t: ''dI ý: ': RRT R X IA R:.,1' ll-ý d ;. k : ti ; A' K..l

BORANG PENGESAHAN STATUS THESIS

Judul: Study Of Flow In A Non-Symmetrical Compound Channel With Rough Flood Plain.

'ýI; ý1 1'I: i'ýte 1 @ä: 1ý 'asd t - ý4bn?.:

Saya CHARLES BONG HIN JOO (HURUF BESAR)

mengaku membenarkan tesis ini disimpan di Pusat Khidmat Maklumat Akademik, Universiti Malaýsia Sarawak dengan svarat-syarat kegunaan seperti berikut:

1. Hakmilik kertas projek adalah di bawah nama penulis melainkan pcnulisan sebagai proiek bersama dan dibiavai oleh UNIMAS, hakmiliknya adalah kepunvaan UNIMAS.

2. Naskhah salinan di dalam bentuk kertas atau mikro hama botch dibuat dengan kebenaran bertulis daripada penulis.

3. Pusat Khidmat Malaumat Akademik, UNIMAS dibenarkan membuat sahnan untuk pengajian mcreka. 4. Kertas projek hanya botch diterbitkan dengan kebenaran penulis. Bayaran royalti adalah mengikut kadar

yang dipersetujui kelak. 5. * Saya membenarkan/tidak membenarkan Perpustakaan membuat salinan kerlas projek ini sebagai bahan

pertttkaran di antara institusi pengajian tinggi. 6. ** Sila tandakan ( ý ) di kotak yang berkenaan.

SULIT (Mengandungi makhtmat yang berdarjah keselamatan atau kepentingan Malaysia seperti yang termaktub di dalam AKTA RAHSIA RASMI 1972).

ý TERHAD (Mengandungi makltunat TER-HAD yang telah ditcntukan oleh organisasi; badan di mana penvclidikan dijalankan).

TIDAK TERHAD

Disahkan oleh

k ý .

(T . DATANGAN PENULIS) (TAN JATANGAN PENYELIA)

Alamat tetap: No. 99, Lorong AS-A, Taman

13i)C' Stampin, 93350, Kuching, SARAWAK. Associate Prof. Dr. NABIL BE. SSAIH

( Nama Pctnclia Tcl : 082-455537

Tarikh: 24 - 3- 2c)03 Tztriklt:

CATATAN *,.

Potong yang tidak berkenaan. Jika Kertas Projek ini SULIT Mau TERHAD, sila lampirkan surat daripada pihak herkuasa/

organlwºsi berkenaan dengan menyertakan sekali tempoh kertas projek. Ini perlu dikelaskan

sebagai SULZT atau TERHAD.

f'ksi_`UO3

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

This Final Year Project report is as follows:

Title : Study Of Flow In A Non-Symmetrical Compound Channel With Rough Flood Plain

Author's Name : Charles Bong Hin Joo

Matric No. : 5185

had been read and confirmed by:

ý. ? - ý-

(Associate Prof. Dr. Nabil Bessaih) Date Project Supervisor

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

Study Of Flow In A Non-Symmetrical Compound Channel With Rough Flood Plain

P. KHIDMAT MAKLUMAT AKADEMIK UNIMAS

IIIIIIIIIIIIIIIIIIIIIIIIIIIII 0000120243

Charles Bong Hin Joo

A Project Report Submitted in Partial Fulfilment for the Bachelor Degree of Engineering (Civil) With Honours in the Faculty of

Engineering Universiti Malaysia Sarawak 2003

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

Specially Dedicated to : Mum, Dad

to all who had guided, helped, taught, advised and motivated me along the way

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

Ill

ACKNOWLEDGEMENT

The author would like to express his profound appreciation and gratitude to the

following persons whom either direct or indirectly helped to complete his project.

1. Associate Prof. Dr. Nabil Bessaih, project supervisor, for his guidance, advice

and comments, and also encouragement throughout this project.

2. Civil Engineering Program, Faculty of Engineering, Universiti Malaysia

Sarawak in providing all laboratory facilities and materials.

3. Tan Haji Affandi B. Hap Osman, Mohd. Hafiz and Nawawi, (technicians of

Civil Engineering Program, Faculty of Engineering) for preparing all the

laboratory facilities and materials.

4. To all my friends, especially Albert Lai and Phang Hock Lim xyho have always

been there whenever needed.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

I%

ABSTRAK

Kajian tentang aliran air dalam saluran gabungan yang terdiri daripada saluran utama

dan dataran banjir di sebelahnya telah menjadi topik utama kajian kebelakangan ini.

Minat dalam topik kajian ini timbul akibat daripada keperluan untuk menganggar

isipadu aliran air semasa kejadian banjir danjuga untuk mendapatkan hubungan antara

aras kedalaman air dengan isipadu untuk tujuan kawalan banjir. Telah didapati bahawa

cara hidraulik lama untuk menganggar isipadu di mana saluran gabungan dibahagikan

kepada sub-bahagian adalah kurang tepat kerana tidak mengambil kira kesan interaksi

di antara aliran air dalam saluran utama dengan aliran pada dataran banjir. Walau

bagaimanapun, kebanyakan kajian yang dijalankan setakat mi adalah melibatkan

saluran gabungan yang berbentuk simetri. Oleh sebab itu, matlamat utama kajian ini

adalah untuk memberi pengenalan tentang ciri-ciri aliran dan anggaran isipadu bagi

saluran gabungan yang berbentuk tidak simetri dan mempunyai dataran banjir kasar.

Untuk tujuan ini, eksperimen dijalankan ke atas model saluran gabungan tidak

bersimetri berskel kecil. Daripada keputusan eksperimen, didapati untuk saluran

gabungan tidak bersimetri berskel kecil, cara hidraulik lama masih boleh digunakan

untuk menganggar isipadu aliran walaupun kurang tepat.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

ABSTRACT

The study of flow in compound channel sections has been the subject of considerable

research in recent years. The practical interest in the problem arises from the necessity

for accurate discharge predictions during flood events and for a reliable stage-

discharge relation for flood control measures and management schemes. It has been

long realized that traditional hydraulic methods of channel subdivision are inadequate

for discharge calculation due to the significant interaction between main channel and

flood plain. However, most of the research work was done for symmetrical compound

channel. Therefore, the intention of this project is to study the flow characteristics and

the discharge estimation for a non-symmetrical compound channel with rough flood

plain. In order to do this, experimental investigations are carried out on a small scale

non-symmetrical compound channel model. Results show that for a small scale non-

symmetrical compound channel, the traditional hydraulic method can still be used to

predict the discharge although less accurate.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

CONTENT

TITLE PAGE

DEDICATION

ACKNOWLEDGEMENT

ABSTRAK

ABSTRACT

CONTENT

LIST OF TABLES

LIST OF FIGURES

LIST OF NOTATIONS

CHAPTER I INTRODUCTION

1

II

III

iv

v

vi

IX

X

Xill

1.0 GENERAL OVERVIEW 1

1.1 OBJECTIVE OF THIS PROJECT

CHAPTER 2 LITERATURE REVIEW

I

2.0 BACKGROUND OF RESEARCH-] 3

2.1 CONVENTIONAL METHODS FOR 4 ESTIMATING DISCHARGE

2.2 THE CHARACTERISTICS OF FLOW IN A 6 COMPOUND CHANNEL

2.3 FACTORS AFFECTING THE INTERACTION 10 EFFECT BETWEEN THE MAIN CHANNEL AND THE FLOOD PLAIN

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

2.4 THE DISCHARGE ADJUSTMENT FACTORS 12

2.5 APPARENT FRICTION FACTOR ON THE 19 FLOOD PLAIN-MAIN CHANNEL INTERFACE OF COMPOUND CHANNEL SECTION

2.6 WEIGHTED DIVIDED CHANNEL METHOD 26 (WDCM)

CHAPTER 3 METHODOLOGY

3.0 INTRODUCTION 32

3.1 EXPERIMENTAL ARRANGEMENTS 32

3.2 EQUIPMENTS USED FOR THIS EXPERIMENT 35

3.3 PROCEDURE OF EXPERIMENT 38

CHAPTER 4 RESULT AND DISCUSSION

4.0 INTRODUCTION 47

4. 1 DISCHARGE CHARACTERISTICS OF NON- 47 SYMMETRICAL COMPOUND CHANNEL

4.2 EFFECT OF FLOOD PLAIN ROUGHNESS 50

4.3 EFFECT OF RELATIVE DEPTH OF THE FLOOD 52 PLAIN FLOW TO THE MAIN CHANNEL FLOW

4.4 EFFECT OF FLOOD PLAIN WIDTH TO MAIN 53 CHANNEL WIDTH RATIO

4.5 COMPARISON OF DIFFERENT METHODS FOR 56 PREDICTING VELOCITIES

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

4.6 THE WEIGHTED DIVIDED CHANNEL METHOD (WDCM) AND ý VALUE

CHAPTER 5 CONCLUSION

5.0 CONCLUSION

CHAPTER 6 RECOMMENDATIONS FOR FUTURE STUDIES

6.0 THE SUGGESTIONS

REFERENCES

APPENDIX A

APPENDIX B

61

67

69

70

71

107

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

I .

LIST OF TABLES

Table Page

I Main geometrical and flow characteristics of experiments analyzed 23

2 Comparison of % error with depth for different B/b ratio 54

3 Summary of the trial ý value with R` 62Dem

o (Visi

t http

://www.pd

fsplit

merger.

com)

ý

LIST OF FIGURES

Fig. Page

I Compound channel section 1 2 The Divided Channel Method (DCM) with a) vertical division 4

and b) horizontal division 3 Stage discharge relationship 6 4 Ratios of main channel and flood plain discharges to full cross- 7

sectional values 5 Ratios of main channel and flood plain velocities to full cross- 7

sectional values 6 Variation of Manning's resistance coefficients with depth 8 7 Variation of the Darcy-Weisbach resistance coefficients with 9

depth and Reynolds number for the compound section, main channel and flood plains

8 Sample test results from FCF: plot of DISADF (ratio of measured l 3 discharge to sum of zonal calculated discharges), also coherence COH to same scales

9 Notation sketch 20 10 Variation of apparent friction factor with relative depth and width 24

ratio in the Flood Channel Facility experiments 11 Dependence of the apparent friction factor on the flood plain 25

Reynolds number and the width ratio (HRW = HR Wallingford, 1992; KD = Knight 7 Demetriou, 1983; see Table 1)

12 Variation of main channel mean velocity with relative depth 27 13 Variation of flood plain channel mean velocity with relative depth 28 14 Comparison of observed and predicted main channel velocity for 29

asymmetric flood plains 15 Comparison of observed and predicted flood plain velocity for 30

asymmetric flood plains 16 Comparison of observed discharges with values predicted by the 30

WDCM for asymmetric flood plain 17 Side view of the experimental arrangements ( not to scale) 33 18 Plan view of the experimental arrangements (not to scale) 34 19 Cross-section of the non-symmetrical compound channel 34

(not to scale) 20 View of the channel with the supply wooden tank as seen from 35

the end of the channel 21 The valve use to control the flow into the supply wooden tank 35 22 The collecting tank at the end of the channel 36 23 The pump used to supply water to the wooden tank 36 24 The steel tank. As can be seen above, the water is transfer 37

from the collecting tank to the steel tank via a 4" PVC pipe

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

xi

Fig. Page

25 Electronic depth meter 37 26 Current meter with recorder 38 27 The wire mesh installed in main channel to determine the value 39

of Manning's n 28 Location to take the point velocities (not to scale, all unit in m) 40

for flood plain width _ 0.20 m (B/b=9). For the number of vertical points, it varies with the stage of water, please refer to TABLE A. 2 in APPENDIX A for the observed depth where the point velocities is taken

29 Location to take the point velocities (not to scale, all unit in m) 41 for flood plain width = 0.14 m (B/b=6.6). For the number of vertical points, it varies with the stage of water, please refer to TABLE A. 3 in APPENDIX A for the observed depth where the point velocities is taken

30 Location to take the point velocities (not to scale, all unit in m) 41 for flood plain width = 0.09 m (B/b=4.6). For the number of vertical points, it varies with the stage of water, please refer to TABLE A. 4 in APPENDIX A for the observed depth where the point velocities is taken

31 The roughened floodplain for the floodplain width of 0.20 m (B/b=9) 42 32 The needle of the electronic depth meter touches the water surface 43 33 Subsection in the midsection method (not to scale) 44 34 The floodplain wall which is clenched in place to reduce the 46

floodplain width to 0.14 m (B/b=6.6) 35 The floodplain wall which is clenched in place to reduce the 46

floodplain width to 0.09 in (B/b=4.6) 36 Discharge-stage relationship for non-symmetrical compound 48

channel with smooth main channel and flood plain for Bib=9 37 Discharge-stage relationship for non-symmetrical compound 49

channel with rough flood plain but smooth main channel for B/b=-9 38 Comparison of discharges for rough flood plain with smooth flood 50

plain for B/b=9 39 Comparison of discharges for rough flood plain with smooth flood 51

plain for B/b---6.6 40 Comparison of discharges for rough flood plain with smooth flood 51

plain for Bib=4.6 41 Ratios of main channel and flood plain discharges to full cross- 53

sectional values 42 Relative Depth Vs (Q est -- Q obs /Q obs) for different B/b ratio 55 43 Variation of flood plain mean velocity with relative depth for Bb-9 57 44 Variation of main channel mean velocity with relative depth for 58

Bib--9 45 Variation of flood plain mean velocity with relative depth for 58

Bib=6.6

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

SII

Fig. Page

46 Variation of main channel mean velocity with relative depth 59 for B/b=6.6

47 Variation of flood plain mean velocity with relative depth 59 for B/b=4.6

48 Variation of main channel mean velocity with relative depth 60 for B/b=4.6

49 V mean WDCM Vs V mean observed for 4=0 (B/b=9) 63 50 V mean WDCM Vs V mean observed for 2=1 (B/b=6.6) 63 51 V mean WDCM Vs V mean observed for i? =l (B/b==4.6) 64

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

\111

LIST OF NOTATIONS

A - cross-section area

B - half total width of main channel plus flood plain

b - half bed width of main channel

- weighting factor for weighted division channel method (WDCM)

fp - flood plain

H - total flow depth

H. - relative depth (H-h/t-U)

h - depth of main channel below berm level

me - main channel

n - Manning's roughness coefficient

Qes, - estimated discharge using channel division method

Qons - observed discharge

V - mean velocity

Valve - observed mean velocity for the whole cross-section

Vmc-I) M-II - calculated mean velocity in the main channel region using the DCM with horizontal division

V,,, c-1X: M-V - calculated mean velocity in the main channel region using the DCM with vertical division

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

I

CHAPTER 1

INTRODUCTION

1.0 GENERAL OVERVIEW

The study of flow in compound channel sections has been the subject of

considerable research in recent years. The practical interest in the problem arises from

the necessity for accurate discharge predictions during flood events. The term

`compound' or two stage covers channel cross-sections having berm(s) or flood

p'ain(s) that come into action at high flows but which are normally dry (see fig, 1). It

has long been realized that traditional hydraulic methods are inadequate for discharge

calculation due to the significant interaction between main channel and flood plains.

&kwr cronrw. Ipn, c>.

Fig. ]. Compound channel section.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

Many experimental studies have been carried out addressing various aspects of

the problem, ranging from the boundary shear distribution to the structure of

turbulence in compound section and various methods as well as empirical formulas

have been proposed for discharge calculation. The available studies on flow in

compound channels include Myers (1975 & 1978), Knight & Shiono (1990),

Wormleaton & Merritt (1990) and Lambert & Myers (1998). Despite the progress

achieved so far, no consensus has been reached for the estimation of discharge in

compound channel.

1.1 OBJECTIVE OF THIS PROJECT

Since most of the research work was done for symmetrical compound channel,

thus, the objective of this project is to study the flow characteristics and the discharge

estimation for a non-symmetrical compound channel with rough flood plain.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

I

CHAPTER 2

LITERATURE REVIEW

2.0 BACKGROUND OF RESEARCH

The apparently simple problem of determining the discharge capacity of a

compound channel under uniform flow conditions has proved to be difficult. Sellin1

(1964) first identified the modification of the velocity distribution and the resulting

changes in the discharge capacity caused by the turbulent interaction between the main

channel and the flood plain. Compound channels have traditionally been analysed by

dividing the compound cross-section into relatively large homogeneous sub-areas

which are easier to analyse. This approach is termed the divided channel method

(DCM). However, this approach assumes no interaction between the subdivided areas

despite the existence of mean velocity discontinuities at the assumed internal

boundaries.

The main features that affect the interaction and hence loss of discharge

capacity when the flow is above bank according to P. Ackers(1992)are

a) Relative depth of the flood plain flow to the main channel flow.

b) Roughness of the flood plain compared with the roughness of the main

channel.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

4

c) Ratio of the flood plain width to the main channel width.

d) The number of flood plains.

e) The side slope of the main channel.

f) The aspect ratio of the main channel.

2.1 CONVENTIONAL METHODS FOR ESTIMATING DISCHARGE.

In 1933, Lottetx proposed the `divided' channel method which is based on the

subdivision of the flood plain from the main channel, as shown in Fig. 2.

B

i -».,. r . ý h ý

f... w

b (a) Yerticai diVWons.

---, ----_ ý------

-sý"

.. ý- - r- ,... awý

(b) Honzont, at divisions.

Fig. 2. The Divided Channel Method (DCM) with a) vertical division and b) horizontal division.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

In the divided channel method (DCM), it is necessary to split the compound

channel cross-section into subsections, either by vertical division or horizontal

division. Then, Manning's formula (Q = s` AR2/ 3 ! n) will be applied to each

subsection and the discharges for all the subsection will be summed to estimate the

overall discharge of the compound channel. However, since this approach assumes no

interaction between the subdivided areas, where the interaction between the slower

moving berm flows and the main channel flow increases head losses significantly, so,

the discharge calculated by this method will overestimate significantly the true

channel capacity.

Another conventional method for estimating discharge though not as popular as

the divided channel method (DCM) is the single channel method (SCM). In this

method, the discharge for the compound channel is calculated as a whole using the

Manning's formula without any subdivision. However, the single channel method

(SCM) tends to underestimates the discharge capacity.

Since the conventional methods used for predicting discharge did not take into

account the interaction between the slower moving berm flows and the main channel

flow, thus, a more reliable methods of predicting the discharge for compound channel

is needed.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

6

2.2 THE CHARACTERISTICS OF FLOW IN A COMPOUND CHANNEL.

In order to develop a reliable method for estimating discharge in compound

channel, first of all, the characteristics of flow in compound channel must be known

In order to better understand the flow characteristics, experiments to study the flow

resistance has been carried out by Myers and Brennan3 (1990) at the Flood Channel

Facility (FCF) in the United Kingdom. The results of this studies have shown that : -

i_ The primary stage discharge relationship for compound channel are as shown

in Fig. 3. The most notable feature of these relationship is the discontinuity at

bankfull depth.

I

i I W

Y ý r

_ ý ý= . r. ý _ . ý

- -- ý. t

inrhU ý ý ý _ _. . ý ý ý. ý ý _ . ý ý ý __ . _ . _ _ ý

_ý_ ý . _ _ý_.. -. ý_ý ý ý_ __ ... ý. ý. ý_ý _

ý__. -+ I

-_

Trapuobal, JHlC__. _. __ . _. ___-----_-_i-... _-ý

. . . ___-. _.. . . _. ______ . ___. I

. . _. . _. . _____ I _ . __ . _rý. _. _..... __ . .. .. _ __''_.. _ I

, : s

I1 's ý !

OISCN11R6E flý/Ne

h o 1 9 ý

Fig. 3. Stage dischargu relationship.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

7

ii. There is a rapid increases in flood plain discharge and velocity as depth rises,

to a point where main channel and flood plain are roughly equal in carrying

capacity (as shown in Fig. 4 and Fig. 5). This equalization of discharge and

velocity results in a consequent decrease in momentum transfer from main

channel to flood plain and may lead to a reversal in the direction of momentum

transfer at larger depths.

56/e Qwe/o Oio/O 667 O " 4Q 7 "

ý zt

- - L_

" i{

Yý

ýý-=, F'ý _----- ------ _j

týl, ' ý ý '

- - ý'ý- ---

o1 os0-(

+ 0 0

I

Fig. 4. Ratios of main channel and flood plain discharges to full cross-sectional values.

4

ý

g 0

2

c

ý r ., /

__/ f _ ýi_,

ý-ý --ý y 11

,

ý " ý ý. _ .. w ý - "' .

�'ý It, lý °1 6,.

------E%-rr /ý.

ýt-ý-----ý v- 'ý". _ 10 i

--F'ýý ý "ýý. ý. r.. c ý

"lrý ý , ýtiý ;

ý L-

4 6Vf a VnK V ý-

t ý j e ý : ý ý ý ý

II

Fig. 5. Ratios of main channel and flood plain velocities to full cross-sectional values.

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

8

iii. The flow resistance has been presented in terms of Manning's and Darcy-

Weisbach resistance coefficients. The compound channel resistance

coefficients show a significant reduction in value at depths just above bankfull,

but increase to simple channel values with increasing depth (see Fig-6 and

Fig. 7). The main channel and flood plain resistance coefficients are increased

and decreased respectively by the presence of the momentum transfer

mechanism.

aL

Fig 6. Variation ofManning's" resistance coefficients with depth

Demo (

Visit h

ttp://

www.pdfsp

litmerg

er.co

m)

9

)so

l: sm, )5

v ýý

4

12$

9 0

i

N-lq! r- -

" _- ----------. - N -ý . :.. o; -.. a. : - . \ 1, ý.....,..,, - ---- -\ -\ ' - _- _---- ..,.I n , __,

ýý . : ýý_Nrý«1Ný'. ý

ýý\\

/

:

, -. ý-- -6

'S

ea

w I I a

i_

IM I. n

5

PI. rni0 .1Ne. R.

Bib E-87 0 "

. 20

2 70 . "

ý . , " aý

Fig. 7. Variation of the Darcy-Weisbach resistance coefficients with depth and Reynolds number for the compound section, train channel and flood plains.Dem

o (Visi

t http

://www.pd

fsplit

merger.

com)