‘Inwil...
(far: :1

‘
«2.15.»:
.:l> . 38'5“?
.32...:...:. ‘
3:53....
. ..
a
m. .3 2.,
3...: .31..
3.. 3. 3.3::
$39,... § :.
r

1
£131..)
53‘ .1

3:
a;
r

3.31.3.3.
39..., A.
‘. Hp

.13.... $
5‘ 21+: 1
.J

i:
.35.

0...? . .1
.01. I .
a... .

 

Cl “L‘Ki‘b

9:30)

 

LIBRARY
Michigan State
University

 

 

 

This is to certify that the

thesis entitled

Active Flow Control for Maximizing Performance of
Spark-Ignited Stratified Charge Engines

presented by
Alvin Chun Hun Goh

has been accepted towards fulfillment
of the requirements for

 

 

Master degree in Mechanical Engineering

 
   
 

H.
Date Wan]

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

Majo professor

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6'01 c-JCIRC/DateDuo.p86-p.15

 

ACTIVE FLOW CONTROL FOR MAXIMIZIN G PERFORMANCE OF
SPARK-IGNITED STRATIFIED CHARGE ENGINES

By

Alvin Chun Hun Goh

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Mechanical Engineering

2001

ABSTRACT

ACTIVE FLOW CONTROL FOR MAXIMIZING PERFORMANCE OF
SPARK-IGNITED STRATIFIED CHARGE ENGINES

By

Alvin Chun Hun Goh

Cycle-to-cycle variation in stratified charge engines is a serious problem in the
automotive industry. Due to cycle-to-cycle variation in engines, fuel injectors need to
inject more fuel than necessary in order to have complete combustion in the engines. By
reducing cycle-to-cycle variation, 3 better fuel economy could be achieved.

The present study investigates the possibility of acoustically perturbing the intake
flow at different frequencies and amplitudes to control the in-cylinder flow field. This
research work concentrates mainly on an engine speed of 600 rpm and 270 crank angle
degree (CAD), with the interest of changing the flow field in compression before
ignition. Molecular Tagging Velocimetry (MTV) was used to measure the 2D-velocity
field at the mid-plane of the engine cylinder.

The experimental results showed that perturbing the flow acoustically has an
effect in reducing the cycle-to—cycle variation. The level of cycle-to-cycle variation is
measured based on U ms and V nns values. The free-run perturbation method and fixed-
time perturbation method managed to reduce the U ms and V nns by about 30% in

certain regions of the flow.

This thesis is dedicated to
My parents and my girlfriend, Sarah Keok

For their love and moral support throughout my studies

iii

ACKNOWLEDGEMENTS

I would like to thank all the people who have helped and assisted me in this
research work. It would be impossible to complete this work without the help from all of
them. I would like to thank my advisor, Dr. Manoochehr Koochesfahani, for his
assistance, support and patient guidance throughout my research and studies. A special
thanks to Dr. Harold Schock and Dr. Giles Brereton for being on my committee and for
their guidance.

I would like to thank Tom Stuecken, Edward Tim and Hooman Rezaei for all
the help in setting up the engine and all the technical assistance. I would also like to thank
my fellow friends at MSU TMUAL and MSU ERL, Chee Lum, Doug Bohl, Dr. Hu Hui,
Adeel Khalid, Andrew Fedewa, Andrew Sasak, Anthony Christie, Boon-Keat Chui, Mark
Novak, Dr. Kyle Judd, Yuan Shen, Mahmood Rahi, Jan Chappell and Bobbie Slider, for
all the help and moral support. This work would also not have been possible without the
financial supports of the Ford Motor Company and Department of Energy (DOE) under
Research Grant # DE-FC02-99EE50574 and also made use of shared facilities of the
MRSEC Program of the National Science Foundation.

Finally, I would like to thank my parents again for their love and prayers. Last but
not the least, I would like to thank my girlfriend, Sarah Chui San Keok, for her love and
patience throughout my Masters Program. This day would not have been possible without

your support. Thank you!

TABLE OF CONTENTS

LIST OF TABLES ..................................................................... vii

LIST OF FIGURES ..................................................................... viii

LIST OF SYMBOLS AND ABBREVIATIONS ................................. xvi

Chapter 1

Introduction ..................................................................... 1
1.1 Motivation ............................................................ 1
1.2 Previous work ............................................................ 1
1.3 Thesis arrangement ................................................... 6

Chapter 2

Experimental measurement techniques and procedures ........................ 7
2.1 Optically accessible research engine ................................. 7
2.2 Molecular Tagging Velocimetry (MTV) set-up ........................ 12
2.3 Data acquisition ............................................................ 15
2.4 Data processing ............................................................ 17
2.5 Experimental procedures ................................................... 18

Chapter 3

Results and discussion for 270 CAD ................................................... 19
3.1 Introduction ............................................................ 19
3.2 Number of realizations needed for averaging ........................ 21
3.3 Amplitude effects ............................................................ 26
3.4 Free-run perturbations ................................................... 31

3.5 F ixed-time perturbation: perturbation from 0 CAD to 180 CAD . . 40
3.5.1 Perturbation from 0 CAD to 180 CAD with no phase shift. 40
3.5.2 Perturbation from 0 CAD to 180 CAD with 90° phase shift 53
3.5.3 Perturbation from 0 CAD to 180 CAD with -900 phase shift 58

3.6 Fixed-time perturbation: perturbation from 0 CAD to 90 CAD ...... 63
3.7 Sweeping perturbation: sweep frequencies starting from lOOHz 66
3.8 Sweeping perturbation: sweep frequencies starting from 30Hz ...... 77

Chapter 4

Results and discussion for 90 CAD and 180 CAD ................................. 86
4.1 Introduction ............................................................ 86
4.2 Fixed-time perturbation for 90 CAD: forcing from 0 CAD to 90 CAD 88
4.3 Fixed-time perturbation for 180 CAD: forcing from 0 CAD to 90 CAD

....................................................................................... 96
Chapter 5
Results and discussions for 270 CAD running at 1200 rpm ........................ 104
5.1 Introduction ............................................................ 104
5.2 Fixed—time perturbation for 270 CAD: forcing from 0 CAD to 180 CAD
....................................................................................... 106
Chapter 6
Conclusions .............................................................................. l 14
Chapter 7
Recommendations ..................................................................... 1 16
Appendix A
Supplementary plots and graphs for chapter 3 ................................. 118
A.l Supplementary graphs for free-run perturbation ............... 118
A2 Fixed-time perturbations: perturbation from 0 CAD to 180 CAD
with no phase shift ................................................... 121

A.3 Sweeping perturbations: sweep frequencies from lOOHz to 2000Hz 124

Appendix B
Experimental equipment and devices ................................................... 129
Appendix C
Zero-delayed or reference realizations investigation ................................. 136
Bibliography .............................................................................. 1 3 8

vi

LIST OF TABLES

Table 1. Prototype Engine Specifications ............................... 7

Table 2. Cases investigated in the sweeping perturbation experiment 67

vii

Figure 1.

Figure 2.

Figure 3.
Figure 4.
Figure 5.
Figure 6.
Figure 7.

Figure 8.

Figure 9.

Figure 10.

Figure 11.

Figure 12.

Figure 13.

Figure 14.

Figure 15.

LIST OF FIGURES

Examples of cycle-to-cycle variation at 270 CAD with engine running at

600 rpm ....................................................... 3

Flow visualization of a natural and excited (St=0.48) round jet at a
Reynolds number of 5,500 ..................................... 5
Schematic and optical arrangement of the experiment setup . 9
Schematic of the research engine setup ................... 10
Engine setup viewing from the front and back .......... l 1
Examples of zero-delayed (A) and delayed (B) images ....... 13
Field of view for MTV measurement in the engine cylinder . l4

Intensified CCD video camera, Xybion model ISO-350, and together with
Nikkor 50mm fl .2 lens ..................................... 15

Field of view in the engine cylinder ............................ 20

U mean (cm/s) and V mean (cm/s) for vertical line taken at x = 4.5 cm
from cylinder wall .............................................. 22

U nns (cm/s) and V rms (cm/s) for vertical line taken at x = 4.5 cm from
cylinder wall ....................................................... 23

U mean (cm/s) and V mean (cm/s) for horizontal line taken at y = 1.2 cm
below the intake and exhaust valves ............................ 24

U rms (cm/s) and V rms (cm/s) for horizontal line taken at y = 1.2 cm
below the intake and exhaust valves ............................ 25

U mean (cm/s) and V mean (cm/s) of different amplitudes for vertical line
taken at x = 4.5 cm from cylinder wall .................. 27

U rms (cm/s) and V rms (cm/s) of different amplitudes for vertical line
taken at x = 4.5 cm fiom cylinder wall .................. 28

viii

Figure 16.
Figure 17.

Figure 18.

Figure 19.
Figure 20.

Figure 21.

Figure 22.
Figure 23.

Figure 24.

Figure 25.

Figure 26.

Figure 27.

Figure 28.

U mean (cm/s) and V mean (cm/s) of different amplitudes for horizontal
line taken at y = 1.2 cm below the intake and exhaust valves 29

U rrns (cm/s) and V rms (cm/s) of different amplitudes for horizontal line
taken at y = 1.2 cm below the intake and exhaust valves ...... 30

Ensemble averaged results showing rrns velocity (cm/s),U mean, without
perturbation and with perturbation of 50Hz, 100Hz, 200Hz, 300Hz and
400Hz at 270 CAD .............................................. 32

Effect of perturbations on the U mean and U ms at cylinder centerline,
about 4.5 cm from cylinder wall ............................. 34

Effect of perturbations on the U mean and U rrns at horizontal line about
1.2 cm below the intake and exhaust valves .................... 35

Ensemble averaged results showing nns velocity (cm/s),V mean, without
perturbation and with perturbation of 50Hz, 100Hz, 200Hz, 300Hz and
400Hz at 270 CAD ............................................... 36

Effect of perturbations on the V mean and V rms at cylinder centerline,
about 4.5 cm from cylinder wall ............................. 38

Effect of perturbations on the V mean and V ms at horizontal line about
1.2 cm below the intake and exhaust valves .................... 39

Ensemble averaged results showing rms velocity (cm/s),U mean, without
perturbation and with perturbation of 30Hz, 50Hz, 100Hz, 200Hz and
300Hz at 270 CAD .................................................. 41

Ensemble averaged results showing rms velocity (cm/s),U mean, without
perturbation and with perturbation of 400Hz, 600Hz, 800Hz and 10001-12
at 270 CAD .......................................................... 42

Effect of perturbations on the U mean and U ms at cylinder centerline,
about 4.5 cm from cylinder wall

Effect of perturbations on the U mean and U rms at horizontal line about
3.1 cm below the intake and exhaust valves ................... 45

Ensemble averaged results showing rrns velocity (cm/s),V mean, without
perturbation and with perturbation of 30Hz, 50Hz, 100Hz, 200Hz and
300Hz at 270 CAD ................................................. 47

Figure 29.

Figure 30.

Figure 31.

Figure 32.

Figure 33.

Figure 34.

Figure 35.

Figure 36.

Figure 37.

Figure 38.

Figure 39.

Figure 40.

Ensemble averaged results showing rms velocity (cm/s),V mean, without
perturbation and with perturbation of 400Hz, 600Hz, 800Hz and 1000Hz
at 270 CAD .......................................................... 48

Effect of perturbations on the V mean and V rms at cylinder centerline,
about 4.5 cm from cylinder wall ............................. 50

Effect of perturbations on the V mean and V rrns at horizontal line about
3.1 cm below the intake and exhaust valves .................... 51

Ensemble averaged results showing rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C): 100Hz,
(D):200Hz at 270 CAD ...................................... 54

Ensemble averaged results showing rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):500Hz at 270 CAD ...................................... 55

Ensemble averaged results showing rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C): 100Hz,
(D):200Hz at 270 CAD

Ensemble averaged results showing rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):500Hz at 270 CAD .................................... 57

Ensemble averaged results showing rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C): 100Hz,
(D):200Hz at 270 CAD .................................... 59

Ensemble averaged results showing rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):500Hz at 270 CAD .................................... 60

Ensemble averaged results showing rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C): 100Hz,
(D):200Hz at 270 CAD

Ensemble averaged results showing rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):500Hz at 270 CAD .................................... 62

Ensemble averaged results showing rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):600Hz, (C):800Hz,
(D): 1000Hz at 270 CAD

Figure 41.

Figure 42.

Figure 43.

Figure 44.

Figure 45.

Figure 46.

Figure 47.

Figure 48.

Figure 49.

Figure 50.

Figure 51.

Figure 52.

Ensemble averaged results showing rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):600Hz, (C):800Hz,
(D): 1000Hz at 270 CAD

Ensemble averaged results showing rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B)2100Hz-200Hz,
(C): 100Hz-300Hz, (D): 100Hz-400Hz at 270 CAD .......... 68

Ensemble averaged results showing rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B): 100Hz-500Hz,
(C): 100Hz-600Hz, (D): 100Hz-800Hz at 270 CAD ........... 69

Ensemble averaged results showing rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B): 100Hz-1000Hz,
(C): 100Hz-1200Hz, (D): 100Hz-1400Hz at 270 CA ............ 70

Ensemble averaged results showing rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B): 100H2-1600Hz,
(C): 100Hz-1800Hz, (D): 100Hz-2000Hz at 270 CAD .......... 71

Ensemble averaged results showing rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B): 100Hz-200Hz,
(C): 100Hz-300Hz, (D): 100Hz-400Hz at 270 CAD ............ 72

Ensemble averaged results showing rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B): 100Hz-500Hz,
(C): 100Hz-600Hz, (D): 100Hz-800Hz at 270 CAD ............ 73

Ensemble averaged results showing rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B): 100Hz-1000Hz,
(C): 100Hz-1200Hz, (D): 100Hz-1400Hz at 270 CAD ......... 74

Ensemble averaged results showing rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B): 100Hz-1600Hz,
(C): lOOHz-l 800Hz, (D): 100Hz-2000Hz at 270 CAD ......... 75

Effect of perturbations on the U mean and V mean at horizontal line about
2 cm below the intake and exhaust valves ...................... 76

Ensemble averaged results showing rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):30Hz-200Hz,
(C):30Hz-300Hz, (D):30Hz-400Hz at 270 CAD ........... 78

Ensemble averaged results showing rrns velocity (cm/s),U mean (cm/s),

(A): without perturbation and with perturbation (B):30Hz-600Hz,
(C):30Hz-800Hz, (D):30Hz-1000Hz at 270 CAD ........... 79

xi

Figure 53.

Figure 54.

Figure 55.

Figure 56.

Figure 57.

Figure 58.

Figure 59.

Figure 60.

Figure 61.

Figure 62.

Figure 63.

Figure 64.

Ensemble averaged results showing rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):30Hz-1200Hz,
(C):30Hz-1400Hz, (D):30Hz-1600Hz at 270 CAD ............ 80

Ensemble averaged results showing rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):30Hz-1800Hz,
(C):30Hz-2000Hz at 270 CAD .............................. 81

Ensemble averaged results showing rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):30Hz-200Hz,
(C):30Hz-300Hz, (D):30Hz-400Hz at 270 CAD ............ 82

Ensemble averaged results showing rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):30Hz-600Hz,
(C):30Hz-800Hz, (D):30Hz-1000Hz at 270 CAD ............ 83

Ensemble averaged results showing rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):30Hz-12001-lz,
(C):30Hz-l400Hz, (D):30Hz-l 600Hz at 270 CAD ............ 84

Ensemble averaged results showing rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):30Hz-1800Hz,
(C):30Hz-2000Hz at 270 CAD ........................... 85

Cycle-to-cycle variation at 90 CAD ........................... 87

Ensemble averaged results showing U rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C): 100Hz,
(D):200Hz at 90 CAD ............................................. 89

Ensemble averaged results showing U rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):600Hz at 90 CAD .................................... 90

Ensemble averaged results showing U rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):800Hz, (C): 1000Hz at
90 CAD ...................................................... 91

Ensemble averaged results showing V nns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C): 100Hz,
(D):200Hz at 90 CAD ..................................... 92

Ensemble averaged results showing V rrns velocity (cm/s),V mean (cm/s),

(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):600Hz at 90 CAD ..................................... 93

xii

 

 

Figure 65.

Figure 66.

Figure 67.

Figure 68.

Figure 69.

Figure 70.

Figure 71.

Figure 72.

Figure 73.

Figure 74.

Figure 75.

Figure 76.

Ensemble averaged results showing V nns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):800Hz, (C): 1000Hz at
90 CAD

Effect of perturbations on the U mean and V mean at horizontal line about
2 cm below the intake and exhaust valves ...................... 95

Ensemble averaged results showing U rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C):100Hz,
(D):200Hz at 180 CAD ..................................... 97

Ensemble averaged results showing U rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):600Hz at 180 CAD ...................................... 98

Ensemble averaged results showing U rrns velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):800Hz, (C): 1000Hz at
180 CAD ........................................................ 99

Ensemble averaged results showing V rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):50Hz, (C): 100Hz,
(D):200Hz at 180 CAD ...................................... 100

Ensemble averaged results showing V rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):300Hz, (C):400Hz,
(D):600Hz at 180 CAD .................................... 101

Ensemble averaged results showing V rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):8001-Iz, (C): 1000Hz at
180 CAD ...................................................... 102

Effect of perturbations on the U mean and V mean at horizontal line about
2 cm below the intake and exhaust valves ..................... 103

Cycle-to-cycle variation at 270 CAD and 1200 rpm ......... 105
Ensemble averaged results showing U rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B):600Hz, (C):800Hz,
(D): 1000Hz at 270 CAD and 1200 rpm .................. 107
Ensemble averaged results showing U nns velocity (cm/s),U mean (cm/s),

(A): without perturbation and with perturbation (B): 1200Hz, (C): 1400Hz,
(D):l600Hz at 270 CAD and 1200 rpm .................. 108

xiii

Figure 77.

Figure 78.

Figure 79.

Figure 80.

Figure 81.

Figure Al.

Figure A2.
Figure A3.

Figure A4.

Figure A5.

Figure A6.
Figure A7.

Figure A8.

Figure A9.

Figure A10.

Ensemble averaged results showing U rms velocity (cm/s),U mean (cm/s),
(A): without perturbation and with perturbation (B): 1800Hz, (C):2000Hz
at 270 CAD and 1200 rpm .................................... 109

Ensemble averaged results showing V rrns velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B):600Hz, (C):800Hz,
(D): lOOOHz at 270 CAD and 1200 rpm .................. 1 10

Ensemble averaged results showing V rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B): 1200Hz, (C): 1400Hz,
(D): 1600Hz at 270 CAD and 1200 rpm .................. 111
Ensemble averaged results showing V rms velocity (cm/s),V mean (cm/s),
(A): without perturbation and with perturbation (B): 1800Hz, (C):2000Hz
at 270 CAD and 1200 rpm .................................... 112

Effect of perturbations on the U mean and V mean at horizontal line about
2 cm below the intake and exhaust valves ....................... 1 13

Dwdz RMS and Vorticity RMS for horizontal line, about 1.2 cm below the
valves ............................................................... 118

Reynolds stress for horizontal line, about 1.2 cm below the valves 119
Dwdz rrns at cylinder centerline, about 4.5 cm from cylinder wall 119

Vorticity ms and Reynolds stress at cylinder centerline, about 4.5 cm
from cylinder wall ............................................. 120

Dwdz RMS and Vorticity RMS for horizontal line, about 3.1 cm below the
valves ............................................................... 121

Reynolds stress for horizontal line, about 3.1 cm below the valves 122
Dwdz rms at cylinder centerline, about 4.5 cm from cylinder wall 122

Vorticity rrns and Reynolds stress at cylinder centerline, about 4.5 cm
from cylinder wall ............................................. 123

U RMS and V RMS for horizontal line, about 2.75 cm below the valves
........................................................................ 124

Vorticity RMS and dwdz RMS for horizontal line, about 2.75 cm below
the valves ...................................................... 125

xiv

Figure A] 1. Reynolds stress for horizontal line, about 2.75 cm below the
valves ...................................................... 126

Figure A12. U rrns at cylinder centerline, about 4.5 cm from cylinder wall 126

Figure A13. V rrns and vorticity rrns at cylinder centerline, about 4.5 cm from cylinder

wall ............................................................... 127
Figure A14. Dwdz rrns and Reynolds stress at cylinder centerline, about 4.5 cm from

cylinder wall ...................................................... 128
Figure Bl. Engine from different point of view ........................... 129
Figure BZ. Delay generators, digital multimeter and monitor ......... 130
Figure B3. Gateway E-5200 pentium III computer .................. 130
Figure B4. Kenwood RFC-1077 4-ohm speaker .. ......................... 131
Figure BS. Hafler P-1000 amplifier .................................... 131
Figure B6. HP function generator .................................... 132
Figure B7. Biacetyl seeding chamber and acetone seeding chamber . 132
Figure B8. Beam blocker ....................................................... 133
Figure B9. Very-long-focal-length lens (VF L) ............................ 133
Figure B10. Engine encoder ......................................... ' ..... l 34
Figure B1 1. Nitrogen bank ....................................................... 135
Figure C 1. Zero-delay: U mean and V mean along horizontal line at the center of the

cylinder .............................................................. 136

Figure C2. Zero-delay: U mean and V mean along vertical line at the center of the
cylinder .............................................................. 137

XV

Smbol

CAD
MTV

Re
St
TDC

rpm
u,v,w
x,y,z

Greek

0)
I“
‘I’

Overscores

LIST OF SYMBOLS AND ABBREVIATIONS

Description

Crank Angle Degree

Molecular Tagging Velocimetry
Number of Engine Cycles
Reynolds Number

Strouhal Number

Top Dead Center

X Component of Velocity

Y Component of Velocity

Z Component of Velocity
Revolutions per Minute

X, Y, Z Components of Velocity
Coordinate System Axis

Description
Vorticity

Circulation
A Variable of Interest

Description

RMS
Mean

xvi

 

 

 

 

Chapter 1

INTRODUCTION

1.1 Motivm

In the area of automotive engines, cycle-to-cycle variation is of main concern of
many researchers. In recent years, a significant amount of work has been done to improve
engine’s performance and fuel efficiency. Recent research supports the concept of a
significant interaction between the mean flow and the fluctuating parts of the in-cylinder
velocity [1]. Due to cycle-to-cycle variation, the engine fuel injectors are sometimes
programmed to inject more fuel than required. In order to design better lean burn engines,
one has to have a better understanding of the flow patterns during compression stroke and
one also has to find ways to reduce cycle-to-cycle variation. If cycle-to-cycle variation is
reduced and the in-cylinder flows are made more predictable, one can design a leaner fuel
injector. In return, this would improve fuel economy in engines. The ultimate desire of
this study is to find a way of controlling the flow and reducing the cycle-to-cycle
variation. As a result, it would provide control of the mixing and burning rates of an

internal combustion engine.

1.2 Previous Work

Cycle-to-cycle variation in internal combustion engines has always existed and
has become an interest of automotive researchers since the invention of the internal
combustion engines. The results of previous works have shown that the flow fields inside

the combustion chamber are unsteady with a wide range of spatial and temporal

 

 

fluctuations[2]. Zhang et al., l995[2] stated that the fluctuations of the flow fields are
result from cycle-to-cycle variation and the center of swirl changes its location from one
cycle to the next. Dai, Trigui et al., 2000[5] found that in nature, cyclic variations are
superposition of a non-chaotic deterministic process on a stochastic process. There are
deterministic effects from previous cycles on the combustion process in subsequent

cycles.

It is true that cycle-to-cycle variation does exist. Throughout this research,

numerous data and images have shown that cycle-to-cycle variation is very significant in

 

internal combustion engines. Several examples of cycle-to-cycle variation phenomena
velocity field are shown in Figure 1. All the instantaneous velocity fields were taken at

successive cycle at 270 CAD with the engine running at 600 rpm.

 

l‘é
ti

 

 

 

 

 

 

 

//////, -,.,.....\ iii/z; ......... \\
/////’ ...... \\ .15‘Iito’ ,._._._‘\\\\
Ill/’5.-- ,,..\\ 2 \\\ -,.,._._..-e\\\
///’....-.,,,..\\ _2.51\\\ ...... .......x\\\
////’;-,I,..~o~\\ 'I\ ........ -xx‘s‘xx
fl/a--..,,..\\\ '3Ii\....._._._._..._..\\\\\
III)..-“‘,..\\\ >.3,51\-..._._._._....o.\\\\\\
I’ll/I..‘,-\\\ 4 s--s._¢_._«\\\\\\\\
Iiii\\\,,,.\\‘ ........._._«k\\\\\\\\
\\\\\\\....,--\.. '45 ,....._._~.s\\\\\\\\\
\\\\\\kq_._._..\\ _5 «x\\\\\\\\\\\\\\
.‘s‘s‘s‘s‘sxwmwmsm. '55 shhh‘skhrsxxxxmm.
3 4 5 6 7 2 3 4 5 6 7
x x
5m -
ffff//////-‘?\ 1/////////-.,,,-:1”:
fff///////.-,‘ '1'5////////.-,, ,,,,,
lff/f///x..,.- -2///////ia,, ,,,,,,
lf/‘l‘f/l, ..... . _25f/////r\..,..._..._,,,
fff(f"\-,,-\‘ ’ff/l/f/i‘---._._..,,,
ff;1\\\-,,..\\ .3f///,I\\._._._._k"”
I; ‘\\\x..,r._\\ ’135 {ff f i \\....._._._._rr,,
i I\\\\k._._._a_\\ fi\\..._..._.._..._._‘__'._...._
\ I I A \ \\k\k«%\\\\ .4 \ \Wk a...
‘4-5 \ \ \ \ \\\\\k\\\\\\ ‘45 \\M\\\\w_«k\
.5 \\\\\\\\\\\\\\\\ _5 \\“\K\\\\\k\k\\ \
\\\\\\\\\\\\\\ \\ \\\\\\\\\\~\\\ \ \
.\\1\7.\‘.\.‘.\.\‘S13\\I\ ix .\ 5. xxxisxym‘mxxx xx .\ K.
2 3 4 5 6 7 2 3 4 5 6 7
x it
Figure 1. Examples of cycle-to-cycle variation at 270 CAD with engine
runningat600rpm

In addition, excited flows have been studied in other geometries. Several studies
have shown that the natural flows’ behaviors could be altered or controlled by exciting
the natural flows at certain frequency. For example, in round jets, coaxial jets and two-
stream mixing layers, it had been shown that perturbing the flow can modify the amount
of the mixing downstream of the wake and the turbulence intensities [4, 7, 8, 9]. As for
the case of an engine, the enhancement of turbulence in the flow promotes mixing of
unburned gas, reacting gas, quenched gas and burned gas in the late compression stroke.
An example of the round jet perturbation is Koochesfahani’s experiment. Koochesfahani
et al., 1997 [6] indicated that in round jet experiment, the forcing increased the grth
rate of the shear layer vortices, the roll-up occurred much closer to the jet nozzle and the

vortical structures were larger in size. Results of the experiment are shown in Figure 2.

Recent studies by Ambrose [3] had indicated that by exciting the intake flow in an
internal combustion engine water analog model, the mean circulation and the mean
kinetic energy increased 40% and 30% respectively. Ambrose also found that the amount
of amplification of the measured parameters as a function of frequency generally showed
the preferred frequency at 40 Hz or a Strouhal Number of 1.19 based upon the maximum
velocity of the flow entering through the valve inlet [3]. Since water was used in
Ambrose’s experiment, the compression stroke was not examined. He only investigated
the effects during the intake stroke. Therefore, Ambrose’s research motivated this current

study to investigate the effect of perturbation in a real internal combustion engine.

 

 

IE' ‘

 

   

Natural Flow Excited Flow

Figure 2. Flow visualization of a natural and excited (St=0.48) round jet at a
Reynolds number of 5,500

1.3 Thesis Arrangement

This thesis is arranged as follow:

Chapter 2 reviews and discusses the experimental measurement techniques and
procedures. The first part of Chapter 2 describes the specification of the engine used in
the experiment, Molecular Tagging Velocimetry (MTV) measurement method, data
acquisition and data processing method. In the last part of Chapter 2, the experimental

procedure for atypical run is described briefly.

The results and discussions are in Chapter 3. This chapter describes the results of
all the cases studied and discusses the effects of the forcing parameters on the mean flow
and the RMS of the flow at 270 CAD and 600 rpm. In Chapter 4, some results from the
experiment for 90 CAD and 180 CAD are presented. As for Chapter 5, the results of all
the effects of the forcing on the mean flow and the RMS of the flow at 270 CAD and
1200 rpm are presented. Finally, the conclusions of this research can be found in Chapter

6.

Chapter 2

EXPERIMENTAL MEASUREMENT TECHNIQUES AND
PROCEDURES

2.] ticall Accessible Research En ine

The schematic of the experiment and optical arrangement are shown in Figure 3
and 4. Two close-up pictures of the experiment were shown in Figure 5. This study was
conducted with a 1999 model year Ford cylinder head with 4 valves and double overhead
cams. The engine cylinder head was part of the left bank of a V8 prototype engine with
90.2 mm bore and 900 degrees bank angle. The engine specifications are listed in Table 1.

Table 1. Prototype Engine SJ)ecifications.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Model and Make Ford 4-Valve 4.6L
Bore and Stroke 90.2 mm / 90.0 mm
Connecting Rod Length 150.7 mm

Valve Activation DOHC

Intake Valve Diameter 37.0 mm

Exhaust Valve Diameter 30.0 mm

Maximum Valve Lift 10.02 mm at 120 CAD
Zero CAD Intake TDC

Intake Valve Opening 6 CAD Before TDC
Intake Valve Closure 250 CAD After TDC
Exhaust Valve gaming 126 CAD After TDC
Exhaust Valve Closure 16 CAD After TDC
Compression Ratio 9.85 : l

Piston Top Flat

 

A flat-head, optically accessible piston was used in the experiment. The bore and
stroke were 90.2 mm and 90.0 mm respectively. The length of the connecting rod was
150.7 mm. The cylinder used in the experiment was made from quartz so that it would be
optically accessible. The cylinder was mounted on a single-cylinder research engine that

served as a reciprocating mechanism. The engine was connected to a lO-HP electrical

 

motor through a rubber-damped coupling. Two different belt drive setups were connected
to the exhaust camshaft that would allow interchanging between the two engine stroke
variants. The exhaust camshaft was linked by a roller-chain to the intake camshaft. The
intake valve diameter was 37 mm and the exhaust valve diameter was 30 mm. The
maximum valve lifl was 10.02 mm at 1200 crank angle. Only the valves of the
investigated cylinder were activated. A 4-ohm, 4'/2 inch mid-woofer speaker was
mounted at the modified intake manifold (refer Figure 4 and 5 for the location of the
speaker). The speaker had wattage of 40 W and was driven by an amplified sine wave
signal through a Hafler P1000 amplifier. The amplifier had a full power bandwidth of 0.1

Hz to 100 kHz. The sine wave was generated originally from a function generator.

Nitrogen was used as the working fluid in the experiment. The reason was that
MTV measurement methods in gas-phase applications rely on the phosphorescence of
biacetyl and it requires an oxygen-free environment due to the phosphorescence

quenching by oxygen (refer section 2.2 for detail description on MTV) [2]].

  
   

Beam Blocker

 

Engine Cylinder

Beam Blocker

Vertical Laser

Engine Cylinder G

Horizontal Laser Grid

‘— 50mm Cylindrical Lens Beam Blocker

‘— 150mm Cylindrical Lens
V

Splitter

4

50mm Cylindrical Lens

Incoming Laser Beam

Figure 3. Schematic and optical arrangement of the experiment setup

NIT! o 0 en + Blocelyl

 
 

Intake

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

”" , Exhaust
Speaker . ' ‘
lcco Camera
UV Lenses
{1 {1 Loser l / Mirror
* Motor Shaft 4 6L 1
1 . en ne
U Q
‘ .-._’. 31L. { .'.'. .'.. .,.-‘ m-‘t-T ' : -.-. Ir :gf“ 4“» --'-- ‘ r
l l 1
Figure 4. Schematic of the research engine setup

Nitrogen In

 

CCD Camera

.-

 

Figure 5. Engine setup viewing from the front and back

*Images in this thesis are presented in color.

2.2 Molecular Tagging Velocimetry (MTV) Set-1Q

The velocity measurements were done using Molecular Tagging Velocimetry
(MTV). MTV is a whole-field optical diagnostic that allows for the non-intrusive
measurement of the velocity field in a flowing medium. This technique has been
previously used by several authors such as Cohn et a1. l995[10], Stier and Koochesfahani
1998[l9], Koochesfahani l999[20], Hill and Klewicki 1996[15], Koochesfahani et a].
1996[13], Cohn and Koochesfahani 1997[11], and Gendrich, Bohl, and Koochesfahani
1997[12], Gendrich, Koochesfahani, and Nocera 1997[14] to make measurements in a
wide variety of flows. For gas-phase application, Molecular Tagging Velocimetry, MTV,
uses nitrogen as a working fluid andbiacetyl as a long-lived luminescence. Nitrogen is
used due to the phosphorescence quenching by oxygen. A pulsed UV laser is used to tag
the regions of interest, and those tagged regions are interrogated at two successive times
within the lifetime of the tracer. The measured Lagrangian displacement vector and the
time over which the displacement occurred provide the velocity vector [20]. This
velocimetry approach yields whole-field, instantaneous maps of velocity vectors over a

plane allowing the derivation of various kinematic quantities from the velocity field [21].

For MTV, a Lambda Physik LPX 200 (308nm) laser capable of operating at
approximately 290 m] provided the energy to excite phosphorescence from a working
fluid mixture of nitrogen and biacetyl. In order to measure the velocity field throughout a
region, a series of vertical and horizontal laser lines were generated. These lines were
orthogonal to each other. Each intersection was a “location” where a velocity

measurement was made. At this point, images were recorded by the intensified CCD

12

camera. These images were considered as reference images or zero-delayed images. At
certain delay later, normally it was about 200 microseconds for 270 CAD measurement,
the tagged region was recorded again using the intensified CCD camera. These images
were considered as delayed images. Post-processing by correlation-based techniques
allowed the instantaneous planar velocity field to be deducted the reference and delayed

images. Examples of reference and delayed images are shown in Figure 6.

A

 

B

Figure 6. Examples of zero-delayed (A) and delayed (B) images

The MTV experiments were done with a field of view of approximately 5 cm x 5
cm. It was located about 1 cm below the intake and exhaust valves and about 2 cm from
the cylinder wall (refer Figure 7). The original laser beam was first split into two laser
beams (horizontal and vertical beams). Each laser beam was focused using a combination
of two cylindrical lenses with focal lengths of 150mm and 50mm. By going through these
two lenses, a very thin laser sheet was generated for each beam. Beam blockers with
series of thin lines cut through them were used in each beam path to generate a series of

12 lines for the horizontal and 15 lines for the vertical.

 

 

t Sparkl’lug :32”;
Intake Valve Exhaust Valve 1.653,

>- 144.286

1 l— 142 041
l- 139 796

,_ 137 551

r- 135 306

._ 133.061
mm mm 130.816

§§
aim
um
MN
‘..

 

 

 

 

 

 

 

Figure 7. Field of view for the MTV measurement in the engine cylinder

2.3 Data Acquisition

An intensified CCD video camera, Xybion model ISO-350, and together with
Nikkor 50mm fl .2 lens were used to record images in the experiment (refer Figure 8). An
engine encoder was hooked up at the camshaft to produce a 5-V TTL signal at the desired
crank angle. The signal was then used to trigger the laser. At that time, the signal
triggered the intensified camera and acquisition board to record zero-delay images. The
digital delay generators were then used to delay the original signal from the engine
encoder to trigger the intensified camera to record the delayed images. The delayed time
of 200 microseconds would provide approximately 5 to 10 pixels of displacement of the

phosphorescent lines at 270 CAD.

 

 

Figure 8. Intensified CCD video camera, Xybion model ISG-350, together
with Nikkor 50 mm fl .2 lens

The image acquisition board used in the experiment was the M-Vision 1000 by
MuTech Corporation. The M-VISION 1000 (MV-1000) is a single slot video digitizer
board for the PCI (Peripheral Component Interconnect) bus. It digitizes standard or non-
standard analog camera video into 8 bits per pixel at a rate up to 40 million samples per
second. The digitized video is stored in on-board VRAM or transferred in real time to
system memory and/or the VGA card for display. The software used together with the
acquisition board was the MV-1000 Grab Sequence Application version 1.3 for Windows
NT.

A total of 500 instantaneous realization of the velocity field was measured for
each measurement condition except for the free-run perturbation method. In the free-run
perturbation method, only 200 instantaneous realization of the velocity fields were
measured for each frequency condition (refer section 3.4 for more information). An
experiment Was conducted to determine how many instantaneous realizations were
needed to produce a reliable average. The results showed that 500 instantaneous
realizations were sufficient to produce a reliable average (refer section 3.2 for more

information).

2.4 Dela Processigg

The data processing process was performed using an in-house code developed
over the past few years at MSU Turbulent Mixing and Unsteady Aerodynamics
Laboratory (TMUAL). At the earlier stage, a four-processor Silicon Graphics Origin 200
was used as the primary data processing device. Recently, the in-house code has evolved

and become available in PC version too.

When processing the data, the main objective was to determine the displacement
vector of the tagged regions with the highest possible sub-pixel accuracy in order to
, increase the dynamic range of the velocity measurements. The displacement of the tagged
regions was determined using a direct digital spatial correlation technique. The details of
this approach can be found in Gendrich and Koochesfahani l996[l3]. The basic idea of
this approach is that a small window called a source window is selected from 3 tagged
region in the reference or zero-delay image. It is then spatially correlated with a larger
roam window in the delayed image. A well-defined correlation peak occurs at the
location corresponding to the displacement of the tagged region by the 2-component of
the velocity flow. The displacement peak is found by using a multi-dimensional

polynomial fit.

2.5 Experimental Procedures

Typical procedures for running an experiment are given. First, start the water
chiller for the UV laser and then start the UV laser. The UV laser will need about 10
minutes to warm up. While waiting for the laser to warm up, turn on all the necessary
devices namely, crank angle encoder, delay generators, function generator, amplifier, and
TV monitor. Next, check the nitrogen bank and make sure that there is enough gas
pressure to run the experiment. The nitrogen pressure has to be above 400psi for a typical
experimental run. Once the laser is warmed up, turn on the oil pump and vacuum pump
for the engine. After that, open a little bit of the valve regulator of the nitrogen bank and
let a small amount of nitrogen flow into the engine cylinder. Turn off the room lights and
start the engine. At the same time, open the valve regulator of the nitrogen bank
completely. Let the engine run for while to reach its stability. Before turning on the
camera, make sure that the intensifier of the camera is at minimum position. Turn on the
camera and run the MV-1000 Grab Sequence Application program. Remove the camera
cap and slowly increase the camera intensifier. Key in the appropriate delay time at the

delay generators and start taking data and images.

Chapter 3

RESULTS AND DISCUSSION FOR 270 CAD

3. 1 Introduction
The results of the experiments will be presented in this section and are organized
in chronological order starting with the early stages of the experiment and progressing

through the end stage of the experiment.

The experiment was done at an engine speed of 600 rpm, focusing on 270 CAD.
The reason 270 CAD was chosen is because 270 CAD is in late compression stroke.
Furthermore, the ability to control the flow before ignition would improve fuel economy
in engines because one can design a leaner fuel injector if the in-cylinder flows are more
predictable. However, in the beginning of this research, 90 CAD and 180 CAD were
investigated too. These results will be presented after all the 270 CAD results are
presented. The experimental parameters were the number of velocity realizations needed

for averaging, the amplitude, the forcing frequencies, and the duration of forcing.

The field of view of the MTV measurement forcing experiment was about 4.5cm
x 5.0cm which located at the center of the cylinder, about 2cm from each side of the
cylinder wall (refer Figure 9). The intake line was modified and the speaker was put at
the intake line with the purpose of perturbing the shear layer of the flow at the intake

valves.

19

 

 

-1

>4’2

-3

-5

-6

Spark Plug

Intake Valve Exhaust Valve

IIIIIIIII

_

 

 

 

 

 

 

 

Figure 9.

Field of view in the engine cylinder

20

 

3.2 Number of realizations needed for averaging

The purpose of this experiment was to determine the number of velocity
realizations needed to give a reliable average of the velocity profile and fluctuating level
that would represent stability and convergence. The experiment was done for the natural
flow (unforced) case with the engine running at 600 rpm without any perturbation. 1200
instantaneous images were taken at 270 CAD. They were then processed and averaged
using different numbers of images namely, the average of 200 images, 500 images, 750

images and 1200 images.

The velocities for the experiment were extracted along horizontal and vertical
lines. For the horizontal line, it was located about 1.2 cm below the intake and exhaust
valves. As for the vertical line, it was located at the center of the cylinder and about 4.5
cm from the cylinder wall. The locations of these lines extracted were arbitrary and with

no specific reason.

The results showed that the average of 200 realizations is the minimum number
for a reliable average. However, in order to make it more accurate, the 500 realizations
were chosen as the number of realizations needed for the all experiment except for the
free-run perturbation experiment. In the free-run perturbation experiment, only 200

realizations were used in each average.

21

Natural Flow. Average. U component. Vertical Line

 

 

 

 

 

 

 

 

 

 

500 - - . . a - - - . - - - -
A——-—- A Avg of 1200 .
’ v- — —v Avg of 500
[3 - - — +3 Avg of 200
250 .
0 r .
j 1
-250 > <
l
-500 - - + - - -
-7.5 -5.0 -2 5 0
Y-axis
Natural Flow, Average. V component. Vertical Line
300 — - - - . - - - - T

 

A--——-— A Avg of 1200 t
G-——eAvgof750 .
200, V"—VAVQOI500
B——- +3 Avgof200

 

 

 

100-

 

 

 

 

Y-axis

Figure 10. U mean (cm/s) and V mean (cm/s) for vertical line taken at x = 4.5
cm from cylinder wall

22

Urms

Natural Flow, Average. U RMS, Vertical Line

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

150 - - - a . - - - - . - . -
i 1
130 t
110 .
w b
? A— A Avg of 1200
G— — -0 Avg of 750
v— - —v Avg of 500
70' B---«E1Avgot200
50 L - - - ‘ ‘ L L - ‘ - - L
-7.5 -5.0 -2.5 0
Y-axis
Natural Flow. Average. V RMS. Vertical Line
120 - e - - . - - f - . . . -
. EJ
110 » ‘. .
. ' t
t . 4
> l
100 .
m .
r IE"
. // A——— A Avg of1200 J
80 . Q o— — -e Avg of 750 ,
{3 ~— — _ El Avg of 200
70 L - - - ‘ L L - ‘ ‘ - - A L
-7.5 -5.0 -2.5 0

Figure 1 1. U rrns (cm/s) and V rrns (cm/s) for vertical line taken at x = 4.5 cm
from cylinder wall

23

400 - - . - - - - -
A—«-—— A Avg of 1200 .
G— —— —0 Avg of 750 ,
300’ v—-—vAvgof500
[3 ~ -- ~ E1 Avgof200
200 ~
1
100 -
o A A A A A A
2 4 6 8
X-axis
Natural Flow. Average, V component. Horizontal Line
300 ~ ~ T ~ - ~ . - . -
r 4
. Ar— A Avg of 1200
200, G———€Avgof750 i
, v— - —7 Avg of 500
, B—--5Avgof200
100 ~ .
o .

i r
-100 . .
-200 L L L L L L L L L L

2 8

X-axis
Figure 12. U mean (cm/s) and V mean (cm/s) for horizontal line taken at y 2

Natural Flow. Average, U component. Horizontal Line

 

 

 
 
 
     

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1.2 cm below the intake and exhaust valves

24

U.

v.

Natural Flow, Average. U RMS. Horizontal Line

 

 

 

 

 

 

 

 

 

 

 

 
 
  

 

 

 

 

 

 

 

140 - - - . - A - -
i
A————- A Avg of1200
. A e———--e Avgof750 ~
437- \\ v—-—V Avg of 500
. K/ K E-HBAvgofzoo .
\ r
120 » i i
100 > i
' l
80 A A L A A A A A A AA
2 4 6 8
X-axis
Natural Flow. Average, V RMS. Horizontal Line
120 . ~ T ~ - e - a
i A—— A Avg of 1200
, G— — —0 Avg of 750
, £3 — ~ - Cl Avg of200
100 - .
go . .
80 u 4
70 L L L L L L L
2 4 8

X-axis

Figure 13. U rrns (cm/s) and V rrns (cm/s) for horizontal line taken at y = 1.2

cm below the intake and exhaust valves

25

3.3 Amplitude effects

The next experiment was performed to determine the amplitude effect of the
acoustic perturbations. The amplitude here was referred to the amplitude of the sine wave
coming out from the amplifier. This experiment was done at three different amplitudes
with the engine running at 600 rpm and the crank angle of investigation was at 270 CAD.
The three different amplitudes investigated were a small amplitude of 675 mV, 3 medium
amplitude of 10 V and a large amplitude of 28 V (28 V is the maximum amplitude that
the amplifier can produce). The perturbation frequency was at 50 Hz. This frequency was
arbitrary chosen which later proven that it was not the optimal reduction frequency.
However, the trend of the results found in this experiment is expected to the same if it

was done with the optimal reduction frequency which was at 400 Hz.

The results of the experiment indicated that as the amplitude of the perturbation
increased, the U rrns and V rrns were reduced. In other words, the trend of the experiment
was the higher the amplitude, the lower the cycle-to-cycle variations. For U mean and V
mean, the results showed that there were some changes in the mean as the amplitude
increased, but they were small. Thus, from the experiment, the highest amplitude (28V)

was chosen as the best parameter for the rest of the research and experiments.

26

U mean

V mean

Different Amplitudes at 50Hz Forcing. Center Vertical Line

 

 

 

500 - - ' ' ' T ' ' - - - I
' :3 -. - - a Unforoed
. 9. — -0 Small Ammitude (575 mV)
_ V- - —v Medium Amplitude (10 V)

250 . [3,...— A High Amplitude (28 V)

 

 

 

~250

 

 

.5004# W.
-6.0 4.5 .3.0 -1.5

 

Y-axis

Difierent Amplitudes at 50Hz Forcing. Center Vertical Line

 

300 - - - - . - - - r - L

 

— ~ ~ {1 Unforoed ‘
—— -0 Small Amplitude (675 mV) .
- —7 Medium Amplitude (10 V)
—— A High Amplitude (28 V)

T ?rfl

I?

 

 

200

 

 

 

 

 

 

100 r i
P 81
/. L;
. )3 fig.
0 i- //// 4
’ ”5742/ A
. wfiexg/ .
-100 - - e . - - A - 4 - e - -
-6.0 -4.5 -3.0 -1.5

Y-axis

Figure 14. U mean (cm/s) and V mean (cm/s) of different amplitudes for
vertical line taken at x = 4.5 cm from cylinder wall

27

Urms

Vrrns

Different Amplitudes at 50Hz Forcing. Center Vertical Una

 

 

 

 

 

 

 

 

 

l3 - - - {3 Unforeed ‘
160 G— — -0 Small Amplitude (675 mV) J
* v—-—v MediumAmplitude(10V)
L A——A-A Hig-lAmplitude(28V) ’ '8 __ E; ~r‘=
, ‘3 . e'/ - ‘ a E 5%
/ ‘ t
i 7' g, X \x\\\‘\bi
,L‘ / //.’ K \
h I I \- 1
120 i /‘2/ fl 3.
. / l
i ,m / / ‘
i / i
gait/em!
1\ D _. ;//’
80 ’V/ 1
i i
-6.0 4.5 -3.0 -1.5
Y-axis
Different Amplitudes at 50Hz Forcing. Center Vertical Line
120 v - 1 , a
i , B— "3* _ fl 1
/ ’9 ‘
i g / \\ i
110 l ,
. B, d anew-A Qua/xi \:
' / / ’ ‘\ \1 iii
, 2/ \ L‘l
, )3 /O/// , \ 1
1m , / /0/ (AI// ‘ 1
i

/ v ‘
m b \ 1
? \ [fif/ ’v._, k e “‘z/ ‘
i“ /" ‘ 1
A \ ll / ‘
80 E - .21 B ' “ ’ {3 UMOI'OOd 1
. \/ o— —— —0 Small Amplitude (675 mV) ,
. v— - —-v Medium Amplitude (10 V) .
, 12,—— d High Amplitude (28 v) 1
i i
70 A - . A A A . - - A
-6.0 4.5 .3.0 -1.5
Y-axis
Figure 15. U rms (cm/s) and V rms (cm/s) of different amplitudes for vertical

 

 

 

 

 

 

 

line taken at x = 4.5 cm from cylinder wall

28

U mean

V mean

Different Amplitudes at 50Hz Forcing. 2nd Horizontal Line

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

400 - ~ A . , A
(3 - ~ - a Unforoed
b o- — -0 Small Amplitude (675 mV)
V- - —7 Medium Amplitude (10 V)
. A—— A High Amplitude (28 V)
i
300 ’ x \ 1
l ~\\&
200 t
100 -
; c» a ‘
0 A 4 A A
2 4 6 8
X-axis
Different Amplitudes at 50Hz Forcing. 2nd Horizontal Line
300 A - - . , + - , v a
G - - 8 Unforoed
. 9g o~ -— -0 Small Amplitude (675 mV) i
200 , a >gg\ v— - —7 Medium Amplitude (10 V) i
, f3 _‘-\ x \0\ A~—--——— A High Amplitude (28 V) ‘
» ‘ " EBN‘ \
N
_ \
100 - \‘\‘ \
\ . 1
: Km l
c-\ ‘
o \‘3 ‘65
h \ \\ .4
. RR
C \“a?;§ ‘
\AAx
-100 - ‘3 A
-200 A A 4 - + A
2 4 6 8

X-axis

Figure 16. U mean (cm/s) and V mean (cm/s) of different amplitudes for
horizontal line taken at y = 1.2 cm below the intake and exhaust
valves

29

Urms

Vrms

Diflerent Amplitudes at 50Hz Forcing. 2nd Horizontal Line

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

130
i 3 - - - +3 Unlorced ‘
' o— — -9 Small Amplitude (675 mV) ‘
v. - —7 Medium Amplitude (10 V) ‘A ‘
120 A “ ‘
’ i
L l
110 > .
. J
I /’ I
. - d/I ‘
100 - ' b ’1 ‘
’ 1 , ’ )5// i
L ’U fl ‘
. ,, ,x/
90 _ .-’/g
[r7 / i
//
. , ,e
80 A A A A A A A A A A e
2 4 6 8
X-axis
Different Amplitudes of Forcing at 50 Hz
130 ~ A v . m ' ' ' ‘ ' r
i [3 - ~ Q {3 _ - — {3 Unlorced
O— -— ——0 Small Amplitude (675 mV)
\ v—-—V MediunAmplitude(10V)
. 121—— -A High Amplitude (28 V)
l‘ t . 5* ' a
’ E \ l
110 A A K J
\ /_ §\/ \
\\ g\\ \ \ \\
\ \ \o
\e.‘ M t)
90 \ \. ‘
\\ \ w - e
\73 \ i
b _ U
L K - V ‘:%\\ 9 l
\e- ‘V
> x M‘ ‘
\a
70 A A A A A A A A A
2 4 8
X-axis

Figure 17. U rms (cm/s) and V rms (cm/s) of different amplitudes for
horizontal line taken at y = 1.2 cm below the intake and exhaust
valves

30

3.4 Free-run perturbations

At early stages of this research, the perturbation was done on free run mode. In
this mode, the speaker was turned “on” continuously with the engine at respective
frequency. The number of realizations averaged was 200 realizations for each frequency
case in the experiment. This was the only experiment with ZOO-realization average. The
rests of the experiments in the thesis were with SOO-realization average.

In the experiment, the range of frequencies investigated was from 50Hz to
4000Hz. Nevertheless, from this experiment, it was found that the best reduction
occurred at low frequency, namely frequency below 400Hz. Beyond 400Hz, there was no
significant reduction in velocity rms or cycle-to-cycle variation. This is consistent with
what Ambrose had found. Ambrose found that the preferred frequency was at 40 Hz or a
Strouhal Number of 1.19 [3]. According to Ambrose, a Strouhal number (StD = fl)/U) for
the piston-cylinder configuration could be based on the maximum mean velocity of the
flow through the intake valve opening, the diameter of the valve and the forcing
frequency [3]. In this research, with the valve diameter of 37 mm and the maximum mean
velocity of 10 1er, the preferred frequency would be at 321.62 Hz.

The results are organized in the following manner; starting with U mean and U
rrns, then followed by V mean and V rrns. Figure 18 showed the results for U mean and
U rrns taken from running the engine at 600 rpm and 270 CAD. The color contour
showed the U rms. In other words, the cycle-to-cycle variation is reduced if the rms value
is reduced. The trend of the results was as the frequency of the perturbation increased, the

U rms decreased.

3]

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

# ___fi__ ,7fi7 __ ___J

 

Figure 18.

Ensemble averaged resulmzhowmg rrns velocity (cal/s),U-mean, without
perturbation and with perturbation of 50Hz, 100Hz, 200Hz, 300Hz and
400Hz at 270 CAD

32

Figure 19 and 20 show lines plots of U mean and U rrns. From Figure 19 and 20,
it is obvious that the perturbation did not have effect on the U mean flow. The U mean
stayed about the same values throughout the whole frequencies. However, the
perturbation had huge effect on the U rms that represents the cycle-to-cycle variation and
fluctuating velocity field. As the frequency increased, the U rrns decreased. The largest
reduction among all was the 400Hz case. It showed about 25% to 30% reduction at

certain region of the flow field.

The results for V mean and V rms are shown in Figure 21 with the same way of

organizing the results; starting with color contour plots and then line plots.

33

(cm/s)

U
avg

(cm/s)

U

"“8

U mean at cylinder centerline

 

 

 

 

 

 

 

 

 

400-e.-,.-..,-.-....-....--
Elm—e 400Hz
o---¢ 300Hz
A—A-A ZOOHz
200 A v—-—v 100Hz .
G———o 50Hz
L r- —- Unforced
O A .
-200 A -
400 k; . . l . . A . l A . I L L .........
-5 -4 -3 -2 -1 0
Y-axis(cm)

U rrns at cylinder centerline

 

175m vvvvvv IYT'VI'V"

L lain—a 400Hz
o---0 300Hz fifilr-A‘,
150 . A———A ZOOHZ // \ .. .-
F-q 1OOHZ /0/’j v.8::%\

- G—ue 50Hz g/JA 42;

I
‘9
’ ‘ \ \V ‘
Uf d A ‘ ~
125 ' ' "owe A" fl -. \8 «

1oo- /}/{ ” ~
7.6-“ ,I ,, 3‘ ~

 

 

 

 

 

 

Y-axis (cm)

Figure 19. Effect of perturbations on the U mean and U rrns at cylinder
centerline, about 4.5 cm from cylinder wall

34

(cm/s)

avg

U

(cm/s)

U

rrns

U mean, horizontal line, about 1.2 cm below the valves

 

 

   

 

 

 

 

 

 

 

 

 

 

 

400 . A
is ~ ~ » e 400Hz
AX o---o 300Hz ‘
_ / -_ A—— A 200Hz ,
300 ‘ g , v—-—v 100HZ
\ \\ o — ~e 50Hz
I — _ — a Unforced
200 . A
100 - -
0 . .
-100 ‘ ‘
0 2 4 6
X-axis (cm)
U rrns, horizontal line, about 1.2 cm below the valves
160 r .
:::1' § E A - i,
140 F-.. 1 g Ix} A \ ‘l -
O— —o 5 Z V. \S \
n A - a Unforced // [:LTVW \s ‘ i
120 - y ,x 73% \\0\ -
I ’ (I , , ’3‘ ::‘\ \\
100 ’7 / \ A3
.. / E \ \ \ ‘
,’ I / K ’ EN‘Q
80 V7 5/1 / \ ~ \I
V
60 . b -
4O ' a L L r
0 2 4 6
X-axis (cm)
Figure 20. Effect of perturbations on the U mean and U rms at horizontal line

about 1.2 cm below the intake and exhaust valves

35

 

WM Fora-g

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 21. Ensemble averaged results showing rms velocity (cm/s),V mean, without
perturbation and with perturbation of 50Hz, 100Hz, 200Hz, 3001-12 and
400Hz at 270 CAD

36

 

Figure 21 showed the results for V mean and V rrns taken from running the same
engine condition which was at 600 rpm and 270 CAD. The color contour showed the V
rrns. The trend was as the frequency increased, the V rrns was reduced. There were huge
reductions for 200Hz, 300Hz and 400Hz cases. One could notice huge reduction
difference between lOOHz and 200Hz. The red-and-yellow-color region in the lOOHz
case disappeared when the frequency was increased to ZOOHz. In order words, the V rrns
was reduced to smaller value by increasing frequency from lOOHz to 200Hz. The line

plots for V mean and V rrns are shown in Figure 22 and 23.

37

Vavg (cm/s)

(cm/s)

V

V mean at cylinder centerline

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

800 T D f T I v v 1 r v v v v I x v . . I
B-A-fl 3 QHZ 4
ou-o 2
H z
600 " F--" % HZZ ‘
-- U
400 A .
200 A .
0 - r
-200 4-4--.4a---44
-5 -4 -3 -2 -1 O
Y-axis(cm)
V rms at cylinder centerline
175 ....................
one g z
0-". z
150 - L -2 i .
o——e z
---a Unforced .
125 A .
” ‘.r.._
,1 —-9~ \
’ /I”"*>"‘Q/m\‘L \
//8"/&,4ee __.e"
100 * iii-3:9”. VA 1" "F’ ‘Ae‘k ‘
fies-tame A
, )1ZPPVH'J
75 - 2!? .
b I/ '
50 i l A l 1
-5 -4 -3 -2 -1 O
Y-axis(cm)
Figure 22. Effect of perturbations on the V mean and V rrns at cylinder

centerline, about 4.5 cm from cylinder wall

38

avg (cm/s)

V

V mean, horizontal line, about 1.2 cm below the valves

 

 

   
  

 

 

 

 

 

 

 

 

 

 

 

(cm/s)

V

rms

300 . . . . . . . r T - T
b nifi-a é gal ‘
0-“. Z
AA—A Z
200 - ¥§ IL: 5 22 ~
. i‘ . .__. Unforced ,
100 A A
i
0 A ..
—100 A -
-200 - ‘ A A ‘
O 2 4 6
X-axis (cm)
V rms, horizontal line, about 1.2 cm below the valves
160 . . . , . . , ,
a—A—a 4 Z ‘
14o - Z_.."‘ g g i «
v—--v 2
F4 3 f2 d i
120 - 91:%:9~e\ "“ "me .
GA ' ~Z~.H-"‘\3\
A3323:\~v:\.
100 - wk. .\ .
xi“: ‘ ~— 4::
80 A A 4;?
i
60 A
40 A A 1
0 2
X-axis (cm)

 

 

 

Figure 23. Effect of perturbations on the V mean and V nns at horizontal line
about 1.2 cm below the intake and exhaust valves

39

3.5 Fixed-time perturbations : Perturbation from 0 CAD to 180 CAD

In all these experiments, the starting and ending points of the perturbation were
timed. The timing of the perturbation was done by using the burst mode of a function
generator. The speaker would only start perturbing when the piston reached Top Dead
Center (TDC). It would continue to perturb from 0 CAD to 180 CAD. The duration of

perturbation was 0.05 seconds.

3.5.1 Perturbation from 0 CAD to 180 CAD with no ph_ase shift

The engine condition was the same as in the previous experiments. The
experiment was done at 600 rpm and 270 CAD. The range of frequencies investigated
was fi'om 50 Hz to 4000 Hz. Each frequency case was averaged from 500 continuous,
instantaneous realizations. The reason for choosing this frequency range was based on the
previous experiment (free-run experiment). From the free-run experiment, it was found
that the perturbation had no noticeable effect beyond 4000 Hz and below 50 Hz.
Perturbation with no phase shift here is referring to the sine wave coming out from the
amplifier. The starting point of the sine wave was at 0 degree.

The results had the same trend as the free run results. There was a noticeable
reduction in U rrns and V ms at 300 Hz and 400 Hz. The reduction was about 20%-30%.
As the frequency increased, the U rms and V rms decreased until 400 Hz and it increased
beyond 400 Hz.

The best results of the experiment, namely from 30Hz to 1000Hz are shown in
Figure 24, 25, 28 and 29. The results are organized in following manner; starting with U

mean and U rms, then followed by V mean, V ms.

40

 

 

 

 

 

 

 

 

 

mayors-meta
mm<c~-~_)

go,h
mca‘n—M

 

 

mmmm
or.—

 

 

 

 

 

 

 

Figure 24. Ensemble averaged results showing rrns velocity (cm/s),U mean, without
perturbation and with perturbation of 30Hz, 50Hz, 100Hz, 200Hz and
300Hz at 270 CAD

‘41

 

when Fauna

 

Sm/s

 

 

 

 

 

 

 

 

 

fill

152
1514
165.5
183.8
145

141.2
13.4
131.8
126.8
122

111.2
112.4 —
102.6
1112.8

831
Ill

1592
644

54.8

 

 

 

     
 

““:“‘E:""“““'
AzizzritcztsisiiessA A

~. huh

AEEEE'EEEE'ggiésaaag

 

SEISUBIDI:
.eua s.»-

 

 

Figure 25.

Ensemble averaged results showing rms velocity (cm/s),U mean, without

perturbation and with perturbation of 400Hz, 600Hz, 800Hz and 1000Hz

at 270 CAD

42

The trend of the results was as the frequency of the perturbation increased, the U
rms decreased until 400Hz and then it started to increase beyond 400Hz. In the previous
free-run perturbation experiment, the U mean and V mean of the flow did not change for
cases with or without perturbation. However, for this fixed-time perturbation, the U mean
and V mean of the flow decreased in velocity. They appeared to be slowing down with

the perturbation.

The next few line plots of the same conditions, U mean and U rms, are shown in

Figure 26 and 27.

43

Uavg (cm/s)

Urms (cm/s)

U mean, Vertical line at cylinder centerline

 

 

 

 

 

 

       

 

 

 

 

   

 

 

 

 

 

 

 

250 V v v v 1 v t v I v v t
. a---e Hz
0- — o z
x---x z
I - - i Z
’ o - 4 Z 1
_ +---+ z
A———-A 1 z
o __ F". HZ -
i3 _ _. £1 Z
t r- -& Unforced .
b J
-250 _ A
d
L
h ’5’?-
r f/II
-500 L a A L l 4 1 l A 1 1 l 4 41
s .5 .4 -3 -2 -1
Y-axis (cm)
U rms, Vertical line at cylinder centerline
175 T Y I l l V t l V V
,. . .. _. 1
u " I a \ J
. J3 ’ ’ 4
150 A -
E I
125 . -
i- ll
1“) h B---B Hz -l
L o— —0 Z q
L x-‘* Z 4
I - - A! Z
. 6 _ a z .
h +---+ z 1
75 A ’_ / (tr—A z -
. o v—--v Z ‘
. 13 A - -C 2
i o -- 4 Unforced
L
w + 1 L A l L A 1 l A A l A A 4 g
s -5 4 -3 -2 -1
Y-axis (cm)
Figure 26. Effect of perturbations on the U mean and U rrns at cylinder

centerline, about 4.5 cm from cylinder wall

Uavg (cm/s)

U rms (cm/s)

U mean, Horizontal line at 3.1 cm below the valves

 

 

i
100i
1

 

 

 

 

 

 

 

2‘ 3W456 7
X-axis(cm)

U rms, Horizontal line at 3.1 cm below the valves

 

17525
D

150

125

 

100’

75

 

 

 

 

 

 

 

2 3 4 5 6 7

X-axis (cm)

Figure 27. Effect of perturbations on the U mean and U rms at horizontal line
about 3.1 cm below the intake and exhaust valves

45

From Figure 26 and 27, it is obvious that the perturbation had some effects on the
U mean flow. The U mean decreased as the frequency increased. In some places of the
flow, the U mean decreased as much as 70% to 80% of the unperturbed flow. The flow
direction had changed too in some places of the flow. As for the U rms, the perturbation
had huge effect on the U rms that represents the cycle-to-cycle variation and fluctuating
velocity field. As the frequency increased, the U rms decreased until 400Hz and then it
increased beyond 400Hz. The largest reduction among all were at 300Hz and 400Hz. In

300Hz case, it showed about 20% to 30% reduction at certain region of the flow field.

The results for V mean and V rms are shown in Figure 28 and 29 with the same

way of organizing the result; starting from color contour plots and then line plots version

of the results.

46

 

 

 

........ _ ‘ vi

 

 

 

 

 

 

fig

 

 

ggmggmmas
Adz-’09 ”Amm

:J-u
5:...»
M

 

 

 

 

 

 

 

 

 

 

Figure 28. Ensemble averaged results showing rms velocity (cm/s),V mean, without
perturbation and with perturbation of 30Hz, 50Hz, 100Hz, 200Hz and
300Hz at 270 CAD

47

 

 

 

5m/s

 

 

 

 

 

 

 

 

 

'EEEE'EEEE'SEESAEEEE'EEEE'EEEE'Egfigiégégia§§§5§§i§i\

 

 

 

wm Forcing coon:
“
4 5 6
X
I00“:
5 ml.
—.
4 5 6
X

11}

13.0
123.0
1214
117.2
114

110.0
101.0
104.4
101.2

040
01.0 r
004
052

100
150
224
002

020
600

512

 

 

 

 

 

 

 

 

Figure 29. Ensemble averaged results showing rrns velocity (cm/s),V mean, without
perturbation and with perturbation of 400Hz, 600Hz, 800Hz and 1000Hz

at 270 CAD

48

Figure 28 and 29 showed the results for V mean and V rms taken at 600 rpm and
270 CAD. Each of the frequency case had an average of 500 realizations. The results did
not show any clear trend as found in U mean and U rms previously. They appeared to
have less rms reduction compared with U mean and U rms. The best V rms reduction

cases were 400Hz and 600Hz.

The next few line plots of the same conditions, V mean and V rms, are shown in

Figure 30 and 31.

49

V avg (cm/s)

Vrms (cm/s)

V mean, Vertical Line at cylinder centerline

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

200 1 I I ' ' 1 ' l j 7 I
~ e---a % Z
’ o- —e z
x---x Z
. I - - -II 2
> 0 A '0 z
150 A X: g g
. v—--v n2
(3 - - -0 Z
w - a Unforced
. J
100 - .
50 - -
1.
o A 4 4L L l l A A A A l 1 1 A a
-6 -5 .4 -3 -2 -1
Y-axis (cm)
V rrns, Vertical line at cylinder centerline
135 v r f v 1 i T ' T ' r V f r f T T ffi
120 - -
105 -
m " i-
75 A A
m l I A L
45 -3 -2 -1
Y-axis (cm)
Figure 30. Efi‘ect of perturbations on the V mean and V ms at cylinder

centerline, about 4.5 cm from cylinder wall

50

Vavg (cm/s)

V rms (cm/s)

V mean, Horizontal line at 3.1 cm below the valves

 

 

   
   

 

 

 

 

 

 

 

 

      

 

 

 

 

 

 

400 . . . - e - - . - - - - - - - - a
: . ._ \ e-—-a Hz
’ O o— ——e z
’ -‘ x---x z
. e——a z
300 ' ‘ o - - 4 z
’ +---+ n2 ‘
i . A—A Z 4
. \ v—--v Z .
i ‘ G ' ' 'U z l
200 . \ nforced .
100 .
i
0 >
1
-100 A A - - - - e L - A - A A
2 3 4 6 7
X-axis (cm)
V rms, Horizontal line at 3.1 cm below the valves
14o . - - - - - - -
i B--'B % z
o——e z
‘ x---x z
b I--- -I Z
is _ - a z
1 +---+ z
b—A z
110 . '"4 Z
[3 - - AD 2
«A --4 Unforced
so . 4
w + A A A A A A A A A A A 4A A k A
2 3 4 6 7
X-axis (cm)
Figure 31. Effect of perturbations on the V mean and V rrns at horizontal line

about 3.1 cm below the intake and exhaust valves

5l

Some observations could be made from the previous line plots. The perturbation
had some effects on the V mean flow, the same effects that could be found on U mean
flow. The V mean decreased as the frequency increased. In some places of the flow, the
V mean decreased as much as 40% to 50% of the unperturbed flow. As for the V rms, the
perturbation had some effects on the V rrns too. As the frequency increased, the V nns
decreased until the frequency reached between 400Hz and 600Hz. It increased beyond
600Hz. The largest reduction among all were at 400Hz and 600Hz. In 400Hz case, it

showed about 20% to 30% reduction at certain region of the flow field.

52

3.5_.; Perturbzgion from 0 CAD to 180 CAD with 900 phage shift

Like the previous experiment, the starting and ending points of the perturbation
were timed. The perturbation was turned “on” from 0 CAD to 180 CAD with an input
sine wave of 28 V amplitude. Nevertheless, in this experiment, the input sine wave was

shified.+90° in phase. The starting point of the input sine wave was at 900 instead of 00.

The range of frequencies investigated was from 50Hz to 4000Hz.
The main observation of the experiment was that the perturbation did not reduce
the U rrns and V rms. The main effect of the perturbation was that it increased the U rms

and V rms instead to about 10% to 20% increment.

The color contour plots of the results with frequency range from 50Hz to SOOHz

are shown in Figure 32, 33, 34 and 35.

53

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5 m/s [1'
wmrm sou: 51v: 150
15 116
2 142
25 138
.3 131
>_3_5 138
4 125
4.5 122
.5 118
111
“55 . . 118
2 3 4 x 5 6 106
182
—_ "W T 98
mm zoom 5m 9‘
.1 5 _. 91]
.2 85
2 5 82
18
3 74
>‘ 3 5 10
4 85
A4 5 52
5 58
55 51
2 a 4 x I 5 1 6 50
C WT D
Figure 32. Ensemble averaged results showing rrns velocity (cm/s),U mean

(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 270 CAD

54

 

 

 

 

 

 

 

 

 

 

 

5 m/s u'
WM PM 5:: : 30.“! 5:1. 150
15, 118
2 f- 112
25%. 133
3 f- 131
>-3.5 _— 130
4; 128
g 122
*5 118
5 ’ 111
55 . 118
2 3 4 x 5 6 105
182
38
5””: 2': 94
15 90
_2 85
-2.5 82
_3 18
> . l1
3 5 In
4 88
4 5 52
5 58
5 5 S1
2 3 4 x A 5 A 6 I 50
C D T
Figure 33. Ensemble averaged results showing rms velocity (cm/s),U mean

(cm/s), (A): without perturbation and with perturbation (B):300Hz,
(C):400Hz, (D):500Hz at 270 CAD

55

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5 m/S 1'
wmrm : m: 5:: 130
121.1
123.1
121.1
111.2
a”; 111
4; 111,1
* 111.1
"5 111.1
5 ' 111.2
55 98
2 3‘1‘L4L;Hsluls‘ 9‘13
911
11.1
: 100“: 5:1. 852
12
2; 18.8
.25; 751
12.1
> g 11.2
3 5 :- 56
4 i- 12.1
45 51.1
s - 51.1
55 53.2
2 3 4 ‘x' 5 ' a 51
c D

Figure 34. Ensemble averaged results showing rrns velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 270 CAD

56

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5 m/s 1'
WMFM : SOON! 5:: 130
152- 121.1
.2_ 123.8
.25; 128.1
.3; 111,2
m
4:_ 118.8
’ 181.8
4.5: 181.1
'5: 181.2
2 3 4 x s 6 91.8
91.8
88,1
AWN: 500M: 5;": 852
.15 82
_2 18.8
.25 15.8
_3 12.1
> 882
-3:5 56
4 82.8
45 58.8
-5 . 58.1
.55 53,2
2.11.31.114L;A‘5.IA‘6H 511
C D
Figure 35. Ensemble averaged results showing rrns velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation (B):300Hz,
(C):400Hz, (D):500Hz at 270 CAD

57

3.5.3 Perturbation from 0 CAD to 180 CAD with —90°pl;ase shift

In this experiment, the start of the perturbation sine wave was shifted —90°. Like
the previous experiment, the starting and ending points of the perturbation were timed
and the perturbation was performed from 0 CAD to 180 CAD. The range of frequencies

investigated was from 50Hz to 4000Hz.

The results did not show a clear trend as in the cases of free-run and timed
perturbation without phase shift. In the previous cases, free-run and timed perturbation
without phase shift cases, both cases showed that the perturbation reduced the U rrns and
V rms as the frequency increased until 400Hz. Beyond 400Hz, the perturbation increased

the U ms and V rrns.

As for this experiment with -900 phase shift, the best U rrns reduction case was at
50Hz. It showed a reduction of about 10% to 20% at certain region of the flow. As for the
V ms, the best reduction case was at 500Hz. It also showed a reduction of about 10% to

20% at certain region of the flow.

The color contour plots of the results with frequency range from 50Hz to 500Hz

are shown in Figure 36-39.

58

 

 

 

 

 

 

 

 

 

 

 

5 m/s U'

“"‘ _. 151

15 118

2 112
25 138
_3 131
>-3.5 130
4 128

4 5 122
118

5 111

55 . . . . . 118
2 3 4 ‘ 5 6 105
182

98

zoom 2': 94

15 88

2 88
.25 11
> ‘3 11
3 5 70

4 88

-4 5 62

5 58

5 5 51

2 a 4 x 5 e 50

C D
Figure 36. Ensemble averaged results showing rms velocity (cm/s),U mean

(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 270 CAD

59

 

 

Sm/s u'

SOON: —. 150
118

112

138

131

138

128

122

118

111

.. . .. ... 118
4 ‘ 5 6 105
182
88
5:,- 91
98

88

82

18

14

18

88

82

58

51

 

 

 

 

 

 

 

 

 

 

 

Figure 37. Ensemble averaged results showing rms velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):300Hz,
(C):400Hz, (D):500Hz at 270 CAD

60

 

 

 

 

 

 

 

 

 

 

 

 

 

w s 1 5 m/s 1'
m M 5". WI —. 130
*‘55 1288
-2 1238
52.5 128.1
.3 111.2
>45 111
4 11118
45 181.8
' 181.1
'5 181.2
'5'5 .. 99
2 3 4 x 5 6 91.8
91.8
88.1
mm 5;": zoom 2": 852
.15 82
.2 18.8
45 15.8
_3 12.1
>55 88.2
' 88
4 82.8
"5 59.8
‘5 58.1
.55 53.2
2ll‘13ll..4|‘x1151‘161ll 50
C D
Figure 38. Ensemble averaged results showing rms velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 270 CAD

61

 

 

 

 

 

 

 

 

 

 

 

 

 

5 m/s 8'

wmruung 3:: 301m: —. 1311
~15 128.8
.2 123.8
.25 128.1
_3 111.2

’135 111
4 1188
181.8
4.5 181.1
'5 181.2

-5.5 98

2 3 4 x 5 6 91.8

81.8

88.1

500": fly: 85.2

.15 82

.2 18.8

45 15.8

‘3 12,1

892

>415 58

4 82.8

4.5 595

‘5 58.1

.55 53.2

2....3..-.4.rx..5....6... 50

C D
Figure 39. Ensemble averaged results showing rms velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation (B):300Hz,
(C):400Hz, (D):500Hz at 270 CAD

62

3.6 Fixed—time perturbations : Perturbation from 0 CAD to 90 CAD
After doing the fixed-time perturbation (0 CAD to 180 CAD) experiment, it was
interesting to see what effect the flow would have if the duration of the perturbation were

reduced. With this in mind, the flow in this experiment was only perturbed from 0 CAD

to 90 CAD.

In this experiment, the starting and ending points of the perturbation were fixed. It
started to perturb when the piston reached TDC and stopped when the piston reached 90
CAD. The experiment was done on 600 rpm and 270 CAD. The range of frequencies
investigated was fi'om 50 Hz to 4000 Hz. Each frequency case had an average from 500

continuous, instantaneous realizations.

For this experiment, the results did not have a clear trend of U rms and V rrns
reduction. The reduction range was from 10% to 20% for both U rms and V rms. The
best case of the experiment was the 600Hz case for maximum U rrns reduction. As for V

rrns, the best reduction case occurred at 1000Hz.

The color contour plots of the results with frequency range from 6OOHz to

1000Hz are shown in Figure 40 and 41.

63

 

 

 

 

 

 

 

 

 

 

 

Figure 40. Ensemble averaged results showing rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):600Hz,
(C):800Hz, (D): 1000Hz at 270 CAD

 

 

 

 

 

1000"! 852

 

   

 

 

 

 

Figure 41. Ensemble averaged results showing rrns velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):600Hz,
(C):800Hz, (D): 1000Hz at 270 CAD

65

3.7 Sweepingperturbations : sweep frequencies starting from lOOHz

 

The main reason that motivated this experiment was that the intake valves opened
in a linear manner during the intake stroke. If the perturbation were fixed at certain
frequency, the flow would not be perturbed at that frequency in the beginning of the
valves opening period. To improve the maximum effect of the perturbation to the flow,

the perturbation was done by sweeping the frequencies linearly.

This part of the experiment was done by utilizing the sweeping mode of the
function generator. As in the previous few experiments, all the perturbation frequencies
were fixed. For example, when the flow was said to be perturbed at 400Hz, it was done
by perturbing the flow constantly at 400Hz throughout the experiment using free-run
method or fixed timing method. However, in this experiment, the perturbed frequencies
increased linearly but at a fixed starting point of perturbation which was at TDC. Each
case of experiment was done by perturbing the flow starting at 100Hz and then swept
linearly to the respective end frequency. Nevertheless, the duration of the sweeping was
still only from 0 CAD to 180 CAD. In the case of engine running at 600 rpm, the duration

of sweeping was 0.05 seconds.

The range of frequencies investigated in the experiment was from 200Hz to
2000Hz. The first case of sweeping frequencies was sweeping from lOOHz to 200Hz and
the last case was sweeping from lOOHz to 2000Hz. The cases investigated were sweeping
from 100Hz to 200Hz, sweeping from lOOHz to 300Hz, sweeping from lOOHz to 400Hz,

sweeping from IOOHZ to 600hz, sweeping from lOOHz to 800Hz, sweeping from 100182

66

to 1000Hz, sweeping from 100Hz to lZOOHz, sweeping from 100Hz tol4OOHz, sweeping
from 100Hz to 1600Hz, sweeping from 100Hz to 1800Hz and sweeping from 100Hz to

2000Hz (refer to Table 2).

Table 2. Cases investigated in the sweeping perturbation experiment.

 

100Hz to 200Hz 100Hz to 300Hz 100Hz to 400Hz 100Hz to 600Hz

 

100Hz to 800Hz 100Hz to 1000Hz 100Hz to lZOOHz 100Hz to 1400Hz

 

 

100Hz to 1600182 100Hz to 1800Hz 100Hz to 2000Hz

 

 

 

 

The range of U rrns and V rrns reductions was from 20% to 25% in the
experiment. The best reduction case for U rms was sweeping from 100Hz to 1200Hz case
and the best reduction case for V rrns was sweeping from 100Hz to 800Hz case. The
color contour plots and line plots are shown in Figure 42-50. From the line plots in Figure
50, the U mean and V mean did not change much if compared with the fixed-time

perturbation experiment.

67

 

 

 

 

 

 

 

 

 

 

 

 

 

A B
5 m/s [1'
wmrm tom-b20011:
.... g A 111
15;- 155.1
-2;- 151.2
2.5;— 111.1
.3- 112,1
>‘3‘5 :_ 13B
4; 138.8
45} 129.2
' g 1211
'5 ' 121.1
-5.5 4 L1 1 . . A 115
2 3 4 x 5 6 1115
181.2
1828
100": h 300M: 100“! b40001: 5;": 98‘
1.5 91
.2 89.8
.25 85.2
88.8
-3
A 18.1
-3.5 n
4 81.8
"5 83.2
-5 58.8
.55 51.1
2 3 I 4 t x 5 6 5[]
C # *fifi D F—_

Figure 42.

 

 

Ensemble averaged results showing rms velocity (cm/s),U mean

(cm/s), (A): without perturbation and with perturbation (B): 100Hz-
200Hz, (C): 100Hz-300Hz, (D): 100Hz-400Hz at 270 CAD

68

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

wmr roon soon 5 m/s "I
m 5m no a _. 160
”-5 155.8
-2 151.2
.25 118.8
.3 112.1
5,, 121
4 123.1
45 121.2
' 121.8
'5 128.1
-5.5 r. ......... .... 118
2 3 4 x 5 6 111.6
, i _A_ -_ m 111.2
182.8
room booouz room bloom E 981
.15 81
.2 88.8
,5 15.2
_3 11.1
> 11.1
-3.5 12
4 81.8
“5 83.2
-5 58.8
~55 5“
2.l .I311114‘l‘19511116‘l 50
C *— w" _T)”—*'

Figure 43. Ensemble averaged results showing rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B): 100Hz-
500Hz, (C): 100Hz-600Hz, (D): 100Hz-800Hz at 270 CAD

69

A B

 

 

 

 

 

 

 

    

 

 

 

 

 

 

5 m/s {1'
WM Form 100": h 1000":
5"“ _. 188
"-5 155.8
-2 151.2
.25 118.8
.3 112.1
>135 138
4 133.8
45 129.2
' 121.8
'5 1211
A55 115
2 3 4 x 5 6 1116
181.2
182.8
10011! ”1200"! - 10081: b 1‘00": 2 98‘
4.5 _— 91
.2; 89.8
215; 85.2
‘3; 888
> E 18.1
-3.5 g- 12
4 5' 81.8
“E 83.2
-5 '- 58.8
.55: 51.1
2H L3.1‘14llx1‘51‘116 51]
C D
Figure 44. Ensemble averaged results showing rms velocity (cm/s),U mean

(cm/s), (A): without perturbation and with perturbation (B): 100Hz-
1000Hz, (C): 100Hz-1200Hz, (D): 100Hz-1400Hz at 270 CAD

70

B

 

 

 

5 m/s
mom-aim":

_.

 

llllllll

 

 

 

 

MON: b 1 IN":

10011: b 2000":

 

 

 

 

 

188
188.8
181.2
118.8
112.1
138
133.8
128.2
121.8
128.1
118
111.8
181.2
182.8
88.1
81
88.8
88.2
88.8
18.1
12
818
83.2
88.8
81.1
88

 

 

Figure 45. Ensemble averaged results showing rms velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B): 100Hz-

1600Hz, (C): 100Hz-l 800Hz, (D): 100Hz-2000Hz at 270 CAD

7l

A B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100Hz hZWHI 5 W8 Y.
_. 138
‘5 128.8
2 123.8
2 5 128.1
.3 111.2
’135 111
4 118.8
4.5 181.8
181.1
5 101.2
5 5 L . ._._L 98
2 3 4 x 5 6 9‘8
__ -fi, __,,_A ___.______w 918
88.1
100": h JOON! fr: 100": b40081! L“: 852
15 82
2 18.8
_2_5 18.8
_3 12.1
>135 89.2
' 88
4 82.8
4 5 89.8
5 88.1
55 #4 . . . . . 83.2
2 3 4 x 5 6 50
C D
Figure 46. Ensemble averaged results showing rrns velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation (B): 100Hz-
200Hz, (C): 100Hz-300Hz, (D): 100Hz-400Hz at 270 CAD

72

 

 

 

 

 

 

 

 

 

 

 

 

 

5 m/s 1'
wm PM :U: 100"! I) 50°"! —: 13“
‘5 128.8
2 123.8
2 5 128.1
3 111.2
>65 111
4 118.8
4 5 181.8
181.1
5 181.2
5 5 98
2 3 4 x 5 6 918
91.8
88.1
100": h 000M: 100"! b I”!!! LT 85]
15 82
.2 18.8
45 18.8
.3 12.1
> 89.2
-3 5 55
4 82.8
4 5 89.8
5 88.1
5 5 83:2
2 3 4 I x I 5 A 6 1 51]
C D
Figure 47 . Ensemble averaged results showing rrns velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation (B): 100Hz-
500Hz, (C): 100Hz-600Hz, (D): 100Hz-800Hz at 270 CAD

73

A B

 

 

 

 

 

 

 

    

 

 

 

 

 

 

ioombimuz 5 m/s V.
_. 138
‘5 128.8
2 123.8
2 5 128.1
.3 111.2
>45 111
4 118.8
4.5 181.8
181.1
5 101.2
5" 2 . . . . 93
2 3 4 x 5 6 9‘8
91.8
88.1
100Nxb|mfll : 100N1b1mlu L": 852
15f- 82
2- 18.8
15.6
_3E_ 12.1
>35; 89.2
' g 88
‘ :' 828
4 5 89.8
5 ' 88.1
55 . _ . . 83.2
2 3 4 x 5 6 50
C D
Figure 48. Ensemble averaged results showing rms velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation (B): 100Hz-
1000Hz, (C): 100Hz-1200Hz, (D): 100Hz-1400Hz at 270 CAD

74

 

 

 

 

 

 

“OH! I) 1mm 100“! I: 1000“!

 

 

 

   

 

 

 

Figure 49. Ensemble averaged results showing rrns velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B): 100Hz-
1600Hz, (C): 100Hz-1800Hz, (D): 100Hz-2000Hz at 270 CAD

U mean (cm/s)

U mean, Horizontal line at 2 cm below the valves

 

 

 

 

 

 

 

 

 

 

 

 

 

 

V mean (cm/s)

 

 

 

400
300 - o, X?" if Sig-.1
o ',. ’ 3:— , i ‘ \ a}
9143; Q— a 9 ‘ g 9 §\~\O\
5’ \$\\178_1.§\
200 3&2
Ԥ \\
\ K QB
\\‘ ‘\\
100 - i .
‘x \w R
:4 3132512 1 8&1811 n: \t‘ 4-
7 f1 \\ “ :
one wee In zto z 3
0 A———A weeBin 1:12 to 1:12 ~’\i\\:‘1'
v—--v weepgn 1 1:12 to 1§ 1:12 \\ 1
13 ~ 7 a weepln z to z \
e— —9 wee In Hz to Hz ‘3
s - . Unfor e
-100 A - - +
2 3 4 5 6 7
X-axis (cm)
V mean. Horizontal line at 2 cm below the valves
300
x---x weepin 1 Hzt02 Hz
a ., z weepgn 1 Hz to 1 2
9L 0 — — 4 weepgn 1:12 to z
A——A weepgn 1 z to 1 z
3&3‘ v—--v weepin zto1 Hz
200 spy a~c weepgn 1 zto; Hz
\ ‘\:g\ o——9 w epm 1 zto Hz
\ w. .__.. Unorce
\ G— \
8‘5“:
\ {E0
\SS‘!
100 r \ \ *\m\
s x
\\_\£‘\~\
”A \ ~r~
0 ~\.‘\ 1
‘1 ;\ 6 _
-1oo -
2 3 4 5 6 7
X-axis (cm)
Figure 50. Effect of perturbations on the U mean and V mean at horizontal

line about 2 cm below the intake and exhaust valves

76

3.8 Sweeping perturbations : sweep frequencies starting from 30Hz

The last part of 270 CAD experiment was to lower the starting value of the
sweeping frequency to 30Hz. From previous sweeping experiment, it was found that the
sweeping effects were about 20% to 25% reduction. With this in mind, it was desired to
further the investigation of the sweeping effects by lowering the sweeping frequency to
30Hz instead of 100Hz. In each case of the experiment it was done by perturbing the flow
starting at 30Hz and then swept linearly to the respective end frequency. The duration of
sweeping was still 0.05 seconds. The range of the experiment was from 200Hz to
2000Hz. The other words, the first case of the experiment was sweeping from 30Hz to

200Hz and the last case was sweeping from 30Hz to 2000Hz.

The main observation of the experiment was that the reduction effects were not as
strong as the previous 100Hz sweeping cases. The reduction was about 5% to 10% for
both U rrns and V rms. The best reduction case for U ms and V rms was sweeping from
100Hz to 400Hz case. The color contour plots are shown in Figure 51-58. U mean and U

rrns plots are shown first then followed by V mean and V rrns.

77

A B

 

30"! b 200"! ”OF

 

WM Fencing 6 mn
_.

 

iiiiiiiiiiii

 

 

 

30H: b30000: my 5 mil 30“: S400": In" 94

 

 

 

 

 

 

 

Figure 51. Ensemble averaged results showing rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
200Hz, (C):30Hz-300Hz, (D):30Hz-400Hz at 270 CAD

78

A 7 . B

 

 

WM Fm 5qu MHIDMM: mp
—. ‘

 

 

 

 

 

 

 

30K: to .00“: map 5 MI- MN: I0 1000": mop
—.

 

 

 

 

 

 

 

 

Figure 52. Ensemble averaged results showing rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
600Hz, (C):30Hz-800Hz, (D):30Hz-1000Hz at 270 CAD

79

 

 

5 m/s u'

WM PM 5 ml. :0»: h 1200“: DUI-op
_.

 

 

 

 

 

 

 

JON: lo 160001: m 5 ml. 30Hz b10110": my
_.

 

 

 

 

 

 

 

Figure 53. Ensemble averaged results showing rms velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
1200Hz, (C):30Hz-1400Hz, (D):30Hz-1600Hz at 270 CAD

80

 

WM PM 5 ml.
_.

 

 

 

 

 

 

 

 

 

30Hz I: 1000": m 5 an
_.

Eiifiiifiiiiiééiiii‘

 

 

EBKIIIHIQEaltlEBBII'lllll§g§i355:5

 

 

 

Figure 54. Ensemble averaged results showing rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
1800Hz, (C):30Hz—2000Hz at 270 CAD

8i

 

 

 

wm rm 5 w. you; to zoom can 131]

   

 

 

 

 

30”: lo 300“: mop ill: MN: h (OM: "a 5 m 852

 

 

   

 

Figure 55. Ensemble averaged results showing rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
200Hz, (C):30Hz-300Hz, (D):30Hz-400Hz at 270 CAD

82

 

WM Form.

5 ml:
_.

 

 

30Hz to ICON: men 130

 

 

 

 

30H: I: soon: van-p

5mi-
——.

 

 

 

son-mooouamp 53:: 85.2

 

 

 

Figure 56. Ensemble averaged results showing rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
600Hz, (C):30Hz-800Hz, (D):30Hz-1000Hz at 270 CAD

83

 

 

 

 

wmrm 5m

 

 

 

 

 

 

10": lo Mm: "up 6 m

 

30”: to 1800": I'D-p

 

 

 

 

 

Figure 57. Ensemble averaged results showing rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
1200Hz, (C):30Hz-1400Hz, (D):30Hz-l600Hz at 270 CAD

84

 

 

 

 

 

WM PM 5 ml. MN: h 1000”:
——.

 

 

 

 

 

 

:ouxuozooougnu-op Suv-
__.

21522221

EBHBEIIISEIIIlééiiéigsggg
5 u s. .n.» ._ _ _h__ .h

 

 

 

 

 

 

UEIIIII<528383JI:=
unm- .1.»- unns .1.

 

 

Figure 58. Ensemble averaged results showing rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):30Hz-
1800Hz, (C):30Hz-2000Hz at 270 CAD

85

 

 

Chapter 4

RESULTS AND DISCUSSIONS FOR 90 CAD AND 180 CAD

4. 1 Introduction

The purpose of this research was to investigate the cycle-to-cycle variation and
the effect of perturbation during the early intake stroke and the late intake stroke. The
crank angles chosen for this part of experiment were 90 CAD and 180 CAD. The reason
was that 90 CAD is in the middle of intake stroke process and 180 CAD is at the end of
intake stroke process. The velocity vectors at 90 CAD are very high. They are in the
range of 10 m/s to 13 m/s. The U rrns has a range of 100 cm/s to 250 cm/s and the V rrns
has a range of 120 cm/s to 350 cm/s. If compared with the 270 CAD, U ms and V ms
for 270 CAD have only a range of 50 cm/s to 150 cm/s. As for 180 CAD, the velocity has

a range of 1 m/s to 2 m/s. The U rrns and V rrns have a range of 70 cm/s to 170 cm/s.

Several of the instantaneous realizations are shown in Figure 59. All of them were
taken at 600 rpm and 90 CAD. Each of them corresponded to successive cycles at 90
CAD. From the images, it is obvious that cycle-to-cycle variation exists in the early

intake stroke.

86

 

 

 

 

 

 

 

mum wm Pouch.
nub-on n "2'."

 

 

 

 

 

 

mm Who-l Fm
.1 .- nu-uon 3: ‘°_".'"

 

mm Wm PM
Reunion M

 

 

 

 

 

1011
2000
1810
1580
1520
1380
1200
1010
080
120
580
100
210
80

-88
-218
-188
-888
-128
-888
-1818
-1288
-1388
-1828
-1888
-1818
-2888

 

Figure 59. Cycle-to-cycle variation at 90 CAD

87

 

4.2 F ixed-time perturbation for 90 CAD: forcing from 0 CAD to 90 CAD

This experiment was done by running the engine at 600 rpm and 90 CAD. The
perturbation process started when the piston reached Top Dead Center (TDC) and
continued to perturb until it reached 90 CAD. The duration of perturbation was 0.025
seconds. The range of frequencies investigated was from 50Hz to 4000112. Each of the

frequency cases had an average of 500 instantaneous realizations.

The results showed that the perturbation had some effects on the intake flow. The
U rms and V rms had a reduction of about 20%-25% in certain region of the flow. The
range of frequencies that had an effect on the flow was from 50Hz to 600Hz. It was found

that beyond 1000Hz, the effect was less. The same observation was found in 270 CAD.

The best cases results of the experiment, namely from 50Hz to 1000Hz are shown
in Figure 60—65. The best reduction case for this experiment was 400Hz for U rrns and
600Hz for V rrns. The results are organized in following manner; starting with U mean
and U rrns, then followed by V mean and V rms. In Figure 66, the line plots for U and V

mean are shown. From the plots, both U and V mean did not change much.

88

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A B
10 m/s u'
.1 wmrm ‘11" 4 son: _. 2m
15 244
2 238
25 232
3 228
>35 220
4 214
45 208
5 282
5.5 196
6 . l r A i l l A 1 A A l . r i i l l i i A l i A A A 190
2 4 x 5 6 184
118 :
112
,1 room ”—1" .1 _- zoom "—1“ 168
15; 100
2 — 154
25 - 148
3 _ 142
- m
4 138
4 5 - 124
5 _ 118
5 5 _ 112
5'irLLJLLJDHIHHiHH 1136
3 x 5 6 100
C D
Figure 60. Ensemble averaged results showing U rrns velocity (cm/s),U mean

(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 90 CAD

89

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

A B

1 IO m/s u'
.1 WMFm ‘2']. _1 am: _. 2w
1 5 244
2 238
2 5 232
3 228
>. 3 5 221]
4 214
4 5 208
5 2112
5.5 1%
.6 I I I I I I I I I I I I I I I I I I I I I I I I I I 1m
2 3 4 x 6 184

178 :
122
‘1 .- 400»: ”—1" .1 coon: ‘2" 168
V 161]
154
148
142
130
130
124
113
112
l I I L IN
4 ,K ‘ L e ‘4 100

C D
Figure 61. Ensemble averaged results showing U rrns velocity (cm/s),U mean

(cm/s), (A): without perturbation and with perturbation (B):300Hz,
(C):400Hz, (D):600Hz at 90 CAD

90

 

 

 

lOm/s u'

_i N
‘1 WMFM 1° ml- _1 .00 x

244
£35; 220
208

we

 

 

 

 

 

 

 

 

.1 __ 10m: 10m!- 166

 

 

 

 

 

 

 

 

 

Figure 62. Ensemble averaged results showing U rms velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):800Hz,
(C): 1000Hz at 90 CAD

9]

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Figure 63. Ensemble averaged results showing V rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 90 CAD

92

 

WM FM 300“:

4 .w - 350

 

 

 

 

MN]
_1 10min

 

 

   

 

 

 

 

Figure 64. Ensemble averaged results showing V rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):300Hz,
(C):400Hz, (D):600Hz at 90 CAD

93

 

 

Wm Fm

 

 

 

 

 

I .
E

 

 

Figure 65.

 

 

 

 

 

1 iiééiééééfiéééiiéEtfiziSWEEEE‘ESEFE’AHfiEEEEEEE I

 

 

Ensemble averaged results showing V rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):8OOHz,
(C): 1000Hz at 90 CAD

94

 

U mean (cm/s)

V mean (cm/s)

U mean, Horizontal line at 2 cm below the valves

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

800
+-—-+ 1 OHz
x---x 3 Hz
a— - ~I z
o - - a 4 Hz
600 A——A 3 Hz I
v- — —v z
a — ~ 43 1 HZ
0——0 5 Hz
0 a Unforced
400 >
200 -
0 .
2 3 4 5 6 7
X-axis (cm/s)
V mean, Horizontal line at 2 cm below the valves
300 1
0
—300 .
—600 -
-900 r
-1200 ‘ - : -
2 3 4 5 6 7
X-axis (cm)
Figure 66. Effect of perturbations on the U mean and V mean at horizontal

line about 2 cm below the intake and exhaust valves

95

4.3 F ixed-time perturbation for 180 CAD: forcing from 0 CAD to 180 CAD

This experiment was performed in the same condition as 90 CAD except that the
crank angle of interest was 180 CAD. The perturbation process started when the piston
reached Top Dead Center (TDC) and continued to perturb until it reached 180 CAD. The
duration of perturbation was 0.05 seconds. The range of frequencies investigated was
from 50Hz to 2000Hz. Each of the frequency case had an average of 500 instantaneous

cases.

The results did not show any obvious reduction in either U rms or V rms. All the
frequency cases in U rms showed similar U rrns values. There was no obvious best
reduction in U rrns. As for V rms, the best case would be 300 Hz. However, the reduction

was about 5%-10% in certain regions.

The results of the experiment, namely from 50Hz to 1000Hz are in Figure 67-72.
The results are organized in following manner; starting with U mean and U rrns, then
followed by V mean and V rms. From the line plots in Figure 73, the U mean and V

mean did not change much except for 200Hz which increased the mean velocity.

96

 

 

1” CAD mm RPM
WMFM

1” CAD And no III
50”:

    

 

 

 

 

 

 

 

m cm .Iucoo mu no can mum an
1m: s In:- 2m: am 114
—. V

 

 

 

 

 

 

 

Figure 67. Ensemble averaged results showing U rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 180 CAD

97

 

_.

I 5 m/s u'
noun-mm". ‘McADoMIOORPI
WMFM 5m r 100": 170

 

 

 

   

 

 

 

 

uocmmcoomn "OCADIMMRPI
mu: 5 III!- soon: 5 Inn 1 1 4
—. V ——O

 

 

 

 

 

 

 

 

 

Figure 68. Ensemble averaged results showing U rms velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):300Hz,
(C):400Hz, (D):600Hz at 180 CAD

98

 

 

 

 

IIOCAD-HOMIPI IMm-IIMIPI
WMFM coon:

    

 

 

 

 

 

 

 

lfiofi?“”°"' 5m 114

 

lll
CID
N

 

 

 

 

 

 

 

Figure 69. Ensemble averaged results showing U rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):800Hz,
(C): 1000Hz at 180 CAD

99

 

5 m/s v'

"OCADIMMIPI INCH-fin".
wmrm z

 

 

 

 

 

 

 

 

 

mew-mm»: mcmmmnrn
mm 5m , zoom 5m 1235

 

 

 

 

 

 

 

 

Figure 70.

Ensemble averaged results showing V rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):50Hz,
(C): 100Hz, (D):200Hz at 180 CAD

 

 

1|. CAD “MORP-
Witt-aim

 

 

 

 

 

‘00 CAD IMO” I'-
SOON!

 

 

 

 

no CAD Mud I00 I'-
4m:

 

 

 

 

 

 

 

 

 

Figure 71.

Ensemble averaged results showing V rms velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation (B):300Hz,

(C):400Hz, (D):600Hz at 180 CAD

lOl

A B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

"0 CAD and!“ "I 1” CAD MM IPI I S m/S VI
wmrm 5:: : mm _. 160
151.4
154.5
152.2
149.5

141
144.4
141.5
135.2
135.5

134
131.4
125.5
2 125.2
:aarwmm m. 123.6
* 121
; 115.4
2; 115.5
113.2
115.5

we
155.4
152.5
3‘: 155.2

g“: 51.5

, Ii? 95

 

 

Figure 72. Ensemble averaged results showing V rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):800Hz,
(C): 1000Hz at 180 CAD

102

U mean. Horizontal line at 2 cm below the valves

 

 

      

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

200
+—--+ é OHZ
x——-x Hz ,/
a «I: 4 H;
o — — a
100 - H 3 Hz
v——-v Hz
a . » 4: 1 Z
0, o 5 z
0 U
’47? .
E 0
3
c _
(U
0
g -100 - 1
-200 - 4
-300 ‘
1 3 5 7
X-axis (cm)
V mean, Horizontal line at 2 cm below the valves
300 . T
150 -
E
E
8
1: 0
(U
o
E
>
-150 -
-300 '
1 3 5 7
X-axis (cm)

Figure 73. Effect of perturbations on the U mean and V mean at horizontal
line about 2 cm below the intake and exhaust valves

l03

Chapter 5

RESULTS AND DISCUSSIONS FOR 270 CAD RUNNING
AT 1200 RPM

5. 1 Introduction

The purpose of this experiment was to investigate the cycle-to-cycle variation and
the effect of perturbation at 270 CAD with the engine running at 1200 rpm. In an earlier
chapter, the 270 CAD and 600 rpm experiments indicated that when perturbed with 300
Hz or 400 Hz, the U rrns and V rrns were reduced to about 25%-30%. With this
promising result, it gave a motivation to further investigate the effect of perturbation at

higher engine speed. Thus, the experiment was repeated at 1200 rpm.

At 1200 rpm, U rms ranges from 100 cm/s to 250 cm/s and V rms ranges from
120 cm/s to 230 cm/s. If compared with the 600 rpm, U rrns and V rrns for 270 CAD
have only a range of 50 cm/s to 150 cm/s. In other words, the U ms and V rms values are

doubled at 1200 rpm.

At 1200 rpm and 270 CAD, there is a significant cycle-to-cycle variation. A few

of the instantaneous, continuous realizations are shown in Figure 74.

104

 

 

 

 

 

 

 

 

 

 

Sm/s
-0.5
Sun Sun
.. _1 ..
////’.-.._.....““\‘ . //////’/”’..-“ | .
//////’a-—'--‘\\ \ \ \ 4'5 ////////"”’-‘\ ‘ -
////////’.._““‘\ _2 ////////"a".o_.‘\‘-
/////////’--‘,\,, ////////’,,--,,,-
/////////,--,,,\‘ 25////////----\\\,.
/////////,---,.., a ////////a--~IIII-
////////,,-.-,.., ’a/IIIIIII..-,,,,,,,
//////11.. .,,,. ,3_5//III//I. ,,,,,,-
III!/It\~--,,,,,, IIIIIIII\-,,,,,,-
\lllit\\---,,,,,, 4ll\\\\\\\-,,,,,--
\\\\\\\\‘----,,,, \\\\\\\\\......----~
\\\\\\s‘\\------ 4.5\\\\~.\\\\.—----~.\
\\\\\\\\\\\\s~\\ -5 \\\\\\\\s~-~ss\\
L1111.1114111L41111111111] llAlAllAlLlI1111111111411]
2 3 4 5 6 7 2 3 4 5 6 7
X X
-0.5
San 5m
.. _1 ..
"'—'-‘-'-°--"'”-'-~ \ \\ 45 ”"'""""~~\\\ \ I \ 0
’l”.‘.—o—"”’o- . \ \ \ ///”—oa—o—..‘\\ ‘ , , I
///””’/’—o-. . . , \ -2 l/II’III—o““ . , , .
////—"””-“ \ . \ \ ////////—o..‘\‘ .-- \
////’............._‘\ , , , ‘ -2.5 I’ll/I/l—o-“‘,, , ,
////”......_._“,,,. > ////////a--‘\,,,,
III/l”a......“.,.‘ -3/IIIIIIII-“",’.
I’ll/IIIIO-‘.",‘ _3.5’/II//”..,,,,,,,
"’I”ol’lo..”’. 'IIIII’ .,,,,,,,.
\111..,... ,,,,,, 411114..-,,,,-,,,,
\\\\4..\..,,,, ,, 45\\\\\\-,,,,----,,
\\\~-----,,,, -, '\\\\\..._,,,,..-~-,,
1§.\II\I\I:I:TI‘IT..ITI::1-I 7.12 '5 1.\.\2>1\2::TIZ.’.172‘ITITTITI
2 3 4 5 6 7 2 3 4 5 6 7
X X

 

 

 

 

 

Figure 74. Cycle-to-cycle variation at 270 CAD and 1200 rpm

1%

5.2 Fixed-time perturbation for 270 CAD: forcing from 0 CAD to 180 CAD

This experiment was done by running the engine at 1200 rpm and 270 CAD. The
perturbation process started when the piston reached Top Dead Center (TDC) and
continued to perturb until it reached 180 CAD. The duration of perturbation was 0.025
seconds. The range of frequencies investigated was from 50Hz to 2000Hz. Each of the

frequency cases had an average of 500 instantaneous realizations.

The results showed that the perturbation had some effects on the flow. The U rrns
and V rms had a reduction of about 20%-25% in certain region of the flow. The range of
frequencies that had an effect on the flow was from 600Hz to 2000Hz. It was found that

below 600 Hz and beyond 2000Hz, the effect was less.

The best cases results of the experiment, namely from 600Hz to 2000Hz are
shown in Figure 75-80. The best reduction cases for this experiment were 1400Hz for U
rrns and 1800Hz for V rrns. At 600 rpm, the best reduction cases were 300 Hz and 400
Hz. The results are organized in following manner; starting with U mean and U rms, then
followed by V mean and V rms. From the line plots in Figure 81, the U mean decreased
about 30% - 40% as the frequencies of the perturbations increased. V mean did not

changed much with frequencies.

106

 

wm Fm.

SIN.

 

 

 

 

 

 

100001:

218

263.2
256.1
219.6
212.8
236

229.2
222.1
215.6
288.8
282

195.2
188.1
181.6
111.8
168

161.2
151.1
111.6
118.8
131

121.2
128.1
113.6
186.8
188

 

 

Figure 75. Ensemble averaged results showing U rms velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation (B):600Hz,

(C):800Hz, (D): 1000Hz at 270 CAD and 1200 rpm

107

 

 

 

 

 

 

 

 

 

 

 

 

 

5 m/s u'
.05 "MP!!!“ 05 1200"! 210
-. ‘1" .. ‘3' 253.2
_,., 255.4
_, 249.5
'25 212.8
g ' 236
'3 229.2
‘3'5 222.1
4 215.6
4.5 288.8
'5I....I....I....I....I.... 202
2 3 4 5 6 195.2
" 159.4
1400": 1.00"! 1816
41.5 -o.5 111.8
. ‘1' .. “T 155
_,_, 151.2
_2 154.4
45 141.5
,' 145.5
‘3 131
'3-5 121.2
4 128.1
4.5 113.6
'52....1....I.142_II.4.J 11183
2 3 4 x 5 e l A 100
C D

Figure 76. Ensemble averaged results showing U rrns velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation
(B): 1200Hz, (C): 1400Hz, (D):l600Hz at 270 CAD and 1200 rpm

108

 

 

 

 

 

 

 

 

 

 

5 m/s u'
‘05 WM Fm #05 -_ MOON! 210
.. ‘1" .1 3. ‘1' 253.2
,5 5. 255.4
_, 2_ 249.5
3 242.5
>:25 _- 235
‘3 ;‘ 229.2
95 2224
4 : 215.5
.25 - 259.9
.5 f 222
2 3 4 x 5 s 195]
_ 155.4
2...... gig, 191.5
5.5 255.4 124.5
.1 ‘1" $233 159
333.2 191.2
202 111.8
133:3 142.9
131:: 124
123.2 1212
133% 122.4
1 3;: 113.5
120:4 1118.8
152:: 199
1 00

 

 

Figure 77. Ensemble averaged results showing U rms velocity (cm/s),U mean
(cm/s), (A): without perturbation and with perturbation
(B):1800Hz, (C):2000Hz at 270 CAD and 1200 rpm

109

 

 

5 m/s 9'

42.5 mm m” 42.5 _- mm 225
.1 ‘1“ .1 ‘1‘ 2212.9
5 216.6

212.1

288.2

199.8
195.6
191.1
181.2

 

 

 

118.8
111.6

 

 

 

118.1
166.2
162

151.8
163.6
119.1
115.2

136.8
132.6
128.1
121.2
128

 

 

 

 

 

 

 

 

Figure 78. Ensemble averaged results showing V rrns velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation (B):600Hz,
(C):800Hz, (D): 1000Hz at 270 CAD and 1200 rpm

 

llO

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

A B

5 m/s 14'

05 WMFM .05 ’- IIIONI 225
I ‘1" .. 3. ‘1" 229.5
,5; 216.5
_23_ 212.1
25 - 332
3 ' 199.8
35 ' 195.6
4 191.1
45 181.2

5 l 1 ._L J_. I 183
2....3425 4LU_‘5 6 118.8
‘ 114.5
noon; noon. ”0‘
-0 5 —0.5 ? 166.2

1 5:: .1 E— sav- 182
,5; 151.8
2; 153.6
253_ 119.1
,. ‘ : 116.2

3 ' 111
35 ‘ 136.8
4 132.6
45 128.1
5 I I . I . . . I . I . 1242

2 3 4 x 5 e 120

C D

Figure 79. Ensemble averaged results showing V rms velocity (cm/s),V mean

(cm/s), (A): without perturbation and with perturbation
(B): 1200Hz, (C): 1400Hz, (D):l600Hz at 270 CAD and 1200 rpm

lll

 

VIM Fm noon:

5 ml- 5m!-

 

 

 

 

 

 

zoooux 220.3

 

 

 

 

 

Figure 80. Ensemble averaged results showing V rms velocity (cm/s),V mean
(cm/s), (A): without perturbation and with perturbation
(B):1800Hz, (C):2000Hz at 270 CAD and 1200 rpm

“2

 

 

 

 

 

 

 

 

U mean. Hon'zontal line at 2 cm below the valves

 

 

 

 

 
 

 

X-axis (cm)

”3

Effect of perturbations on the U mean and V mean at horizontal

line about 2 cm below the intake and exhaust valves

 

 

 

 

750

7
zzzzzzzzw
000000HHm
00000000...
08642000n
211-1118nbU
4a .9; 5
W41. Dar“.— .6 W
__ _ . . '
+xmokwaw. w
e
.m
t m
\\\r m
c .5
\ MW\\ \xxxx ) m
‘ five; 3» m m
\ \ .\\ \ § I.“ a
x .3 xx m m
x \x \ \ «.6 u
o: \Nwfiau *\ X m
: ‘3 ~ ‘4 n
n . m
I 3% 1 i .m
_ /. ~
/ . ~ H
o a w / e» n.
_ . z: a
. Ba x? m ::
k% 7/m/ 3 V . \VM
0
flip-z \ \x\\
3 z
X
.. a a. :3 .\\\\__\
. I _ :
o a a 32+ Kg
0 0 O 0 02 W W w my W
m w w w 6 4 2 9..

3E3 some 3 3E8 cmoE >

Figure 81.

Chapter 6

CONCLUSIONS

The objective of this study was to investigate the effects of perturbing the flow in
the engine cylinder using a 4-ohms speaker. This study was conducted on a 1999 model
year Ford cylinder head with 4 valves and double overhead cams. The engine cylinder
head was part of the left bank of a V8 prototype engine with 90.2 mm bore and 900
degrees bank angle. Molecular Tagging Velocimetry (MTV) was used to measure the
velocity field at a tumble plane which is located at the center of the cylinder, about 2 cm

from each side of the cylinder wall. The main conclusions from this study are listed.

1. The amplitude effect experiment showed that the largest amplitude (28 V) of the

perturbation gave the maximum U ms and V rms reduction.

2. Free-run perturbation at 270 CAD and 600 rpm showed that the effective
perturbation frequency range is from 50Hz to 400Hz. For U rms, the best
reduction case was 400Hz. It showed about 25%-30% reduction at certain region
of the flow field. As for V rms, the best reduction case was 200Hz. It showed

about 25%-30% reduction at certain region of the flow field too.

3. Fixed-time perturbation from 0 CAD to 180 CAD, at 270 CAD and 600 rpm
showed that U ms was reduced about 20%-30% when perturbed at 300Hz and

400Hz. As for V rrns, the best reduction cases were at 400Hz and 600Hz.

ll4

The sweeping perturbation experiment (sweeping frequency started at 100Hz)
showed that the U rrns and V ms were reduced to about 20%-25%. The best
reduction case for U ms was sweeping from 100Hz to lZOOHz. The best

reduction case for V rrns was sweeping from 100Hz to 800Hz.

The rest of experiments for 270 CAD and 600 rpm did not show any clear trend
nor huge U rms and V rms reduction compared with free-run and fixed-time

perturbation experiments.

For 600 rpm experiment, with f=‘-300Hz, D=37mm and UZIOm/s, the St=l.l 1.
With f=400Hz, the St=1 .48. Thus, 1.11<St<1.48. This is consistent with
Ambrose’s finding of Strouhal number, St = 1.19 (StItD/U) as a preferred

frequency.

At the early crank angles such as 90 CAD, fixed-time perturbation showed about

20%-25% reduction for both U rrns and V rms.

When the engine speed was increased to 1200 rpm, the fixed-time perturbation

showed about 20%-25% reduction for both U rms and V rms at certain locations

in the engine cylinder.

llS

 

Chapter 7

RECOMNIENDATIONS

In future, the speaker could be placed closer to the intake valves. In this study, the
speaker was mounted about 10 inches from the intake valves. If the speaker could
be moved closed to the intake valves, it would increase the strength of the

perturbation.

Next step would be to take the best perturbation cases and try them on a firing

engine.

”6

 

APPENDICES

ll7

 

 

Appendix A

Supplementary Plots and Graphs for Chapter 3

A.1 Supplementary ggiphs for free-run perturbation

 

dwdz RMS. hon'zontal line. 1.2 cm below the valves

 

 

 

 

 

 

 

 

 

180 - - a . - C
I-—-I Z
. o——o
A—«A
\ v—--v
o——-e Hz
160* I -I nforced
7 1
140» ‘ 9 I,
‘0 \ I,
1. Q
g \ II" /
u 120 . K \I. ,’ /
t a‘\‘ f? ‘9\ I / //
‘.l ISX‘ \0\ II/ ./
\ TIE ‘ , ‘W‘ \d//
K” i \ \ \\ / l
\(l R\\ / J/
80 A A A L + k
0 2 4 6
Xaxis(cm)
Vorticity RMS. horizontal line. 1.2 cm below the valves
250
i l— I I--—-I z
1 \ 23.12: E? l
a f R v—--v
/9.. o——-e 1
‘2 / 9’ \ \ -———-I u creed
\ l' A \

 

 

 

200 v\\ / \‘1

Vorucny RMS
/
V
a.
a
a
a
y
6%
y
‘3: \o

 

 

 

. \\ / \\ a
150» \ \ /’ /’
’ h/ \ 44/
.x /
/
1WD 4 a 7 6

X axis (cm)

Figure A]. Dwdz RMS and Vorticity RMS for horizontal line, about 1.2 cm
below the valves

”8

Reynolds Stress

dwdz RMS

Reynolds Stress, horizontal line. 1.2 cm below the valves

 

 

 

 

 

 

 

 

 

 

1500 - - - f - -
I x
-1500 ' ._ B " " ..\
-..b\ .g
is. ‘&
{xi}
4500 »
.7500 - - - -
O 2 4
X axis (cm)
Figure A2. Reynolds stress for horizontal line, about 1.2 cm below the valves
dwdz RMS at cyinder centerline
1351 -- fijfiv-.-----f C
1
’ , A
120 b 2 ll’
I I] Z”

105:

 

 

 

 

 

 

 

 

P -
,.///j‘_!/ /A’ ’ ’ \\
,' ,1, 7
.0. if 425'"
:Kc’:
’ \ ' I *i
7 j on".
75> PH“;
o——e Hz
o— a nforced
m A 4 A A A AA A A A #4 +
-5 «4 3 2 1

Y axis (cm)

Figure A3.

119

 

Dwdz ms at cylinder centerline, about 4.5 cm fiom cylinder wall

Vorticity RMS

Reynolds stress

Vorticity RMS at cylinder centerline

 

 

 

 

 

 

 

 

 

 

3°° "W731" ” "
. xg’g \\‘-\
f _ 7 \ \ x
/ I" x i v '
. [H "g ‘\ -. ‘
/" “ ~.
. \
. Ex \ I
/ g -
A/fiLJ‘ \\ a‘%&
,4’ . ‘ \
\
fl—I e\ j“ *
‘.- ‘to
a
I—~-I
one
&—-—A
F--fl
o—-——e Hz
IA—a nforced 1
1
100 A A4 A A . A A A- A AA-
-5 .4 -3 2 1 o
Yaxis(cm)

Reynolds stress at cylinder centerline

 

 

 

 

 

1500 UAW-
.~--.
til-2‘ :i {r "1 .3--..n a; 4
{ax—V" “ ‘§“‘I\\ {1' z 1
”“fl \9 \J}(Q\ I—AAIUnfomed
"“a\& \ \fi
-1“. “« \O/K ‘ \\\ 4
a a fl\&\~8-,
A\\ E“ a
‘- \O ,I a
. .B’
x. V ‘8

4500 1 \\ ‘XA :——/

 

 

 

A A A A A A A A—A A A A A A A A A A k L
n
u o g

Y axis (cm)

Figure A4. Vorticity rrns and Reynolds stress at cylinder centerline, about 4.5
cm from cylinder wall

l20

A.2 Fixed-time erturbations : Perturbation from 0 CAD to 180 CAD with no hase shifi

 

dwdz RMS

Vorticity RMS

dwdz RMS. horizontal line. 31 cm below the valves

 

 

250

 

   

N

5551:1111
NNNNNNI

ii

 

 

 

 

 

X axis (em)

Vorticity RMS. horizontal line, 3.1 cm below the valves

 

 

 

 

 

 

 

 

X axis (cm)

Figure A5.
below the valves

l2]

 

 

Dwdz RMS and Vorticity RMS for horizontal line, about 3.1 cm

Reynolds stress

dwdz RMS

Reynolds stress. horizontal line, 3.1 cm below the valves

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2000 - - - A~ -
i i
2“” r ‘7
3‘}\*
9 4815*.
“A. ‘+\
K; {7&2 :\
K a .. $13k
\E E. x
q. k
45000 - '\;j_ ‘
.10000 A A A A A A A A A A A A A
2 3 4 5 6 7
xaxis (cm)
Figure A6. Reynolds stress for horizontal line, about 3.1 cm below the valves
dwdz RMS at cylinder centerline
150AA --AA---CC- fffff -A-
125 i
100 - 1
. ’ l 1
L /
, 1’ 1
75 > ’
w A A A A A A A A AA A
-6 -5 4 3 .2 .1
Y axis (cm)
Figure A7. Dwdz rrns at cylinder centerline, about 4.5 cm from cylinder wall

122

 

 

 

 

Vorticity RMS at cylinder centerline

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

300 fi - - - A . - - - -
s
“\
\
. RNA-
250 9 i 3‘3 .
a)
2
m 1
§
1
o
>
2m> {
150-AAAAAAAAAAAAAAAAA-AA
e -5 .4 -3 -2 .1
Yaxis(em)
Reynolds stress at cylinder centerline
2500~vAvvvvvvvv~~~vAv -
’ snag Hz 1
o-—e z
x--. a2 .
0» I--~I z .
l oA-e z ‘
+---+ Hz
. &——-A Hz ‘
1 v—--v Hz 1
a) 1 13---D Hz
«a , ’-.-.. Unforced 2
-2500
g
(n
m P
E l
2 .
3‘ .5000» .\
rz , :CK
. \X“:
l
I ‘\."'
-10000 A A A A A A A A A A A A
-6 -5 4 -3 -2 1
Yaxis(em)

Figure A8. Vorticity ms and Reynolds stress at cylinder centerline, about 4.5 cm
from cylinder wall

123

A.3 Sweepingperturbations : sweep frequencies from 100Hz to 2000Hz

U nns. horizontal like, 2.75 cm below the valves

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1752*- .v---.f--,----A-A-
150' 4
E 125*
E
g .
g l
D 100.
75+
l
l
50
2 7
Xaxis(cm)
V rrns. horizontal line. 2.75 cm below the valves
130----.---.--
-"--
l k. ‘
1 Ii. 11!.
110-
o—_oA‘9 &\
13‘
E
3
E 90
>
+--—-+ J
70- x---x
I-A-l
» .--.
b—A
v—--v
o——o
'..-..
w A A A; A A A A AA
2 3 4 5 6 7

X axis (cm)

Figure A9. U RMS and V RMS for horizontal line, about 2.75 cm below the
valves

Vorticity RMS

dwdzrms

250

150'

100

170

150

13):

110

Vorticity RMS. horizonld line. 2.75 cm below the valves

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

+---+ sweep Hz- n

x--« sweep z- 11,,» r

I ~£ sweep z-‘ III

o---o sweeo z-‘ 11,11: <

A——-——A sweeo z-l l1~l

v—--v sweelo 2- {MHz J

OA—e a z z

e- a Emlorced .
3 4 5 6 7

X axis (cm)

dwdz rrns. horizontal line, 2.75 cm below the valves

 

 

 

 

 

 

 

 

 

 

 

+---r» sweep1 z- t
x---x swee z-‘ 11,11,» 2
' i SW96 " llhllnl
I on". swee z-‘ 1,11,)
A—A—A swee -‘ ’lnllill-lz
1 v—--v swee z-‘ “NJ-12 1
0- —o w z-tl .1le

I — I n 1

 

 

 

   

 

 

 

X axis (cm)

Figure A10. Vorticity RMS and dwdz RMS for horizontal line, about 2.75 cm
below the valves

l25

Reynolds stress

U rrns (cm/s)

Reynolds stress. horizontal line, 2.75 cm below the valves

 

 

 

 

 

 

 

 

 

1500 e-
1
. . A”? 4
r.:*:‘£t§§
,fifl'oly ,9.- e
-1500 ~ .1251" '
. I ’.,’ ’ .
. I..- 65%.
GA . \ :7 ,
, X-\x“I ekxflf‘} fir} +---+ sweep 100Hz-2000Hz
1 F‘F’K‘x 3 J" ’1 x----x sweep 100Hz-1800Hz
‘ ‘1‘ 3 eff / a “a sweep100Hz-1600Hz
4500 ~ ‘1 1A 233,37 9---. sweep 100Hz-1400Hz
\A‘ ‘E a,’ I I A._,A sweep 100Hz-1200Hz
h ,xl /. ,I‘ VA-—v sweep1OOHz-1000Hz
-_' ‘V e—e sweep 100Hz-800Hz
L‘IAAI I—~ IUnforced
-7500 I I I I I I I I I I I I I I I I I I I I I I
2 3 4 5 6 7
Xaxis (cm)
Figure A] l. Reynolds stress for horizontal line, about 2.75 cm below the valves
U rrns, vertical line, at cylinder centerline
180 .....-..j,..e.....fi.....
r I I
1.,
160 " ,/ .I

140~ -' 247’” QM ~
y .

 

 

 

 

 

 

 

120 ‘ ,l / ' . + - - - + sweep 100Hz-2000Hz ‘
, / ’1 ,a x----x sweep 100Hz-1800Hz
'1‘ X) a - — - a sweep 100Hz-1600Hz 4
'x. e-"e sweep 100Hz-1400Hz
100 _ x’wj ctr—— A sweep 100Hz-1200Hz .
4/ ,j/ v— -—v sweep 1001124000112
g/A/ @- — -e sweep 100Hz-800Hz
/ I ~ — — I Untamed
80 I I I I I I I A A I A A A A A A ALA A I I I I
-6 -5 -4 -3 -2 -1
Y axis (cm)

Figure A12. U ms at cylinder centerline, about 4.5 cm from cylinder wall

126

 

Vrrns.vertieellhe.atcyllndereenterlilne

 

 

 

 

 

 

120 A
+---+ 1
:---x
-‘ a
O---. . ‘l I
I
3?": '
a”! f‘q ‘
b .' "
«0° 167

 

 

 

 

 

 

 

 

 

 

 

 

 

’13
E ’ %/
a I - ' y I /
E " ,ué/e/
> I :Affl/
Ix--*‘: "'r?
806 ,I *4 c‘ ’ /A
2 I, ” I’iF :ar’
m *4 A A A
-6 4 -3 -2 1
Yaxis(em)
Vorticity rrns. vertical line. atcyllnder curtarllne
300»
E 250»
i
>
P
2001 1
l 1
150 * * ‘ ‘ ‘ ‘ * ‘ * ‘ *
-6 -5 -4 -3 -2 -1

Y axis (cm)

Figure A13. V rms and vorticity rrns at cylinder centerline, about 4.5 cm from
cylinder wall

127

 

 

 

 

 

vif r v

 

4.33
. . _ _
. . .

.
+¥lo

A

 

dwdz rrns. vertical line. at cylinder centerline

 

 

 

125 r

75’

 

-1
-1

 

.2
-2

Vfi ff fit

128

 

 

 

 

-3

Yaxis (cm)
Yaxis (cm)
Dwdz ms and Reynolds stress at cylinder centerline, about 4.5 cm
from cylinder wall

Reynoldsstress.verticalline.stcyllnder0enterline

-5

 

 

 

 

Figure A14.

 

2W

Appendix B

Experimental Equipment and Devices

 

Figure Bl. Engine from different point of view

129

 

Figure 82. Delay generators, digital multimeter and monitor

 

Figure B3. Gateway E-5200 Pentium III computer

130

 

Figure B4. Kenwood KFC-1077 4-ohm speaker

 

Figure B5. Hafler P-1000 amplifier

13]

. . .. a:':; n " “.05.-

 

Figure B6. HP function generator

 

Figure B7. Biacetyl seeding chamber and acetone seeding chamber

I32

lll‘lllllHll I'M] l I. ‘,

 

Figure B8. Beam blocker

 

Figure B9. Variable-focal-length lens (VFL)

I33

 

Figure BIO. Engine encoder ‘

134

 

Figure B11. Nitrogen bank

135

Velocity (cm/s)

Appendix C

Zero-delayed or reference realizations investigation

U mean and V mean along horizontal line at the center of the cylinder

 

 

 

 

 

 

 

 

2 - - - f 1 - v . - , - - - - T - ,
i o— —e U cm/s l
L :3 ~- — a V cm/s 1
1 - r
a {a}
- « ’ a _ E
B \ .. ' ’ /G\ ,0 l
0 / )8—A _ —9 / \ \ \ \ / / 'l
i 2/ _ \3 7 “ \c/
l / h - 3
/ ~ \ x
l g/ E
-1 > J
_2 A A r A x 4
3 4 5 6 7
X Axis (cm)
Figure C1. Zero-delay: U mean and V mean along horizontal line at the center
of the cylinder

I36

 

U mean and V mean along vertical line at cylinder centerline

 

 

 

 

 

 

 

 

2 r . T - . . - - . Y . V f- . a , , - . - -1 ,
l
o——o U cm/s .
BUG V cm/s J
1 >
Q
A \
Q \ [Cl-s
E \\ I. [3 a
3 P [33 -+1 EJ/ \\ EY/ \ [I l
0 i B r I
g / \e /e»- ”*3\ /%\ 9>\’O/ r:
d) \ /
> v E‘ l
-1 . 4
_2 - x - - . - - A A . a . . - i - A . - i . - . -
-6 -5 -4 -3 -2 -1
YAxis(cm)

Figure C2. Zero-delay: U mean and V mean along vertical line at the center of
the cylinder

The purpose of this experiment is to investigate the stability of the zero-delayed

average and instantaneous realizations. The results indicated that both U and V mean

were very close to zero velocity.

137

 

BIBLIOGRAPHY

I38

 

10.

11.

Bibliography

Hascher, H. G., Jaffri, K., Noval, M., Lee, K., Schock, H., Bonne, M., Keller P.,
“An Evaluation of Turbulent Kinetic Energy for the In-cylinder Flow of a F our-
Valve 3.5L SI Engine using 3-D LDV Measurements.” SAE 970793, 1997

Zhang, L., Ueda, T., Takatsuki, T., Yokota, K., “Cycle-to-cycle Variations of In-
cylinder Flow in a Motored Engine.” JSME International Journal, Series B.
Vol.38. No.3. 1995

Ambrose, G., “A Study of the Effects of Exciting the Intake Flow in an Internal
Combustion Engine Model.” Master’s thesis, Michigan State University, 1997

Tang, S. K. and Ko, N. W. M., “Experimental Investigation of the Structure
Interaction in an Excited Coaxial Jet.” Experimental Thermal and Fluid Sciences,
Vol 8, pp. 214-229, 1994

Dai, W., Trigui, N., Lu, Y., “Modeling of Cyclic Variations in Spark-ignition
Engines.” SAE International, 2000

Koochesfahani, M. M. and Stier, B., “Molecular Tagging Velocimetry in Gas
Phase and its Application to Jet Flows.” ASME Fluids Engineering Division
Summer Meeting, FEDSM’97, 1997

Zaman, K. M. B. Q. and Hussain, A. K. M. F., “Vortex Pairing in a circular jet
under Controlled Excitation. Part 1. General Jet Response.” Journal of Fluid
Mechanics, Vol. 101, part 3, pp. 449-441, 1980

Zaman, K. M. B. Q. and Hussain, A. K. M. F., “The ‘Preferred Mode’ of the
Axisymmetric jet.” Journal of Fluid Mechanics, Vol. 110, pp. 39-71, 1981

Koochesfahani, M. M. and MacKinnon, C. G., “Influence of Forcing on the
Composition of Mixed Fluid in a two-stream Shear Layer.” Physics of Fluids A,
Vol. 3, pp. 1135-1142, 1991

Cohn, R. K., Gendrich, C. P., Mackinnon, C. G., and Koochesfahani, M. M.,
“Crossflow Velocimetry Measurements in 3 Wake Flow.” Bulletin of the
American Physical Society, p. 1962, 1995

Cohn, R. K. and Koochesfahani, M. M., “Structure of the Velocity and Vorticity

Field in a Confined Wake.” Bulletin of the American Physical Society, p. 2197,
1997

139

 

12.

13.

14.

15.

16.

l7.

l8.

19.

20.

21.

Gendrich, C. P., Bohl, D. G. and Koochesfahani, M. M., “Whole-Field
Measurements of Unsteady Separation in a Vortex Ring/Wall Interaction.” AIAA
Paper #97-1780, 1997

Gendrich, C. P. and Koochesfahani, M. M., “A Spatial Correlation Technique for
Estimating Velocity Fields Using Molecular Tagging Velocimetry.” Experiments
in Fluids, Vol. 22, pp. 67-77, 1996

Gendrich, C. P., Koochesfahani, M. M., and Nocera, D. G., “Molecular Tagging
Velocimetry and Other Novel Applications of a New Phosphorescent
Suprarnolecule.” Experiments in Fluids, Vol. 23, pp. 361-372, 1997

Hill, R. B. and Klewicki, J. C., “Data Reduction Methods for Flow Tagging
Velocity Measurements.” Experiments in Fluids, Vol. 20, pp. 142-152, 1991

Cohn, R. K., “Effect of Forcing on the Vorticity Field in a Confined Wake.” PhD.
Thesis, Michigan State University, 1999

Heywood, J. B., “Internal Combustion Engine Fundamentals.” McGraw-Hill,
New York, 1998

Hascher, H., “The Influence of Speed, Stroke and Load Variations on the Major
In-Cylinder Flow Patterns of a Four-Valve Spark-Ignition Engine.” PhD. Thesis,
Michigan State University, 1999

Stier, B. and Koochesfahani, M. M., “Whole Field MTV Measurement in a
Steady Flow Rig Model of an IC Engine.” Presented at the 1998 SAE
International Congress & Exposition, Detroit, MI, February 23-26, 1998, SAE
Paper 98-0481, 1998

Koochesfahani, M. M., “Molecular Tagging Velocimetry (MTV): Progress and
Application.” AIAA Paper no. AIAA-99-3786 (Invited), 1999

Stier, B. and Koochesfahani, M. M., “Molecular Tagging Velocimetry (MTV)

measurements in Gas Phase Flows.” Experiments in Fluids, 26(4), pp. 297-304,
1999

140

 

IIIIIIIIIIIIIIIIIIIII

llllljl11111111311!llljllllllflllfllll