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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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:

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Solution:

20

* *
* *
16 * *
* *
* *
* *
12 * *
* *
* *
* *
8 * *
* *
* *
* *
4 * *
* *
*

0.0 6.0 12.0 18.0 24.0 30.0

Distance from reference (m)
Natural channel

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

∂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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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:

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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
Godavari Western Main Canal - Mile 5/2

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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).

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

__ 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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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 .

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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
Godavari Western Main Canal - Mile 5/2

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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
:

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

⎛ 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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

8.2 Application of Specific energy

Transition Problem:

__ y1
V1

C1 y2

Sluice gate

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

__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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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).

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 .

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

b1 b2
bx

x
0
L
Plan

Specific energy line

High stage
Sub critical flow

Critical depth line

yc Super critical flow
Low stage

Water surface profile with hydraulic drop

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Specific energy line
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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Specific energy line

for upstream
Specific force line for low stage

Specific energy line

for downstream

__2
Vx
____
2g
yx
zx

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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:

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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).

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

⎛ 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).

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 )

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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 )

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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 ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 )

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

⎛ 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 )

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎟
⎝ ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Labyrinth weir
corner angle: 117
242

20.4 m

11.7 m

Polygonal weir
corner angle: 133 , 152 , 255

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

⎛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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ;

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 )}

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ) }

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎥⎦

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎦

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Noise cancellation probes

Buried
Coil

Coil Signal Probes Signal

Drive Recovery

Noise cancellation probes

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.
Liquid flow measurement in open channels -
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.
Liquid flow measurement in open channels -
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.
Liquid flow measurement in open channels by
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.
Liquid flow measurement in open channels -
ISO: 4359 1978
flumes.
Liquid flow measurement in open channels by
ISO: 4360 1978
weirs and flumes - triangular profile weirs.
Liquid flow measurement in open channels by
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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎤
⎣ ⎦

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. 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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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).

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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"

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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):

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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
C= (in metric units)
n+ R

12. GIBSON FORMULA:

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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
C= (in metric units)
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
C= (in metric units)
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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎥⎦

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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,

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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⎦

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

⎛ 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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

−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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 *.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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?

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

⎡ 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

Indian Institute of Technology Madras

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Radius of curves for lined canals

Discharge (m3/s) Radius (minimum) in m

280 and above 900
Less than 280 to 200 760
Less than 200 to 140 600
Less than 140 to 70 450
Less than 70 to 40 300

Note: Where the above radii cannot be provided, proper super elevation in bed shall be

provided.

The width of the banks along a canal are usually governed by a number of

considerations which include the size of the need for maintenance roads. Where roads

are needed, the top widths for both lined and unlined canals are designed so that

precipitation will not fall in to the canal water and, to keep percolating water below the

ground level beyond the banks.

21.1.1 Hydraulically Efficient Channel

It is well known that the conveyance of a channel section increases with increases in

the hydraulic radius or with decrease in the wetted perimeter. Therefore, from the point

of hydraulic aspects, the channel section having the least wetted perimeter for a given

area has the maximum conveyance; such a section is known as the Hydraulically

efficient channel. But this is popularily referred as Best Hydraulic section. The semicircle

has the least perimeter among all sections with the same area; hence it is the most

hydraulically efficient of all sections.

The geometric elements of six best hydraulic section are given in Table. It may be noted

that it may not be possible to implement in the field due to difficulties in construction and

use of different materials. In general, a channel section should be designed for the best

hydraulic efficiency but should be modified for practicability. From a practical point of

view, it should be noted that a best hydraulic section is the section that gives the

minimum area of flow for a given discharge but it need not be the minimum excavation.

The section of minimum excavation is possible only if the water surface is at the level of

the top of the bank. When the water surface is below the bank top of the bank (which is

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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:

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

(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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

can resist erosion.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 )

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 .

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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;

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 .

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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 .

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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:

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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).

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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,

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

braced type

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Typical cross sections of Power canals

Covered duct R.C.C

on piles when there is
land slide problem

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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
Equating equations 1 and 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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Thandaveswara

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
Equating equations 3 and 4
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.

Indian Institute of Technology Madras

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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 ⎠

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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)

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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 .

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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 -

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

HORIZONTAL

CDL
CDL

NONE
H2
ADVERSE H3
CDL

Bed

NONE
A2 A3

Figure 23.8: Water Surface Profiles

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

• 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).

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

• 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.

Indian Institute of Technology Madras

Hydraulics Prof. B.S. Murty

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

Indian Institute of Technology Madras