order of 20% . The brick linings
have 25 years of life.
Total losses from unlined watercourses are known to be more than those from the main
system, but they dont get the same attention during a lining programme. Lining
programmes are divided into main system lining and watercourse lining. The main
system canals (main, distributaries, and minors) are large channels supplying several
watercourses.
A typical value for the seepage rate in an unlined channel in clayey loam is 1.8 m3/s per
M m2 and through a rendered brick line water course or canal reduces to 0.1 m3/s per M
m2. However, if the impervious canal lining has few crack what would happen? The
experiments indicate it as good as that of the original unlined case when the ground
water level is high.
Inspite of the above observation, lining can significantly reduce conveyance losses.
Lined channels have a smaller surface area for a given discharge than unlined

channels. Typically a lined channel will have 40% of the unlined surface area for a given
discharge. Therefore even at the same loss rate per unit area there will be a saving in
water. When estimating the reduction in losses from a lining programme, this should be
based on the combination of a reduced cross-section and a reduced seepage rate per
unit area.
In the Indian Punjab, measurements on lined distributaries and watercourses between
four months and seven years old showed that seepage rates from the distributaries
rapidly became comparable to seepage rates from unlined canals, whilst seepage rates
from watercourses were highly variable.
Lining of the distributary canal seemed to have had a beneficial effect on the equity of
supply between watercourses. However, the effects of lining watercourses is still to be
established.
A sample of 15 out of 130 watercourses were tested on the 30,000 ha Mudki subsystem
of the Sirhind Feeder in Ferozpur district using ponding tests and inflow-outflow
methods. The mean and variability of seepage losses increased dramatically for lining
more than four years old. Some lining older than four years performed as well as new
linings, with losses as low as 0.4 m3/s per Mm2 but others has losses of up to 11.5 m3/s
per Mm2. Overall conveyance losses were significantly greater than seepage losses
alone.
The variability of conveyance losses was observed to be related to the condition of the
channels. Losses from raised watercourses with cracked or broken linings appeared as
surface leakage causing waterlogging of adjacent fields and localised crop damage.
This was due to poor quality control during construction, particularly earth compaction
behind sidewalls, and a lack of subsequent maintenance.
The design life of concrete and brick lined channels is generally assumed to be 25-50
years. Major repairs of lined channels are sometimes required within a few years of
construction. At the Kraseio Scheme in Thailand, completed in 1981, long lengths of the
concrete lined main canals have needed to be replaced each year.
In one 26-year lining test, a complete repair of the drained test channel was required
every 22 months.

There is no watertight case for or against lining.
If lining goes ahead, a high standard of construction is essential, especially of water
courses which must withstand a great deal of wear and tear. Without adequate
supervision, poor construction of channels will lead to reduce life and higher
maintenance costs.
Earthen watercourse in the Bikaner area of Rajasthan discharge an average 28.3 l/s,
but begin to seep and leak badly after little more than a year. These cracks and the slow
movement of shallow water favours developement of thick aquatic weeds, which
encourages the drying and the cracking process, and strcuturally weaken the banks.
This obviously adds significantly to the cost of maintenance.
The cracks opened in dry periods do not close fully when saturated by water flows, and
losses can be up to 25% of the water diverted into the system.
The cycle of swelling, heaving, shrinkage and settlement leads to progressive bank
deterioration. Shear strength of clays depends on cohesion between particles. In a
newly-formed compacted clay masses the interparticle cohesion is high. On first drying,
the cracks appear and close up again on wetting, but do not regain their original
interparticle cohesion. This means a reduction in shear strength after a few drying and
wetting cycles.
The reinforced concrete lining would reduce the seepage loss drastically and has lower
operation and maintenance costs.
Reference
World Water, April 1989, pp 16, 19, 21.
Appendix (i)
Table - 1 Salient Aspects of Seepage through Lined Canals [7]:
Sl.No State Adopted practice

1 U.P. Ql = 1 / 200 ( B + D )
2/ 3
Practice
Ql = Seepage losses in cumecs/km length B and D are bed
width and depth.
2 Punjab & Kl = 1.9Q1 / 6
Haryana
K l = Seepage loss in cumecs / 106 sq.mof wetted area.
Q = Main discharge in the canal.

Bhakra Lined canal Kl = 1.25Q 0.056 (F.P.S)

3
Nangal Unlined canal ql = 0.00928Q 0.5265 (M.K.S)
4 Andhra lined canal: Ql = 0.60 cumecs / 106 m 2 of wetted surface
Pradesh
Unlined canal: Ql = 1.85 to 2.40 cumecs / 106 m 2 of wetted
surface and 20% extra for distributory system.
5 CBIP Main canal - 7% of head discharge
Distributory and minors - 8% of head discharge
Water surface - 20% of head discharge
Table-2 Salient Aspects of Seepage through Lined Canals (After Yu. M. Kosichenko [8]):
Type of lining and

Sl.
character of Permeability Kl m/s seepage loss Q m3/sec/m2
No.
disturbance
I Concrete-film lining
i. Smooth 0.705 * 10-9 / 0.141 * 10-7 0.334 * 10-4 / 0.685 * 10-5
ii. Crack with rough
0.652 * 10-9 / 0.136 * 10-7 0.318 * 10-6 / 0.665 * 10-5
walls
iii. Silted cracks 0.137 * 10-9 / 0.156 * 10-7 0.318 * 10-8 / 0.76 * 10-7
iv. Without cracks 0.590 * 10-11 / 0.114 * 10-9 0.228 * 10-8 / 0.556 * 10-7
II Concrete lining
i. Smooth 0.225 * 10-8 / 0.331 * 10-6 1.243 * 10-6 / 1.516 * 10-4
ii. Crack with rough
0.867 * 10-9 / 0.202 * 10-6 0.423 * 10-6 / 0.982 * 10-4
walls
iii. Silted cracks 0.706 * 10-10 / 0.752 * 10-10 0.344 * 10-7 / 0.367 * 10-7
iv. Without cracks 0.694 * 10-10 0.388 * 10-7
III Soil film lining
i. Protective layer of
0.981 * 10-9 / 0.189 * 10-7 0.564 * 10-7 / 1.087 * 10-6
local soils
ii. Protective layer
0.221 * 10-11 / 0.425 * 10-10 0.127 * 10-9 / 0.245 * 10-8
of compacted clay
IV Soil Membrane 1 * 10-8 0.575 * 10-6
V Unlined channel - 0.295 * 10-3
Plastics as a lining material
Introduction
Water suitable for human consumption, livestock and irrigation is limited in many parts
of the world. Since precipitation is the only source of all fresh water and since it varies
from place to place, from year to year and from season to season, water must be
collected, stored and transported, if it is to be available when and where it is required.
The loss of irrigation water in a canal system occurs during its conveyance through
canal, sub-canal, distributary, minor, water course and finally during application in the

field. In a study made by the Central Water and Power Commission (1967), the losses
during the various phases of water conveyance was found to be as shown in Table
Water use efficiency in percentage

System Number of Water Field Total Net
Characteristics Canals Distributaries courses Losses Loss Utilization
Entire system
15 7 22 27 71 29
unlined.
Only Canal
4 7 25 30 66 34
lined.
Canal and
Distributary 4 2 26 32 64 32
lined.
Whole system
4 2 6 42 54 46
lined.
It may be observed that, in a completely unlined canal irrigation system the loss of
irrigation water is as much as 71%. This loss is attributed to both evaporation and
seepage losses; evaporation losses are a function of temperature, humidity and wind
velocity. It is not practicable to prevent evaporation loss (in some reservoirs this loss
has been prevented by floating a thin PE film on the water surface) in running water.
However, seepage losses could be effectively minimised by the use of an impervious
medium between the porous soil and the water flowing in the system.
Conventional materials like clay, tiles, cement-concrete, have been used for prevention
of seepage losses. Most of the conventional methods are either too expensive or not
very effective.
Table 2 shows the state wise break-up of irrigation resources and the length of major
irrigation canals operating in each state.
Irrigation and Power Research Institute at Amristar has particularly conducted intensive
research in this field and a suitable lining technique has been evolved for existing
earthen canals and channels.
This technique is known as "Combination Lining" and in this, a low cost Polyethylene
film is laid at the bed of the canal or distributary whereas the sides are lined with pre
cast brick-tile blocks. This method can be conveniently adopted for lining distributaries
during short closures. This technique has many advantages over the other methods of
lining such as laying new lined parallel channels which involve construction of new

bridges, regulators, falls and outlets. Moreover, in case of combination lining, the
existing section, regime and stability of the channel are not changed. Irrigation supplies
are also not interrupted as lining is done during short closures.
Table 2 Statewise distribution of Irrigation area under major irrigation canals.
Name of Average Net Percentage of net area irrigated in Length

State/ Unionrainfall irrigated 1970-71 by various sources of major
Territory in cm area canals
'000 in km
hectares Canals Tanks Wells Others
1970-71
Andhra 900 3313 48 34 15 3 1,242.8
Assam 240 572 63 - - 37 137.2
Meghalaya 241 37 - - - 100 -
Nagaland 193 12 - - - 100 -
Manipur 193 65 - - - 100 -
Sikkim 355 NA NA NA NA NA NA
Tripura 193 22 - - - 100 NA
Bihar 132 2160 38 8 25 29 906.2
Gujarat 81 1209 17 2 80 1 2,437.2
Haryana 76 1532 62 - 38 - 1,098.3
H.P. 184 91 - - 1 99 -
J&K 102 279 98 - - 2 32.0
Karnataka 104 1137 37 32 23 8 2,924.6
Kerala 267 431 49 17 1 33 491.9
M. P. 123 1480 48 9 38 5 238.4
Maharashtra 106 1427 22 16 57 5 2,546.1
Orissa 153 1149 23 51 4 22 613.7
Punjab 64 2888 45 - 55 - 501.2
Rajasthan 46 2132 35 13 51 1 869.4
Tamil Nadu 97 2592 34 35 30 1 561.4
U. P. 117 7190 48 34 15 3 1,683.5
W. B. 174 1489 65 20 1 14 1,666.0
All India 161 31207 40 14 38 8 17,949.9
NA- Not Available (a) Includes Harayana. (b) Included in Punjab. * Source : Irrigation
Commission (1972).
Nearly 30 to 80 percent cost saving could be achieved by using combination type of
lining as compared to double tile lining. The behaviour of this lining laid in different
channels in Punjab has been studied and it has been found that this low cost lining is
durable, efficient and quite suitable for lining the existing distributaries.
Gujarat State Irrigation Department is also using Combination Lining. To render the
canals seepage proof, a 400 gauge (100 microns) thick black LDPE film is used to line

the bed of the canal which is then covered with single layer brick masonry. For further
reduction of cost, soil cover varying from 60 cm to 90 cm could be tried.
A similar trial was also carried out in Azamgarh District of Uttar Pradesh by the UP State
Irrigation Department. In this trial, 600 gauge (150 microns) PE film was used in the
canal bed overlaid with 45 cm soil cover.
The advantages of Polyethylene (PE) film lining are:
(a) Cost factor is very much in favor of PE lining in comparison with conventional lining
methods.
(b) The PE film lining has the advantage of being a superior moisture barrier than any
other construction material including cement-concrete, tiles, tar-felt, etc. Irrigation and
Power Research Institute (IPRI), at Amristar and Central Building Research Institute at
Roorkee have conducted extensive trials on this type of lining. Permeability tests on this
type of lining. Permeability tests by IPRI showed that a 400 gauge PE film subjected to
a hydraulic head of 3.65 meter has stayed water tight for about 15 years.
(c) Another advantage of this type of lining is the speed with which it can be laid at site.
By using this type of lining, the total time involved in lining work can be substantially
reduced, thus facilitating wider coverage during short closure of canals.
The Anand Irrigation Division, Gujarat, have found that with double tile lining, the work
progress is around 30 m of canal length a day (15 m canal width) whereas for the
combination type of lining using PE film with a cover of single tile masonry, the progress
is as much as 90 m. This could be still further increased by using LDPE film lining with
60 cm to 90 cm soil cover.
Distributaries:
A Canal Irrigation system comprises of a network of Canals, Sub-Canals, Distributaries,
Minors, with water courses forming the last link. In Government Canals, the jurisdiction
of the Irrigation Department is generally up to the minors. The water courses fall under
the jurisdiction of the individual cultivators.

Use of Plastics for Lining of Canals

Introduction
Canal lining with LDPE film was undertaken in the country in 1959, as a measure to
reduce seepage losses. Punjab and Gujarat States were first to use this type of lining
during 1959 to 1962 followed by other states subsequently.
The Low Density Polyethylene (LDPE), Polyvinyl Chloride (PVC), High Molecular High
Density Polyethylene (HDPE), Linear Low Density Polyethylene etc., are the plastics
used for canal lining. In India, mostly LDPE film has been used for canal lining.
The characteristics of the plastic film are:
Light in weight, impermeable to liquids and gases, chemically inert, flexibility, resistance
to microbiological attacks, ease of transportation and installation etc.
Black LDPE film lining over laid with soil cover or protective hard cover had been widely
used during the last three decades. IPCL's Petro-Chemical Industry in Maharashtra, will
be able to provide the adequate supply of this plastic film for lining of canals.
The LDPE films presently available are too smooth and glossy to hold mortar below
blocks, tile or brick lining indicating the necessity of developing rough or serrated films.
The purpose of the canal lining which helps in reduction in seepage to save water for
additional irrigation, and to increase the velocity for reduction in canal sectional area
due to smooth lined surface. Lining with bricks tiles etc., is fraught with the problem of
seepage through the joints, cracks etc. This naturally leads us to the use of
impermeable plastic film to use as a cut-off. The LDPE film is used with soil cover/ C.C.
Cover/ P.C.C. Cover. C.C. in-situ is less costly and suitable for bed but not for sides.
P.C.C. ensures quality control in case of canal lining on large scale. It is costly but
suitable for both bed and sides. In using P.C.C. pointing needs special attention. LDPE
film is susceptible for puncturing by uneven soil surface, impact of feet, weed growth
etc. A layer of sand would help as a leveling course. Sand with a little silt content or
sand wetted would stand well on slopes. If the surface of the film is rough, it would keep
the mortar in position. The hard cover over the film is meant to be the load over the film
to keep it in position and also to be smooth surface for better flow conditions.

Vandalism with regard to LDPE film is due to lack of awareness of the farmers as to the
value of water. Burrowing by rodents is a menace. The easy and practicable solution is
a sand layer. A burrow does not stand in sand, and the hole gets closed.
Lined canal costs 20 - 25% more relative to cost of unlined canal. If LDPE film is used,
the cost of lining would be 25 - 30% (an increase of 5% due to film cost). This additional
cost due to canal lining using LDPE film is rather a wise investment as the benefits of
water saved, additional irrigation and food production, and reduced canal section would
more than offset that additional cost.
Therefore canal lining is obligatory and the use of LDPE film is bound to spread once its
efficacy commensurate with economy is realized.
Films of different thickness varying from 100 to 250 micron are used in bed or on sides.
Cement concrete cover over LDPE has been used on canals as steep as 1.3:1 on Ravi
Canal (J&K), and 1:1 on Malaprabha canal.

F.S 1 1
1 1
LDPE Film
Free board
LDPE Film
Single tile lining

in Cement Mortar
LDPE Film
LDPE Film
Lining of Canal with LDPE film

Reference
"Proceeding of the workshop on Use of Plastics for Lining of Canals", Central Board of
Irrigation and Power, New Delhi, Vadodara, 7-8, May 1987
Venkatasubramanian K.N, Kamal Nanavaty, Prasanta Misra - Role of plastics in water
management published by Indian Petro chemicals Corporation Limited, Government of
India undertaking, august 1968.

Table 3 - Performance of different Canal Linings
The details regarding the design, specifications and performance of various types of
linings as provided in the case of the following projects.
S.No. Canal Type of lining

1 Nangal Hydel Channel (Punjab) Concrete and Tile Lining.
Sundernagar Hydel Channel, Beas Project
2 Cement Concrete.
(Punjab)
Concrete, Brick and
3 Yamuna Power Channel (U.P)
Concrete Tile.
4 Gandak Canal (U.P) Brick Tile Lining.
5 Lower Ganga Canal-Link Canal (U.P) Sandwiched Brick Lining.
Cement Concrete Tile
6 Kosi Feeder Channel (U.P)
Lining.
Tiles Lining with plasting,
7 Rajasthan Feeder Channel (Rajasthan) Single tile lining in bed and
double tile lining in the side.
Banaskantha Left Bank Main Canal of
8 Cement Concrete Lining.
Dantiwada Project (Gujarat)
Banaskantha Left Bank Main Canal of
9 Brick Lining.
Dantiwada Project (Gujarat)
Sandwiched Brick Tile
10 Mahi Right Bank Canal (Gujarat)
Lining.
11 Shetrunji Canal Project (Gujarat) Masonry Lining.
12 Shetrunji Canal Project (Gujarat) Lime Concrete Lining.
13 Karad Project Canal (Gujarat) Precast Blocks.
Cement Concrete and
14 Tungabhadra Project (A.P)
Rubble Masonry.
15 Krishanarajasagar Canal (Karnataka) Cement Concrete Lining.
Krishnarajasagar, Right Bank Low Level Canal
16 Soil Cement Lining.
(Karnataka)
Krishnarajasagar Right Bank Low Level Canal Lime Surkhi Concrete
17
(Karnataka) Lining.
In situ Cement Concrete
18 Manniaru Canal (Tamil Nadu)
Lining.
Cement Concrete and
19 Pattanamkal Main Canal ( Tamil Nadu)
Metal.
Cement Concrete and
20 Rethapuram Channel (Tamil Nadu)
Metal.
Random Rubble Masonry
Left Bank Main Canal, Pamba Irrigation Project
21 and Cement Concrete
(Kerala)
Lining.
Precast Cement Concrete
22 Left Bank Canal, Kuttiady Irrigation (Kerala)
Slabs.
Right Bank Main Canal, Peechi Irrigation Rubble Masonry in Cement
23
Scheme (kerala) Mortar.
24 Left Bank Canal (Kerala) Precast Slabs and Blocks.
25 Kuthanur Branch (Kerala) Bitumen and Cement.

KERS, Krishnarajasagar had tried HMHDPE film (Tuflene) covered with 1:4:8 or 1:5: 10
Cement Concrete in a length of 140 m on 38.4 km distributory of Visveswaraya Canal
and reported about 90% reduction in seepage losses when compared with unlined
reaches. Likewise Fibre glass Reinforced Plastic Tissue Asphaltic Canal Liner had been
experimented.
Backing Material for Swelling Soil

I.I.T, Bombay after conducting field studies with various types of backing materials in
swelling soil (especially black cotton soil) area in Karnataka has suggested that a 1 m
thick cohesive non-swelling (CNS) layer covered with P.C.C. slabs would suffice the
need for lining in swelling soil. The minimum specification suggested for backing
material is
Gravel - Less than 10 percent,
Sand - 30 - 40 percent
Silt - 45 - 50 percent
Clay - 20 - 25 percent
Compaction - 96 percent of standard proctor density, in layers not exceeding 20

cm.
Moisture content - OMC ± 1.5 percent.
Shear Parameters - 0.2 kg/cm2 25° to 0.5 kg/cm2 for 2 to 15°
KERS, Krishnarajasagar had conducted experiments with different lining materials, viz.,
(1) Size stone pitching over 22.5 thick layer of murrum backing, (2) Rough stone
pitching over 15 cm thick layer of murrum backing, and (3) 15 cm, 22.5 cm and 30 cm
murrum lining without any protective cover for small channels only.
The size stone pitching and rough stone pitching have functioned well and have
recommended size stone pitching for small channels in view of lower rugosity coefficient
and lesser maintenance cost though it is slightly costlier initially.

PCC lining with

1 m thick CNS layer
drainage arrangement
IP 2.43
3.04
3.04
3.04
2.74 3.04
1.82 1
15 cm thick
CNS layer 2.74 FSD
1
2 __
2
Typical section in deep cutting of Malaprabha Right Bank Canal kilometer number 35.354
2.103 m
PCC lining
FB 3.04
3.04
3.65
3.04
2.74
15 cm thick
CNS layer 2.74 Full Supply Depth
1 m thick CNS layer
Typical section in cutting at MRBC kilometer number 35.354 (a) Canal section in cutting
(a) Canal section in cutting
PCC slab lining 2:1

1 m thick CNS
FSD 2.74 IP 2.43
Black Cotton soil
SR 4.26
30 cm thick CNS
2 m and below Rock toe

1m
B
__ L
+ 2y 30 cm thick sand blanket
30 cm thick CNS 2
2.74
Bed filling 2 m and below
PCC slab lining 2:1

SR 4.26 1 m thick CNS 2.43 Black Cotton soil
30 cm thick CNS
60 cm thick CNS
2.74 FSD
2.74 Rock toe
below 2 m bed filling 2 m and above
1m
B
__ L
30 cm thick sand blanket + 2y
2
60 cm thick CNS
Bed filling 2 m and below
(b) Canal section in embankment
Construction of Malaprabha Right bank Canal in expansive soil area using Cohesive
Non-swelling Soil (CNS) layer of 1 m thickness as suggested by IIT-Bombay

Geo synthetic Reinforced Canal Lining

Currently, in India the lining systems adopted generally consist of three components.
The top is usually a lining cover of either concrete slabs or brick tiles. Sometimes
flagstone or soil as cover has also been used, depending on suitability. Below the cover
a LDPE lining is provided underlain by a graded sand filter normally placed to account
for planned drainage. Film of 175 micron thickness has been used in the bed lining of
Indira Gandhi Main Canal in Rajasthan. The thickness of the LDPE film depends on
whether it is the primary water barrier constituent of the lining (in that case the thickness
varies from 150-250 micron) and the rigid cover is provided to protect it against damage
and heaving. The rigid cover may also be provided as the main lining and the film is
provided as a secondary back up specification (generally 100 micron film is provided for
this purpose). Figure shows typical details of LDPE lined canal system used in India.
A
B
C
D
E
Details of Sandwitched Brick Lining
Index:
A = Top layer of 50.8 mm Flat Brick (C.H.1:3)
B = A layer of (1:3) Red Cement Mortar 6.35 mm thick
C = 15.875 mm layer of (1:3) Red Cement Mortar
D = 50.8 mm thick Flat Brick layer with joints of Red Cement Mortar of proportion (1:3)
E = 9.525 mm thick Red Cement Mortar layer (1:5)

FLAGSTONE LINING 40 to 50 mm thick

on a bed of cement mortar 1:6 and
pointing with 1:3 cement mortar
TRANSVERSE SLEEPERS
.30 m centre to centre in
cement concrete 1:3:6
with 10 mm metal rod
100 MICRON LDPE

FILM AND GAUGE
20 mm THICK BEDDING
IN CEMENT MORTAR 1:6
FLAGSTONE LINING
LDPE 600/800/1000 Gauge
Brick lining 76.2 mm thick

in cement mortar 1:3
Dressed sub grade
Brick Lining

Tenax Mint 100

Composite membrane
as laid in profile
HDPE Membrane
30 mm thick concrete
size in flush point or
brick lining for
counter weight Geotextile filter
Graded soil slope
Tenax Mint 100

Composite liner Perforated drainage pipe
and Drainage wrapped in Geotextile
(Pressure relief drainage
with laminar drainage path)
SECTIONAL VIEW SHOWING DETAILS OF CANAL LINING
It has been ascertained through field seepage tests that lined canals save a
considerable amount of water loss as shown in Table
Percentages savings in
Average seepage loss water due to lining
State
m3/s/million sq.m compared with unlined
canal
Haryana
a) Unlined canal 2.25 -
b) Conventional lined canal
0.43 81
without LDPE film
c) Lined Canal with LDPE
0.16 93
film 2.5 mm
GUJARAT
a) Unlined Canal 2.50 -
b) Lined Canal without
1.00 60
LDPE film
c) Lined Canal with LDPE
0.80 68
film

WEST BENGAL
a) Unlined Canal
i) in rocky strata 2.94 -
ii) in lateritic strata 20.00 -
b) Conventional lined Canal
0.30 rocky strata 90
without LDPE film
c) Only LDPE film with soil
0.12 rocky strata 96
cover
Seepage studies - Ponding method Semi field Studies - KERS
Minimum value of
Sl. No. Type of lining material seepage after two years
in m3 / s / M m2
1 Tuflene with brick in cement mortar cover. 0.0008
2 Tuflene with earth cover. 0.0057
Tuflene with concrete cover (1 : 4 : 8
3 0.0008
proportion)
Before reconstruction 2.34 to 6.63 m3 / s / M m2
After reconstruction 1.53 m3 / s / M m2
LDPE 400 gauge (0.1 mm) Fibre glass Reinforced Plastic Tissue
LDPE 1000 gauge (0.25 mm) Reinforced asphaltic canal liner (2.5 kg / m2)
HMHDPE 700 gauge (0.1 mm)
Inspite of the advantages offered by conventional lining system with and without LDPE
film lining, they have several shortcomings. For example, in case of lined canals with
PCC concrete tiles as cover complete impermeability is not ensured as construction
joints are a must between two concrete panels, where the LDPE film below serves the
lining purpose. The ill achieved compaction of side slope of a canal in most cases give
rise to local shear failure due to lining overburden, with excessive slope settlement. In
such case of slope stability failure the unreinforced concrete slab cracks, consequently,
the unreinforced weak LDPE liner also gets damaged. LDPE membranes do not
possess suffficient strength and hence under sub grade reaction, or dead load and live
load on cover or hydrostatic pressure, the membrane gets ruptured on many occasions.
Under steep slope conditions and high flow discharge in the canal stringment reinforced
cover specifications are required which often makes the construction process of the
lined canal very tedious and time consuming.

Geo synthesis for Irrigation Structures

A wide range of Geo synthetic products find application in irrigation structures. They
include geo membrane, geo textiles, geo composites, geo grids, geo drains, gabions
and mattresses. Gabions are constructed using bi-oriented geo grids.
Geo synthetic Canal Lining and Drainage

As a fabric layer this system is extremely easy to install. The lining geo composites are
found in combinations of membrane-grid-textile, membrane-grid or grid-textile
sandwiches which can be used depending on the requirement of the project.
In case of the canal lining systems Geo composites having HDPE membrane-grid-
textile combination is the most suitable. HDPE membranes composites now available in
India, eliminates the shortcomings of the LDPE lined canal systems. The grid eliminates
the requirement of graded filter for drainage resulting in considerable reduction of time
and costs, increasing canal capacity of discharge. Drainage Geo nets, a few millimetres
thick have the same capacity of drainage water as a graded sand layer a few hundred
millimetres thick. The textile layer provided in these composites prevents the fines from
interfering with the drainage path. The Geo composites are used in canal lining system
with the water proof membrane in contact with the cover (required to weigh down the
geo composite on the slope) and the geo textile is laid against the well graded slope.
Geo composites are available in light weight roll form which can be easily laid on slopes
as well as on prepared canal beds. Geo composites have strength ranging from 6.5
kN/m to 23 kN/m which can adequately resist soil pressure on slopes and distribute
concentrated forces preventing local shear failure conditions from developing. The Geo
net with its high drainage capacity prevents high pore pressure development on the
slope and provides an efficient flow line.
Saving Geo synthetic lining vs Conventional lining 19%.
Reference
Som, S. Sarkar and Ranjana Majumder, Geo synthetic Reinforced Canal Systems and
Irrigation Structures, Proceedings Water Energy 2004, International R&D Conference
1995, New Delhi, India, pp 262 to 274.

22.2 Seepage Prevention with Impermeable Membranes

The use of flexible, impermeable membranes to prevent the seepage of water and
wastes is increasing, with application to a wide variety of seepage situations. The
materials used as membranes are synthetic, organic, high polymers in the form of
relatively thin, flexible, impermeable sheeting. Most used are plasticized polyvinyl
chloride (PVC), polyethylene, and Butyl rubber. They all possess low permeability
which, when combined with the strength of soil or other base material, will prevent
seepage of water or aqueous solutions. While the membranes differ somewhat from
each other, they differ completely from rigid materials such as concrete, asphalt and
planking in that they are not expected to contribute structural strength- only
impermeability.
The earth itself will generally support high hydrostatic loads, but the sheeting employed
as a membrane is often required to follow earth movements and settlements while
retain-ing its impermeability. This adds several other considerations to that of
impermeability. The membrane must resist puncturing during Installation, and must be
able to adapt to soil deflections without tearing or puncturing. Vinyl (PVC) has the
highest puncture resistance, with Butyl in second place. Polyethyelene ranks as poor in
this respect. Both vinyl and Butyl have high elastic strain and will yield with soil
deflections. Polyethylene does have a high elongation, but it becomes thin in localized
areas under stress and hence is less desirable than the other sheetings is formulated to
meet specifications for long life and resistance to soil attack.

Typical properties of these three membranes are:
Tensile strength, Elastic strain, percent, Puncture resistance, Available gauges,
Available widths, Method of joining, Weight, Density, Cost, Installation.
Installation of liners for preventing seepage is quite simple after the earth has been pre-
pared. In designing structures to utilize membranes, a side slope of at least 3:1 is
necessary if the liners are to be covered for protection. For uncovered membranes, a
normally stable slope can be used. In both cases, all sharp rocks, stones, roots, and
other.

22.3 Failure of Canal Lining

In connection with the mini-hydel power plants being constructed on the Gunrur Banch
Canal near Nekarikallu, AP, it was proposed to raise the water level in the canal by
about 2.5 m over the previous level. The canal banks were raised by about 2 m during
the summer of 1996. The canal banks were lined with 100 mm thick concrete panels.
The slope of the inner canal bank is 1:1 whereas the slope of the outer bank is 1.5 : 1
(H:V). When the water level was raised in the canal, the canal banks have exhibited
some instability and seepage through the banks and during a thunder storm, the canal
linings over some portion have slipped away from the banks. When the water level was
raised and lowered, the right bank has shown some distress. In order to minimise the
seepage, a secondary berm was constructed along the right bank at two locations. The
secondary berm has sunk into the foundation soil.
0.6 m
FSL
0.15 m
100 mm thick RCC 1:2:4 Lining 1.5 m
G Level Backing
BED LEVEL
100*300 Leg Beam
0.5 m
0.6 m 300*300 Model Section
0.6 m
24.995 m




Case:1 Q = 62 m3/s
109.345
Q = 62 m3/s
RL 113.7 0.969
60 cm
1.25 0.969 m
1.2 m 1.3 1.00
1
1.211.169 1.00 0.969
RL 112.5 1/1250 60 cm
2.1 m
2
75 cm
1/ 800
0.76 m 2.75 1.65 m
800 m 3
1/1500
2000 m
Water Surface profile

Case:2 Q = 27 m3/s
109.345
Q = 27 m3/s
0.562
1.00 0.562 m
1.2 m 0.75
1 1.25
1.17 1.00 0.75 0.562 m
RL112.5 1/1250
2.1 m
2
75 cm
1/800
0.4045 m 4.7
1.65 m
3
800 m
1/1500
2000 m
Water Surface Profile
Weir at the downstream end removed

Q = 54 m3/s
106.58 105.85
2.741 m 2.34 m
Draft Tube
102.853
500
30.5 m
Water Surface Profile

600 mm φ Pipe
Flow
397 m 70 m
Slope 1: 1.5
Right Bund
Slope 1:1
By-Pass 600
900
BL Q = 54 m3/s
Flow BL + 103.84 Flow
BL + 106.58
+ 103.767 900
Power 600
House
Slope 1:1
Bund 6m wide
Side Slope 1: 1.5
Section
Pit Details
106.58 600 mm 150 mm
750 mm
900 mm 600 mm
A Pit of
900*600 mm
600 150 mm
Plan View of the Pipe Line for supplying water for Irrigation

22.4 Most efficient Hydraulic section

For a given cross section determine the hydraulic section.
Hydraulically Best section (Hydraulically Efficient Section)
1. Rectangular Channel:
P = b + 2y
A
b=
y
A = by
A
P= + 2y
y
dp
= − Ay −2 + 2 = 0 , A = 2 y 2
dy
∴ by = 2 y 2
or b = 2 y
b = 2y
Hydraulically efficient rectangular
channel is half of a square.
2. Trapezoidal Section:
P = b + 2 y 1 + m2
A
A = ( b + my ) y or b = = − my
y

A
P= − my + 2 y 1 + m 2 (For a given area of flow)
y
Differentiate with reference to y assuming A and m to be constant.
dp
= − Ay −2 − m + 2 1 + m 2 = 0
dy
Substituting for area, the above equation can be rewritten as
− y −2 ( b + my )y − m + 2 1 + m 2 = 0
b + my + my
+ = +2 1 + m 2 = 0
y
b + 2my
= y 1 + m2
2
Half the top width = side slope distance (for given side slope)
b = 2 y 1 + m − 2my = 2 y ⎡ 1 + m 2 − m ⎤
⎢⎣ ⎥⎦
Substitute this value of b into the equation A and P and simplifying
P = 2 y ⎡2 1 + m2 − m⎤
⎢⎣ ⎥⎦
A = y 2 ⎡ 2 1 + m2 − m ⎤
⎢⎣ ⎥⎦
0.5
⎡ A ⎤
y=⎢ ⎥
⎣⎢ 2 1 + m − m ⎦⎥
2
Substitute the value of y into P
P=2
A1 / 2
(2 1+ m − m) 2
(2 1+ m − m)
0.5
2
(
P = 2 A ⎛⎜ 2 1 + m 2 − m ⎞⎟
⎝ ⎠ )
which is the m value that makes P least?
D.w.r to m and equate it to zero
dP
=0
dm

3
m= = 3
3
y 1 3
tan θ = = =
my m 3
∴θ = 60
2
∴b = y
3
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.
m= 3
60 60
2
b = __ y
3
Half Hexagon - inscribed circle
of radius equal to depth is
tangential as shown in figure

Triangular Section: If θ is half angle
⎡ 2a sin θ ⎤
Area = ay sin θ ⎢∵ y = ay sin θ ⎥
⎣ 2 ⎦
ay sin θ a cos θ sin θ a
R= = = sin 2θ
2a 2 4
⎡ y ⎤
⎢⎣∵ a = cos θ ⎥⎦ a sinθ
a a
θθ
Hydraulically efficient channel

Half a square on vertex
R should be max.
dR 2a
= 0, cos 2θ = 0
dθ 4
∴θ = 45
o o
45 45
Free surface width is equal to the diagonal
Half Square on its apex
It is a half square resting on its apex and maximum width is equal to diagonal.
Alternative derivation for Triangular Section:
A = y 2 tan θ
y = A / tan θ
P = 2 y 1 + m2

l
y
m θ
= 2 y sec θ
A
=2 sec θ
tan θ
dp d ⎧⎪ A ⎫⎪
= ⎨2 sec θ ⎬
dθ dθ ⎪⎩ tan θ ⎪⎭
dp ⎡ sec θ tan θ sec3 θ ⎤

=2 A⎢ − ⎥=0
dθ ⎢⎣ tan θ 2 ( tan θ ) ⎥⎦
3/ 2
sec3 θ
sec θ tan θ − =0
2 ( tan θ )
3
2 sec θ tan θ − sec3 θ = 0
∴ Solve for θ
∴ θ = 45

Hydrostatic Caternary (Linteria)
Equation for hydrostatic catenary is given by

y ⎡⎛ 3 2 15 4 ⎞ ⎛3 5 ⎞ 5 4 ⎤
x1 = ⎢ ⎜1 − k − k ⎟ φ + ⎜ k 2 + k 4 ⎟ sin 2φ − k sin 4φ ⎥
2k ⎣⎝ 4 864 ⎠ ⎝8 32 ⎠ 256 ⎦
y1 = y cos φ
x1 , y1 are measured from mid point of the surface
θ
k= sin ; θο = slope angle at the point x 1 y 1 . θο varies from 0 at the
2
bottom of the curve to θο at the ends.
⎡ φ⎤
⎢ sin ⎥
φ = sin −1 ⎣
2⎦
, θ is slope at any point ( x, y)
k
For the hydraulically efficient channel
θο = 35° 37' 7" , y= 3.5 m. Find A, R, D, Z at full depth.
Also plot the cross section of the channel
θο 35° 37 ' 7 "
solution: = = 17°48 ' 33.5 "
2 2
θο
k= sin = 0.30585
2
φ φ
sin sin
sinφ = 2 or k = 2
k sinφ
φ
sin
2 1
k= =
φ φ φ
2 sin cos 2 cos
2 2 2
φ 1
or cos =
2 2k
φ
y1 = y cos φ φ = 90 cos = cos 45 ∴ k = 0.707
2
A
A = 17.0992 m 2 P= 10.443 m R= = 1.6374 m
P
T = 6.7114 m D = 2.5478 m Z = A D = 22.293 m5 / 2
y1 (m) 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

φ ( in
90 81.78 73.39 64.62 55.15 44.4 31.0 0
deg )
k 0.7071 0.6614 0.6235 0.5916 0.564 0.5400 0.51887 0.5

Exercise: Plot the graph using the above data
x1 x
y1 y
2 A
A = 17.0992 m P= 10.443 m R= = 1.6374 m
P
5/2
T = 6.7114 m D = 2.5478 m Z = A D = 22.293 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
T
( )
A = y 2 m 2 cot −1 m
P = 2 y ( m + cot m ) −1
1
A r y
r r m y 1 R= = =
θ0 P 2 2
θ0
T = 2r (1 + m )2
cot θ0 = m
__ m
1
θ0 = cot-1m
Hydraulically efficient sections could be derived using Lagrange Multiplier approach.

22.5 Design of Unlined Channels

In 1895 Kennedy evolved an equation for V0 , nonsilting and non-scouring 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.
V0 = 0.546 y 0.64 ( 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 (m3s-1), the surface
width B (m) and the depth of flow y (m):
V = 0.57 y 0.57 ( in mks )

V = 0.2646 B 0.335 ( in mks )
B = 7.86 y1.61 ( 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:
P = 4.825 Q1 / 2 ( in mks )
R = 0.4725 Q1 / 3 / f 1 / 3 (Same for both units)
S = 0.0003015 f 5 / 3 / Q1 / 6 ( in mks )
1
V= R3 / 4 So1 / 2 ( in mks )
Na
N a = 0.0225 f 1 / 4 (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 f = 1.76 m on the implicit
assumption of the 'regime charge' being carried by a channel in regime. Chitale
analysed the data subsequently in 1966.

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.

22.6 Examples
Design a triangular lined channel section with rounded bottom as prescribed by CBI&P
given Q =30 m3s-1, n=0.015, bed slope S0
1:1800, side slope m = 1.25.
Solution :
The side slope of the channel m =1.25,
i.e., cotθ = 1.25,

∴ θ = 0.6747 radian = 38.6598D
Area of the channel A = y 2 (θ + cotθ )
= 1.9247 y 2
Perimeter of the channel P = 2y ( θ + cotθ ) = 2y(0.6747+1.25)
Hydraulic radius R= A / P
=y/2
8
1.9247 y 3 8
therefore AR 2 / 3 = 1.9247 y 2 * (y / 2)2 / 3 = 2
=1.2125y 3 --->1
2 3
nQ 0.015 * 30
and = = 19.092 --->2
So 1
1800
Equating equations 1 and 2
( )8
3
∴ y= 19.092 1.2125
3
= (15.7458 ) 8
y=2.81 m
therefore, Area of the cross section of the channel A= 1.9247 * ( 2.81 )2
= 15.197 m 2
and a free board = 34 cm
therefore the total height of the channel section H = 2.81 +0.34 = 3.15 m
Top width of the channel T= 2* m * H
= 2* 1.25 * 3.15 = 7.875 m
Perimeter of the channel P = 2H 1+ m 2
= 10.085 m

2. Design a lined channel to carry Q =125 m3s-1 , n = 0.015, bed slope S0 1:2500, side
slope m = 1.25. Maximum permissible velocity is 2.5ms-1
Solution:
The side slope of the channel m = 1.25,
i.e., cot θ = 1.25,

therefore θ = 0.6747 radian
Area of the channel A = y 2 (θ + cot θ ) + by
= 1.9247 y 2 + by -->1
and Area A=Q/V
= 125 /2.5= 50 m 2
= 50 m 2 . --->2
50 = 1.9247 y 2 + by --->A
Perimeter of the channel P = 2y (θ +cotθ ) + b --->3
= 3.8494y + b
A 1.9247 y 2 + by
R= =
P 3.8494 y + b
1
2.5 = R 2/3 1 2500
0.015

3
⎡ ⎤ 2
1 .5
⎢ 0.015*2.5 ⎥ ⎛ 0.0375 ⎞
∴ R= =⎜ ⎟
⎢ 1
( )
0.5 ⎥
⎝ 0.02236 ⎠
⎣⎢ 2000 ⎦⎥
= 2.1719
A 50
P= = = 23.0213 m
R 2.1719
23.0213 = 3.8494y + b
therefore B = 23.0213 - 3.8494y --->5
put equation 5 in equation A, then
50 - 1.9247 y 2 + (23.0213 − 3.8494y) * y = 0 --->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 m3/s, with n = 0.013, bed slope S0 = 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 m3/s, n
= 0.014, bed slope S0 = 0.0016, side slope m = 1.25.

Hydraulics Prof. B.S. Murty
GRADUALLY VARIED FLOW

23.1 Introduction
The flow in an open-channel 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 open-channels 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, α , 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.

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.
αv2
___
dH
2g
Energy grade line
Water Surface
Flow
Slope = S0
Channel bottom
dx
Z
Datum
Figure: 23.1 Derivation of the gradually varied flow equation
The total head above the datum at a section is
αv 2
H=z+y+ (23.1)
2g
where H is the total head; z is the elevation of the channel bottom; y is the depth of flow;
α 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.
dH dz dy d ⎛ V2 ⎞
= + +α ⎜ ⎟ (23.2)
dx dx dx dx ⎝ 2g ⎠
As the slope of the channel bottom is assumed small, Sin θ ≈ tan θ ≈ θ , in which θ 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,
dH dz
Sf = − , and slope of the channel bottom, S0 = − .
dx dx
Equation (23.2) becomes
dy d ⎛ V2 ⎞
+α ⎜ ⎟ = S0 − Sf (23.3)
dx dx ⎝ 2g ⎠

Velocity, V can be expressed in terms of the flow rate, Q and area of the cross
section, A.
Q
V= (23.4)
A
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
dy αQ 2 ⎛ -2 ⎞ dA
+ ⎜ ⎟ (23.5)
dx 2g ⎝ A 3 ⎠ dx
However, for a prismatic channel
dA dA dy dy
= =T (23.6)
dx dy dx dx
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,
dy So − Sf
= (23.7)
dx αQ 2 T
1-
gA 3
Equation (23.7) is a non-linear first-order differential equation. In this equation, slope of
the energy grade line, Sf may be estimated using the Manning's equation.
n 2Q2
Sf = (23.8)
A 2 R 4/3
where n is the Manning roughness coefficient; and R is the hydraulic mean radius.

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
αQ 2 T
F2 = ( 23.9 )
gA 3
where F = Froude number. Substitution of Equations (23.8) and (23.9) in Equation
(23.7) leads to
n 2Q2
S0 -
dy A 2 R 4/3
= ( 23.10 )
dx 1 − F2
For a specified value of Q, both F and Sf are functions of the depth, y. In fact, both F
and Sf will decrease as y increases. Recalling the definitions for the normal depth, y n ,
and the critical depth, y c , the following inequalities can be stated
Sf > S0 when y < yn

( 23.11)
Sf < S0 when y > yn
F>1 when y < yc

( 23.12 )
F<1 when y > yc
A gradually varied flow profile is classified based on the channel slope, and the
magnitude of flow depth, y in relation to y n and y c . The channel slope is classified
based on the relative magnitudes of the normal depth, y n and the critical depth, y c .
• y n > yc : "Mild slope" (M)

• yn < yc : "Steep slope" (S)
• yn = yc : "Critical slope" (C)
• S0 =0 : "Horizontal slope" (H)
• S0 <0 : "Adverse slope" (A)

It may be noted here that slope is termed as "sustainable" slope when S0 > 0 because
flow under uniform conditions can occur for such a channel. Slope is termed as
"unsustainable" when S0 ≤ 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.
Zone - 1
NDL
2
CDL
Yc Yn
3
Bed
(a) Mild Channel
1
2 CDL
Yc
Yn NDL
3
Bed
(b) Steep Channel
NDL: Normal depth line
CDL: Critical depth line
Figure 23.2: Profile Classification
The space above both the CDL and the NDL is designated as zone-1. The space
between the CDL and the NDL is designated as zone-2. The space between the
channel bed and CDL/NDL (whichever is lower) is designated as zone-3. 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 zone-1 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
zone-1.
It may be noted that an M1 profile indicates a subcritical flow since flow depth, y is
greater than the critical depth, y c .

Water Surface
M1
NDL
CDL
Bed
Figure 23.3: M1 Profile
Similarly, an S2 profile (Figure 23.4) indicates the water surface lies in zone-2 in a steep
channel. It may be noted that a S2 profile indicates a supercritical flow since flow depth,
y is lower than y c .
CDL
S2
Water Surface
NDL
Bed
Figure 23.4: S2 Profile
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 y n = yc .
Table 23.1: Types of Flow Profiles (Sc: Critical Slope)
Slope Profile Designation Relative Type of Flow

zone - 1 zone - 2 zone - 3 position of y
Adverse S0 = 0 None
A2 y > yc Subcritical
A3 y < yc Supercritical
None
Horizontal S0 = 0 H2 y > yc Subcritical
H3 y < yc Supercritical
M1 y > yn > yc Subcritical
Mild 0<S0<Sc = 0 M2 yn > y > yc Subcritical
M3 yn > yc> y Supercritical

Subcritical
C1 y > yc = yn
Critical S0 = Sc > 0 C2
uniform -

y = yc = yn critical
C3 yc = yn > y Supercritical
S1 y > yc> yn Subcritical
Steep S0 > Sc > 0 S2 yc > y > yn Supercritical
S3 yc > yn > y 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:
• y > y c ; flow is subcritical; F<1 ; denominator is positive.
• y < yc ; flow is supercritical; F>1 ; denominator is negative.
• y = y n ; flow is uniform; Sf = S0 ; numerator is zero.
• y > y n ; Sf < S0 ; numerator is positive.
• y < y n ; Sf > S0 ; numerator is negative.
• As y → y n (y tends to y n ); Sf → S0 ; Sf → S0 ; numerator approaches zero;

dy
→ 0; the surface profile appraches normal depth asymptotically.
dx
• As y → yc ; Flow tends to critical conditions; F → 1; denominator tends to zero;
dy
→ ∞; water surface profile approaches the critical depth vertically.
dx
It is not possible to have a vertical water-surface 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

longer hydrostatic in those regions. Thus equation (23.10) is not valid whenever flow
depth is close to the critical depth.
dy
As y → ∞; Sf → 0; F → 0; → S0 ; Water surface profile becomes horizontal as flow
dx
depth becomes very large.
q2
For a wide channel, hydraulic mean radius R ≈ h and F2 = . Equation (23.10) can be
gy3
simplified to
dy gy ( S0 y -q n )
3 10/3 2 2
=
dx y10/3 ( gy3 -q 2 )
dy
where q = flow rate per unit width. It can be seen from the above equation that →∞
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 water-surface profile may be studied using
the points discussed earlier. For example, consider an M1 profile. For an M1 profile,
y>y n >yc . y > yc implies that F<1 and y > y n implies that Sf < S0 .
Therefore,
dy S0 -Sf +
= = =+
dx 1-F2 +
This means that flow depth increases with distance x. On the downstream side, as y
dy
keeps increasing tends to S0 and the water surface becomes horizontal. On the
dx
upstream side, as y approaches the normal depth, y n , it approaches asymptotically. The
sketch of an M1 profile is shown in Figure 23.5.

ApproachesNDL becomes horizontal

asymptotically
Water Surface
NDL
CDL
x
Bed
Figure 23.5: Sketch of an M1 profile
Similarly, consider an M2 profile. In an M2 profile, y n >y>yc . y > yc implies that F<1 and
the denominator is positive. On the other hand, y<y n implies that Sf > S0 . Therefore,
dy S0 -Sf − Ve
= = = − Ve
dx 1-F2 + 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.
Water Surface
NDL
CDL
Bed
Figure 23.6: Sketch of an M2 profile
Now, Consider an S2 profile. In an S2 profile, y c > y > y n . y < yc implies that F>1 and
the denominator is negative. y > y n implies that Sf < S0 . Therefore,
dy S0 -Sf + Ve
= = = − Ve
dx 1-F2 −Ve
This means that flow depth decreases with distance x. On the downstream side, as y
decreases towards y n it approaches NDL asymptotically. On the upstream side, as y
increases toward y C , it approaches CDL almost vertically. The sketch of an S2 profile is
shown in Figure 23.7.

Water Surface
CDL
NDL
Bed
Figure 23.7: Sketch of an S2 profile
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
MILD Zone -2 Zone -3
NDL
CDL NDL
CDL NDL
CDL
M1
M2
M3
CRITICAL
NDL /
CDL
C1
C2
STEEP C3
CDL
NDL
S1 NDL
S2
S3

HORIZONTAL
CDL
CDL
NONE
H2
ADVERSE H3
CDL
Bed
NONE
A2 A3
Figure 23.8: Water Surface Profiles

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.

Sluice gate
M1
NDL
M2
M3
Hydraulic Jump CDL
Bed (mild slope)
Free over fall

(a) M1, M2 and M3 profiles
Hydraulic Jump Sluice gate

S1
S2
CDL
S3
NDL
Bed (steep slope)
(b) S1, S2 and S3 profiles
Sluice gate
H2
CDL
Hydraulic Jump
H3
Bed (horizontal)
Free over fall

(c) H2 and H3 profiles
Figure 24.1: Real-life cases of water-surface profiles

24.2 Sketching of Composite Water - Surface Profiles

Characteristics of water-surface 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.

• 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).

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.
NDL
Reservoir
CDL
CDL
Channel-1
Channel-2 NDL CDL
Channel-3
Figure 24.2: Channel for Example 24.1
Solution
• Channel-1 is a MILD channel since NDL is above CDL.
• Channel-2 is a STEEP channel since NDL is below CDL.
• Channel-3 is a HORIZONTAL channel since NDL does not exist.
• Critical flow conditions occur at the downstream end of Channel-3 since it is not a
steep channel and there is a free overfall.
• Critical flow conditions occur at the junction of Channel-1 and Channel- 2 since
the uniform flow in Channel-1 is sub critical while uniform flow in Channel-2 is
supercritical.
• Flow is uniform in both Channel-1 and Channel-2 far away from the junction
point, since the channels are long. Thus flow depths in Channel-1 and Channel-2
fall between NDL and CDL. Therefore, flow profile in Channel-1 is M2 type, while
flow profile in channel-2 is S2 type
• In Channel-3, downstream portion would have sub critical flow conditions (critical
depth occurs at the downstream end), while in the entrance region, flow would be

supercritical as it enters from the steep Channel-2. Therefore, a hydraulic jump
should occur in Channel-3. The composite profile is shown in Figure 24.3.
NDL
Control Point
Reservoir CDL
M2
Hydraulic Jump Control Point
S2 CDL H2
Channel-1
CDL
NDL
Channel-2 H3
Channel-2
Figure 24.3: Solution for Example 24.1
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.
Sluice Gate
Reservoir
NDL
CDL
NDL
NDL CDL
Channel-1 CDL
Steep
Channel-2 Overfall
Mild
Figure 24.4: Channel for Example 24.2
Solution
• Critical flow conditions occur at the entrance to Channel-1 because Channel-1 is
steep and the reservoir water level is above the CDL.
• Flow depth in the Channel-1 varies from critical depth at the entrance to the
uniform flow depth far downstream. This is an S2 profile.

• At the entrance to Channel-2, flow depth is equal to the normal depth in Channel-
1. This depth is below the CDL for Channel-2. Therefore, flow in the upper
reaches of Channel-2 is supercritical. An M3 profile occurs in this region.
• The sluice gate in Channel-2 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 Channel-2, 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
Channel-2.
• 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 Channel-2 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 Channel-2. Therefore,
critical depth occurs at this location. Note that the Channel-2 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.

Control Point Control Point

Reservoir Sluice Gate
S2 NDL M1
CDL
NDL HJ NDL
HJ M2 Control Point
Steep CDL
M3 M3 CDL
Mild Over fall

Mild
Control Point
Figure 24.5: Solution for Example 24.2

Nptel PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nptel PDF

Uploaded by

Copyright:

Available Formats

Hydraulics Prof. B.S.

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" - JOHN.F. KENNEDY.

1.3.1 Growth of Hydraulics and Irrigation Research In India-

Introduction (CBI&P 1979)

problems were being tackled mostly by engineering judgement based on experience.

professional papers on Indian Engineering (1863-1886) contain many original and

in the techniques of the measurements of discharges and determination of velocities. By

taken up for implementation.

Indian Institute of Technology Madras

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

Sir Arthur Cotton".

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).

port. Truly was he a farsighted engineer.

Indian Institute of Technology Madras

idle for a single moment. Remember Time is short, Eternity is near."

on hearing of her husband's death.

Madras Musings, October 1 - 15, 1999.

of irrigation, these included reopening of Western and Eastern Yamuna canals,

Indian Institute of Technology Madras

construction of canal in 1842.

supported the proposal of Ganga canal.

functioning well and have stood the test of time.

Indian Institute of Technology Madras

Works Department on May 15th, 1858.

presidency. In May 1822, he was posted as an Assistant to the Superintending

Engineer of the Tank department, Southern Division.

beginning of the Sethusamudram Project we have been talking of for a century ! .

Indian Institute of Technology Madras

and much more imperceptible and intangible.

Major Arthur Cotton, was to come soon after.

was bringing in irrigation-cum-navigation facilities in the Delta by building an anicut

across the wide river.

Madras Musings-September 16-30, 1999.

relation to weir design. Col.Clibborn's other contribution was on investigations on the

Indian Institute of Technology Madras

reliable design of hydraulic structures, design of stable channels, efficient distribution

devices, weed control, anti-water logging measures and land reclamation.

Lacey's sustained and pioneering work led to the development of comprehensive

to be a subject of sustained interest in India.

notable investigations in the same field.

India's contribution of the development of subsoil flow hydraulics in relation to the

(1936) and others.

Indian Institute of Technology Madras

conducted on broad-crested weirs were utilised by Burkitt in developing the 'Head-less

Bharat Rathna Dr. Sir. Dr. Mokshagundam Visvesvaraya (1861 - 1962)

September 15 is a memorable day in the annals of the engineering community in

except judicious use of the results of technological innovations. In recognition of his

Indian Institute of Technology Madras

presentation of the country's highest award - Bharat Ratna - in 1955.

To perpetuate the memory of this great engineer-statesman, the Council of Institution of

evolved guidelines for celebrating the Day.

rivers, evaporation and seepage losses, etc,.

standing wave flume type of fall was developed by Inglis by 1930.

immensely in tackling various hydraulic problems. Important investigators on this

Indian Institute of Technology Madras

Losses by evaporation and percolation in canals were investigated by Kennedy on the

research. During 1916-1928, valuable investigations were made on the problems of