Alnahhal Mohammed - PHD Thesis PDF
Alnahhal Mohammed - PHD Thesis PDF
Alnahhal Mohammed - PHD Thesis PDF
BY
OF
DOCTOR OF PHILOSOPHY
IN
AT
Patras, 2010
CANDIDATES DECLARATION
It is hereby declared that this thesis or any part of it has not been submitted elsewhere
for the award of any degree or diploma.
.
Dr. Thrassos Panidis
Assistant Professor
Laboratory of Applied Thermodynamics
Division of Energy, Environment and Aeronautics
Department of Mechanical Engineering and Aeronautics
University of Patras
Rio 26 504
Greece.
Email: [email protected]
.
Mr. Mohammed Alnahhal
PhD Candidate
Laboratory of Applied Thermodynamics
Division of Energy, Environment and Aeronautics
Department of Mechanical Engineering and Aeronautics
University of Patras
Rio 26 504
Greece.
Email: [email protected]
[email protected]
PUBLICATIONS FROM THE PRESENT WORK
A) Journals
B) Conferences
.
1. Dr. Ioannis Ekaterinaris
Professor, Department of Mechanical Engineering and Aeronautics
University of Patras, Patras, Greece
.
2. Dr. Apostolos Goulas
Professor, Department of Mechanical Engineering
Aristotle University of Thessaloniki, Thessaloniki, Greece
.
3. Dr. Ioannis Kallinderis
Professor, Department of Mechanical Engineering and Aeronautics
University of Patras, Patras, Greece
.
4. Dr. Panayiotis Koutmos
Associate Professor, Department of Mechanical Engineering and Aeronautics
University of Patras, Patras, Greece
.
5. Dr. Dionisios Margaris
Associate Professor, Department of Mechanical Engineering and Aeronautics
University of Patras, Patras, Greece
.
6. Dr. Thrassos Panidis
Assistant Professor, Department of Mechanical Engineering and Aeronautics
University of Patras, Patras, Greece
.
7. Dr. Kostas Perrakis
Lecturer, Department of Mechanical Engineering and Aeronautics
University of Patras, Patras, Greece
To Allah
To my parents, my sisters and brother
To my wife, and my parents in law
CONTENTS
Acknowledgment i
Abstract ii
Abstract-Greek iii
List of symbols iv
List of figures v
List of tables vi
1. Introduction
1.1 Background 1
1.2 Contribution of present thesis 5
1.3 Open issues and further work 6
1.4 Thesis outline 7
2. Theory 8
2.1 Turbulent flows 8
2.2 Free shear flows 8
2.3 Boundary layer flows 10
2.4 Turbulent jets 11
2.5 Turbulent jets issuing from rectangular nozzles 13
2.5.1 Description of the velocity field 14
2.5.1.1 Mean velocity fields 14
a) Mean streamwise velocity 14
b) Mean lateral velocity 15
c) Mean spanwise velocity 15
d) Spread of rectangular jets 16
e) Decay of centerline mean velocity 17
f) Type of initial boundary layer 18
2.5.1.2 Turbulent velocity fields 18
a) Turbulent velocities 18
2.5.1.3 Self-similarity concept 19
2.6 Related equations 20
a) Equation of motion 20
b) The integral momentum equation 23
c) The integral energy equation 24
d) Entrainment hypothesis 26
2.7 Initial and boundary conditions of rectangular jets 26
a) Reynolds number 26
b) Aspect ratio 27
c) Nozzle type 27
d) Sidewalls and/or endplate 28
3. Review 30
3.1 Literature review 30
4. Experimentation 40
4.1 Experimental Set-up 40
4.1.1 Nozzle exit 40
4.1.2 Settling chamber 40
4.1.3 Blower 42
4.1.4 Endplates and sidewalls 42
4.2 Hot wire anemometry 43
4.2.1 Constant temperature hot-wire anemometry (CTA) 43
4.2.1.1 Working principle 44
4.2.1.2 Noise 45
4.2.1.3 Advantages of hot wire anemometry technique 46
4.2.1.4 Disadvantages of hot wire anemometry technique 46
4.2.1.5 Selection of hot wire sensors 46
4.2.1.6 Calibration of hot wire anemometer 47
4.2.1.7 Measuring velocity components 48
4.2.1.8 Specifications of the hot wire anemometry system and
experimental conditions of the present investigation 49
4.2.1.9 Experimental uncertainty 50
4.3 Experimental procedure 50
6. Conclusions 160
Bibliography 163
i
ACKNOWLEDGMENT
I would like to thank almighty Allah for everything. Almighty Allah provided
me the strength, patience and kind people who helped me to complete my thesis.
Those people are:
First and foremost, I am deeply grateful to my supervisor Prof Thrassos Panidis,
Department of Mechanical Engineering and Aeronautics, University of Patras for the
invaluable assistance, guidance and encouragement that he provided me during the
preparation of this thesis and during my stay in Greece. His depth of insight,
understanding, involvement and dedicated supervision at every critical stage made
possible the completion of this study. Many thanks to the departed Professor D. Papailiou
for his help, encouragement and introducing me to Professor T. Panidis. Also, I would
like to express my appreciation to the members of the supervision and examination
committees.
I would like to thank Prof Kostas Perrakis for his help to me during my stay at
LAT. Many thanks to Dr. Alexandros Romeos for the continuous cooperation, help and
encouragement. Also, thanks to Dr. Thanasis Giannadakis, my roommate for the
motivation and help and Dr. Alexander Cavo and Dr. Alexandros Vouros for their help
during my PhD thesis. Mr. Skouras Giannes for introducing me to hot wire anemometry
technique, Mr. Andreas Vouros, Mr. Zafiris Sotiris and Mrs. Papadopoulou Sofia, for
their cooperation during my stay here. I would like also to thank Dr. Kalogirou Ioannis.
Assistant Professor, Technological Educational Institute of Patras for the assistance that
he provided me. Many thanks to everyone who stood beside me during this journey.
I would like to thank the State Scholarship Foundation (I.K.Y) in Greece for the
financial support that provided me the ability to complete this research.
I would like also to thank my father Hamed, my mother Amina, my father in
law Said, my mother in law Asma, my sisters Ruba and Reem, my brother
Amjad, and all of my relatives for their wish and encouragement; and for their
everlasting support. I want to thank my wife Raghda; she successfully compromised
between home and her PhD work, thank you for your unlimited love, patience and
support, good luck in your PhD thesis. Also, thanks to all of my friends for their
encouragement.
Finally, my appreciation is highly directed to Greece and Greek people for their
hospitality and generosity.
ii
ABSTRACT
Turbulent jet flows issuing from rectangular nozzles are used in many technological
and practical applications. Understanding their development and mechanics is
important to the design and performance improvement of these applications.
Therefore, rectangular jets have been studied extensively over the past decades.
Previous investigations have extensively studied the effects of various initial and
boundary conditions such as nozzle type, nozzle aspect ratio, nozzle exit turbulence
intensity, jet exit Reynolds number. Nozzle type (e.g. smooth contraction nozzle,
channel, radially countred nozzle) has been found to affect the development of these
jets. While varying Reynolds number affects the mixing of the jet with its ambient
either in the near field or in the far field. Nozzle aspect ratio is also an important
boundary condition. It has been experimentally shown that aspect ratio influences the
dimensionality of flow field, and hence entrainment between the jet and its ambient.
Some of these investigations have used sidewalls (two parallel plates attached to the
nozzle short sides) to improve the two-dimensionality of the flow fields, and/or
endplate (a wall flush at the nozzle exit) to prevent mixing upstream of the nozzle.
Despite this fact very few investigators have specifically considered the effect of
endplate or sidewalls on jet development.
In the present investigation, the effects of endplates and/or sidewalls on turbulent jets
issuing from a rectangular of aspect ratio, AR= 15 have been studied. The mean
streamwise and lateral velocity and turbulent characteristics of four different
rectangular jets, namely,
1. jet with no endplate and no sidewalls, NENS,
2. jet with endplate and no endplate, WENS,
3. jet without endplate and with sidewalls, NEWS and
4. jet with endplate and with sidewalls, WEWS,
have been measured, with x-sensor hot wire anemometry, up to an axial distance of 35
D under identical inlet conditions. Centreline measurements for the four
configurations have been collected for three Reynolds number, ReD=10,000, 20,000
and 30,000. For ReD=20,000 measurements in the transverse direction were collected
at 13 different downstream locations in the range of x= (0, 1, 2, 2.5, 3, 4, 5, 10, 15, 20,
25, 30 and 35) nozzle widths.
The jet with no endplate and no sidewalls (NENS) and the jet with endplate and no
sidewalls (WENS) produce nearly similar mean and turbulent velocity profiles
indicating insignificant effect of the endplates on their development in the absence of
the sidewalls.
At Re=20,000, larger mean streamwise velocity values were observed at the edges of
the jets (NEWS or WEWS) at distances from nozzle in the range of x/D=3-30 whereas
the presence of the endplate has an insignificant effect.
The presence or absence of sidewalls is also key factor determining the distributions
of the lateral mean velocity component. The presence of sidewalls is associated with
lower outward velocities within the edges of the jets and higher inward ones outside
the edges. The absence of sidewalls makes the presence of the endplate insignificant
and the lateral velocity attains outside the edges low negative values of almost the
same level in both cases (NENS, WENS) The presence of an endplate has again some
significance only when the sidewalls are present alleviating their effect.
Absence of the endplate in jet with sidewalls, leads to lowest spread rate compared to
the case with endplates (comparing jet with no endplate and with sidewalls, NEWS
and jet with endplate and with sidewalls, WEWS). The presence of an endplate in the
absence of sidewalls has again insignificant effect on the jets spread rate (comparing
jet with no endplate and no sidewalls, NENS and jet with endplate and no sidewalls,
WENS).
The effects of endplates and/or sidewalls on the decay rates have been investigated for
ReD=10,000, 20,000 and 30,000. The presence of the endplate in jets with no
sidewalls has insignificant effect on decay rate as was observed for all Reynolds
numbers. The presence of the endplate in jet with sidewalls, leads to higher decay
rates for all Re tested compared to all cases. The decay rates of the jets with no
sidewalls in the presence and absence of the endplate (NENS, WENS) seem to reach
their asymptotic values at around ReD=20,000. But the jets with sidewalls have not
reached an asymptotic behaviour even at the highest Re tested.
The implementation of sidewalls has been found to lead to a decrease of the
streamwise turbulent velocity, u, at the exit shear layer. Current results indicate that
the presence of an endplate has an insignificant effect and the estimates for the root
mean square of the streamwise velocity fluctuation u, at the exit shear layer can be
grouped according to the absence or presence of sidewalls (3.8% of the mean
centreline velocity for NENS and WENS, and 3.1% for NEWS and WEWS).
For all turbulent terms, the presence of the endplate has an effect in the presence of
sidewalls and the nondimensional streamwise turbulent velocity attains always higher
values in its presence (comparing WEWS with NEWS). The profiles of the
streamwise turbulent velocity, u profiles for the jet with no endplate and no sidewalls,
(NENS) and the jet with endplate and no sidewalls, (WENS) are almost
indistinguishable in the range x/D=15-35 but the profiles of the jet with no endplate
and with sidewalls, (NESWS) and jet with endplate and with sidewalls, (WEWS)
indicate a monotonic increase of the values in the central area of the jets. This
indicates that the observed trends are mainly due to the presence of the sidewalls and
presence or absence endplate has insignificant effect.
iii
,
.
.
(
, , , )
( ,
Reynolds, )
.
(sidewalls)
/ (endplate)
. ,
,
.
15
:
1. -
(No endplate - no sidewalls, NENS),
2. - (With
endplate - no sidewalls, WENS),
3. - (No
endplate - with sidewalls, NEWS),
4. - (With endplate
- with sidewalls, WEWS).
(Hot
wire anemometry, HWA)
.
Reynolds, ReD = 10,000, 20,000 30,000 35
, D, . ReD = 20000
x/D = 0, 1, 2, 2.5, 3, 4, 5, 10, 15, 20, 25, 30 35.
.
.
iv
LIST OF SYMBOLS
Greek letters
Kinematic viscosity of air, m2/s
Boundary layer thickness, mm
1 Displacement thickness, mm
Shear stress, N/m2
Circumferential coordinate
m Momentum thickness, mm
Dynamic viscosity, N.s/m2
Mass density of fluid, kg/m3
Abbreviations
CTA Constant Temperature Hot-Wire Anemometry
CCA Constant Current Hot-Wire Anemometry
HWA Hot Wire Anemometry
LDA Laser Doppler Anemometry
NENS jet without (No) Endplate and without (No) Sidewalls (free jet)
NEWS jet without (No) Endplate and With Sidewalls
OHR Over heat ratio
WENS jet With Endplate and without (No) Sidewalls
WEWS jet With Endplate and With Sidewalls
2D 2- Dimensional
3D 3-Dimensional
v
LIST OF FIGURES
Chapter 1
INTRODUCTION
1.1 Background
Turbulent rectangular jets are used in many technical applications in aerospace,
mechanical and chemical engineering. They can be found in the propulsion units in
both conventional and V/STOL aircrafts, dispersion of pollutant effluents, and gas
turbine plants of electric power utilities (Quinn, 1992). In combustion, the mixing
between the reactants in diffusion flames, or between the hot reaction products and
the fresh reactants needed to sustain combustion, is a crucial part of the combustion
process. The mixing occurs in two stages. The initial stage brings relatively large
amounts of the reactants together (large-scale stirring). The second stage involves
small-scale turbulent mixing, which accelerates the molecular contact between the
reactants. Noncircular jets can be used to improve the small and large scale mixing in
combustion process (Gutmark & Grinstein, 1999).
The flow field of a rectangular jet is characterized by the presence of three distinct
regions (Figure 1.1), defined by the centreline mean streamwise velocity, Uc, decay
(Krothapalli et al., 1981), namely: (a) the potential core region, with almost constant
Uc, which ends when the two shear layers along the short dimension of the nozzle
meet, (b) the two-dimensional region in which the velocity decays at a rate roughly
the same as that of a planar jet, and (c) the axisymmetric region, in which the velocity
decays as in an axisymmetric jet, the location where the two shear layers along the
long dimension of the nozzle meet. A rectangular jet may approximate the behaviour
of a plane jet1, given that its aspect ratio, AR=L/d (where L is the long side of the
nozzle in z direction and D is the short side of the nozzle in the y direction, see Figure
1.1) is large enough, to minimize the importance of entrainment in the spanwise
direction, z, and hence establish a statistically two-dimensional flow (Pope, 2002).
1
The ideal plane jet is statistically two-dimensional, in the dominant direction of the mean flow, x, and
the cross-stream coordinate, y, whereas statistics are expected to be independent of the spanwise
coordinate z (Pope, 2002).
streamwise
direction, x
zc yc
Axisymmetric
(or 3-D) region
x/D 25-30
Planar
(or 2-D) region
4-5 x/D 25-30
lateral Potential
D direction, y
region
0 x/D 4-5
spanwise
L
direction, z
Nozzle
Figure 1.1: Schematic representation of the flow fields of a rectangular jet. Redrawn
after modification from Krothapalli et al., (1981).
It is well known that initial and boundary conditions strongly affect the development
of turbulent jets issuing from rectangular nozzles. Many experimental investigations
have found that the flow fields of rectangular jets are strongly influenced by Reynolds
number (e.g. Namer & tgen, 1988, Deo et al., 2008, Suresh et al., 2008).
Moreover, nozzles shape (or nozzle type) (e.g. rectangular nozzle, rectangular
channel) has been found to affect the evolution of these jets (e.g. Sfier, 1979, Deo et
al., 2007b). Also, aspect ratio, which is the ratio of major axis to minor axis of the
nozzle, is an important boundary condition, since it also influences the flow fields of
rectangular jets (e.g. Quinn, 1992, Mi et al., 2005, Deo et al., 2007a). Krothapalli et
al., (1981) have observed that, the point where the jet first assumes axisymmetric
behavior goes upstream toward the exit of the nozzle with the decrease of the aspect
ratio. In addition, two sidewalls attached to the nozzles short sides extending in the
axial direction (Figure 1.2) have been used in many investigations (e.g. Heskestad,
1965, Gutmarak & Wygnanski, 1976, Browne et al., 1982, Browne et al., 1984, Deo
et al., 2007a) to force the jet to entrain only in the axial and lateral directions and
hence ensure the 2-dimensionality of the flow fields. Endplate (i.e. a plate flush at the
nozzle exit) has been also used many previous investigations (e.g. Browne et al.,
1982, Browne et al., 1984, Quinn, 1992) to prevent entrainment upstream of the
nozzle. Figure 1.2 shows the endplate in the presence of the sidewalls.
During the past years, a great number of studies have been devoted to study turbulent
jets issuing from rectangular nozzles. A review on the experimental set-ups of these
investigations reveals that these set-ups can be classified into four groups based on
the presence and/or absence of the sidewalls and/or endplates, namely:
1. jets without (No) Endplate and without (No) Sidewalls (free jets), NENS
(e.g. Sforza et al., 1966, Trentacoste & Sforza, 1967, Krothapalli et al., 1981,
Namer & tgen, 1988, Papadakis & Staiano, 1993, Lazanova & Stankov,
1998, Mi et al., 2005, Cavo et al., 2007, Deo et al., 2007c).
1,200
sidewalls
Axial direction, x
2,000
Spanwise
direction, z
Lateral direction, y
endplate
640
Figure 1.2: A schematic view of a rectangular nozzle with sidewalls and endplate.
(Dimensions in this figure are the ones that used in the present investigation and all
are in mm).
Although sidewalls and/or endplates have been used in the above mentioned
investigations at various initial conditions (e.g. different aspect ratio, Reynolds
numbers, and nozzle shapes), few experiments have been devoted to explore their
effects on turbulent jets issuing from rectangular nozzles (Hitchman et al., 1990,
Papadakis & Staiano, 1993, and Deo et al., 2007c).
Papadakis & Staiano, (1993) tried to assess the influence of endplate on free plane
jets (without sidewalls, NENS, WENS) of aspect ratio AR=40 at Re=9,800. Their
study has investigated the mean velocity only up to a downstream location x/D18.
Hitchman et al., (1990) and Deo et al., (2007c) investigated the effect of sidewalls on
plane jets with and without endplate respectively. Both investigations used similar
aspect ratio, AR=60 and nearly similar Reynolds number ~ 7,000. At this aspect ratio
it is expected to have a longer region of 2-dimensional behaviour. Moreover, their
measurements focused on the axial velocity component. In fact measuring two or all
the velocity components is essential when evaluating the effects of endplates and/or
sidewalls. According to our knowledge there is no available information on the
combined effect of the endplate and sidewalls (WEWS).
Using x-wire probe hot wire anemometry, the development of mean longitudinal and
lateral velocities is studied to enable direct comparisons among the characteristics
(e.g. spread rate, velocity decay, momentum, entrainment, self-similarity) of the four
configurations. The turbulence field is also investigated to evaluate the influence of
endplate and/or sidewalls on turbulent velocities, shear stresses and third moments.
The present results are discussed and carefully compared to published data to
investigate the possible contributions of the influences of the endplate and/or
sidewalls.
1) For the present jet, the effect of the endplate in the absence of the sidewalls is
considered to be insignificant.
2) In the presence of the sidewalls, the jets produce statistically different flow
fields compared to the jets without sidewalls, as clearly indicated by the
measured velocity field. The presence of the sidewalls seems to improve the
two-dimensionality as indicated by the measured flow field e.g. formation of
sharp peaks of the skewness factor at the edges. But again, their presence
seems to alter the three-dimensional development of the mean streamwise
velocity in the spanwise direction.
1) In the present work, x-wire probe hot wire anemometry was used to measure
the 2-dimensional flow fields of all configurations and no visualization of the
flow field was made. Therefore, visualizing the flow fields of the four
configurations is recommended for better insight and a clearer picture of the
presence of the sidewalls/or endplates.
Chapter 2 starts with brief introductions on turbulent flows, free shear flows,
boundary layer flows, and turbulent jets. Next, the descriptions of the velocity flow
fields of rectangular jets are given. A review on the formulation of the equation of
motion, the integral momentum equation, and the entrainment hypothesis are also
provided. Finally, various initial and boundary conditions of rectangular jets are
defined.
Chapter 3 reviews the literature devoted to study the turbulent jets issuing from
rectangular nozzles. The need of the present investigation is discussed based on the
available information.
Chapter 4 begins with a description of the jet facility used in the present investigation.
Following this, hot wire anemometry is introduced (i.e. constant temperature
anemometry, working principle, noise, advantages, disadvantages, selection of hot
wire probes, calibration, measuring the velocity components, a specification of the
measuring system used in the present investigation, and the experimental
uncertainty). Finally the experimental procedure is given.
Chapter 5 presents all results of this investigation. The effects of endplate and/or
sidewalls on mean and turbulent velocity fields of the studied jets are discussed.
Comparisons with previous studies available in the literature are also provided.
Chapter 2
THEORY
Turbulent flows enhance the ability to transfer momentum, energy and species of the
fluid particles in comparison to laminar flow which is desirable for the mixing
processes i.e. the demand usually is to have as rapid mixing as possible. Such
improvement will reduce the impact of pollutant streams released from chimneys of
industries into atmosphere, and for the mixing of various reactants in combustion
devices (Pope, 2002).
Figure 2.1 illustrates four types of free shear flows, these are, jets, mixing layer,
wake, and boundary layer. In plane jet flows (Figure 2.1-a), fluid leaves the nozzle to
larger area. The excess of jets velocity compared to the surrounding is responsible
for spreading the jet into that surrounding. In mixing layers, as illustrated in Figure
2.1-b, the fluid layers flows at different velocities which are brought together. The
wake flows generally develop downstream of streamlined or bluff body (Figure 2.1-
c). In these flows, turbulence is formed around the body which is filled up in the wake
region. Here, the mean velocity behind the object shows a defect in momentum
compared to the outside region (Bernard & Wallace, 2002). The last case is the
boundary layer flow. In this case, fluid flows past a plate can form a boundary layer
as shown in Figure 2.1-d. In such flows, a boundary layer is formed by the fluid on
the plate. The velocity distribution is zero on the plate due to the no-slip condition
and increasing gradually away from the plate (Douglas et al., 1995).
y, V
(a)
Plane jet x, U
y, V
(b)
Mixing layer x, U
y, V
(c)
Plane wake x, U
y, V
(d)
Boundary layer
x, U
Figure 2.1: Illustrations of two dimensional flows. Redrawn after Pope, (2002).
It is may be interesting to differentiate between the viscous and inviscid flow regions
regarding the flow over a flat plate. If we consider the imaginary line where U=0.99
U as base, we will find that the region under that line where the viscous forces are
considered (i.e. boundary layer) is termed as a viscous flow region while above that
imaginary line, the region is termed as invicid flow. In the invicid flow region, the
frictional effects are negligible and the velocity will remain nearly unchanged
(engel, 1997).
In the boundary layer region, the faster fluid layer will drag along the slower layer
due to the friction between the two layers which will result in drag force (or friction)
on it. The drag force per unit area is known as the shear stress, . Experimental studies
have shown that the shear stress is proportional to the velocity gradient. The shear
stress at plate surface can be expressed by
= du/dy (2.1)
where is the dynamic viscosity of the fluid which is a measure of the flow
resistance. Its unit is kg/ (m.s) or equivalently N.s/m2. The viscosities of liquids
decrease with temperature whereas for the gases, it increases (engel, 1997)
Figure 2.2: The development of boundary layer for flow over a flat plate and different
flow regimes. Redrawn after engel, (1997).
A finite thickness region is formed between two jets, responsible for the continuous
distribution of velocity, temperature, and species concentration. This region is called
turbulent jet boundary layer, in rare cases, it can be laminar at very low Reynolds
numbers. Boundary layer in a simple case can be found when a jet flows in a moving
medium at fixed velocity. In this case the thickness of the boundary layer in the initial
stages is equal to zero. And as the jet develops, the thickness of the boundary layer
increases, carrying the fluids particles which results in spreading of the jet
(Abramovich, 1963).
When the jet leaves its exit into an ambient atmosphere, its free surface is exposed to
the atmosphere; hence the pressure of the jet itself is an atmospheric pressure.
However, the pressure at exit can be also different from the atmospheric pressure.
Such boundary condition is maintained only if we take into account the surface
tension effect which is usually not considered, or in supersonic jet which can separate
itself from atmosphere with expansion or compression waves (White, 1994).
Figure 2.3 shows a round jet facility with the defining coordinates. A fluid is
discharged from a nozzle with diameter, Dr producing a nearly constant initial
velocity profile with velocity U. The flow is axisymmetric in the developed region
hence statistics depends on the axial and radial coordinates (x, and r) but not on
circumferential coordinate, and time. The velocity components in the x, r, and
coordinate directions are denoted by U, V, and W (Pope, 2002).
x U
W V
U
r
Dr nozzle
Flow
Figure 2.3: Illustration of the round jet facility and coordinate. Redrawn after Pope,
(2002).
Figure 2.4 shows non-circular nozzles shapes as well as circular one used by Mi et al.,
2000 to produce different jets. In the present investigation we will deal only with the
turbulent jets issuing from a rectangular nozzle.
Plane or two dimensional jets (i.e. the jet entrains fluid only in axial and lateral
directions), are produced from high aspect ratio nozzles. Their velocity fields are two
dimensional. Pope, (2002) stated that aspect ratio must be larger than 50 to obtain two
dimensional flow fields without any assistance from the sidewalls. Whereas in
rectangular or three dimensional jets, smaller aspect ratios are used (Rajaratanam,
(1976). In these jets, the velocity fields will be axisymmetric after the two
dimensional region and the length of the later region was found to be a function of the
aspect ratio (e.g. Rajaratanam, 1976, Krothapalli et al., 1981, Deo et al., 2007a). The
velocity flow fields of a rectangular jet are shown in Figure 1.1.
1.2
1.0
0.8
U/Uc
0.6
0.4
Quinn, 1992
WENS, AR=20
0.2 x/D=5.04
0.0
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
z/zc
The development of the mean streamwise velocity in the spanwise direction (i.e.
along the jet) of the present investigation (for NENS) is shown in Figure 5.3. As one
can see, uniform mean streamwise velocity distributions are observed, but further
downstream undulations are appeared which is more evidenced in the furthest
measured location. Other authors (e.g. Marsters, 1980, Tsuchiya et al., 1986, Quinn,
1992) have observed these undulations which were referred as saddle back shape
mean streamwise velocity profiles. The saddle back velocity profiles of Quinn, 1992
is shown in Figure 2.5. It is also of interest to mention that the present investigation
have revealed that the presence of the sidewalls will result in different mean
streamwise velocity profiles in the spanwise direction, in comparison to the free jet
case (Figure 5.3).
0.04
0.02
W / Uc
0.00
-0.02
Krothapalli et a., 1981
Re=10,000, AR=16.7, NENS
40 D 60 D 100 D
-0.04
0.0 0.1 0.2 0.3
z/x
Figure 2.6: Development of mean spanwise velocity in the spanwise direction. Data
extracted from Krothapalli et al., (1981).
0.5 Uc
y
yc
Flow
Figure 2.7: Illustration of jets development across the minor dimension of a nozzle
indicating the location of yc. Redrawn after modification from Deo, (2005).
The spread rate and the spread virtual origin of rectangular jets have been found to be
influenced by the different boundary and initial conditions (e.g. Deo et al., 2007a,
Deo et al., 2007b, Deo et al., 2007c, Deo et al., 2008).
(U0/Uc)2=B1[x/D+B2] (2.3)
Where U0 is the mean streamwise velocity at the exit plane, Uc is the local centerline
mean velocity, the slope, B1 is a measure of the centreline mean velocity decay rate
and B2 is the jets virtual origin. The values of B1 and B2 have been found to depend
on the various initial and boundary conditions (e.g. Deo et al., 2007a, Deo et al.,
2007b, Deo et al., 2007c, Deo et al., 2008).
(2.4)
(2.5)
a) Turbulent velocities
In turbulent flows, fluctuations will lead the velocity to rapidly vary in a random
function of time and space (White, 1994). If we consider a turbulent velocity, u' in
e.g. streamwise direction, then u' is defined as the deviation of U from its average
value u (u' = u U) as shown in Figure 2.8.
u' u = U + u'
Figure 2.8: Definition of mean and fluctuating turbulent of velocity. Redrawn after
White, (1994).
0.20
Cavo et., 2007
Re=21,000, AR=6, NENS
x/D
0.15
6
w'/Uc 11
0.10
0.05
0.00
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 2.9: Lateral profiles of the turbulent spanwise velocity at different downstream
locations. Data extracted from Cavo et., (2007).
a) Equation of motion
In this section, the equations of motion for the turbulent plane jet are discussed
(Rajaratanam, 1976, Chapter 1).
In a turbulent flow, the velocity of every particle that passed through a point at any
given time is given by V= (u, v and w), Then one would find the actual velocity
fluctuating about a mean value:
(2.7)
(2.8)
(2.9)
(2.10)
Where x-axis defines the streamwise direction, y-axis defines the lateral direction, and
z-axis defines the spanwise direction. In the following derivations, u, v, w and
are the mean and turbulent velocities in the streamwise, lateral and spanwsie
direction. It should be mentioned that throughout this thesis except this part (i.e. these
derivations), the mean velocities in the streamwise, lateral and spanwise directions
are denoted as U, V, and W respectively. Here, p is the mean pressure at any point, v
is the kinematic viscosity, and is the mass density of the fluid and t is the time
variable.
(2.11)
(2.12)
(2.13)
Integrating (2.12) with respect to y from y to a point located outside the jet, we
obtain:
where p, is the pressure outside the jet. Differentiating the above equation and
substituting in (2.11), we get:
( - ) (2.14)
The last term in the above equation is very small and can be neglected. Hence we
obtain the reduced equations of motion as:
(2.15)
and,
where p, is simply written as p for convenience. In (2.15), we could rewrite the last
two terms as:
where 1 and t are, respectively, the laminar and turbulent shear stresses and is the
coefficient of dynamic viscosity. In free turbulent flows, due to the absence of solid
boundaries, t is much larger than 1 and hence it is reasonable to neglect 1 and
rewrite (2.15) as:
(2.16)
Further, because in a large number of practical problems the pressure gradient in the
axial direction is negligibly small and also to study the jet under relatively simpler
conditions, let us set dp/dx = 0. Then (2.16) and (2.13) become:
(2.17)
(2.18)
which are the well-known equations of motion for the plane turbulent free jet with a
zero pressure gradient in the axial direction.
(2.19)
( ) = (by Liebnitz rule)
dy =
=0 (2.20)
Equation (2.20) tells us that the rate of change of the momentum flux in the x-
direction is zero; that is the moment flux in the x-direction is conserved (or
preserved).
Integrating (2.20):
=M0 (2.21)
The momentum flux M0, is an important physical quantity controlling the behavior of
the plane jet.
(2.22)
where D/Dt stands for the particle derivative and DE/Dt is the total rate of change of
the kinetic energy.
(2.23)
which says that the rate of decrease of the kinetic energy is equal to the rate at which
turbulence is produced. For our present purposes, we will rewrite the above equation
in a slightly different form.
dy=
( u) dy =
(2.24)
We can see from equation 2.24 that the rate of decrease of the kinetic energy flux is
equal to the rate at which turbulence is produced.
d) Entrainment hypothesis
If m is the rate of (forward) flow for unit length at any section of the jet:
(2.25)
If m is the flow from the nozzle, it is known from experiments that m/mo is greater
than unity and assumes very large values as x becomes large. That means, the jet
entrains a considerable amount of the surrounding fluid as it travels forward. We now
write:
(2.26)
a) Reynolds number
Experimental investigations have shown that the various characteristics of a jet (e.g.
growth and decay rates, mass and momentum flow rates, turbulent flow fields)
depend on Reynolds number (Namer & tgen, 1988, Deo et al., 2008). For jet
issuing from a rectangular nozzle, the jet exit Reynolds number, Re is defined using
the nozzle opening width D (minor dimension of the nozzle, Figure 1.1), centerline
mean velocity at the exit, U0 and kinematic viscosity, v of the tested fluid using the
expression
ReD=U0D/v (2.27)
b) Aspect ratio
The aspect ratio, AR of a jet issuing from a rectangular nozzle is the ratio of the long
side of the nozzle, L to the short side of the nozzle, D (see Figure 1.1)
AR=L/D (2.28)
Aspect ratio plays a very important role in determining the characteristics of a jet
issuing from rectangular nozzle. Therefore, much effort was given to study the
influence of this parameter (e.g. Quinn, 1992, Deo et al., 200a).
c) Nozzle type
However turbulent rectangular jets are produced from rectangular openings, still there
is a difference among these rectangular openings as long as the exit shape is
considered. The exit type (or shape) can be for example a smooth contraction nozzle,
a rectangular channel exit, a radially countered nozzle, or may be any exit with
sharped-edge. Different shapes of rectangular nozzles are shown in Figure 2.10.
Experimental investigations have shown that the exit type can strongly influence the
development of the velocity flow fields of these jets (e.g. Sfier, 1979, Deo et al.,
2007b).
a)
1,200
sidewalls
Axial direction, x
2,000
Spanwise
direction, z
Lateral direction, y
endplate
640
b)
endplate
640
Figure 2.11: Smooth contraction nozzle: a) with sidewalls and endplate, and b) with
endplate. Dimensions shown in the figure are the ones used in present investigation.
(All in mm).
On the other hand endplate (or front plate) is a plate flush at the nozzle exit plane to
prevent entraining from upstream of the nozzle and hence improve the two-
dimensionality of the flow fields. Endplate can be used either in the presence of the
sidewalls as shown in Figure 2.11a (e.g. Gutmark and Wygnanski, 1976) or in the
absence of the sidewalls as shown in Figure 2.11b (e.g. Quinn, 1992).
Since this thesis investigates the influence of the presence or the absence of the
endplate and/or sidewalls, more information and detailed discussions will be provided
in the following chapters.
Chapter 3
REVIEW
3.1 Literature review
As was demonstrated in previous chapters, turbulent rectangular jets are used in many
engineering fields but are also important to fundamental fluid mechanics. Thus,
numerous studies have been conducted on these jets.
This thesis reports experimental investigations on the effects of the sidewalls and/or
endplate on the development of turbulent rectangular jets. To understand the
mechanics of a rectangular jet, one needs to better understand the mechanics of plane
and round jets. Hence, a review on these jets will be addressed below.
A round jet is produced from a round as shown in Figure 2.3. Three regions of the
round jet can be identified based on the velocity fields; these are the near field, the
intermediate field and the far field. The near field region is also termed as the initial
flow or development region. In this region the velocity is nearly similar. It is named
as named the potential core. It is usually found within 0x/D6. The far field is the
fully developed (or self-similar) region where the thin shear layer approximations
hold. The near and intermediate fields together comprise the developing portion of
the jet (Ball & Pollard, 2007).
The flow fields of plane jets have been studied based on the two dimensional concept
that considers infinite length of the nozzle, though in many practical cases is not
maintained as a finite nozzles length is used (Seo & Kwon, 2005). The ideal plane jet
is statistically two-dimensional, in the dominant direction of the mean flow, x, and
cross-stream coordinate, y, whereas statistics are expected to be independent of the
spanwise coordinate (e.g. Deo, 2005).
The flow field of rectangular jets is characterized by the presence of three distinct
regions (see Figure 1.1), as already discussed, it is defined by the decay of the
centreline mean streamwise velocity, Uc, (Krothapalli et al., (1981), namely: (a) the
potential core region (x/D = 0 to 4 5), with almost constant Uc, which ends when the
two shear layers along the short dimension of the nozzle meet, (b) the two-
dimensional region (x/D = 45 to 2530) in which the velocity decays at a rate
roughly the same as that of a planar jet, and (c) the axisymmetric region, in which the
velocity decays as in an axisymmetric jet, originating at x/D = 2530, the location
where the two shear layers along the long dimension of the nozzle meet. The extent of
these regions has been found to be strongly dependant on the different initial and
boundary conditions such as Reynolds number e.g. Namer & tgen, 1988, Deo et
al., 2008, Suresh et al., 2008, aspect ratio e.g. Krothapalli et al., 1981, Deo et al.,
2007a, sidewalls e.g. Hitchman et al., 1990, Deo et al., 2007c, nozzle type e.g. Deo et
al., 2007b, and type of the initial boundary layer e.g. Hussain & Clark, 1977.
The effects of Reynolds number on the development of turbulent jets issuing from
rectangular nozzles have been extensively studied. Lemieux & Oosthuizen, (1985)
investigated the effects of Reynolds number on the flow characteristics of a turbulent
jet issuing from a rectangular slot of aspect ratio, AR=58 with sidewalls. The mean
velocity and turbulent stresses in the flow field were measured with hot wire
anemometry at six discharge Reynolds number values in the range 700-4,200. Their
results indicated a strong influence of the Reynolds number on the lateral and normal
shear stresses, jet decay and spreading rates, but smaller influence on the longitudinal
normal stress levels and no effects on the normalized mean velocity profiles. The
results of Lemieux & Oosthuizen, (1985) were limited to smaller range of Reynolds
number. Namer & tgen, (1988) extended the range of Reynolds number
investigated by Lemieux & Oosthuizen, (1985). They have studied a free plane air
jets of aspect ratio, AR=56 at Reynolds number over the range of 1,000 Re 7,000
using both HWA and LDA techniques. Namer & tgen, (1988) have found a strong
influence of Reynolds number on the flow fields of their jets. Specifically, they have
observed that increasing Reynolds number has decreased the jet potential core length.
This implies that, with the increase of the Reynolds number, the near field mixing has
been increased. The jet spread and decay rates were found to be decreasing with the
increase of Reynolds number. The effects of Reynolds numbers on the development
of rectangular jet was also assessed by Chua & Lua, (1998) for low aspect ratio,
AR=6 jets. Two Reynolds numbers were tested (2,759 and 5,517). Their hot wire
measurements showed that the rms velocity distributions established approximate
self-preservation by x/D=12 for both jets. But the mean velocity distributions became
self-similar at about x/D=4 and 6 at Reynolds number, Re=5,517 and 2,759
respectively. With the increase of the Re, a slight decrease in the spread rate was also
noticed. They have reported an increase of the mean streamwise velocity values at the
edges of both jets. This trend was more obvious for the jet with lower Reynolds
number. They have attributed this increase in the mean streamwise velocity values to
the presence of the sidewalls. This trend was also evidenced in previous
investigations where sidewalls were present e.g. Daves et al., (1975) and Moum et al.,
(1983).
1981). Trendscoste & Sforza, (1967) investigated a range of aspect ratio between 2.5
and 40. Their study found a decrease in length of the jets potential cores, an increase
in the decay of mean centerline velocity and an increase in jet spreading rate with an
increase in nozzle aspect ratio. Sfeir in 1979 measured different rectangular jets using
x-wire probe to assess the influence of aspect ratio over the range 10 AR 60. He
obtained similar results to Trentacoste & Sforza, (1967). Marsters & Fotheringham,
(1980) also investigated the influence of smaller aspect ratios between 3.4 and 12 at
Reynolds numbers between 33,000 and 57,700, using total head tubes and hot wire
anemometry. Their spanwise velocity distributions showed saddle-back behavior.
This saddle-back behavior is suppressed in the smaller aspect ratio jets. Many other
studies were followed focusing on the effects of aspect ratio, such as those of
Krothapalli et al., (1981), Tsuchiya et al., (1985) and Quinn, (1992). Krothapalli et
al., (1981) provided the mean and turbulent velocity fields using x- sensor hot wire
anemometry over an axial distance of 115 nozzle widths. Four free jets with different
aspect ratio (i.e. 5.5, 8.3, 12.5 and 16.6) were investigated in the absence of the
endplates and sidewalls. They have reported a dependence on the aspect ratio of the
location where the jet may be assumed to be of axisymmetric type. This location goes
upstream toward the nozzles exit with the decrease of the aspect ratio.
Quinn, (1992) also gave attention to the effect of aspect ratio on rectangular jets
issuing from rectangular slots with sharp-edge (with endplate and without sidewalls).
He investigated four aspect ratios (2, 5, 10 and 20). He further confirmed the
dependence of the velocity flow fields of rectangular jets on their aspect ratios. Later,
Lazanova & Stankov, (1998) investigated the effect of aspect ratio on the
development of rectangular jet in the range between 3 and 10. Their results also
confirmed the dependence of the velocity fields of rectangular jets on aspect ratio. In
2005, Mi et al., have undertaken an investigation to characterize the turbulent jets
issuing from high aspect ratio rectangular jets. Their aspect ratio range was varied
between 15 and 120 in absence of endplate and sidewalls. Their measurements were
conducted by a normal hot wire probe over a downstream distance of up to 160 slot
width at Reynolds number of 10,000. They found that these jets can be characterized
by three distinct zones: an initial quasi-plane jet zone, a transition zone and a final
quasi-axisymmetric-jet zone. The extent of the quasi-plane zone was found to
increase with the increase of aspect ratio while other regions decrease. Recently, Deo
et al., (2007a) investigated the effects of nozzle aspect ratios between 15 and 72 on
turbulent jet issuing from rectangular nozzle in the presence of the sidewalls and
absence of endplate. Single hot wire anemometry probe was used to collect
measurements up to 85 nozzle width at Reynolds number, Re= 1.80 x104. Their
results have further showed the dependence of the near and far flow fields of plane jet
flows on the nozzles aspect ratio.
Besides nozzle aspect ratio, experimental evidence revealed that the shape of the
rectangular nozzle (or nozzles type) can influence the jets flow fields. Many efforts
have been devoted to study the differences among different rectangular nozzle shapes.
Sfier, 1979 studied two groups of rectangular nozzles namely rectangular slots and
rectangular channels. He has found differences in the development of the jets
produced from the two nozzle shapes. Mi et al., (2000) investigated centerline mean
flow and turbulence characteristics of different jets issuing from different nozzle
types (i.e. a smoothly contoured circular, an elliptical, a triangular, a square, a
rectangular, a cross-shaped and a star-shaped). Their results have shown increased
mixing levels of the noncircular nozzles in comparison to the round one, particularly
the triangle nozzle has shown the greatest mixing level. Attention was also given to
the influence of nozzle type on turbulent jets issuing from round nozzles (e.g. Mi et
al., 2001, Antonia & Zhao, 2001). Mi et al., (2001) investigated axisymmetric free
jets from a contoured nozzle, an orifice plate and a pipe. They have found the jet
issuing from orifice plate has the highest mixing rate with the ambient. The type of
the round nozzle determines the initial streamwise velocity profile. Figure 3.1 shows
the various initial velocity profiles of different types of round nozzle of Mi et al.,
2001 (i.e. the smooth contraction nozzle produces a top-hat initial velocity profile,
sharped-edge rectangular slot produces a saddle back initial velocity profile and pipe
produces fully turbulent and developed initial velocity profiles). Antonia &
Zhao, (2001) studied two round jets. The first is issued from smoothly contracting
nozzle and the other from pipe. Their measurements confirmed the difference
between these two jet configurations though both jets reached the self-preservation at
the same downstream location.
1.5
1.0
U/Uc
0.5 Orifice
Pipe
Contraction
0.0
-0.4 -0.2 0.0 0.2 0.4
r/Dr
Figure 3.1: Initial velocity profiles from a sharp-edged orifice plate, a smooth
contraction nozzle, and a long pipe for a round jet. Data extracted from Mi et al.,
(2001).
The relationship of turbulent jets and the type of initial boundary layer (e.g. laminar
or turbulent) have also been given attention. Hussain & Clark, (1977) have
investigated the effect of the type of the initial boundary condition on the
development of turbulent jets issuing from nozzle type and channel of 44 aspect ratio
with endplate using x-wire probe. The two laminar boundary layer jets were produced
from the nozzle type at Re=32,550 and 81,400. And at the same Reynolds numbers,
the two turbulent boundary layer jets from the rectangular channel exit were
produced. Their investigation revealed that, the laminar boundary layer lead to higher
growth rate of longitudinal turbulence intensity, and the mass flux. An increase in the
total average streamwise momentum was observed for both laminar and turbulent
boundary condition jets, higher in the former one. Chambers et al., (1985) measured
the mean and turbulent velocity and temperature fields of two plane jet (one issuing
from contraction with initial top hat velocity profile and initial laminar boundary
condition, and the other issuing from two dimensional duct with fully developed
velocity profile) using schlieren photography and HWA. Their results have shown
strong influence of the type of the initial boundary layer on the formation of the
mixing layer structures. The mixing layer structures were strongly organized and
symmetric about the centerline in the laminar boundary condition jet while in the
other configuration, turbulent structures were three dimensional and asymmetric
about the centerline.
The above discussions reveal the importance of the various initial and boundary
conditions to the development of rectangular and plane jets. Boundary conditions
such as sidewalls (two parallel walls confining the jet at the planes of the nozzles
short sides) and endplates (or front plates: a wall flush at the nozzle exit plane), have
been used in many studies. Many investigators have used endplate (e.g. Hussain &
Clark, 1977, Quinn, 1992), sidewalls (e.g. Bradbury, 1965, Deo et al., 2007a, Suresh
et al., 2008) or both together; sidewalls and endplate (e.g. Gutmark & Wygnanski,
1976, Browne et al., 1986) to control the dimensionality of the flow fields.
1) jets without (No) Endplate and without (No) Sidewalls (free jets), NENS
(e.g. Trentacoste & Sforza, 1967, Krothapalli et al., 1981, Namer &
tgen, 1988, Papadakis & Staiano, 1993, Lazanova & Stankov, 1998,
Mi et al., 2005, Cavo et al., 2007, Deo et al., 2007c).
As evidenced from the above classification, sidewalls and/or endplate have been used
in many investigations at various initial conditions (e.g. different aspect ratio and
Reynolds numbers, nozzle types), but few experiments have been devoted to study
their effects on turbulent jets issuing from rectangular nozzles (Hitchman et al., 1990,
Papadakis & Staiano, 1993 and Deo et al., 2007c).
It is worth noting that both of the above investigations have focused on the same
aspect ratio (i.e. AR=60) and similar Reynolds number ~7,000. For this large aspect
ratio, the jet centreline velocity is expected to have planar jet behaviour up to a
considerable downstream distance from the exit nozzle, whereas for rectangular jets
of smaller aspect ratio a weaker planar jet behaviour may be anticipated (Mi et al.,
2005). In other words, the downstream location where the centreline velocity first
assumes axisymmetric characteristics is expected to move upstream, toward the exit
of the jet, with the decrease of the aspect ratio (Krothapalli et al., 1981).Therefore,
there is a need to explore the influence of the sidewalls on the development and
mixing, in the near field (i.e. potential core) and in the two-dimensional region (i.e.
self persevering), for jets issuing from rectangular nozzles of moderate and smaller
aspect ratio. Our investigation will try to provide information on this need.
Studying the influence of an endplate, Papadakis & Staiano, (1993) measured the
flow field of two jets, one with endplate and the other without, issuing from a
rectangular nozzle of aspect ratio, AR=40, at Reynolds number, Re=9,800. Based on
LDA measurements, they have found that both jets produce very similar velocity
profiles indicating that the endplate did not have significant effect on the jet velocity
distribution. On the other hand Abdel-Rahman et al., (1997) investigating the use of
an endplate at the exit plane of an axisymmetric round jet have found that,
introducing an endplate at the nozzle exit caused a reduction in the velocity spreading
rate and reduced kinematic mass flow rate. The reduction in the kinematic mass flow
rate of Abdel-Rahman et al., (1997) investigation is in agreement with Kotsovinos,
(1978) and Schneider, (1985) studies. Kotsovinos, (1978) reformulated the equation
of the momentum flow in a turbulent jet and used it to show that the conservation of
momentum is an inaccurate approximation for a plane jet with an endplate.
Furthermore Schneider, (1985) has shown that in a turbulent jet (round or plane) with
endplate the axial velocity decreases more rapidly and the jet suffers momentum
reduction in the streamwise direction due to the induced outer flow, concluding that
the momentum is conserved only for the case of a jet without endplate. It should be
noted that, the investigation of Papadakis & Staiano, (1993) was limited since most of
the reported profiles were focused on the near mean field of only (i.e. x/D18).
Furthermore, the findings of Papadakis & Staiano, (1993) were based on experiments
conducted in the absence of sidewalls. Since sidewalls were found to strongly affect
the development of rectangular/plane jets (e.g. Deo et al., 2007c), it will be interesting
to study the influence of endplates on jets with sidewalls beside jet without sidewalls.
Information on the influence of the endplate in the presence and absence of sidewalls
will be provided in this thesis.
In summary, the above review reveals that there is a shortage of information on the
influence of the 1) endplate, 2) sidewalls, and 3) both endplate and sidewalls on the
development of turbulent rectangular jets although these flows are under investigation
since many decades. Therefore, the present investigation will try to provide
information for these needs by investigating a rectangular jet of aspect ratio, AR= 15
at three Reynolds number, Re=10,000, 20,000 and 30,000 using x-wire probe HWA.
Chapter 4
EXPERIMENTATION
a)
1,200
b) sidewalls
Axial direction, x
2,000
Spanwise
direction, z
Lateral direction, y
endplate
640
c)
endplate
640
Figure 4.1: Experimental apparatus, a) jet facility, and b) rectangular nozzle with
endplate and sidewalls, and c) rectangular nozzle with endplate. (all dimensions in
mm).
4.1.3 Blower
A centrifugal blower air supply system of 1.5 kW power was used to provide the
airflow to the diffuser, which decelerates the flow and converts the dynamic pressure
to static in order to minimize energy losses and preserve the uniformity of the flow
field.
Gutmark & WEWS 13x500 1560 143 2000 2000 2013 1.28 13.93 155 14
Wygnanski (120H)
(1976)
Hithcman WEWS 10x600 748 78 1850 1100 2010 2.47 14.1 201 25.7
et al., (75H)
(1990)
Antonia et WEWS 12.7x254 889 99 1500 700 254 1.69 7.07 20 2.56
al., (1983) (70H)
a)
b)
+
R3 Rprobe
Error signaling
Figure 4.2: Hot wire anemometry components a) typical x-wire probe, and b) the
Wheatstone bridge. Redrawn after AN-1003, Hot wire & film anemometry systems,
Users Manual.
And they are i) the probe, ii) electronic feedback circuit which drives and produces a
voltage proportional to velocity, and iii) signal conditioning circuit which provides
noise reduction and amplification (AN-1003, Hot wire & film anemometry systems,
Users Manual).
The sensitivity of the circuit to changes of probe resistance and adjust the heating or
the frequency response of the circuit depends on three variables 1) overheat ratio
(OHR), 2) loop gain, and 3) cable compensation. The OHR is the ratio of the
electrical resistance of the heated wire to the electrical resistance of unheated wire at
reference atmospheric condition. It is usually set by the operator to fix the
temperature difference between the probe and ambient air. Higher OHR will result in
greater system bandwidth and higher sensitivity of the probe when the fluid flow is
changing. The probe may burn at some high temperature hence it is advised to set
OHR that is recommended by the probes manufacturer. The probe also can be burnt
due to improper damping of the anemometer circuit that produces oscillations. (AN-
1003, Hot wire & film anemometry systems, Users Manual)
The input of the signal conditioning unit is the voltage across the bridge. While the
output can be used and connected to an oscilloscope to give immediate observation of
the fluid velocity. Meanwhile it is recorded by a computer by using data acquisition
system, data storage devices or analog to digital converter. The conversation of output
voltages to fluid velocities needs building a calibration graph since the relationship
between the velocity and voltage is not linear. This relationship can be plotted using
5th degree polynomial. The calibration must be performed for every working channel.
If the voltage signal is acquired by an A/D converter, extra linear conversion must be
performed (AN-1003, Hot wire & film anemometry systems, Users Manual).
Conditioning of the signal which is produced by the fluctuating voltage on the probe
must be done to have useful measurements. The main functions of signal conditioning
are to amplify, offset and filter the signal. This can be done by the experimenter (AN-
1003, Hot wire & film anemometry systems, Users Manual). A signal before and
after conditioning is shown in Figure 4.3.
a) b)
5 10 15 20 25 30 35 40 45 5 10 15 20 25 30 35 40 45
Figure 4.3: Signal a) before conditioning, and b) after conditioning. Redrawn after
AN-1003, Hot wire & film anemometry systems, Users Manual.
4.2.1.2 Noise
To obtain stable signal, experimenter must avoid any noise which can strongly
influence that signal. Among the noises these are due to the dynamic stresses caused
by improper welding or soldering of the sensors; poor electrical connection (e.g. dirty
or faulty BNC connectors) can add much noise due to the sudden voltage fluctuation;
if the wires are not correctly placed, electromagnetic inference is may produced from
external sources; cross talking between the channels due to ground loops; amplifier
circuit will add noise due to heating; poor calibration of A/D conversion units. Noises
can be introduced from other surrounding systems e.g. computers, motors etc (AN-
1003, Hot wire & film anemometry systems, Users Manual).
3.0
2.0
20 40 60 80 100
velocity, m/s
In this investigation, the x-wire was calibrated at nozzle exit in comparison with pitot
tube measurements, using on average, 10 calibration points.
In the present investigation, x-probe was used to measure the two velocity
components (V1, and V2). In this case, two sensors as shown in Figure 4.5 are made
sensitive to V2 by orienting them to the mean flow. The two sensors are oriented at
90o degree to each other (Goldstein, 1996).
V2A,eff=(V1 cos 1-V2 sin 1 - V2 sin 1)2 + KT2(V1 sin 1 + V2 cos 1)2+KN2V32 (4.1)
V2B,eff=(V1 cos 1-V2 cos 1 - V2 sin 1)2 + KT2(V1 cos 1 + V2 sin 1)2+KN2V32 (4.2)
The coordinates are usually selected such that =0, if the sensors are sufficiently
long that KT0 and KN1, the Equations 4.1 and 4.2 will reduce to
In large turbulence intensities, equations 4.3 and 4.4 cannot be further reduced,
however if v3 is small then the last term can be ignored, rearranging equations 4.3 and
4.4 give:
1 V1
V2
V
V3
Measuring system
1) The anemometer used in this investigation was AN-1003 model from A.A.
Lab System LTD.
2) X-wire probe from Dantec Dynamics (5 micron in diameter and 1 mm in
length).
3) Monitoring the signal was done on a GoldStar oscilloscope (OS 9020G-
20MHz).
4) The overheat ratio (OHR) was 1.4.
5) The sampling frequency was 2 kHz for a time series of 16,000 samples.
6) A PCL 814 data acquisition board attached to a personal computer was used.
(100 k Hz, 16 channels A/D converter with 14-bit resolution).
Experimental conditions
a) Centreline
1) Measurements were collected at ReD =10,000, 20,000 and 30,000 up to
x/D=35, for all configurations (i.e. NENS, WENS, NEWS, and WEWS).
Chapter 5
at the edges of the jets NEWS and WEWS at distances from the nozzle in the range
x/D = 3-30 indicate that the presence of sidewalls leads to this increase.
Downstream, focusing on the effect of the endplates on the free jet, comparing the
cases with no sidewalls (NENS and WENS), the two jets exhibit similar trends and
the measurements collapse for all part of the distributions, indicating an insignificant
effect of the endplate. These findings agree with those of Papadakis & Staiano,
(1993) who carried out LDA measurements in a free rectangular jet of aspect ratio
AR=40 with and without endplate. They also found an insignificant effect of the
endplate on the transverse mean streamwise velocity distributions. Similar remarks
hold for the two configurations comprising sidewalls. Again no difference due to the
presence of the endplate can be identified (NENS, WENS). It can therefore be
concluded that the dominant effect on the streamwise velocity profile is due to the
presence or absence of sidewalls whereas the presence of the endplate has an
insignificant effect.
a)
1.0
X/D=0
NENS
WENS
NEWS
WEWS
0.5
U/Uc
0.0
0.0 0.5 1.0 1.5
y/yc
b)
1 .0
X /D = 1
NENS
W ENS
NEW S
W EW S
U /U
c
0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
c)
1 .0
X /D = 2
NENS
W ENS
NEW S
W EW S
U /U
c
0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
d)
1 .0
X /D = 2 .5
N EN S
W EN S
N EW S
W EW S
U /U
c
0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
e)
1 .0
X /D = 3
N EN S
W EN S
N EW S
W EW S
U /U
c
0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.1: Mean streamwise velocity profiles on the central xy plane (contd).
f)
1 .0
X /D = 4
NENS
W ENS
NEW S
W EW S
U /U
c 0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
g)
1 .0
X /D = 5
NENS
W ENS
NEW S
W EW S
U /U
c
0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
h)
1 .0
X /D = 1 0
NENS
W ENS
NEW S
W EW S
U /U
c
0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.1: Mean streamwise velocity profiles on the central xy plane (contd).
j)
1 ,0 X /D = 1 5
NENS
W ENS
NEW S
W EW S
P a p a d a k is & S ta ia n o , ( 1 9 9 3 )
A R = 4 0 , R e = 9 ,8 0 0 , x /D = 1 8
NENS
U /U
c
W ENS
0 ,5
0 ,0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
k)
1 .0
X /D = 2 0
NENS
W ENS
NEW S
W EW S
0 .5
U /U
c
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
l)
1 .0
X /D = 2 5
NENS
W ENS
NEW S
W EW S
U /U
c
0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.1: Mean streamwise velocity profiles on the central xy plane (contd).
m)
1 .0 X /D = 3 0
NENS
W ENS
NEW S
W EW S
U /U
c 0 .5
0 .0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
n)
1 ,0
X /D = 3 5
N E N S
W E N S
N E W S
W E W S
U /U
0 ,5
c
0 ,0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.1: Mean streamwise velocity profiles on the central xy plane (contd).
increasing the Reynolds number from 1,500 to 16,000, the distance where their
mean velocity profiles first assumed self-similar was decreased from 20 nozzle width
to 5 nozzle width. The nozzle type has been found also to affect the similarity of the
streamwise velocity profiles. Deo et al., (2007b) for AR=72 and Re=1.8 x 104 found
that when the ratio of exit radii, r and nozzle width, D (i.e. r/D) was increased from
zero to 3.6, the location where the mean velocity profiles first assumed self-similarity
decreased from 20 nozzle widths to 5 nozzle widths. Besides Reynolds number and
nozzle type, aspect ratio has been found to affect the similarity of the streamwise
profiles. Deo et al. (2007a) have investigated the effect of aspect ratio for Re=1.8 x
104. They found that with the increase of aspect ratio from 15 to 72 the downstream
location where their mean streamwise velocity profiles first assume self-similar was
decreased from x/D= 20 to 5. The location where their mean streamwise velocity
profiles first show self-preservation for AR=15 (i.e. x/D=20) is not in agreement with
what we found for the same aspect ratio (i.e. x/D=5). This difference can be attributed
to differences in the experimental conditions between our and their investigations e.g.
the nozzle shape i.e. we have used smooth contraction nozzle whereas they have used
radially contoured nozzle.
a)
1 ,0
NENS
x /d
0
1
2
2 .5
3
U /U
c
0 ,5 4
5
10
15
20
25
30
0 ,0 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.2: Self similarity profiles of mean streamwise velocity on the central xy
plane for a) NENS, b) WENS, c) NEWS and d) WEWS.
b)
1 ,0
W ENS
x /d
0
1
2
2 .5
U /U
c 3
0 ,5
4
5
10
15
20
25
30
0 ,0 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
c)
1 ,0
NEW S
x /d
0
1
2
2 .5
3
U /U
c
0 ,5 4
5
10
15
20
25
30
0 ,0 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
1 ,0
W EW S
x /d
0
1
2
2 .5
U /U
c
0 ,5 3
4
5
10
15
20
25
30
0 ,0 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
As shown in Figure 5.5a at x/D=2, all jets exhibit similar characteristics having
approximately a uniform mean streamwise velocity profile. For the jet without
sidewalls a rather thin shear layer is observed. At further downstream locations
(x/D=6 and 25) differences are observed between the two test cases. At x/D=6 the
distribution in the cases of the jets with sidewalls presents a local minimum at the
central axis and a peak close to the wall. In the cases of the jets without sidewalls a
local minimum is observed between the centre and the edge of the jets where a
thicker shear layer is developing. At x/D=25 both test cases present a local maximum
at the centre and a local minimum in between the centre and the edges. For the jet
with sidewalls (NEWS) the maximum of the distribution is observed very close to the
wall. For the jets without sidewalls the distributions reached close to the shear layer a
local maximum of the same order as the values at the centre on the wall. In this case
the shear layer thickness has become comparable to the dimensions of the jet in z
direction. Undulations of the spanwise profiles of the mean streamwise velocity have
been also observed in previous investigations, characterised as saddle back profiles
(e.g. Marsters, 1980, Tsuchiya et al., 1986, Quinn, 1992).
a)
X /D = 2
1 ,0 NENS
NEW S
U /U
C
0 ,5
0 ,0
0 2 4 6 8 10 12 14
Z /D
b)
X /D = 6
1 ,0 NENS
NEW S
U /UC
0 ,5
0 ,0
0 2 4 6 8 10 12 14
Z /D
c)
X /D = 2 5
1 ,0 NENS
NEW S
U/UC
0 ,5
0 ,0
0 2 4 6 8 10 12 14
Z /D
0 ,0 2
c
V/ U
0 ,0 0
-0 ,0 2
0 ,0 0 ,5 1 ,0 1 ,5
y /y c
b)
0 .0 6
X /D = 1
N E N S
W E N S
0 .0 4
N E W S
W E W S
V /U
c
0 .0 2
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
c)
0 .0 6
X /D = 2
N E N S
0 .0 4 W E N S
N E W S
W E W S
0 .0 2
V /U
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
d)
0 .0 6
X /D = 2 .5
N EN S
0 .0 4 W EN S
N EW S
W EW S
0 .0 2
V /U
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.4: Mean lateral velocity profiles on the central xy plane (contd).
e)
0 .0 6
X /D = 3
N E N S
0 .0 4 W E N S
N E W S
W E W S
0 .0 2
V /U
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
f)
0 .0 6
X /D = 4
N EN S
0 .0 4 W EN S
N EW S
W EW S
0 .0 2
V /U
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
g)
0 .0 6
X /D = 5
N E N S
0 .0 4 W E N S
N E W S
W E W S
V /U
c
0 .0 2
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.4: Mean lateral velocity profiles on the central xy plane (contd).
h)
0 .0 6
X /D = 1 0
N EN S
0 .0 4 W EN S
N EW S
W EW S
V /U
c
0 .0 2
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
j)
0 .0 6
X /D = 1 5
NENS
W ENS
0 .0 4
NEW S
W EW S
0 .0 2
V /U
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
k)
0 .0 6
X /D = 2 0
NENS
0 .0 4 W ENS
NEW S
W EW S
0 .0 2
V /U
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.4: Mean lateral velocity profiles on the central xy plane (contd).
l)
0 .0 6
X /D = 2 5
NENS
0 .0 4 W ENS
NEW S
W EW S
0 .0 2
V /U
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
m)
0 .0 6
X /D = 3 0
N E N S
0 .0 4 W E N S
N E W S
W E W S
V /U
c
0 .0 2
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
n)
0 .0 6
X /D = 3 5
N E N S
0 .0 4 W E N S
N E W S
W E W S
V /U
0 .0 2
c
0 .0 0
-0 .0 2
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.4: Mean lateral velocity profiles on the central xy plane (contd)
a)
0 ,0 6
N E N S
x /d
0 ,0 4 0
1
2
2 .5
V /U
0 ,0 2 3
c
4
5
1 0
0 ,0 0 1 5
2 0
2 5
3 0
-0 ,0 2 3 5
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
b)
0 .0 6
W ENS
x /d
0 .0 4 0
1
2
2 .5
V /U
c
0 .0 2 3
4
5
10
0 .0 0 15
20
25
30
-0 .0 2 35
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.5: Self similarity profiles of mean lateral velocity on the central xy
plane a) NENS, b) WENS, c) NEWS and d) WEWS.
c)
0 ,0 6
N E W S
x /d
0 ,0 4 0
1
2
2 .5
V / U 0 ,0 2 3
c
4
5
1 0
0 ,0 0 1 5
2 0
2 5
3 0
-0 ,0 2 3 5
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
0 .0 6
W EW S
x /d
0 .0 4 0
1
2
2 .5
V /U
c
0 .0 2 3
4
5
10
0 .0 0 15
20
25
30
-0 .0 2 35
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
Figure 5.5: Self similarity profiles of mean lateral velocity on the central xy
plane for a) NENS, b) WENS, c) NEWS and d) WEWS (contd).
The evolution of the half velocity width for the four test cases for Re=20,000 is
presented in Figure 5.6. All measurements almost collapse in the first stages of
development, within or just outside the potential core, whereas minor differences may
be identified further downstream. The spreading parameters, A1 and A2 deduced from
Figure 5.6 for x/D 10 have been included in table 5.1. The virtual origin is in all
cases close to the actual jet exit, on the negative side for the cases without sidewalls
(NENS, -0.17 D and WENS, -0.18 D) and on the positive one for the cases with
sidewalls (NEWS, 0.14 D and WEWS, 0.15 D).
7 ,5
R e= 20000
N E N S W E N S
N E W S W E W S
6 ,0 D eo et a l. ( 2 0 0 7 c ) , R e = 7 0 0 0 , A R = 6 0
N E N S N E W S
4 ,5
y /D
c
3 ,0
1 ,5
0 ,0
0 10 20 30 40 50
x /d
Considering the effect of the sidewalls on the free jet (NENS and NEWS), it is seen
that the jet with no endplate and no sidewalls (NENS) spreads faster having a spread
rate A1=0.126, compared to the jet with no endplate and with sidewalls (NEWS),
A1=0.105. These findings agree with those of Deo et al., (2007c), in their
investigation on the influence of the sidewalls on a plane jet (without endplate) of
AR=60 and Re=7,000, where similar trends were observed. They found that the jet
without sidewalls (NEWS) spreads at a faster rate of 0.14 compared to 0.11 for the jet
with sidewalls. Results of Deo et al., (2007c) are included in Figure 5.6 for
comparison. These findings are also in good agreement with Hitchman et al., (1990)
who obtained a spread rate of 0.106 for the jet without sidewalls and 0.083 for the jet
with sidewalls.
Studying the effect of endplate on the free jets (NENS, WENS), we have found that,
the two jets spread nearly in the same rate (NENS 0.126, WENS 0.129). This
indicates that insignificant effect of the endplate on the spread of the studied jet.
These findings agree with the findings of Papadakis & Staiano, (1993). Their LDA
measurements have also, shown insignificant effect of endplates on the spreading rate
of the jet (AR=40, Re=9800), whereas Kotsovinos, (1978) studying a plane jet out of
a wall associated jets with endplates with slower spreading rates in line with a
reduction in momentum. Abdel-Rahman et al., (1998) studying the effect of the
endplate on an axisymmetric jet, using LDA concluded that a reduction in momentum,
slower spread rate but similar decay are associated to the jet with endplate.
Based on these findings, it is may be concluded that, the spread rate, A1, seems to
depend more on the presence or absence of the sidewalls and less to the presence or
absence of the endplate, which seems to have an effect only when the sidewalls are
installed.
The initial mass flow rate of all configurations is almost identical. In the first stages
corresponding to the potential region the values collapse and thereafter the trends
distinguish the jets incorporating sidewalls (NEWS, WEWS) from those without
sidewalls (NENS, WENS). In the range x/D=5-25 the jets incorporating sidewalls
entrain ambient fluid more effectively than the jets with no sidewalls whereas this
trend is reversed at successive stations and the mass flow rate increase declines for
the jets with sidewalls (NEWS, WEWS). The differences between the configurations
incorporating sidewalls and those which do not may be attributed to the two
dimensional characteristics imposed by the sidewalls. In the early stages the mass
flow at the central xy plane is increasing slower in the absence of sidewalls due to the
spreading in the spanwise direction. When the 3D jets with no sidewalls are
beginning to attain an axisymmetric structure, the flow rate due to the rearrangement
of the flow is increasing faster than for the two dimensional jets with sidewalls, which
gradually lose their potential to entrain ambient fluid.
4.5
4.0
3.5
3.0
0
m/m
2.5
Re=20,000
2.0 NENS
WENS
1.5
NEWS
1.0 WEWS
0 10 20 30 40 50
x/D
1.4
1.2
1.0
0.8
0
M/M
0.6
Re=20,000
0.4 NENS
WENS
0.2 NEWS
WEWS
0.0
0 10 20 30 40 50
x/D
Assessing the effects of the endplate on jet without sidewalls (NENS and WENS),
when the sidewalls are not present, the endplate seems to have an insignificant impact
on the decay characteristics for all Reynolds numbers tested (configurations NENS
and WENS). The decay rate increases from Re=10,000 to Re=20,000 and then shows
a small reduction from Re=20,000 to Re=30,000 for both configurations, as if it had
reached saturation. The decay curves in Figures 5.9-a, b and c are hardly
distinguishable and the corresponding values of B1 and B2 remain very close for both
configurations (see also table 5.1 and Figure 5.10). Insignificant influence of the
endplate on the decay characteristics has been also found for jets of AR=40, without
sidewalls, for Re=9800 by Papadakis & Staiano, (1993). The results of Papadakis &
Staiano, (1993) are included in Figure 5.9a.
a)
7 ,5 R e=10000
NENS
W ENS
6 ,0 NEW S
W EW S
P a p a d a k is & S ta ia n o , ( 1 9 9 3 )
AR =40, R e=9800
4 ,5 NENS
2
W ENS
(U / U )
c
0
3 ,0
1 ,5
0 ,0
0 5 10 15 20 25 30 35
X /D
b)
7 ,5
R e=20000
N E N S
6 ,0
W E N S
N E W S
W E W S
4 ,5
2
(U / U )
c
0
3 ,0
1 ,5
0 ,0
0 5 10 15 20 25 30 35
X /D
c)
7 ,5
R e=30000
N E N S
6 ,0
W E N S
N E W S
W E W S
4 ,5
2
( U/ U)
c
0
3 ,0
1 ,5
0 ,0
0 5 10 15 20 25 30 35
X /D
Focusing on the effects of the sidewalls on jet without endplate (NENS and NEWS),
we will find that at Re=10,000, the centreline velocity in the case of NENS decays
faster than NEWS after the potential region, in accordance with Deo et al., (2007c)
findings. A switch is observed at Re=20,000 regarding the trends of the two test
cases. The curves in Figure 5.9-b almost collapse, and a scarcely discernible
difference is observed in a small area after the potential core where the centreline
velocity of the jet with no sidewalls takes higher values. At Re=30,000 the relation
observed at Re=10,000 has been completely reversed, and the centreline mean
velocity of the jet with no sidewalls decays at a slower rate than in the presence of the
sidewalls, in agreement with the work of Hitchman et al., (1990), in which slower
decay rate was associated with the jet without sidewalls.
The Reynolds dependence and the effect of the sidewalls in this series of experiments
is notably different to that found by Namer & tgen, (1988) for moderate Reynolds
number jets (Re=1,000-7,000) of AR=56 without sidewalls and Deo et al., (2008) in a
large range of Reynolds numbers (Re=1,500-57,500) for jets of aspect ratios AR=36
or 60 with sidewalls. Both these investigations indicate a decrease in B1 as well as in
A1 with increasing Reynolds number. Taking into account the findings of Deo et al.,
(2007a) the different dependence of B1 on Reynolds number should be mainly
attributed to the low aspect ratio of the currently presented jets and the resulting
interplay between the effects of AR, Re and that of sidewalls. In the presence of
sidewalls the endplate has a significant effect on the Reynolds number dependence of
B1 and B2. As indicated by the values in table 5.1 and the graphs in Figure 5.10a, B1 is
increasing with the Reynolds number in a similar manner to that of the NEWS jet but
the corresponding curve has been shifted to higher values. It is interesting to note that
corresponding although opposite, trends are observed in Figure 5.10-b for the
dependence of B2 on configuration and Reynolds number.
a)
0.20
0.18
1
0.16
Decay rate, B
0.14 NENS
WENS
0.12 NEWS
WEWS
0.10
10000 20000 30000
Reynolds number, Re
b)
6
5
NENS
WENS
2
Virtual Origin (Decay), B
4
NEWS
3 WEWS
2
-1
Table 5.1 Jet characteristics including far field mean stream wise velocity decay, B1,
virtual origin, B2, spread rate, A1, spread virtual origin, A2 and nozzle type of the present
investigation compared to other published data
Source Aspect Reynolds Jet Decay Virtual Spread Spread Nozzle
ratio, number, conditio rate, Origin, rate, Virtual type
AR Re n B1 B2 A1 origin,
A2
The lateral distributions of the streamwise turbulent velocity, u, on the central xy-
plane, for all four configurations at Re=20,000, are shown in Figures 5.11-a, b, c, d, e,
f, g, h, j, k, l, m and n 6.8-a, b, c and d at x/D=0, 1, 2, 2.5, 3, 4, 5, 10, 15, 20, 25, 30
and 35 respectively. Peak values are within the range 16-24 % of Uc, increasing
slightly with the distance from the nozzle. In the early stages of development peaks
are close to y/yc=1 moving towards the centre of the jet at subsequent locations.
Values at the centre are quite low in the potential zone increasing gradually
downstream and reaching values close to peak values after the merging of the two
shear layers.
The presence of the endplate has an insignificant effect when the sidewalls have not
been implemented and the measurements for NENS and WENS jets almost collapse
at all stations. In the results of the configurations comprising sidewalls (NEWS,
WEWS), in the near field, both jets exhibit smaller values at the shear layers and at
edges of the jets in comparison to the free jets (NENS, WENS). In the far field, the
a)
0.05
x/D=0
NENS
0.04 WENS
NEWS
WEWS
u'/Uc
0.03
0.02
0.01
0.0 0.2 0.4 0.6 0.8 1.0 1.2
y/yc
b)
0 .2 5
X /D = 1
N E N S
0 .2 0
W E N S
N E W S
0 .1 5 W E W S
u '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
c)
0 .2 5
X /D = 2
N E N S
0 .2 0
W E N S
N E W S
0 .1 5 W E W S
u '/U
c 0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
d)
0 .2 5
X /D = 2 .5
0 .2 0 N E N S
W E N S
N E W S
0 .1 5 W E W S
u '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
e)
0 .2 5
X /D = 3
N E N S
0 .2 0 W E N S
N E W S
0 .1 5
W E W S
u '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
f)
0 .2 5
X /D = 4
N E N S
0 .2 0
W E N S
N E W S
0 .1 5
W E W S
u '/U
c 0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
g)
0 .2 5
X /D = 5
N E N S
0 .2 0 W E N S
N E W S
0 .1 5
W E W S
u '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
h)
0 .2 5
X /D = 1 0
N E N S
0 .2 0 W E N S
N E W S
0 .1 5
W E W S
u '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
j)
0 .2 5
X /D = 1 5
0 .2 0 N E N S
W E N S
N E W S
0 .1 5 W E W S
u '/U
c 0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
k)
0 .2 5
X /D = 2 0
N E N S
0 .2 0 W E N S
N E W S
0 .1 5
W E W S
u '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
l)
0 .2 5
X /D = 2 5
0 .2 0
N E N S
W E N S
N E W S
0 .1 5 W E W S
u '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
m)
0 ,2 5
X /D = 3 0
N E N S
0 ,2 0
W E N S
N E W S
0 ,1 5 W E W S
u '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
n)
0 ,2 5
X /D = 3 5
NENS
0 ,2 0 W ENS
NEW S
W EW S
0 ,1 5
c
u '/U
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
a) 0 .2 5
N E N S
0 .2 0 x /d
0
1
0 .1 5
2
2 .5
u '/U 3
c 4
0 .1 0 5
10
15
0 .0 5 20
25
30
35
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
b)
0 .2 5
W EN S
0 .2 0 x /d
0
1
0 .1 5 2
2 .5
u '/U
c
3
4
0 .1 0 5
10
15
0 .0 5 20
25
30
35
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
c)
0 .2 5
NEW S
0 .2 0 x /d
0
1
2
0 .1 5
2 .5
u '/U
c
3
4
0 .1 0 5
10
15
0 .0 5 20
25
30
35
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
d)
0 ,2 5
W EW S
0 ,2 0 x /d
0
1
0 ,1 5 2
2 .5
u '/U 3
c
4
0 ,1 0
5
10
15
0 ,0 5 20
25
30
0 ,0 0 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
a)
0 ,2 0
x /D = 1
N EN S
0 ,1 5
W EN S
N EW S
W EW S
0 ,1 0
u '/U
0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
b)
0 ,2 0
x /D = 4
N EN S
W EN S
0 ,1 5 N EW S
W EW S
0 ,1 0
u '/U
0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
c)
0 ,2 0
x /D = 1 5
N EN S
0 ,1 5 W EN S
N EW S
W EW S
u '/U
0 ,1 0
0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
0 ,2 0
x /D = 3 0
N EN S
0 ,1 5
W EN S
N EW S
W EW S
0 ,1 0
u '/U
0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
a)
0.30
0.25
0.20
u'/Uc
0.15
Re=10000
0.10
NENS
WENS
0.05 NEWS
WEWS
0.00
0 10 20 30 40
x/D
b)
0.30
0.25
0.20
u'/Uc 0.15
Re=20,000
NENS
0.10
WENS
NEWS
0.05 WEWS
0.00
0 10 20 30 40
x/D
c)
0.30
0.25
0.20
u'/Uc
0.15
Re=30,000
NENS
0.10
WENS
NEWS
0.05 WEWS
0.00
0 10 20 30 40
x/D
To study the effect of Reynolds number on u for each configuration, the data
presented in Figures 5.14-a, b and c are rearranged based on configuration and
presented in Figures 5.15-a, b, c and d. The influence of Re for most configurations is
similar except for the cases that have sidewalls and endplate, are associated with the
previously discussed monotonic increase.
Deo et al. (2008) have studied the relationship between Reynolds number and vortex
shedding. They found that with the increase of Reynolds number from 1,500 to
16,500, the vortex passage frequency (or Strouhul number, St) was slightly increased
from 0.201 to 0.218 indicating weak dependence. The same team, estimated the St for
two jets one with sidewalls and the other without (NENS, NEWS) and found that the
St was 0.22 and 0.36 for jet with sidewalls (NEWS) and free jet (NENS) respectively.
Their findings implied that Reynolds number and nozzle condition affect the vortex
formation and its interaction. Although we have not estimated the St, we may suspect
that the variations among the configurations at various Re are due to the differences
in the vortex formation and its interaction.
a)
0 .2 5
0 .2 0
0 .1 5
NENS
u '/ U
c
1 0 ,0 0 0
0 .1 0 2 0 ,0 0 0
3 0 ,0 0 0
0 .0 5
0 .0 0
0 10 20 30 40
x /D
b)
0,25
0,20
/ cUc
0,15
uu'/U 0,10
WENS
10,000
20,000
30,000
0,05
0,00
0 10 20 30 40
y/yc
X/ D
c)
0,25
0,20
/ cUc
0,15
uu'/U
0,10 NEWS
10,000
20,000
0,05 30,000
0,00
0 10 20 30 40
y/yc
X/ D
d)
0.25
0.20
0.15
u'/Uc
0.10
WEWS
10,000
20,000
0.05
30,000
0.00
0 10 20 30 40
x/D
The distributions of the four configurations have the expected shapes which follow in
general the trends observed in turbulent longitudinal velocity. And the magnitudes of
v are smaller than the corresponding u.
a)
0.05
x/D=0
NENS
WENS
NEWS
WEWS
v'/Uc
0.00
b)
0 .2 0
X /D = 1
N E N S
0 .1 5 W E N S
N E W S
W E W S
v '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
c)
0 .2 0
X /D = 2
N E N S
W E N S
0 .1 5
N E W S
W E W S
v '/U
c
0 .1 0
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
d)
0 .2 0
X /D = 2 .5
N EN S
0 .1 5
W EN S
N EW S
W EW S
v '/U
0 .1 0
c
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
e)
0 .2 0
X /D = 3
N E N S
0 .1 5
W E N S
N E W S
W E W S
0 .1 0
v '/U
c
0 .0 5
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
f)
0 ,2 0
X /D = 4
N E N S
0 ,1 5 W E N S
N E W S
W E W S
v '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
g)
0 ,2 0
X /D = 5
NENS
W ENS
0 ,1 5 NEW S
W EW S
v '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
h)
0 ,2 0
X /D = 1 0
NENS
0 ,1 5 W ENS
NEW S
W EW S
v '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
j)
0 ,2 0
X /D = 1 5
N E N S
0 ,1 5 W E N S
N E W S
W E W S
v '/U
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
k)
0 ,2 0
X /D = 2 0
NENS
0 ,1 5 W ENS
NEW S
W EW S
v '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
l)
0 ,2 0
X /D = 2 5
NENS
0 ,1 5 W ENS
NEW S
W EW S
v '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
m)
0 ,2 0
X /D = 3 0
N E N S
0 ,1 5 W E N S
N E W S
W E W S
v '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
n)
0 ,2 0
X /D = 3 5
NENS
W ENS
0 ,1 5 NEW S
W EW S
v '/U
c
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
a)
0 ,2 0
NENS
x /d
0 ,1 5 0
1
2
2 .5
v '/ U
c
0 ,1 0 3
4
5
10
0 ,0 5 15
20
25
30
0 ,0 0 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
b)
0 .2 0
w ens
0 .1 5 x /d
0
1
2
2 .5
v '/ U
c
0 .1 0 3
4
5
10
0 .0 5 15
20
25
30
0 .0 0 35
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5
y /y c
c)
0 ,2 0
NEW S
x /d
0 ,1 5 0
1
2
2 .5
3
v '/U
0 ,1 0
c
4
5
10
0 ,0 5 15
20
25
30
35
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.17: Self similarity of turbulent intensity distributions of the lateral velocity
component on the central xy plane for different test cases: a) NENS, b) WENS, c)
NEWS and d) WEWS.
d)
0 ,2 0
W EW S
0 ,1 5 x /d
0
1
2
2 .5
v '/U
c
0 ,1 0 3
4
5
10
0 ,0 5 15
20
25
30
0 ,0 0 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.17: Self similarity of turbulent intensity distributions of the lateral velocity
component on the central xy plane for different test cases: a) NENS, b) WENS, c)
NEWS and d) WEWS (contd).
a)
0 ,2 0
x /D = 1
N E N S
W E N S
0 ,1 5
N E W S
W E W S
v '/U
0
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.18: Turbulent intensity distributions of the lateral velocity component on the
central xy plane normalised to the exit velocity at different downstream locations.
b)
0 ,2 0
x /D = 4
N E N S
W E N S
0 ,1 5 N E W S
W E W S
v '/U
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
c)
0 ,2 0
x /D = 1 5
N E N S
0 ,1 5 W E N S
N E W S
W E W S
v '/U
0
0 ,1 0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
0 ,2 0
x /D = 3 0
N E N S
0 ,1 5 W E N S
N E W S
W E W S
v '/U
0 ,1 0
0
0 ,0 5
0 ,0 0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.18: Turbulent intensity distributions of the lateral velocity component on the
central xy plane normalised to the exit velocity at different downstream locations
(contd).
0.15
v'/Uc
0.10 Re=10000
NENS
WENS
NEWS
0.05
WEWS
0.00
0 10 20 30 40
x/D
b)
0.20
0.15
v'/Uc Re=20,000
0.10
NENS
WENS
NEWS
0.05
WEWS
0.00
0 10 20 30 40
x/D
c)
0.25
0.20
0.15
Re=30,000
v'/U0
NENS
0.10 WENS
NEWS
WEWS
0.05
0.00
0 10 20 30 40
x/D
These measurements imply that, for all configurations, Reynolds number is found to
have an effect on the turbulent velocity field.
a)
0 .2 5
0 .2 0
0 .1 5
v '/U
c
0 .1 0
NENS
0 .0 5
1 0 ,0 0 0
2 0 ,0 0 0
3 0 ,0 0 0
0 .0 0
0 10 20 30 40
x /D
b)
0 .2 5
0 .2 0
0 .1 5
v'/Uc
0 .1 0
W ENS
0 .0 5 1 0 ,0 0 0
2 0 ,0 0 0
3 0 ,0 0 0
0 .0 0
0 10 20 30 40
x /D
c)
0 .2 5
0 .2 0
0 .1 5
c
v '/U
0 .1 0
NEW S
0 .0 5 1 0 ,0 0 0
2 0 ,0 0 0
3 0 ,0 0 0
0 .0 0
0 10 20 30 40
x /D
d)
0 .2 5
0 .2 0
0 .1 5
v '/ U
c
0 .1 0
W EW S
1 0 ,0 0 0
2 0 ,0 0 0
0 .0 5
3 0 ,0 0 0
0 .0 0
0 10 20 30 40
x /D
X/D=0
NENS
WENS
NEWS
( <uv> / Uc ) X 100
0.025 WEWS
2
0.000
Figure 5.21: Reynolds shear stress profiles on the central xy plane normalized to
the centreline velocity at different downstream locations.
b)
1,5 X/D=1
NENS
WENS
NEWS
(<uv>/Uc )X100
1,0 WEWS
0,5
0,0
c)
1,5 X/D=1
NENS
WENS
NEWS
(<uv>/Uc )X100
1,0 WEWS
2
0,5
0,0
d)
1,5 X/D=2,5
NENS
WENS
NEWS
(<uv>/Uc )X100
1,0 WEWS
2
0,5
0,0
Figure 5.21: Reynolds shear stress profiles on the central xy plane normalized to
the centreline velocity at different downstream locations (contd).
e)
1,5
X/D=3
NENS
WENS
NEWS
1,0
WEWS
(<uv>/Uc )X100
2
0,5
0,0
f)
1,5 X/D=4
NENS
WENS
NEWS
(<uv>/Uc )X100
1,0 WEWS
2
0,5
0,0
g)
X/D=5
1,5
NENS
WENS
NEWS
WEWS
(<uv>/Uc )X100
1,0
2
0,5
0,0
Figure 5.21: Reynolds shear stress profiles on the central xy plane normalized to
the centreline velocity at different downstream locations (contd).
h)
X/D=10
1,5
NENS
WENS
NEWS
(<uv>/Uc )X100
1,0
WEWS
0,5
0,0
j)
X/D=15
1,5
NENS
WENS
NEWS
WEWS
(<uv>/Uc )X100
1,0
2
0,5
0,0
k)
1,5 X/D=20
NENS
WENS
NEWS
(<uv>/Uc )X100
1,0
WEWS
2
0,5
0,0
Figure 5.21: Reynolds shear stress profiles on the central xy plane normalized to
the centreline velocity at different downstream locations (contd).
l)
1,5 X/D=25
NENS
WENS
NEWS
1,0 WEWS
(<uv>/Uc )X100
2
0,5
0,0
m)
1,5 X/D=30
NENS
WENS
NEWS
WEWS
(<uv>/Uc )X100
1,0
2
0,5
0,0
n)
1,5 X/D=35
NENS
WENS
NEWS
(<uv>/Uc )X100
1,0 WEWS
2
0,5
0,0
a)
1.5
NENS
x/d
0
b)
1.5
WENS
x/d
0
( < u v > /Uc ) X 100
1
1.0 2
2.5
2
3
4
5
0.5 10
15
20
25
0.0 30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
c)
1.5
NEWS
x/d
0
( < u v > / Uc ) X 100
1
1.0 2
2.5
2
3
4
5
0.5 10
15
20
25
30
0.0
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 5.22: self similarity profiles of Reynolds shear stress on the central
xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d) WEWS.
d)
1.5
WEWS
x/d
0
Figure 5.22: self similarity profiles of Reynolds shear stress on the central
xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d) WEWS
(contd).
0 ,0
X /D = 1
-0 ,5 N E N S
W E N S
N E W S
W E W S
-1 ,0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.23: Reynolds shear stress profiles on the central xy plane normalized
to (uv) at different downstream locations.
b)
0 ,5
X /D = 4
-0 ,5 N E N S
W E N S
N E W S
W E W S
-1 ,0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
c)
0 ,5
< u v > /( u 'v ')
0 ,0
X /D = 1 5
N E N S
-0 ,5
W E N S
N E W S
W E W S
-1 ,0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
0 ,5
< u v > /( u 'v ')
0 ,0
X /D = 3 0
-0 ,5 N E N S
W E N S
N E W S
W E W S
-1 ,0
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.23: Reynolds shear stress profiles on the central xy plane normalized
to (uv) at different downstream locations (contd).
a)
0.050
X/D=0
NENS
WENS
0.025
0.000
3
3
-0.025
-0.050
b)
6
5
X/D=1
NENS
4 WENS
( < u > / Uc ) X 1000
NEWS
3
WEWS
2
3
1
3
-1
-2
c)
6
5
X/D=2
NENS
4 WENS
NEWS
( < u > / Uc ) X 1000
3
WEWS
2
3
0
3
-1
-2
Figure 5.24: Profiles of triple velocity product of <u3> on the central xy plane at
different downstream locations.
d)
6
5
X/D=2,5
NENS
4 WENS
3 1
0
3
-1
-2
e)
6
X/D=3
5 NENS
WENS
4
NEWS
( < u > / Uc ) X 1000
3 WEWS
2
3
0
3
-1
-2
f)
6
X/D=4
5 NENS
WENS
4
NEWS
( < u > / Uc ) X 1000
3 WEWS
2
3
0
3
-1
-2
Figure 5.24: Profiles of triple velocity product of <u3> on the central xy plane at
different downstream locations (contd).
g)
6
X/D=5
5
NENS
4 WENS
-1
-2
h)
6
X/D=10
5 NENS
4 WENS
NEWS
( < u > / Uc ) X 1000
3 WEWS
2
3
0
3
-1
-2
j)
6
X/D=15
5 NENS
WENS
4
NEWS
( < u > / Uc ) X 1000
3 WEWS
2
3
0
3
-1
-2
Figure 5.24: Profiles of triple velocity product of <u3> on the central xy plane at
different downstream locations (contd).
k)
6
X/D=20
5 NENS
WENS
4 NEWS
3
3 1
-1
-2
l)
6
5
X/D=25
NENS
4 WENS
( < u > / Uc ) X 1000
NEWS
3
WEWS
2
3
1
3
-1
-2
m)
7
6 X/D=30
NENS
5
WENS
( < u > / Uc ) X 1000
4 NEWS
3 WEWS
3
1
3
-1
-2
Figure 5.24: Profiles of triple velocity product of <u3> on the central xy plane at
different downstream locations (contd).
n)
7
6 X/D=35
5 NENS
WENS
4
0
3
-1
-2
Figure 5.24: Profiles of triple velocity product of <u3> on the central xy plane at
different downstream locations (contd).
a)
8
7
NENS
6 x/d
( < u > / Uc ) X 1000
0
5 1
4 2
2.5
3
3
3
4
2
5
1 10
3
15
0 20
-1 25
30
-2 35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
b)
8
7
WENS
6 x/d
5 0
( < u > / Uc ) X 1000
1
4 2
2.5
3
3
3
2 4
5
1 10
3
0 15
20
-1 25
-2 30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 5.25: Self similarity profiles of triple velocity product <u3> on the
central xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d)
WEWS.
c)
8
7
NEWS
6 x/d
0
5
0 20
-1 25
30
-2 35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
d)
8
7
WEWS
6 x/d
0
5
( < u > / Uc ) X 1000
1
4 2
2.5
3 3
3
4
2
5
1 10
15
3
0 20
-1 25
30
-2 35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 5.25: Self similarity profiles of triple velocity product <u3> on the central
xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d) WEWS (contd).
In the early stages of jet development <v3> is increasing closer to the centreline,
attaining zero values in the potential core. Further downstream (15 D and 30 D) the
decreasing distributions reach zero on the jet axis (the symmetry properties of this
variable enforce a profile symmetric around (0, 0) point and zero values on the central
axis). From the location of maximum v, the <v3> values increase towards the edge of
the jet reaching a positive maximum outside y/yc=1, and decreasing to zero at larger
distances from the centreline. Peak values generally increase with x/D. The observed
similarities and differences between the results of the four configurations follow the
trends discussed for the turbulent velocities.
The presence of the endplate has an insignificant effect when the sidewalls have not
been implemented and the measurements for NENS and WENS jets are very close at
all stations. In the results of the configurations comprising sidewalls, the endplate has
a clear effect, and the nondimensional streamwise turbulent velocity attains always
higher values in its presence (comparing WEWS with NEWS). The self-preservation
of <v3> for all configurations is shown in Figure 27.
a)
0.006
X/D=0
NENS
WENS
0.003
NEWS
( < v > / Uc ) X 1000
WEWS
0.000
3
3
-0.003
-0.006
0.0 0.2 0.4 0.6 0.8 1.0 1.2
y/yc
b)
2
X/D=1
NENS
WENS
( < v > / Uc ) X 1000
1
NEWS
WEWS
3
0
3
-1
Figure 5.26: Profiles of triple velocity product of <v3> on the central xy plane at
different downstream locations.
c)
2
X/D=2
NENS
WENS
NEWS
3
0
3
-1
d)
2
X/D=2,5
NENS
WENS
( < v > / Uc ) X 1000
1
NEWS
WEWS
3
0
3
-1
e)
2
X/D=3
NENS
WENS
1 NEWS
( < v > / Uc ) X 1000
WEWS
3
0
3
-1
Figure 5.26: Profiles of triple velocity product of <v3> on the central xy plane at
different downstream locations (contd).
f)
2
X/D=4
NENS
WENS
1
3
0
3
-1
g)
2
X/D=5
NENS
WENS
1 NEWS
( < v > / Uc ) X 1000
WEWS
3
0
3
-1
h)
2
X/D=10
NENS
WENS
NEWS
1
WEWS
( < v > / Uc ) X 1000
0
3
3
-1
Figure 5.26: Profiles of triple velocity product of <v3> on the central xy plane at
different downstream locations (contd).
j)
2
X/D=15
NENS
WENS
1 NEWS
3
0
3
-1
k)
X/D=20
2 NENS
WENS
NEWS
( < v > / Uc ) X 1000
1 WEWS
3
0
3
-1
l)
2
X/D=25
NENS
WENS
NEWS
1
( < v > / Uc ) X 1000
WEWS
3
0
3
-1
Figure 5.26: Profiles of triple velocity product of <v3> on the central xy plane at
different downstream locations (contd).
m)
2
X/D=30
NENS
WENS
1 NEWS
3
0
3
-1
n)
2 X/D=35
NENS
WENS
NEWS
( < v > / Uc ) X 1000
1 WEWS
3
0
3
-1
Figure 5.26: Profiles of triple velocity product of <v3> on the central xy plane at
different downstream locations (contd).
a)
2
NENS
x/d
0
1
1
(<v >/Uc )X1000
2
2.5
3
3
0 4
5
10
3
15
20
-1 25
30
35
0,0 0,5 1,0 1,5 2,0 2,5
y/yc
Figure 5.27: Self similarity profiles of triple velocity product <v3> on the central
xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d) WEWS.
b)
2
WENS
x/d
0
1
1
15
20
-1 25
30
35
0,0 0,5 1,0 1,5 2,0 2,5
y/yc
c)
2
NEWS
x/d
0
1
1
2
(<v >/Uc )X1000
2.5
3
0 4
3
5
10
15
3
20
-1 25
30
35
0,0 0,5 1,0 1,5 2,0 2,5
y/yc
d)
2
WENS
x/d
0
1
1
(<v >/Uc )X1000
2
2.5
3
3
0 4
5
10
3
15
20
-1 25
30
35
0,0 0,5 1,0 1,5 2,0 2,5
y/yc
Figure 5.27: Self similarity profiles of triple velocity product <v3> on the central xy
plane for all test cases: a) NENS, b) WENS, c) NEWS and d) WEWS (contd).
a)
0.02
X/D=0
NENS
( < u v > / Uc ) X 1000
0.01 WENS
NEWS
WEWS
3
0.00
2
-0.01
-0.02
0.0 0.2 0.4 0.6 0.8 1.0 1.2
y/yc
b)
2
X/D=1
NENS
WENS
( < u v > / Uc ) X 1000
1 NEWS
WEWS
3
0
2
-1
Figure 5.28: Profiles of triple velocity product of <u2v> on the central xy plane
at different downstream locations.
c)
2
X/D=2
NENS
WENS
3
2 0
-1
d)
2
X/D=2,5
NENS
WENS
( < u v > / Uc ) X 1000
1 NEWS
WEWS
3
0
2
-1
e)
2
X/D=3
NENS
WENS
( < u v > / Uc ) X 1000
1 NEWS
WEWS
3
0
2
-1
Figure 5.28: Profiles of triple velocity product of <u2v> on the central xy plane
at different downstream locations (contd).
f)
2
X/D=4
NENS
WENS
3
2 0
-1
g)
2 X/D=5
NENS
WENS
( < u v > / Uc ) X 1000
NEWS
1 WEWS
3
0
2
-1
h)
2 X/D=10
NENS
WENS
NEWS
( < u v > / Uc ) X 1000
1 WEWS
3
0
2
-1
Figure 5.28: Profiles of triple velocity product of <u2v> on the central xy plane
at different downstream locations (contd).
j)
2
X/D=15
NENS
WENS
3
2 0
-1
k)
X/D=20
NENS
2
WENS
NEWS
( < u v > / Uc ) X 1000
WEWS
1
3
0
2
-1
l)
2 X/D=25
NENS
WENS
NEWS
( < u v > / Uc ) X 1000
1
WEWS
3
0
2
-1
Figure 5.28: Profiles of triple velocity product of <u2v> on the central xy plane
at different downstream locations (contd).
m)
2 X/D=30
NENS
WENS
3
2 0
-1
n)
2 X/D=35
NENS
( < u v > / Uc ) X 1000
WENS
NEWS
1
WEWS
3
0
2
-1
Figure 5.28: Profiles of triple velocity product of <u2v> on the central xy plane
at different downstream locations (contd).
a)
2
NENS
x/d
0
( < u v > / Uc ) X 1000
1 1
2
2.5
3
3
4
0
5
10
15
2
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 5.29: Self similarity profiles of triple velocity product <u2v> on the
central xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d)
WEWS.
b)
2
WENS
x/d
0
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
c)
NEWS
x/d
( < u v > / Uc ) X 1000
0
1 1
2
2.5
3
3
0 4
5
10
15
2
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
d)
2 WEWS
x/d
0
( < u v > / Uc ) X 1000
1
1 2
2.5
3
3
4
0 5
10
15
2
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 5.29: Self similarity profiles of triple velocity product <u2v> on the
central xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d) WEWS
(contd).
0.0000
3
2
-0.0025
-0.0050
b)
2
X/D=1
NENS
WENS
( < u v > / Uc ) X 1000
1 NEWS
WEWS
3
0
2
-1
c)
2
X/D=2
NENS
WENS
( < u v > / Uc ) X 1000
1 NEWS
WEWS
3
0
2
-1
Figure 5.30: Profiles of triple velocity product of <uv2> on the central xy plane
at different downstream locations.
d)
2
X/D=2.5
NENS
WENS
3
0
2
-1
y/yc
e)
2
X/D=3
NENS
WENS
( < u v > / Uc ) X 1000
1
NEWS
WEWS
3
0
2
-1
f)
2
X/D=4
NENS
WENS
1 NEWS
( < u v > / Uc ) X 1000
WEWS
3
0
2
-1
y/yc
g)
2
X/D=5
NENS
WENS
NEWS
3
0
2
-1
h)
2 X/D=10
NENS
WENS
NEWS
( < u v > / Uc ) X 1000
1
WEWS
3
0
2
-1
j)
2 X/D=15
NENS
WENS
NEWS
1
WEWS
( < u v > / Uc ) X 1000
3
0
2
-1
k)
X/D=25
2 NENS
WENS
NEWS
0
2
-1
l)
X/D=25
2
NENS
WENS
NEWS
( < u v > / Uc ) X 1000
1 WEWS
3
0
2
-1
m)
X/D=30
2 NENS
WENS
NEWS
( < u v > / Uc ) X 1000
1 WEWS
3
0
2
-1
n)
X/D=35
2 NENS
WENS
3
0
2
-1
Figure 5.30: Profiles of triple velocity product of <uv2> on the central xy plane
at different downstream locations (contd).
a)
2
NENS
x/d
0
( < u v > / Uc ) X 1000
1
1
2
2.5
3
3
0 4
5
10
2
15
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
b)
2
WENS
x/d
0
( < u v > / Uc ) X 1000
1
1
2
2.5
3
3
0 4
5
10
2
15
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 5.31: Self similarity profiles of triple velocity product <uv2> on the
central xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d)
WEWS.
c)
2
NEWS
x/d
3
3
0 4
5
10
2
15
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
d)
2 WEWS
x/d
0
( < u v > / Uc ) X 1000
1
1 2
2.5
3
3
4
0 5
10
2
15
20
-1 25
30
35
0.0 0.5 1.0 1.5 2.0 2.5
y/yc
Figure 5.31: Self similarity profiles of triple velocity product <uv2> on the
central xy plane for all test cases: a) NENS, b) WENS, c) NEWS and d)
WEWS (contd).
The transverse distributions of the skewness factor of the longitudinal velocity in the
xy plane, i.e. Su=<u3>/ u3 are shown in Figures 5.32-a, b, c, d, e, f, g, h, j, k, l, m and
n for downstream locations x/D=0, 1, 2, 2.5, 3, 4, 5, 10, 15, 20, 25, 30 and 35
respectively. Su usually give insight into the flow field which is a measure of the
statistical symmetry of u (where u is fluctuating streamwise velocity). As evidenced,
the distributions of NENS and WENS are nearly identical at all downstream location
indicating the insignificant effect of the endplate when sidewalls are not present.
When the sidewalls are present (NEWS, WEWS), sharp peaks are found at the outer
region of the jets. These peaks which are produced indicate to presence of 2-d rollers
which are formed in jets with sidewalls. Figure 5.33 demonstrates the self-
preservation behaviour of Su for the four test cases which indicates similar trends as
other turbulent quantities.
a)
5
X /D = 0
4
N E N S
W E N S
N E W S
3
W E W S
2
u
S
-1
0 ,0 0 ,2 0 ,4 0 ,6 0 ,8 1 ,0 1 ,2
y /y c
b)
5
X /D = 1
4
N E N S
W E N S
N E W S
3
W E W S
2
u
S
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
c)
5
X /D = 2
4 NENS
W ENS
3 NEW S
W EW S
2
u
S
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
5
X /D = 2 ,5
4 N EN S
W EN S
3 N EW S
W EW S
2
u
S
1
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
e)
5
X /D = 3
4
NENS
W ENS
NEW S
3
W EW S
2
u
S
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
f)
5
X /D = 4
4 NENS
W ENS
3
NEW S
W EW S
u
2
S
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
g)
5
X /D = 5
4
N EN S
W EN S
N EW S
3
W EW S
2
u
S
1
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
h)
5
X /D = 1 0
4 N EN S
W EN S
3
N EW S
W EW S
2
u
S
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
j)
5
X /D = 1 5
4 N EN S
W EN S
3 N EW S
W EW S
2
u
S
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
k)
5
X /D = 2 0
4 N EN S
W EN S
3 N EW S
W EW S
2
u
S
1
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
l)
5
X /D = 2 5
4 N EN S
W EN S
3 N EW S
W EW S
2
u
S
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
m)
5
X /D = 3 0
4
NENS
W ENS
NEW S
3
W EW S
2
u
S
-1
0 ,0 0 ,5 1 ,0 y / y c 1 ,5 2 ,0 2 ,5
n)
5
X /D = 3 5
4 N E N S
W E N S
3 N E W S
W E W S
2
u
S
1
-1
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
a)
5
4 N E N S
x /d
0
3 1
2
2 .5
2
3
u
S
4
1 5
10
15
0
20
25
-1 30
35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
b)
5
4 W ENS
x /d
0
3 1
2
2 2 .5
3
u
S
4
1 5
10
15
0
20
25
-1 30
35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
c)
5
4 NEW S
x /d
0
3 1
2
2 2 .5
3
u
S 4
1 5
10
15
0
20
25
-1 30
35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
5
4 W EW S
x /d
0
3 1
2
2 2 .5
3
u
4
S
1 5
10
15
0
20
25
-1 30
35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
The upper peak may be produced due to the growth and merging of large scale
structures which are present at the end of the potential core region, whereas the lower
peak is probably associated with the induction of low velocity ambient fluid, due to
lateral oscillation of the jets potential core (Deo, 2005). At Re=10,000, all
configurations display a local maximum at x/D=4 and a local minimum at x/D=9. At
Re=20,000, all configurations (NENS, WENS, NEWS and WEWS) display similar
local maximum at x/D=2.5 whereas local minimum at x/D= 6 and 5 for jets without
sidewalls (NENS, WENS) and jets with sidewalls (NEWS and WEWS) respectively.
At Re=30,000 all the jets produce upper peak at the same location i.e. x/D=2.5 but
different location for the lower peak i.e. x/D=6 for NENS and WENS, x/D=5 for
NEWS and WEWS. Although the location of positive peak for each configuration at
all Reynolds number is found to be similar, the value of this peak is found to depend
on the configuration. We have found that the positive peak in the jets with sidewalls
(NEWS, WEWS) present lower values in comparison to jets without sidewalls
(NENS, WENS). These trends are probably due to the more organized structures
which produced in jets with sidewalls. This is compatible with the sharp peaks which
were found at the edges of the jets with sidewalls (see the transverse distributions of
Su, Figure 5.32). Nevertheless, the location of the negative peak (or local minimum)
is found to vary in the presence or the absence of the sidewalls. It is observed that the
presence of the sidewalls (NEWS and WEWS) result in early establishment of these
negative peaks, i.e. x/D=7, in comparison with free jets (NENS, WENS) i.e. x/D=9.
These findings further confirm the effects that are imposed due to the presence or
absence of the sidewalls and/or endplates.
a)
1.5
Re=10000
NENS
1.0
W ENS
NEW S
0.5 W EW S
Su
0.0
-0.5
-1.0
-1.5
0 10 20 30 40
x/D
b)
1.5
Re=20,000
NENS
1.0
WENS
NEWS
0.5 WEWS
0.0
Su
-0.5
-1.0
-1.5
0 10 20 30 40
x/D
c)
1.5
Re=30,000
NENS
1.0 WENS
NEWS
0.5 WEWS
Su
0.0
-0.5
-1.0
-1.5
0 10 20 30 40
x/D
It may be of interest to study the effect of high Reynolds number on the development
of skewness factor e.g. Su. Most of studies available in literature presenting the effect
of Reynolds numbers on skewness factor, were confined to lower range of Reynolds
number than ours (e.g. Deo et al., 2008, 1,500 Re 16,500). Therefore, the data of
Figure 5.34 are rearranged based on configuration and are presented in Figure 5.35 to
demonstrate any possible effects of the present range (10,000 Re 30,000). As one
can see, the effect of high Reynolds number on Su still exists. For all configurations,
at the initial stages, three different distributions corresponding to Re=10,000, 20,000
and 30,000 are observed. The location and the value of the local maximum are found
to be related to Reynolds number. When Re increases, the distance where the positive
peak is displayed increases in the axial direction whereas its value decreases. These
findings are compatible with published results on lower range Re (e.g Deo et al.,
2008). But in the presence of the sidewalls are found to suppress these peaks, as
discussed earlier (see Figure 5.34). Similarly, as for the positive peak (i.e. local
maximum), the location and value of the negative peak (i.e. local minimum) is found
to depend on Reynolds numbers. As shown in Figure 5.35, for all configurations, with
the increase of Reynolds number the value of the local minimum also increases.
Beyond x/D=10, the values of Su vary towards their asymptotic value.
a)
2
0
Su
-1 NENS
10,000
20,000
30,000
-2
0 10 20 30 40
x/D
b)
2
0
Su
WENS
-1
10,000
20,000
30,000
-2
0 10 20 30 40
x/D
c)
2
Su NEWS
-1 10,000
20,000
30,000
-2
0 10 20 30 40
x/D
d)
2
1
Su
-1 WEWS
10,000
20,000
30,000
-2
0 10 20 30 40
x/D
a)
4
X /D = 0
N E N S
3
W E N S
N E W S
2 W E W S
1
v
S
-1
-2
0 ,0 0 ,2 0 ,4 0 ,6 0 ,8 1 ,0 1 ,2
y /y c
b)
4
X /D = 1
N E N S
3 W E N S
N E W S
2 W E W S
1
v
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.36: Skewness profiles of the lateral velocity on the central xy plane at
different downstream locations.
c)
4
X /D = 2
NENS
3
W ENS
NEW S
2 W EW S
v
S
1
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
4
X /D = 2 ,5
N EN S
3
W EN S
N EW S
2 W EW S
v
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0
y /y c
e)
4
X /D = 3
NENS
3
W ENS
NEW S
2 W EW S
1
v
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5. 36: Skewness profiles of the lateral velocity on the central xy plane at
different downstream locations (cont).
f)
4
X /D = 4
N E N S
3
W E N S
N E W S
2 W E W S
1
v
S
0
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
g)
4
X /D = 5
N EN S
3
W EN S
N EW S
2 W EW S
v
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
h)
4
X /D = 1 0
NENS
3
W ENS
NEW S
2 W EW S
v
1
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5. 36: Skewness profiles of the lateral velocity on the central xy plane at
different downstream locations (cont).
j)
4
X /D = 1 5
N E N S
3
W E N S
N E W S
2 W E W S
1
v
S
0
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
k)
4
X /D = 2 0
3
NENS
W ENS
NEW S
2
W EW S
v
1
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
l)
4
X /D = 2 5
NENS
3
W ENS
NEW S
2 W EW S
v
1
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5. 36: Skewness profiles of the lateral velocity on the central xy plane at
different downstream locations (cont).
m)
4
X /D = 3 0
N E N S
3
W E N S
N E W S
2 W E W S
v
S
0
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
n)
4
X /D = 3 5
N E N S
3
W E N S
N E W S
2 W E W S
v
S
-1
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
Figure 5.36: Skewness profiles of the lateral velocity on the central xy plane at
different downstream locations (contd).
a)
4
N E N S
3
x /d
0
2 1
2
2 .5
1 3
v
S
4
5
0 10
15
20
-1 25
30
35
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
b)
4
W ENS
3
x /d
0
2 1
2
2 .5
1 3
v
S
4
5
0 10
15
20
-1 25
30
35
-2
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
c)
4
NEW S
3
x /d
0
2 1
2
2 .5
1 3
v
S
4
5
0 10
15
20
-1
25
30
-2 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
d)
4
3 W EW S
x /d
0
2 1
2
2 .5
1 3
v
S
4
5
0 10
15
20
-1
25
30
-2 35
0 ,0 0 ,5 1 ,0 1 ,5 2 ,0 2 ,5
y /y c
a)
1.5
1.0
0.5
0.0
Sv
-0.5 Re=10000
NENS
WENS
-1.0 NEWS
WEWS
-1.5
0 10 20 30 40
x/D
b)
1.5
1.0
0.5
0.0
Sv
-0.5 Re=20,000
NENS
WENS
-1.0
NEWS
WEWS
-1.5
0 10 20 30 40
x/D
c)
1.5
1.0
0.5
0.0
Sv
-0.5 Re=30,000
NENS
WENS
-1.0
NEWS
WEWS
-1.5
0 10 20 30 40
x/D
Figure 5.39 demonstrates the effect of Reynolds number on Sv for all the
configurations. Surprisingly, the trend of Sv for each configuration varies with
Reynolds number. For NENS and WENS, at Re=10,000, Sv doesnt display local
maximum whereas at Re=20,000 and 30,000, a local maximum is observed at
x/D=2.5. While the local minimum is observed only at Re=10,000 and 30,000.
Beyond x/D=10, Sv varies dramatically with higher values at the highest Reynolds
number. For jets with sidewalls (NEWS, WEWS), at all Reynolds numbers, a local
maximum is observed. While local minimum was only observed at the highest Re.
Beyond x/D=10, Sv varies closely to the Gaussian value of zero, compared to the free
jets (NENS, WENS) which deviate from the Gaussian. This is probably due to the
improved two-dimensionality of the flow in the presence of the sidewalls. The lack of
information regarding the evolution of Sv in the literature makes any further
assessment on its development impossible.
a)
2
Sv
NENS
-1 10,000
20,000
30,000
-2
0 10 20 30 40
x/D
b)
2
0
Sv
-1 WENS
10,000
20,000
30,000
-2
0 10 20 30 40
x/D
c)
2
0
Sv
-1 NEWS
10,000
20,000
30,000
-2
0 10 20 30 40
x/D
d)
2
Sv
0
-1 WEWS
10,000
20,000
30,000
-2
0 10 20 30 40
x/D
Sidewalls (with or without endplate) have been used in many previous investigations
(e.g. Bradbury, 1965, Gutmark & Wygnanski, 1976, Browne et al., 1982, Everitt &
Robins, 1978, Browne et al., 1984, Deo et al., 2007a, Suresh et al., 2008), in most of
the cases as a plausible means to increase the two dimensionality of the jet and
generate a flow field resembling that of an infinite plane jet. The present results
indicate that, the latter objective is only partly satisfied. The sidewalls indeed seem to
increase the two-dimensionality, but establish distinct boundary conditions due to the
shear on the wall and also impose a length scale equal to the length of the exit slot.
The entrainment process is also closely related to the spreading of the jet (Namer &
tgen, 1988). The initial momentum is redistributed to account for the mobilization
of the entrained fluid resulting in lower velocities of the flow masses imparting
momentum. Thus the original mean mass flow emanating from the nozzle along with
the entrained fluid need to be constantly redistributed downstream, through an
outward mean transverse velocity component within the edges of the jet, to even
larger cross sectional areas.
During the initial stages while the shear layer is thin the internal jet flow remains
unaffected in the potential core region. There is evidence that at this stage the shear
layer has in fact a confining effect on the internal flow (Grinstein et al., 1995). The
mean streamwise velocity remains constant in the core region, or can even increase in
a similar way that the velocity increases in the potential core in the entrance section
of a pipe, to preserve continuity, due to the development of the boundary layers and
the consequent mass flux reduction close to the boundaries. In a jet this effect is
competing with the tendency of the jet to spread and increase its effective cross
section. If the spreading is not sufficiently quick the velocity may increase
downstream in the potential core. Thereafter, once the long side shear layers meet at
the centre of the jet, the spreading becomes dominant and the centreline velocity is
decreasing downstream (Namer & tgen, 1988). In the case of the jet with no
sidewalls (NENS, WENS), the development of the shear layers on the short sides
forces the internal fluid towards the central part of the jet. Again the jet spreading
assisted by the outward increased lateral velocity component is the competing
mechanism maintaining lower velocities at the central area. It is interesting that these
effects have not a clear and monotonic impact on the decay and spreading
characteristics.
Regarding the mean flow field the most characteristic differences due to the presence
of the sidewalls (cases NEWS, WEWS) are the larger streamwise velocities
established in the outer field (especially just after the potential region), along with the
systematically smaller transverse velocities at all stations (Figure 5.1 and Figure 5.4).
The combination of these characteristics is rather surprising since the larger outward
transverse velocities of the jet without sidewalls would be expected to transfer
momentum to the outer field more effectively assisted by the larger turbulent
diffusion indicated by the turbulence terms (u, v, <uv>). The only mechanism
capable to produce the observed differences, counter fighting those mentioned above
is the convection of longitudinal momentum by 2D roller structures established more
stably in the cases of the jets with sidewalls (NEWS, WEWS). These rollers seem to
constantly carry fluid from the inner field to the outer and vice versa, pumping
momentum from the inner jet to the outer field, whereas the net outward flow is
maintained lower than that of the free jet. In this latter case of the jets without
sidewalls (NENS, WENS) the shear layers developing on the sort side edges of the jet
are responsible for the destruction of the 2D rollers, increasing turbulence levels and
decreasing the efficiency of momentum exchange by convection. The concept
described above finds support in the work of Davies et al., (1975) and Moum et al.,
(1983) who have concluded that organized motions that exist in the fully developed
plane turbulent jet play an important role on the spread of the flow and transport
processes.
Based on the work of Grinstein, (2001), 3D ring like vortices can be suspected to
develop at the edges of a rectangular jet. The presence of such structures is also
evidenced by the undulations of the streamwise velocity profiles on the xzplane at
x/D=6, 25 (Figure 5.3). The sidewalls seem to alleviate but not completely suppress
these characteristics, indicating that even in the presence of sidewalls the flow field is
not two dimensional. According to Marsters, (1980) the saddle back shape, observed
in the spanwise distributions, may be due to the pressure field associated with curved
streamlines, which would have a peak value along the centreline (xzplane) in the jet
flow. This centrally located high-pressure region would drive secondary flows that
could move high velocity fluid from the central part to the outer ends of the flow
field. Tsuchiya et al., (1986) postulated that the saddle back profile is probably due to
the formation of a retarded mixing region. In contrast to previous investigations the
velocity undulations in the present work incorporate a lower peak at the central area
of the profile.
Besides the presence of ring like structures a plausible mechanism explaining these
attributes could be the instability of the jet front. The jet impulse is not sufficient to
maintain a uniform velocity along the jet length. In an infinite jet the instability length
scale would be dictated by the jet and fluid characteristics, and the instability effects
would have been smeared out in the mean field. In rectangular jets either bounded by
sidewalls or not, the length of the jets long side imposes an additional length scale.
Besides the shear layers developing at the sort sides of the jet, either as a boundary
layer on the sidewall or as a free shear layer entraining fluid in the unbounded case,
provide specific locations for the stabilization of instability undulations. The presence
of a secondary peak in the present results for higher aspect ratio jet in comparison to
the clear saddle back profiles of the previous lower aspect ratio investigations is
compatible with this concept.
Regarding the spreading, the results presented in Figure 5.6 for Re=20,000 indicate
very similar behaviour for the four test cases in the first stages of development. Given
that in the far field the roller structures have been destroyed even for the case of a jets
with sidewalls (NEWS, WEWS), this effect may be attributed to the accumulated
result of the higher outward transverse velocities, which eventually spread
longitudinal momentum in a larger transversal area in the jet with no sidewalls (Deo
et al., 2007c).
Both components of the rms velocity (Figure 5.11 and Figure 5.16) present higher
values in the shear layers for the case of the jets with no sidewalls in the early stages
of development. This trends may be associated with the more intensive outward
transverse velocity observed in these areas although they are of a more limited and
progressive nature. Close to the centreline, values may also indicate a small increase
of the potential core in the case of the jets with sidewalls, since both rms components
increase (i.e. penetrate) at a slower pace in this area.
Adopting the dependence of the transverse pressure distribution to the lateral velocity
rms component one, predicted by Tennekes & Lumley, (1972) based on order of
magnitude analysis in developed flows, a plateau of lower pressure is established in
the area within the shear layers after the potential region. Initially values are lower for
the jets with no sidewalls but this relation is alleviated or even inversed downstream.
The triple velocity products represent turbulent transport of the Reynolds normal
stresses by the turbulent velocities (Figure 5.24, Figure 5.26, Figure 5.28, and Figure
5.30) whereas the corresponding correlation is the skewness factor (Figure 5.32 and
Figure 5.36). The monotonic increase of the skewness factor across the shear layers
from negative values to positive and especially the formation of sharp peaks at the
outer edges of the jet are illustrative of the presence of well defined roller structures
in the case of the jets with sidewalls (NEWS, WEWS). On the contrary in the free jets
(NENS, WENS) cases the early destruction of the 2D rollers and the increase of
turbulence, smooth out the sharp peaks.
Chapter 6
CONCLUSIONS
A systematic study has been performed of the influence of endplate, sidewalls and
endplate and sidewalls on the development of turbulent rectangular jets of Reynolds
number, Re=10,000, 20,000 and 30,000 issuing from a smooth contraction nozzle of
aspect ratio, AR=15. For each condition, the statistical properties of rectangular jet,
arising from changes in the initial and boundary conditions were quantified.
The present study on the influence of endplate on free rectangular jets (without
sidewalls, cases NENS and WENS), over an axial distance of 35 D in the lateral
direction revealed no significant effects of the presence or the absence of the
endplate. It is almost impossible to identify differences larger than the uncertainty and
of a systematic nature between the results of the two configurations with no sidewalls.
It can almost be stated that the results of both NENS and WENS jets collapse in all
the presented distributions.
The effect of sidewalls on the velocity field of turbulent rectangular air jet from a
nozzle of aspect ratio 15 and of Reynolds number, Re=10,000, 20,000 and 30,000
was investigated using hot wire anemometry over an axial distance 35 nozzle width.
In this study, the two configurations NEWS and WEWS indicate that sidewalls have a
clear effect on the development of the mean and turbulent flow field, which is
different in the presence or absence of an endplate in the regions investigated. In the
presence of the sidewalls, a systematic increase of the mean streamwise velocity
values at the jets edges was observed. The presence or the absence of the endplate
seems not to have an effect on these trends. These trends have been evidenced in
previous investigations (e.g. Chua & Lua, 1998, Davies et al., 1975, Moum et al.,
1983). These trends may be attributed to difference in the development of the
structures (vortices) formed in the presence or absence of the sidewalls. In the
presence of the sidewalls, these 2-D rollers seem to constantly carry the fluid from the
inner field to the outer field, pumping the momentum from the inner jet to the outer
jet. The presence of sidewalls is characterized by lower outward velocities where the
presence of the endplates (WEWS) seems to have an effect resulting in larger values
than the one without endplate (NEWS) in some locations. The presence of sidewalls
and endplates has an effect on the development of the negative outward velocities
formed at the edges of the jet. The presence of sidewalls is associated with higher
negative velocities in these areas. Again the presence of the endplate does play a role
(WEWS) leading to lower negative velocities in comparison to the case with no
endplate (NEWS).
The presence of sidewalls has a significant effect on the jet spreading rates. We have
found that the jets without sidewalls spread in nearly similar rates (NENS 0.126,
WENS 0.129) but higher than when the sidewalls were implemented (NEWS 0.105,
WEWS 0.112). The smaller outward velocities observed in the presence of the
sidewalls are probably responsible for the slower spread rates in cases NEWS and
WEWS. While the larger outward velocities found in the absence of the sidewalls as
well as the higher rate of 3-D effects seem to spread the longitudinal momentum to a
larger transverse area.
The presence of the sidewalls and endplates has a clear effect on the mass flow rate of
the studied jets leading to higher values in the region x/D=5-25 compared to the jets
without sidewalls. The differences between the configurations incorporating sidewalls
and those which do not may be attributed to the two dimensional characteristics
imposed by the sidewalls. In the early stages the mass flow at the central xy plane is
increasing slower in the absence of sidewalls due to the spreading in the spanwise
direction. When the 3D jets with no sidewalls are beginning to attain an axisymmetric
structure, the flow rate due to the rearrangement of the flow is increasing faster than
for the two dimensional jets with sidewalls, which gradually lose their potential to
entrain ambient fluid. The presence of the sidewalls was found to result in higher
initial momentum probably due to the confinement of the jet, preventing the
expansion of the jet and the diffusion of momentum in the spanwise direction (i.e.
long dimension).
The influence of Reynolds number on the decay rate for each configuration was
assessed. The decay rates of (NENS and WENS) where found nearly the same at all
tested Reynolds numbers. The decay rates for these cases (NENS, WENS) were
increased from Re=10,000 to Re=20,000 and then showed a small reduction from
Re=20,000 to Re=30,000 for both configurations. This indicates that the decay rates
of the tested jets in the absence of the sidewalls attain their asymptotic values at
Re=20,000. But in the presence of the sidewalls (with or without endplate i.e. NEWS
and WEW) the decay rates trends were found to be different when Reynolds number
was varied. The presence of endplate and sidewalls (WEWS) result in the highest
decay rates at all Reynolds number tested compared to the case without endplate
endplates (NEWS). The increase of the decay rates of both jets with sidewalls
(NEWS, WEWS) seems not have the saturation even at the highest Reynolds number
i.e. Re=30,000.
The presence of the sidewalls and endplates was found to strongly affect the turbulent
field of the rectangular jet at Re=20,000. In the near field, it was found that the
presence of the sidewalls lead to smaller turbulent velocities in the axial and lateral
direction at the shear layers as well as towards the edges of the jet. But in the far field,
a systematic increase of the rms velocities for cases incorporating sidewalls was
observed with a higher rate in the presence of the endplate.
The Reynolds shear stresses (uv) as well as third moment (i.e. u3, v3, u2v, uv2) lateral
profiles show similar trends to that on the corresponding fluctuating velocities
moment, indicating the significant effects of the endplate and sidewalls on the
development of the studied jets.
In the presence of the sidewalls, a monotonic increase of the skewness factor across
the shear layers from negative values to positive and especially the formation of sharp
peaks at the outer edges of the jet are illustrative of the presence of well defined roller
structures.
Our investigation support the previously established understanding that these jets are
strongly influenced by the various initial and boundary conditions.
Bibliography
Antonia R. A., Browne L. W. B., Chambers A. J. and Rajagopalan, S.: 1983a, Budget
of temperature variance in a turbulent plane jet, Int. J. Heat Mass Transfer 26, 4148.
AN-1003 user manual (Hot wire and film anemometry system) from A.A. Lab
System LTD.
Beavers G. S. and Wilson T. A.: 1970, Vortex growth in jets, J. Fluid Mech. 44(1),
97112.
Bernard P. S. and Wallace J. M.: 2002, Turbulent flow, John Wiley and Sons, New
Jersey.
Bradbury L. J. S.: 1965, The structure of self-preserving turbulent planar jet, J. Fluid
Mech. 23, 3164.
Browne L. W. B., Antonia R. A., Rajagopalan S. and Chambers A. J.: 1982, Structure
of complex turbulent shear flows, IUTAM Symposium, Marseille, pp. 411419.
Browne L. W. B., Antonia R. A. and Chambers A. J.: 1984, The interaction region of
a turbulent planar jet, J. Fluid Mech. 149, 355373.
Cavo, A., Lemonis G., Panidis, Th. and Papailiou, D.D.: 2007, Performance of a 12-
sensor vorticity probe in the near field of a rectangular turbulent jet, Exp. Fluids 43,
pp. 1730.
engel Y. A.: 1997, Introduction to thermodynamic and heat transfer, The McGraw-
Hill Companies, Inc. United States of America.
Chua L.P. and Lua A.C.: 1998, Measurements of a confined jet, Phys. Fluids 10 (12)
Chambers A. J., Antonia R. A. and Browne L. W. B.: 1985, Effect of symmetry and
asymmetry of turbulent structures on the interaction region of a plane jet, Exp. Fluids
3, 343348.
Crow S. C. and Champagne F. H.: 1971, Orderly structure in jet turbulence, J. Fluid.
Mech. 48(3), 547591.
Daves A. E., Keffer J. F. and Baines W. D.: 1975, Spread of heated plane turbulent
jet, Phys. Fluids 18 (7), pp. 770775
Deo R. C., Nathan G. J. and. Mi J.: 2007a, The influence of nozzle aspect ratio on a
plane jet. Exp. Therm. Fluid Sci. 31 pp 825-838.
Deo R. C., Nathan G. J. and. Mi J.: 2007b, The influence of nozzle exit geometric
profile on a statistical properties of a turbulent plane jet. Exp. Therm. Fluid Sci. 32 pp
545-559.
Deo R. C., Nathan G. J. and Mi J: 2007c, Comparison of turbulent jets issuing from
rectangular nozzle with and without sidewalls. Exp. Therm. Fluid Sci. 32 (2) pp 596-
606.
Deo R. C., Mi J., and Nathan G. J.: 2008, The influence of Reynolds number on a
plane jet. Phys. Fluids (20), 075108.
Douglas J. F., Gasiorek J. M. and Swaffield J. A.: 1995, Fluid Mechanics, 3rd edition,
Longman Scientific & Technical. Printed in Singapore.
Everitt K. W. and Robins A. G.: 1973, The development and structure of turbulent
plane jets, J. Fluid. Mech. 88, 563583.
Frank M. W.: 1994, Fluid Mechanics, International edition, 3rd edition, McGraw-Hill
companies, Inc. Printed in the United States of America.
Goldstein R. J.: 1996, Fluid Mechanics Measurements, second edition, Taylor &
Francis, Printed in the United States of America.
Grinstein F.F.: 2001, Vortex dynamics and entrainment in rectangular free jets, J.
Fluid Mech. 437, pp. 69101
Grinstein F. F., Glauser M. N. and George W. K.: 1995, Vorticity in Jets, S.I.Green
Ed., Kluwer Academic Publisher, Netherland, chapter III, Fluid Vorticity, pp. 6594.
Gutmark E. and Wygnanski I.: 1976, The planar turbulent jet, J. Fluid Mech. 73(3),
465495.
Gutmark E.J. and Grinstien F.F.: 1999, Flow control with non circular jets, Annu.
Rev. Fluid Mech. 31:239-272.
Heskestad G.: 1965, Hot-wire measurements in a plane turbulent jet, Trans. ASME, J.
Appl. Mech. 32(721-734).
Hitchman G. J., Strong A. B., Slawson P. R. and Ray G.: 1990, Turbulent planar jet
with and without confining walls, AIAA J. 28(10).
Hussain A. K. M. F. and Clark A. R.: 1977, Upstream influence on the near field of a
planar turbulent jet, Phys. Fluids (20) 9, 1416-1426.
Krothapalli A., Baganoff D. and Karamcheti K.: 1981, On the mixing of rectangular
jet, J. Fluid Mech. 107, 201220.
Moum J. N, Kawall J. G. and Keffer J. F.: 1983, Coherent structures within the plane
turbulent jet, Phys. Fluids 26 (10), pp. 29392945
Mi J., Nathan G. J. and Luxton R. E.: 2000, Centerline mixing characteristics of jets
from nine differently shaped nozzles, Exp. Fluids 28, 9394.
Mi J., Deo R. C. and Nathan G. J.: 2005, Characterization of turbulent jets from high-
aspect-ratio rectangular nozzles, Phys. Fluids 17(6).
Namer I. and tgen M. V.: 1988, Velocity measurements in a planar turbulent air jet
at moderate Reynolds numbers, Exp. Fluids 6, 387399.
Norman B.: 1967 Hot-wire anemometer calibration at high subsonic speeds, DISA In-
formation 5.
Quinn W. R.: 1992, Turbulent free jet flows issuing from sharp-edged rectangular
slots: The influence of slot aspect ratio, Experimental Thermal and Fluid Science 5,
203215.
Romeos A., Lemonis G., Panidis Th., and Papailiou D.: 2007, Multisensor hot wire
vorticity probe measurements of the formation field of two corotating vortices, Flow,
Turbulence and Combustion (2009) 83:153183.
Sato, H.: 1960, The stability and transition of a two-dimensional jet, J. Fluid Mech.7,
5380.
Schneider W: 1985, Decay of momentum flux in submerged jets. J. Fluid Mech. 54,
91-110.
Sfeir A.: 1979, Investigation of three dimensional turbulent rectangular jets, AIAA J.
17(10), 10551060.
Suresh P. R., Srinivasan K., Sundararajan T., Das K.: 2008, Reynolds number
dependant of a plane jet development in the transition regime, Phy. Fluids 20,
044105.
Tennekes H., Lumley J. L.: 1972, A first course in turbulence. MIT Press.
Trendscoste, N. and Sforza, P.: 1967, Further experimental results for three-
dimensional re jets, AIAA J. 5(5), 885890.
Tsuchiya Y., Horikoshi C. and Sato T.: 1985, On the spread of rectangular jets, Exp.
Fluids 4, 197204.
Winant C. D. and Browand F. K.: 1974, Vortex pairing: the mechanism of turbulent
mixing-layer growth at moderate Reynolds number, J. Fluid Mech., 63:2:237-255.