. ”m.-.
. . 2&9.
..3. firm:

t. F1!!!
. .

mg. mm _.

23..

.
3 .I.
a .

1. ‘ 3.3.: T.

(1.3): 1 .‘rx
‘\

awn»...
. .

I.

.939..- .I .
A) S.‘ . i; .
s at; -
9.3m.
.

my“

a .
IKE... H...nw1m(u.... _.~
: v . t, a. ..
$3.» .- .. a

a. . .
. a
in“? , 1.9.! g
. . n
l...

v. ‘7
s... [I wotii‘.
.9;‘€.§1.l 1..

3..

 

H..£..,.a.. . . . .._....

 

 

UNIVERSTY LIB

2 IIIIIIIIIIIIIIIIIIIIIIIIIII IIIIIIIIIIIIIIIIII

31293 01688 0449

 

 

 

 

 

 

 

 

This is to certify that the
thesis entitled

EXPERIMENTAL MEASUREMENTS OF
THE FLOW FIELD INSIDE OF AN
AUTOMOTIVE TORQUE CONVERTER

presented by

Lawrence Joseph Dalimonte

has been accepted towards fulfillment
of the requirements for

 

 

M.S. degree in Mechanical Engineering

Mew

Major professor

Date {jg/73

 

0.7639 MS U is an Affirmative Action/Equal Opportunity Institution

 

 

LIBRARY

Michlgan State
Unlverslty .

 

 

E

 

PLACE IN RETURN BOX
to remove this checkout from your record.
TO AVOID FINES return on or before date due.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ma WM“

 

 

EXPERIMENTAL MEASUREMENTS OF
THE FLOW FIELD INSIDE OF AN
AUTOMOTIVE TORQUE CONVERTER
By

Lawrence Joseph Dalimonte

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
MASTER OF SCIENCE

Department of Mechanical Engineering

1998

ABSTRACT
EXPERIMENTAL MEASUREMENTS OF THE FLOW FIELD INSIDE
OF AN AUTOMOTIVE TORQUE CONVERTER USING LASER
DOPPLER VELOCIMETRY
By

Lawrence Joseph Dalimonte

The automotive torque converter flow fields are complex, in that
they contain three-dimensional flows, which are viscous and unsteady. To
add to the complexity, the difference in rotor speeds between the impeller
and turbine compound the flow effects. The purpose of this study was to
characterize the internal flow experimentally in a production torque
converter from Ford Motor Company.

The flow characteristics were quantified by Laser Doppler
Velocimetry (LDV) measurements. The LDV system was employed for
this study because understanding the fluid flow is crucial for
characterizing the flows at the measured points within the torque
converter. LDV measurements were conducted at a total of 9 planes at
two different speed ratios. Flows at the impeller exit and gap region were
affected by the turbine blade as it passed. Fluid flow was also found to be
leaving the impeller and flowing up behind the turbine, negatively
effecting converter efficiency. Velocity profile animations were created
using the measured velocity data and custom software to allow Viewing on

a unix workstation.

ACKNOWLEDGEMENTS

I would like to take this Opportunity to express my Sincere
appreciation to those of whom made this study possible. First and
foremost, tO my wife Kim, and my children Lawrence 11 and Leann, for
their patience and understanding during my academic career. Also to my
parents Al and Fran Dalimonte, for their support and encouragement that
continues to this day. TO Tom Stuecken on his expertise in designing and
building of experimental equipment; Dr. Keunchul Lee for his guidance
with the LDV system and his knowledge of research in general; to Matt
Foster for his help with the data management and processing, among other
things; tO Mark Novak for his work with Explorer making the animation
Of the velocity profiles possible; to Dr. Giles Brereton, Jon Darrow,
Mahmood Rahi, Mikhail Ejakov, and Hans Hascher for their knowledge
and advice on numerous other matters; and finally, to Dr. Harold Schock

for making this opportunity not only possible, but a reality.

iii

TABLE OF CONTENTS

LIST OF TABLES ........................................................................... vii
LIST OF FIGURES ......................................................................... viii
LIST OF SYMBOLS ........................................................................ xii
CHAPTER 1
INTRODUCTION .............................................................................. 1
1.1 Motivation .......................................................................... 1
1.2 Torque Converter Fundamentals ............................................ 2
1.3 Problem Statement ............................................................... 4
1.4 Literature Review ................................................................ 4
CHAPTER 2
EXPERIMENTAL EQUIPMENT .......................................................... 9
2.1 Test Facility: Test Rig ......................................................... 9
2.2 Transmission Oil Supply System ......................................... 10
2.3 Test Rig Controller ............................................................ 12
2.4 LDV System ..................................................................... 12
CHAPTER 3
TORQUE CONVERTER ................................................................... 18
3.1 Modification ..................................................................... 18
3.2 Front Window ................................................................... 20
3.3 Side Window ..................................................................... 21
CHAPTER 4 '
EXPERIMENTAL PROCEDURE ....................................................... 23
4.1 Preliminary Testing ........................................................... 23
4.2 Setup ............................................................................... 23
4.3 Data Acquisition ............................................................... 26
4.4 Measurements with Front Window: First Series ..................... 32
4.5 Measurements with Front Window: Second Series ................. 33
4.6 Measurements with Side Window: Third Series ..................... 34
4.7 Side Window Geometry ...................................................... 36
4.8 Data Post Processing .......................................................... 39
4.9 Animation ......................................................................... 45
CHAPTER 5
RESULTS AND DISCUSSION .......................................................... 47
5.1 Measurements with Front Window: First Series (Planes 1-9) .. 47
5.1.1 Plane 1: 0.8 Speed Ratio ......................................... 48

iv

TABLE OF CONTENTS (CONT .)

5.1. 2 Plane 1: 0.4 Speed Ratio ......................................... 48
5.1.3 Plane 1: Comparison of Both Speed Ratios ................ 49
5.1.4 Plane 2: 0.8 Speed Ratio ......................................... 50
5.1. 5 Plane 2: 0.4 Speed Ratio ......................................... 50
5.1.6 Plane 2: Comparison of Both Speed Ratios ................ 51
5.1.7 Plane 3: 0.8 Speed Ratio ......................................... 52
5.1.8 Plane 3: 0.4 Speed Ratio ......................................... 52
5. 1.9 Plane 3: Comparison of Both Speed Ratios ................ 53
5. 2 Measured with Front Window: Second Series ........................ 53
5. 2.1 Plane 4: 0.8 Speed Ratio ......................................... 53
5.2.2 Plane 4: 0.4 Speed Ratio ......................................... 54
5. 2. 3 Plane 4: Comparison of Both Speed Ratios ................ 54
5. 2. 4 Plane 5: 0.8 Speed Ratio ......................................... 55
5.2. 5 Plane 5: 0.4 Speed Ratio ......................................... 56
5. 2. 6 Plane 5: Comparison of Both Speed Ratios ................ 57
5. 3 Measurements with Side Window: Third Series (Planes 6- 9)...
5.3.1 Plane 6: O. 8 Speed Ratio ......................................... 68
5. 3. 2 Plane 6: 0.4 Speed Ratio ......................................... 68
5. 3. 3 Plane 6: Comparison of Both Speed Ratios ................ 68
5.3.4 Plane 7: 0.8 Speed Ratio ......................................... 69
5.3.5 Plane 7: 0.4 Speed Ratio ......................................... 69
5. 3. 6 Plane 7: Comparison of Both Speed Ratios ................ 70
5.3. 7 Plane 8: 0.8 Speed Ratio ......................................... 70
5.3. 8 Plane 8: 0.4 Speed Ratio ......................................... 71
5.3.9 Plane 8: Comparison of Both Speed Ratios ................ 71
5.3.10 Plane 9: 0.8 Speed Ratio ....................................... 72
5.3.11 Plane 9: 0.4 Speed Ratio ....................................... 73
5.3.12 Plane 9: Comparison of Both Speed Ratios .............. 73
5.4 Data Comparison ............................................................... 83
5.4.1 Front Window: Planes 2 and 4 at Both Speed Ratios .. 83
5.5 Steady State Condition Velocity Analysis ............................. 88
5. 5.1 Plane 1: Comparison of Both Speed Ratios ................ 89
5.5.2 Plane 2: Comparison of Both Speed Ratios ................ 90
5. 5. 3 Plane 3: Comparison of Both Speed Ratios ................ 91
5.5.4 Plane 4: Comparison Of Both Speed Ratios ................ 91
5.5.5 Plane 5: Comparison of Both Speed Ratios ................ 92
5. 5. 6 Plane 6: Comparison Of Both Speed Ratios ................ 93
5. 5. 7 Plane 7: Comparison Of Both Speed Ratios ................ 93
5. 5. 8 Plane 8: Comparison of Both Speed Ratios ................ 94
5.5.9 Plane 9: Comparison Of Both Speed Ratios ................ 95
5.6 Mass Flow Rates ............................................................. 114

TABLE OF CONTENTS (CONT .)

5 6 1 Mass Flow Rates: Plane 1 ..................................... 115
5.6.2 Mass Flow Rates: Plane 6 ..................................... 115
5.6.3 Mass Flow Rates: Plane 7 ..................................... 116
5 6 4 Mass Flow Rates: Plane 8 ..................................... 116
5 6 5 Mass Flow Rates: Plane 9 ..................................... 117
CHAPTER 6
SUMMARY AND CONCLUSIONS .................................................. 125
CHAPTER 7
RECOMMENDATIONS .................................................................. 128
LIST OF REFERENCES ................................................................. 130

vi

LIST OF TABLES

Table 1. Ford production torque converter specifications ..................... 19
Table 2. Measured plane locations in Ford coordinate system ............... 29

Table 3. Test conditions for data measured ......................................... 31

Vii

LIST OF FIGURES

Figure 1. Torque converter assembly .................................................... 3
Figure 2. View Of Test Rig ............................................................... 10
Figure 3. Transmission Oil Supply System .......................................... 11
Figure 4. VI screen display ............................................................... 13
Figure 5. Test Rig and Test Rig Controller ......................................... 14

Figure 6. View of 4-beam probe of LDV system and test rig with

front window converter ...................................................... 14
Figure 7. 4-beam probe with fringe pattern ......................................... 15
Figure 8. View Of impeller half Of converter with front window

installed ........................................................................... 19
Figure 9. View of front window ......................................................... 20
Figure 10. View of impeller half Of converter with side window

installed .......................................................................... 22
Figure 11. View Of test rig apparatus for traverse table reference

point .............................................................................. 25
Figure 12. View of Impeller and turbine positions for encoder

references ....................................................................... 26
Figure 13. Impeller and turbine blades inside the torque converter

with measured planes (Ford coordinate system) ................... 28
Figure 14. View of measured vector components in both front and

side window series .......................................................... 30

viii

Figure 15.
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.
Figure 29.
Figure 30.

Figure 31.

Figure 32.

Figure 33.

LIST OF FIGURES (CONT.)

First Series with 3 planes .................................................. 33
Second Series with 2 planes .............................................. 34
Third Series with 4 planes ................................................. 35
View of Side window ........................................................ 36
View of beam focal point through the side window .............. 38
3-D display of side window (Demmer) ................................ 38
Data acquisition flow chart ............................................... 40
Data processing flow chart ................................................ 41
Plane 1 (mid-chord of impeller) at 0.8 speed ratio ................ 58
Plane 1 (mid-chord of impeller) at 0.4 speed ratio ................ 59
Plane 2 (impeller exit) at 0.8 speed ratio ............................. 60
Plane 2 (impeller exit) at 0.4 speed ratio ............................. 61
Plane 3 (1mm inside turbine) at 0.8 speed ratio ................... 62
Plane 3 (1mm inside turbine) at 0.4 speed ratio ................... 63
Plane 4 (impeller exit) at 0.8 speed ratio ............................. 64
Plane 4 (impeller exit) at 0.4 Speed ratio ............................. 65
Plane 5 (1mm from impeller exit in gap region)

at 0.8 speed ratio ............................................................. 66
Plane 5 (1mm from impeller exit in gap region)
at 0.4 speed ratio ............................................................. 67

Plane 6 (2mm inside impeller) at 0.8 speed ratio .................. 75

ix

Figure 34.
Figure 35.
Figure 36.
Figure 37.
Figure 38.

Figure 39.

Figure 40.

Figure 41.

Figure 42.

Figure 43.
Figure 44.
Figure 45.
Figure 46.
Figure 47.
Figure 48.
Figure 49.

Figure 50.

LIST OF FIGURES (CONT.)

Plane 6 (2mm inside impeller) at 0.4 speed ratio .................. 76
Plane 7 (2mm inside impeller) at 0.8 speed ratio .................. 77
Plane 7 (2mm inside impeller) at 0.4 speed ratio .................. 78
Plane 8 (impeller exit) at 0.8 speed ratio ............................. 79
Plane 8 (impeller exit) at 0.4 Speed ratio ............................. 80
Plane 9 (1mm from impeller exit in gap region)

at 0.8 speed ratio ............................................................. 81
Plane 9 (1mm from impeller exit in gap region)

at 0.4 Speed ratio .............................................................. 82
Comparison of planes 2 and 4 (both located at impeller exit)

at 0.8 speed ratio ............................................................. 85
Comparison Of planes 2 and 4 (both located at impeller exit)

at 0.4 speed ratio ............................................................. 86
Comparison Of planes 1 and 4 at 0.4 speed ratio .................. 87
Steady state condition velocities plane I at 0.8 Speed ratio 96
Steady state condition velocities plane 1 at 0.4 Speed ratio 97
Steady state condition velocities plane 2 at 0.8 speed ratio 98
Steady state condition velocities plane 2 at 0.4 speed ratio 99
Steady state condition velocities plane 3 at 0.8 speed ratio . 100
Steady state condition velocities plane 3 at 0.4 speed ratio . 101
Steady state condition velocities plane 4 at 0.8 Speed ratio. 102

Figure 51.
Figure 52.
Figure 53.
Figure 54.
Figure 55.
Figure 56.
Figure 57.
Figure 58.
Figure 59.
Figure 60.
Figure 61.
Figure 62.
Figure 63.
Figure 64.
Figure 65.
Figure 66.

Figure 67.

LIST OF FIGURES (CONT.)

Steady state condition velocities plane 4 at 0.4 speed ratio .
Steady state condition velocities plane 5 at 0.8 speed ratio .
Steady state condition velocities plane 5 at 0.4 speed ratio .
Steady state condition velocities plane 6 at 0.8 Speed ratio .
Steady state condition velocities plane 6 at 0.4 speed ratio .
Steady state condition velocities plane 7 at 0.8 speed ratio .
Steady state condition velocities plane 7 at 0.4 speed ratio .
Steady state condition velocities plane 8 at 0.8 speed ratio .
Steady state condition velocities plane 8 at 0.4 speed ratio .
Steady state condition velocities plane 9 at 0.8 speed ratio .
Steady state condition velocities plane 9 at 0.4 speed ratio .
Mass flow rate plane 1 at both speed ratios .......................
Mass flow rate plane 6 at both speed ratios .......................
Mass flow rate plane 7 at both speed ratios .......................
Mass flow rate plane 8 at both Speed ratios .......................
Mass flow rate plane 9 at both speed ratios

Mass flow rate comparison planes 1, 6-9 both speed ratios..

xi

103

104

105

106

107

108

109

110

111

112

113

119

120

121

122

123

124

LIST OF SYMBOLS

Fringe spacing

Focal length of transmitting lens

Doppler Shift frequency Of the scattered light

Time taken for the particle to cross a pair Of fringes
Beam half angle

Wavelength of the laser beam

xii

CHAPTER 1

INTRODUCTION

1.1 Motivation

The purpose of this study was to characterize the internal flow
experimentally in a production torque converter from Ford Motor
Company under known macroscopic Operating conditions. Torque
converter flow fields are complex, in that they contain three-dimensional
flows, are viscous, and unsteady. To add to the complexity, the
difference in rotor speeds between the impeller and turbine compound the
flow effects. Few turbomachines, other than the hydraulic torque
converter, Operate with such a complex interface.

Two production torque converters of the same type were modified
for this study. One converter was modified with a front window, and the
other a side window. Both windows were installed in a housing mounted
on the converter shell, which allowed Optical access into the converter
passages. The front window allows Optical access into the impeller
passage, between the impeller and turbine (gap region) and turbine. The
side window allows Optical access into the exit region of the impeller

passage, gap region, and turbine from a different perspective.

The flow characteristics were quantified by Laser Doppler
Velocimetry (LDV) measurements. The LDV system was employed for
this study because a detailed understanding of the fluid flow is necessary
for improved design of the torque converter. Also, the LDV system is a
non-intrusive form Of measuring with no calibration for temperature,
density, etc., necessary. LDV measurements were conducted at a total of
9 planes. The measurements were conducted under two operating
conditions, which were speed ratios of 0.4 and 0.8. The 0.4 speed ratio
has an impeller speed of 1600 rpm and a turbine speed of 633 rpm, while
the impeller speed is 2000 rpm and the turbine speed is 1600 rpm for the
0.8 speed ratio.

Velocity profile animations were created using the measured
velocity data on a Silicon Graphics (SGI) computer. These animations,
which Show the ensemble averaged velocity profiles, compare data in
plane to plane format. The overall efficiency of the torque converter may
be improved by Optimizing flow fields measured using internal flow

diagnostics.

1.2 Torque Converter Fundamentals

The torque converter has three important functions. The first is that

it must transfer power smoothly and efficiently from the engine to the

transmission. Second, the torque converter must have the ability to

multiply engine torque for improved vehicle launching at low speeds.
Third, it must be able to effectively dampen engine torsional vibrations
and Shock loads. The automotive torque converter is made up of several
components (Figure 1). The impeller, also known as a pump, is attached
directly to the cover. The cover, in turn, is connected to a flexplate,
which is attached to the engine, and rotates at engine speed. When the
converter is rotating fluid exits the impeller and impacts the turbine
causing it to rotate. The turbine is driven by the impeller and rotates on
its own axis driving the transmission input shaft. The stator is positioned
between the impeller and turbine and redirects the fluid flow from the

turbine to the impeller.

 

   
    

Du.

‘ J11. 0.. l. ,l_,0__‘_| H

 

Oil Pump
Drive Hub

Figure 1. Torque converter assembly

1.3 Problem Statement

The Objectives of this study centers on design technology
improvements. Specific Objectives include improving performance of the
torque converter, reducing the length of the torque converter for
automobile use with transverse engines, and reducing fuel consumption in
urban use. To be able to achieve these Objectives a thorough
understanding of the fluid flows dynamic phenomena is essential. The use
of LDV measurements make it possible to examine the internal flow

characteristics of the torque converter.

1.4 Literature Review

The LDV system Offers a non-intrusive form of measurement for
flow characterization, thereby having provided data in numerous studies
[1-4]. By and Mahoney (1988) summarized the technology needs for the
automotive torque converter [5]. They suggested that the flow is highly
three-dimensional due, in part, to complex geometries with three
components rotating at different speeds. This makes the understanding Of
general fluid flow characteristics difficult. They conclude that internal
flow analysis can be used to determine the physical flow phenomena

associated with flow losses inside the converter.

Fister and Adrian (1983) were the first to measure the flow field in
a torque converter, even though the converters were larger industrial units
[6]. They used Laser-Two-Focus (L2F) and Spark Tracer methods in two
geometrically identical and hydrodynamically similar torque converters.
The L2F method used Sight-holes and can only be applied to transparent
fluids. The Spark Tracer method can only be applied in gases, requiring
transparent and non-conductive walls. Therefore one converter was used
with water aS the operating fluid and second converter used air. The
results are somewhat limited in comparison since not only are they not
true automotive torque converters, the operating fluids used were not
transmission oil.

By and Lakshminarayana (1995) measured the parameters of a
torque converter pump and the static pressure field using eight static
pressure probes [7]. The importance was stressed to measure static
pressures under constant turbine/impeller speed ratio conditions, avoiding
transient conditions. They found that the static pressure distribution
generally poor and fluid separation at the core section under all speed
ratio conditions.

Lakshminarayana and von Backstrom (1996) discuss different
techniques in investigating the fluid flows inside Of a torque converter
[8]. These include laser velocimetry, hot wire velocimetry, fast response
probes, conventional five-hole probes, as well as blade and wall static

pressure measurements. They conclude that torque converter flows are

extremely complex and that methods developed need to be able to measure
regions of flow separation and secondary flow circulation.

Watanabe et a1. (1997) used a flow Visualization and measurement
Of torque converter stator blades with laser sheet lighting methods and
LDV [9]. The apparatus included a shaft with an open Space housing a
window and mirror SO fluid flow in the stator could be directly Observed.
This study also involved varying the thickness of the stator blades. This
was a proposed new experimental apparatus to visualize and measure the
flow inside of the torque converter; Since no known published case of
applying a flow visualization or measurement technique with an actual
torque converter was available. They concluded that the thickness of the
stator blade does affect the torque capacity coefficient characteristics
more than the efficiency and torque ratio characteristics.

Bahr et a1. (1990) used a torque converter constructed entirely from
Plexiglas and measured flow with an LDV system [10]. The data
measured included the effect Of instantaneous turbine and blade positions
on the average velocity fields through the use of shaft encoders. This was
an attempt to map flow characteristics in torque converter Operating
conditions. The constraint was that the rotor speeds were reduced due to
the nature Of the material used. They measured at two different speed
ratios and found that the instantaneous turbine and impeller blades had a
strong effect on the average velocity field in the stator. They also

concluded that the torque is very poorly distributed at the upper half of

the blade. Therefore the blade should be redesigned for a more uniform
flow and less flow separation.

There have been numerous studies on the Operation Of torque
converters from CFD, pitot tube analysis to LDV measurements [ll-20].
These were designed to understand fluid flow characteristics and improve
torque converter performance. The shortcoming was none of the studies
were performed in an actual automotive torque converter in realistic
Operating conditions.

McCarrick (1993) measured fluid flow fields in an actual
automotive torque converter provided by Ford Motor Company [21]. The
main emphasis of this study was to determine the feasibility and
limitations of this method in mapping flow characteristics. Concentration
began on the apparatus, which would both support and operate the
converter during testing. It was necessary to make comparisons with
numerical simulation and conduct testing under certain operating
conditions that included the correct turbine/impeller ratios, as well as,
proper flow rates and fluid temperatures. For maximum optical access
within the converter the choice of material for the window was important.
The refractive index Of both the window material and Oil used needed to
be as close as possible. Development Of an appropriate seeding method
was also addressed in McCarrick’s work. Reflectivity was as major an
issue as were size and weight. Tests were conducted at 1000 and 1500

rpm at four different speed ratios. Data was recorded and processed. An

animation Of the data helped interpret the results. It was found that the
instantaneous position Of the turbine had little effect on fluid flow inside
the impeller passage. The turbine did, however, have noticeable effects
on the flow in the gap region. It was also recommended that the LDV
system be upgraded to be able to measure all three velocity components
simultaneously. This would better provide information necessary to
understand actual three-dimensional fluid flow.

It is important to point out here that a 5-beam probe was purchased
for the purpose of measuring all three components simultaneously, as
recommended. After much time and effort it was determined that
measuring three components Simultaneously was not a realistic option at
this time. Instead, efforts were then put in to mapping the flow fields in

one and two dimensions. Pertinent details shall be covered in Chapter 4.

CHAPTER 2

EXPERIMENTAL EQUIPMENT

2.1 Test Facility: Test Rig

The test rig assembly consists of a torque converter fixture, a
bedplate, a 50 Hp DC drive motor and a 50 Hp Universal Dynamometer
(Figure 2). The converter fixture supports the modified torque converter
assembly and allows Optical access into the converters for flow studies.
The bedplate is used to support the fixture assembly with a fixture
bracket. Located below the bedplate, used to drive the converter
assembly, are a 50 Hp DC drive motor and a 50 Hp Universal
Dynamometer, which has an eddy current dynamometer motoring feature.
The 50 Hp DC drive motor is used to simulate an automotive engine,
whereas the 50 I-Ip Universal Dynamometer simulates load conditions. A
belt drive mechanism is used to couple the motor, dynamometer and
torque converter fixture.

The DC drive motor can produce a maximum of 2500 rpm, and is
geared at a 2:1 ratio to that of the torque converter pump (impeller),
resulting in a converter speed of 5000 rpm. The Universal Dynamometer

is at a 1:1 ratio and drives the converter (turbine) at 5000 rpm. The

speeds of the motors are controlled independently through the use of

separate controllers.

 

Figure 2. View Of Test Rig

2.2 Transmission Oil Supply System

The transmission oil supply system controls pressure, flow rates,
and temperature (Figure 3). The transmission Oil supply system is divided
up into two separate systems. One system is used to supply transmission
oil to the torque converter, charging the impeller, turbine, and stator
passages. The other system supplies lubrication oil to the support
bearings, which support the torque converter assembly. The pressures and
flow rates are controlled for each system with the use of regulators and

are monitored through the test rig controller. The bearing lubrication oil

 

is filtered and is used in conjunction with a vacuum pump to ensure
efficient flow through the bearings.

The temperature of the transmission oil is controlled through the
use of a water-cooled heat exchanger. This ensures the necessary
temperatures for operating conditions during testing. The transmission oil
is measured using type T thermocouples enabling the monitoring of the oil
as it both enters and leaves the converter assembly. Another important
use of the supply system is that it controls particle distribution in the
transmission Oil. Proper particle distribution is necessary for consistent

flow and laser velocimetry measurements.

 

Figure 3. Transmission Oil Supply System

 

2.3 Test Rig Controller

The test rig controller operates and monitors the test rig assembly
and the transmission Oil supply system. The controls consist of both hard
wired and computer controlled switches with a complete readout Of the
systems functions. LabVIEW is the software used in the test rig
controller for the readout (Figure 4).

The test rig controller measures torque on the shafts of the impeller,
turbine and stator through strain gauges. It also measures Oil
temperatures, pressures, and Oil flow rates. It maintains constant speed
and torque controls, as well as, records all measured parameters. Figure 5

shows the test rig and test rig controller.

2.4 LDV System

The LDV system is capable of measuring inside complex geometries
such as the automotive torque converter (Figure 6). The components used
in this LDV measurement system are an Argon-ion laser, a ColorBurst
(TSI), a 4-Beam Fiberoptic probe, a digital burst correlator (TSI, IFA
750), a traverse mechanism (TSI, Model 9900-1), and a personal computer
(PC). LDV measures fluid velocities by detecting the Doppler shift
frequency of the laser light that has been scattered by small particles (~10

um) flowing within the fluid at the same velocity as the fluid flow. These

12

particles are metallic coated plastic for efficient light scattering qualities.
When two beams cross a fringe pattern is formed (Figure 7). Particles
pass through this fringe pattern and the light is scattered. The scattered
light is collected by a photomultiplier detector and converted into voltage

signals.

.2‘I.‘III .HMI 'I‘Ii
I I 1N0 ‘JHIU ‘UK

M I‘vtt'v KM!!! Hm:
I" I” V1 In now
I :I n:-

(.1 AH rl HVIVv’
vNu '11“! II N ’n ,J—
... '“ I'lll‘Adhi HIva

()1 ELIIN‘...1 AI'HI 3
.lIIJI mu) I-IlIlll JUL-LI lUlJI III In

1216/97 5.24:44 PM

t.‘

n m ....;. ‘nm, ' u. m um-
,. 9:.” IJJII' W ~ 7. l Ital

‘Jnn'llfx' atom a
Him-'1 Al‘nm

 

Figure 4. VI Screen Display

 

Figure 6. View of 4-beam probe of LDV system and test rig with front window converter

 

 

 

 

 

 

:12}.
m d?
IM- ‘1'"
tank
,1
. d =
Particle o f 2sinx

Figure 7. 4-beam probe with fringe pattern

The velocity of the particles in the fluid is proportional to the
Doppler shift frequency and can be determined by the following

relationship:

Vel=df/t1 =deF (2.1)

where
df= fringe spacing
r1 = time taken for the particle to cross a pair of fringes

F = Doppler shift frequency Of the scattered light

15

The fringe spacing, df, can be readily determined by:

df = ,1 / (2 sin 1:) (2.2)

where
2. = wavelength Of the laser light

x = half angle between the beams

The Doppler shift frequency, F, is measured in hertz and is equal to
the number Of fringe patterns a particle crosses in one unit of time. The
frequency shifter allows flow reversals to be measured. A multi-
wavelength laser beam from an Argon-ion laser is separated by the
ColorBurst into green, blue, and violet beams. The Digital Burst
Correlator (IFA 750) distinguishes the burst signal from the noise. The
IFA 750 analyzes the Doppler signals so that the velocity components at
the measured position can be calculated in coincidence mode. These
transmitting beams are linked to the fiberoptic probe through fibers that
connect tO the fiber coupler on the ColorBurst.

A 4-beam fiberoptic probe was employed to measure one or two
velocity components. The function Of the 4-beam fiberoptic probe is to
transmit the laser beams to form the measurement volume and to collect
the light scattered by the particles traveling in the measurement volume.

Single-mode polarization-preserving fibers carry the green, blue, and

16

violet beams to the probe. The diverging cone of light emitted from each
fiber is focused at the measurement volume with a lens. For this study
both green and blue beams were utilized depending upon components
measured. In the measurement volume, constructive and destructive
interference forms the pattern of dark and bright fringes.

The 4 beams consist Of two pairs of Single wavelength beams, which
are green (514.5nm), and blue (488nm) beams. A fifth fiber, known as the
receiving fiber, is used to collect the light scattered by the particles as
they pass through the fringe spacing. This method used is known as the
on-axis backscatter method. Off-axis forwardscatter is more efficient,
however due to the complicated nature of the Optical access to the torque
converter, this method could not be used.

In this study, a pair Of blue beams are located vertically, while the
green beams are aligned horizontally. The 4-beam probe is used to
measure the horizontal and vertical velocity components perpendicular to

the optic axis.

17

CHAPTER 3

TORQUE CONVERTER

3.1 Modification

Two Ford Motor Company production torque converters Of the same
design were modified for optical access for internal flow measurements.
One converter required Optical access into the lower impeller passage up
through the impeller exit (front window) and the second for access at the
impeller-turbine interface (side window). Table 1 shows the
specifications of the production torque converter used. There were two
parameters that needed to be addressed during the modification process.
The first, installation Of the window must not interfere with the fluid flow
inside Of the torque converter during operation. Doing so would greatly
reduce the accuracy of the results. Second, the window must allow a
clear access for the beams to pass through into the converter when
measuring.

The production torque converter was modified at Ford by cutting
the converter in half radially and welding mild steel flanges to each
section of the converter. Holes were drilled and tapped in the flange and
bolts inserted to hold the assembly together. This modification allowed
for quick access to the components inside of the torque converter for easy

alterations and reassembly (Figure 8).

18

Table 1. Ford production torque converter specifications

orque

 

Figure 8. View of impeller half of converter with front window installed

 

3.2 Front Window

For the front window installation a section of the impeller passage
was milled. This included an area from approximately mid-way up the
impeller passage to the impeller exit. This allowed for optical access to
the impeller passage, gap region, and turbine passage. An acrylic
Plexiglas was chosen as a window material since its refractive index was
close to that of the transmission oil used. The inside surface of the
window follows the curvature of the inside surface of the converter
housing, so that fluid flow inside the converter should not be disturbed
(Figure 9). Another property to consider was temperature. The internal
converter temperature must be maintained below the point at which the

Plexiglas will lose strength and possibly deform.

 

Figure 9. View of front window

20

3.3 Side Window

For the side window the modification was Similar to that Of the
front window. The flange was installed in the same manner as the front
window for quick access for easy alterations as needed. Plexiglas was
also used as the material for the side window for the same reasons as the
front window. The side window was installed to gain Optical access to the
impeller passage, the gap region, and the turbine from a different
perspectve (Figure 10). This location allowed for data to be measured in
1-D flow from the impeller passage out into the gap region to the turbine.
This data would provide a third component to the fluid flow measured.

One area of concern was the geometrical requirement Of the side
window to follow the inside curvature Of the converter. In this location
on the converter the side windows inside surface was much more
contoured as compared to that Of the front window. When measuring, the
beams pass through the window and the focal point is inside the
converter. If there is refraction of the beams when they pass through the
window then the true location Of the focal point may not be known. This

issue is addressed in detail in chapter 4.

21

 

Figure 10. View of impeller half of converter with side window installed

22

CHAPTER 4

EXPERIMENTAL PROCEDURE

4.1 Preliminary Testing

As mentioned in the introduction, a 5-beam probe was
purchased from TSI. This device Offered the promise of a three-
component measurement through a single focusing and collecting lens. It
operates under similar conditions as the 4-beam probe in terms of setup
and operation. This was a TSI model 9233 fiberoptic 5-beam probe
capable Of measuring three velocity components simultaneously. One
concern, when used for this project, was the obtainable data rate for all
three pairs of beams in coincidence mode and another was the beam half
angle used relative to the accuracy of the data acquired. After careful
calibrating, testing, and measuring it was determined that the 4-beam

probe was best suited for this study over the 5-beam probe.

4.2 Setup

Periodic maintenance was performed such as the transmission Oil

changes because of Oil oxidation, and its effect on signal degradation

through Optical absorption. Once the converter assembly was installed

23

and the system was ready, a number of steps were performed via the

following checklist.

l. The traverse table zero point is set so that the coordinate system

origin is known.

2. The impeller and turbine zero reference points are set for the shaft
encoders.

3. The beams focal point is moved inside the converter through the
window.

4. Frequency shifters and receiving optics are set and adjusted.

5. The chosen speed ratio is set.

6. The system, such as temperature, pressure, and flow rates are

allowed to stabilize.
7. The data rate and quality are checked.
8. The data acquisition software (TSI Phase) is programmed.
9. Data is acquired and recorded.

The traverse table must be referenced to the torque converter
assembly when moving the focal point to necessary locations to be
measured (Figure 11). The traverse table moves in three-dimensional
Cartesian coordinates and referencing is accomplished through the use Of

the LDV laser beams.

24

Traverse Table Zero
Reference Point

 

 

Test Rig "
Fixture
Window >0
Torque
Converter

 

 

 

 

Figure 11. View of test rig apparatus for traverse table reference point

The encoder zero degree reference points are set by use of the LDV
laser beams, angle encoders and also a Tektronix 2445 oscilloscope
(Figure 12). For the impeller the welding tab on the shell of the converter
of the following impeller blade is used, whereas, for the turbine the
surface of the turbine blade itself is used. Once the beams are oriented
the shaft encoder trigger signal is viewed on the oscilloscope and the

encoder is then set.

25

Impeller Blade

 

 

0° Reference
Torque .

Converter Turbme Blade
0° Reference

V View of turbine blade

through fiont window

CCW
Rotation

Figure 12. View Of Impeller and Turbine positions for encoder references

4.3 Data Acquisition

The internal flows in the torque converter were quantified using the
2-D LDV system previously described. The data measured totaled three
different data series. The first and second series were measured through
the front window, while the third series was measured through the side
window. A total Of 9 planes were measured, 5 through the front, and 4
through the Side window. All three data series were measured at two
speed ratios. The first being a speed ratio of 0.4, with an impeller speed
of 1600 rpm and a turbine speed of 633 rpm and the second speed ratio is
0.8, with the impeller Speed of 2000 rpm and a turbine speed of 1600 rpm.

Coordinates from Ford blueprints were used to generate wire frames Of

26

KI)
-_-’

II

he

he

rh‘

IIII

{If

TOI

both the impeller and turbine in the computer animation series (Figure
13). The LDV’S Cartesian coordinate system has been transformed to
match Ford’s coordinate system (FCS) for convenience. Therefore all of
the results shown will be in FCS.

The blue and green arrows in Figure 13 represent the beams from
the LDV system as they travel through the front window into the
converter along the positive y-axis. For the front window series there
were 2-D components measured. In the side window series there was l-D
component measured, which was done to provide toroidal flow
characteristics. The locations of the vector components for both the front
and side window series are shown (Figure 14).

The Views are side views Of the impeller and turbine and the
rotation of both rotors is counter-clockwise (CCW). The CCW rotation is
viewed from the left Side of the LDV beam perspective, therefore the
rotors rotate into the page at the top of the View (hence the suction and
pressure Side of the impeller blade).

The vectors represented in Figure 14 are an example Of a point Of
the plane measured. For the front window the LDV beams enter through
the window from the left and measure the two components shown. The
blue beam is the vertical component (positive 2 direction) measured and
the green beam represents the horizontal component (positive x direction).

The side window has the LDV beam entering through the window at the

27

top Of the assembly for one component measurement. Here the green
beam indicates the toriodal flow measured (positive y direction).

9 planes were measured at different locations within the converter
(Table 2). The 2 and y-axes are listed without the x-axis. This is because
the planes (2 and y-axes) represent the total area measured and the x-axis
location represents the point measured in each plane. The point measured
varies due to rotation because the measurement was based on cylindrical
coordinate system and was conducted through a circumferential line.
Therefore, the circular line can be defined by two coordinates such as y
(for depth) and z (radius). The arrows indicate the direction in which the

data were measured.

Impeller I Turbine

 

Figure 13. View of impeller and turbine blades inside the torque converter with measured

planes (Ford coordinate system)

28

Table 2. Measured plane locations in Ford coordinate system

 

 

 

 

 

 

 

Data Measured Location (mm)

Front Window Y Z (radius)
Plane 1 -30.42—)-11.78 93
Plane 2 -2.09 105—)112.5
Plane 3 3.88 105.5—)112
Plane 4 -2.09 105-)112.5
Plane 5 -O.43 105—)111

 

Side Window

 

 

 

 

Plane 6 -5.41 112->107.5
Plane 7 -4.42 113.5—)108
Plane 8 -2.29 117.3—)103
Plane 9 -1.29 117.8—)112

 

 

 

 

 

29

LDV"1: ’
. Beam 1 7 .I

I

{Shroud —p

_. {guns ’ 7 . ‘ f I

*Sh‘roud —? I I ‘ 1‘14.— Suction Side

' ‘ II '4— Suction Side -. 1.. »

I . _ Ԥ- Pressure Side

i4..."— Pressure Side . I ..i
. . - . '1]

\ I

IWeb . ~ I‘I‘fWBI-VI' I

Impeller Turbine III}. Impeller Turbine

Front Window 2-D Side Window l-D

 

Figure 14. Views of measured vector components in both front and side window series

The LDV system, including the 4-beam probe, rests on a TSI model
9900-1 traverse table that is capable of moving in a 3-D Cartesian
coordinate system within an accuracy of i 1.0pm. The 4-beam probe is
positioned for measurement locations as necessary.

When data is being measured the data acquisition system collects
the raw velocity and angular position of impeller and turbine information.
This is done using the output Of two rotating machinery resolvers (RMR),
IFA 750, and TSI software known as Phase. Flow data was collected
based on the impeller and turbine angles, which were detected by an angle
shaft encoder installed on each rotor. Each encoder generates one square
pulse per revolution (CPR) and two series of 1024 square pulses per

revolution. A total of 4096 pulses per revolution were used. The RMR’s

30

(one for each rotor) record the angle from the encoders and attach it to the
velocity measured. These measured angles are then passed to the IFA
750, which collects the data in bins and stores it in the PC. The TSI
Phase software monitors parameters such as the data rate, measurement
time, as well as calculating velocities from frequency information.

During the experiments, the equipment was allowed to run for at
least 30 minutes to be sure pressures and temperatures had stabilized.
Preliminary checks are made in association with the data rate to be sure
criteria such as proper particle distribution is correct. Test conditions
utilized for both speed ratios were supplied by Ford and are based on

normal Operating conditions of the torque converter (Table 3).

Table 3. Test conditions for data measured

 

Speed Ratio 0.4 Speed Ratio 0.8

 

Oil Temp (BC) Pressure (PsiG) Oil Temp (°C) Pressure (PsiG)

 

Vent 4o Vent 75.62 Vent 3o Vent 76.16

 

Charge 40 Charge 59.83 Charge 30 Charge 65.16

 

 

Discharge 65 Discharge 35.86 Discharge 50 Discharge 36.79

 

 

 

 

 

 

 

 

31

 

4.4 Measurements with Front Window: First Series

There are three planes where two component velocity vectors were
measured in series 1 (Figure 15). The blue beam pair was used to measure
the vertical component measured along the positive z-axis, and the green
beam pair was used to measure the horizontal component along the
positive x-axis for all three of the planes (FCS). Ford determined the
plane locations that were to be measured. The first plane was measured
mid-chord in the impeller passage. This plane starts at the web side of the
impeller passage and continues along the positive y-axis (FCS), with the
plane extending toward the z-axis, to the shroud side (plane 1, Figure 15).
The purpose of making measurements in this plane was to study the
uniformity Of the flow as it continues toward the impeller exit.

The second plane was measured at the exit of the impeller passage.
This plane begins at the shroud side (at the bottom of the impeller exit),
extending along the positive z-axis to the web side at the top of the
impeller exit (plane 2, Figure 15). The purpose Of making measurements
this plane was to study the fluid flow as it continues from the impeller
exit out into the gap region.

The third plane was measured approximately 1mm inside the turbine
passage. This plane begins at the shroud Side (at the bottom Of the

turbine inlet), extending along the positive z-axis to the web side at the

32

top of the turbine (plane 3, Figure 15). Velocities in the third plane were

measured to map the fluid flow as it entered the turbine.

4.5 Measurements with Front Window: Second Series

For the second series two planes were measured in two components,
where the blue beam was the vertical component measured along the
positive z-axis, and the green beam being the horizontal component along
the positive x-axis (Figure 16). Plane 4 is located at the impeller exit
with the plane extending along the positive z-axis. This area was chosen

to map fluid flow characteristics as the flow leaves the impeller passage.

..~+——.

page: :jj I g 1 Plane 3

II PIanel I I

.' i I Ip-ShrOud
I L f;
I.._ Web '

 

Figure 15. First Series with 3 planes

33

i/Trl
' I

Plane 5

Plane 4 :

 

Figure 16. Second series with 2 planes

Plane 5 was measured 1mm past the impeller passage into the gap region,
also extending along the z-axis. This area was measured to map flow
characteristics as the fluid was leaves the impeller passage and enters into

the gap region.

4.6 Measurements with Side Window: Third Series

For this data series the torque converter with the front window was
removed from the dynamometer test stand and replaced with the torque
converter with the side window. There are four planes where one
component flow was measured from the top of the torque converter

(Figure 17).

Plane 6 NI E‘— Plane 9

Plane 7 J}

Plane 8

 

Figure 17. Third series with 4 planes

This was to represent the 3rd component of the fluid flow (in the y-axis
direction) allowing for additional information of the trends Observed
measured through the front window. Four parallel planes in this region
were studied. Plane 6 is inside the impeller passage 2mm from the exit
and extends along the positive z-axis. Plane 7 is inside the impeller
passage 1mm from the exit also along the z-axis. Plane 8 is at the
impeller exit and plane 9 that is 1mm outside the impeller passage in the
gap region. Two of the planes (planes 8 and 9) are close to the location

of the front window meaSurement series (planes 2,4 and 5).

35

4.7 Side Window Geometry

One area of concern was refraction and reflection associated with
the non-linear curvature Of the side window lens. While it is imperative
that the lens matches the inside geometry of the torque converter (else
fluid flow would be altered), it is equally important that the location of
the focal point be known when actual measurement takes place. Much
effort was put into determining the curvature of the lens therein by
locating the focal point at given intervals along its inside surface (Figure
18).

One point of particular concern was the sharp curvature at the exit
plane of the impeller. The question posed was, is the focal point truly at
the measured location believed, or had it moved based on the refraction

effect on the laser beams as they left the lens?

 

Figure 18. View of side window

A Plexiglas box was constructed with the side window lens installed
in its side and the box was then filled with base stock transmission Oil.
The LDV system was then used to determine the location Of the focal
point as it moved along the lens surface (Figure 19). The result showed a
good approximation Of the focal point as it moved however; it was not
accurate enough to be sure Of the actual distance the focal point varied.
Other methods were attempted, such as Video taping and photographing
the focal point in the torque converter at measured locations, however the
resolution was still poor.

The next step was to have the lens surface carefully measured which
would enable a more exact location of the focal point to be calculated.
Demmer Corporation, with the use of a coordinate-measuring machine
(CMM), measured the lens and mapped the points to .an accuracy of
1.0um. Particular emphasis was placed on the curved areas of the lens.
The result received from Demmer was in an IGS file format that was then
used as a database for a c-spline surface representation, which was
implemented in a computer code (Figure 20). Using this code, the focal
intersection Of the laser beams measured locations was calculated.
Variations were found to be less than 0.2mm of the overall movement of
the focal point for the range of measurement for this study. This was
believed to be the most accurate way to determine the focal point
location. It was found that where the actual measuring took place the

focal point was not far enough beyond the surface of the lens

37

(approximately 5.0mm) to make a significant difference in spatial

location.

 

Figure 19. View of beam focal point through the side window

 

Figure 20. 3-D display of side window (Demmer)

4.8 Data Post Processing

Post processing Of the LDV collected data involves many different
aspects. The data collection process is shown (Figure 21). The test rig
controller, which is used tO record the strain, pressure, and temperature
during testing and this information is then stored in a file. The test rig
controller also determines the speed, torque, etc. Of the test rig. Sensor
feedback occurs from the test rig to the test rig controller. The receiving
optics (located inside of the 4-beam probe) collects the scattered light and
is changed to voltage signals by the photomultipliers. At the same time
the RMR’S are attaching an angle, based on encoder information to each
velocity component measured. The RMR’S pass this information to the
IFA750. The frequency shifters control the frequency and direction
(usually against the mean flow) of the data rate via the Bragg cell and this
information is sent to the IFA750 as well. The IFA750 stores all Of the
information into bins and then relays this information to the PC to be
stored. The PC sets the IFA750’S details through the Phase software at
the onset of the testing. This process continues one spatial point at a time
until a data set for one location is measured. Then the traverse table is
used to move the focal point from one location to another location to be
measured and the process begins again and is repeated until a full plane is

mapped. It is important to mention that each point measured must

39

generate a line Of data due tO the converters rotation, thus as each point is

moved a plane is the result.

 

 

 

 

 

 

 

 

 

IFA 750 RMR’s Test Rig

data acquisition 4— master interface *— encoders, sensors,
processor tmit gauges, motors, and
A I fixture assembly

 

 

 

 

Freq. Bragg Photo Receiving
Shifter Cell Mult. Optics

 

 

 

 

 

 

 

 

Rig Controller
PC stores data records press, temp.
—9 ms software operates test rig
Phase full gauge readout

 

 

 

 

 

 

Figure 21. Data acquisition flow chart

At this point the measured data is ready for processing. For the
torque converter LDV data is collected through an area the same width of
the impeller passage. The shaft encoders allow for the instantaneous
positions of both the impeller and turbine to be recorded each time a
velocity measurement is acquired by the system. The data stored in the
PC is in the form of raw data (Figure 22). The raw data file is in binary
format and contains information such as the measured velocity,
magnitude, encoder values, and cycle number in the frequency frame Of

reference.

 

 

Raw W Animation

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Data
FIOWpturb Averaged
l 1 (code) -> Velocity
Statisical Velocity Output F" Flows
File -> Massflow M333
i t (code) “D Flow
Phtqv l V V TQANLIZR Rates
e .
(code) I" + (code)

 

 

 

 

 

 

 

 

Figure 22. Data processing flow chart

The cycle number is relevant in that the data is measured over time, which
is necessary for data accumulation. Therefore, there is no cycle-to-cycle
variation present in the data.

The raw data file can be processed through Phase, changing the
format from binary to ASCII and produces two output files known as
statistical and velocity. The statistical file contains information such as
standard deviation, number of points collected, and velocity mean, among
other data. One primary advantage of the statistical file is that it plots the
data in angle of degrees versus the points measured. This provides a
quick indication to see if the data recently measured is correct.

The velocity file contains the necessary information for processing,
however it is not in the correct frame of reference needed. The TSI code
Phtqvel basically changes the frame of reference in the velocity file from

the frequency domain to the time domain, with the output file called the

4l

V.V file. This TSI code Phtqvel was modified to handle large files
consecutively. The output file (V.V) can now be processed into the
format necessary for animation.

A FORTRAN code called TQANLIZR was written to produce an
output file in the format required by Ford engineers for comparison with
simulation results and for animation. TQANLIZR is used to sort the data
into bins containing information, such as velocity and encoder angles for
the impeller and turbine through ensemble averaging. The data is
collected in bins of 0.5° from the respective angular positions. This bin
number was chosen to generate a uniform grid suitable for animation.
Since the turbine and impeller rotate at different speeds and the
measurements are made through an impeller window; the turbine can be
any place within the window at any given moment in time, hence the
turbine angular measurement range is 360° of rotation. For that reason at
each impeller angle all of the turbine angles are simultaneously recorded.
All turbine angle measurements are approximately 0° to 13.3°. The
impeller angle value ranges from 0° to approximately ll.6° coinciding
with the width of the window in the impeller side of the converter. An
additional consideration is to ensure that the impeller and turbine angles
are represented from the same perspective since each is measured
independently. This allows for the animation to represent the passing of

the turbine blade with respect to the impeller blade angles recorded.

42

TQANLIZR is also used for coordinate transformation. The data is
collected initially in Cartesian coordinates with respect to the LDV
system axis and is then stored in terms of cylindrical coordinates. The
data is then transferred from the initial coordinate system to the Ford
coordinate system in cylindrical coordinates and grouped into bins. The
fact that the code transfers the coordinates from the MSU LDV coordinate
system to the Ford coordinates for animation adds credibility to the
accuracy of the location of the measured planes. That is, rather than just
place the velocity profile results in locations in the Ford coordinate
system for animation purposes manually, the code uses the LDV
coordinates inputted and transfers the coordinate system into the PCS,
placing the data in the proper locations. The data also is converted into
the impeller frame of reference for when viewing the animation series.
For animation the 861 software explorer coordinate system was set up to
match Ford’s coordinate system.

The process for the side window data is similar to that of the front
window. TQANLIZR was modified for the side window to incorporate the
cubic spline code accounting for window curvature. The front window
was used to gather two component data, whereas the side window was
used to gather one component data for the toroidal flow near the impeller
exit and the gap region near the turbine entrance. This data is gathered
from different perspectives and uses two unique windows for viewing the

flows. TQANLIZR then generates an output file that can be read by the

43

SGI animation software. As with the front window the data information,
measured in the side window is converted into the impeller frame of
reference in order to facilitate the animation process.

When viewing the animation, data is presented frame by frame over
all of the turbine angles. The results section show significant aspects of
these animations. These snapshots were chosen because they show
distinct flow characteristics noted in the animation, however the average
flow rate is only of that particular frame. A second code was written to
provide more information that could be derived from the output files from
TQANLIZR. The Flowpturb code produced an output file containing the
average flow of the entire measured series for each plane in one frame.
Each series was then animated. The turbine is not included in the
animation since the vectors shown are the results of all the turbine angles
for each impeller angle measured. Since the velocities are averaged for
each impeller angle, the turbine would appear as a solid passage. The
advantage here is the average flow rate in meters per second can be
estimated and used to gain a perspective of the overall fluid flow
measured. The disadvantage is distinct flow characteristics, such as
blade-passing effects cannot be clearly seen.

For purposes of assisting the computational fluid dynamics (CFD)
code development, the code Massflow was written to show the mass flow
of the fluid normal to each plane measured. Here all of the turbine angles

were ensemble averaged at each measured impeller position. The output

file contains the position, as well as, the mass flow rate in kg/s. This
result would show the primary characteristics of the toroidal flow and can
be used for flow comparison of experimental results with those of the
CFD results. Not all of the planes measured fit these criteria. The
vertical component for plane 1 was useful for the primary mid-chord flow
of the impeller passage, which was measured in the front window series.
Planes 6-9 provided primary flow characteristics measured in the side
window series. Planes 2-5 were not used since the components measured

provided secondary flow information.

4.9 Animation

The velocity profile animation series was processed through the use
of an 801 workstation and EXPLORER software. There are several
details about the animation that need to be addressed. First, the animation
provides a view of the final results, after the data has been processed from
the output files containing coordinates and velocities of each plane
measured. Second, the animation is constructed with the impeller used as
the reference frame. More specifically, the observer views the results
from the perspective of sitting on the impeller and watching the relative
fluid flow and turbine motion. Third, the figures shown in the results and
discussion of this report are single frames taken from the animation series

and were chosen to show the observer important fluid flow characteristics.

45

These snapshots provide particular details about flow characteristics in
each plane observed during the animation. A complete animation is

available in high 8 and VHS format as part of this work.

CHAPTER 5

RESULTS AND DISCUSSION

5.1 Measurements with Front Window: First Series (Planes 1-3)

When viewing these Figures, the planes are rotated to an optimal
position to best display the results. The picture on the lower right-hand
corner of the Figure shows the planes measured location. The green and
white arrows correspond to the measured and view perspectives,
respectively. The white wire frame generated represents the impeller and
the red wire frame, the turbine. Cartesian coordinates are also provided
to assist in plane orientation when viewing these Figures. All of the
planes are rotated CW on the z-axis, with a small number (of Figures)
tilted slightly for additional data clarity. The upper half of the figure is a
picture that shows the details of the planes flow pattern characteristics
mapped.

The results are discussed by dividing the measured planes into
sections. The first section discusses the plane results for the 0.8 speed
ratio. The second discusses the 0.4 speed ratio and the third discusses
comparisons of both speed ratios. Subsequently there is discussion of

comparison of particular planes for both the front and side window series.

47

5.1.1 Plane 1: 0.8 Speed Ratio

Plane 1 was measured mid-chord of the impeller passage. The flow
pattern for the 0.8 series was similar to that of the 0.4 speed ratio (Figure
23). The measured velocities were slightly lower for the 0.8 speed ratio
case, at an average velocity of 8.5m/s, as opposed to 9.2m/s in the 0.4
case. Here the vectors vertical component (positive 2 direction) was not
generally uniform from the suction side to the pressure side of the
passage. Rather, the flow pattern increased in magnitude (positive 2 and
x direction) as it approached the pressure side of the impeller passage.
Numazawa had similar results for vector magnitudes from the suction side
to the pressure side of the impeller passage [20]. The flow at the shroud
side of the impeller passage was smaller in magnitude. This flow pattern
also moved in a different direction than that of the rest of the plane
(positive x and —z direction), with vectors pointing downward (-z), or
straight across (+x), as it moved from the suction to the pressure side of

the passage.

5.1.2 Plane 1: 0.4 Speed Ratio

The vertical component (+z direction) was generally uniform from

the suction side to the pressure side of the impeller passage (Figure 24).

The average velocity of the fluid flow was 9.2m/s. The fluid flow

48

generated in the animation showed a general uniform trend (+2 and +x
direction) along the impeller passage, with the exception of an area at the
shroud side of the passage. At the shroud side of the passage, a band of
flow smaller in magnitude was observed. This band of flow traveled
downward (-z direction) as it moved to the pressure side from the suction

side of the impeller.

5.1.3 Plane 1: Comparison of Both Speed Ratios

Plane 1 was measured mid-chord of the impeller passage (Figures
23 and 24). For both speed ratios the data was processed and animated.
There were no noticeable turbine blade-passing effects in the flow at this
location. This result is consistent with the findings of McCarrick [21].
The general trend of the fluid flow traveled upward towards the impeller
exit, with the direction of the vector magnitudes from the suction side to
the pressure side of the impeller (positive x direction). This flow pattern
included the whole plane with the exception of a small area near the
shroud side of the passage. At the shroud side of the impeller passage, 3
band of flow smaller in magnitude was found. This flow also moved in a
different direction (positive x and —z direction) than that of the rest of the
plane [21]. This could result by flow from behind the core ring re-

entering the impeller passage through the blade tab slots. Another

49

possible explanation is that due to the geometry of the shroud a type of

shear flow may be occurring. Further studies are necessary to be certain.

5.1.4 Plane 2: 0.8 Speed Ratio

Plane 2 was measured at the impeller exit. The flow pattern moved
in a general uniform direction (+x direction) from the suction side of the
impeller passage to the pressure side (Figure 25). The average velocity of
the fluid flow was 5.5m/s. The flow magnitudes increased (+x direction)
from the suction to the pressure side as in the previous case. As the flow
moved across the passage, the vector angle (from horizontal to vertical)
was generally increasing as the flow continued to the pressure side. There
were some steeper velocity gradients observed along the bottom of the

plane.

5.1.5 Plane 2: 0.4 Speed Ratio

The flow pattern moved in a general uniform direction (+x) from
the suction side of the impeller passage to the pressure side, however this
flow contained steeper velocity gradients than in the 0.8 speed ratio case
(Figure 26). The average velocity of the fluid flow was 3.3m/s. As the
flow moved from the suction side to the pressure side of the impeller

passage, the flow magnitudes were not at a generally increasing angle

50

(horizontal to vertical) as with the 0.8 speed ratio case. Instead, the
overall flow pattern moved up (+z direction), then down (-z direction), in
a type of s-curve, as it continued toward the pressure side. There were
also steeper velocity gradients observed along the lower portion of the

plane.

5.1.6 Plane 2: Comparison of Both Speed Ratios

Plane 2 was measured at the impeller exit (Figures 25 and 26). For
both speed ratios the data was processed and animated. Here the general
flow pattern of the fluid traveled out of the impeller exit, with the
direction of the flow magnitudes (+x direction) from the suction side to
the pressure side of the impeller. The magnitudes of the flow were
smaller on the suction side of the impeller passage than that of the
pressure side for both speed ratios. This would be expected due to the
rotation of the torque converter.

Blade-passing effects of the turbine were found, although this effect
was more prominent at the 0.4 speed ratio then at the 0.8 speed ratio.
McCarrick found this effect in the gap region, however not as far back as
the impeller exit [21]. As the turbine blade passed in front of the impeller
passage the velocity of the flow field was affected. The blade-passing
effect-appears to trail the actual turbine blade, which indicates a time

delay between the turbine blade and the measured plane. This may be

51

caused by a ripple effect from the fluid created from the turbine blade as
it passes. This ripple effect propagates along the gap region back to the
impeller passage. The animation generated for this series clearly shows

this effect.

5.1.7 Plane 3: 0.8 Speed Ratio

Plane 3 was measured 1mm into the turbine passage. The average
velocity of the measured flow field was 6.5m/s (Figure 27). The
magnitudes of the vectors on the pressure side of the turbine were fairly
constant (+z and x direction), with the exception of some higher velocities
at the bottom of the plane near the blade. On the suction side of the
turbine, the vector magnitudes varied somewhat more than on the pressure

side. Steeper velocity gradients were observed along the top of the plane.

5.1.8 Plane 3: 0.4 Speed Ratio

The magnitudes of the flow on both the pressure and suction side of
the turbine varied more than in the 0.8 speed ratio case (Figure 28). The
average velocity was 7.2m/s. Steeper velocity gradients were observed
along the edge of the suction side of the turbine and along the bottom of
the pressure side. On the upper left-hand region there is a band of blue

vectors moving upward (+2 and x direction) at a speed of 6.0m/s.

52

5.1.9 Plane 3: Comparison of Both Speed Ratios

Plane 3 was measured 1mm into the turbine passage. This plane
showed a distinct band with no data that directly correlates to the turbine
blade as it passes (Figures 27 and 28). This is clearly evident in the
animation. Since this data was measured approximately 1mm in the
turbine passage the band is the from the actual turbine blade itself. The
overall velocity of the fluid was also slowing down, which is expected
since the fluid is now doing work on the turbine blade. The fluid flow
field impacts on the pressure side of the turbine blade causing the turbine
to rotate. On the suction side of the turbine the flow appears to follow

the blade (+2 and x direction).

5.2 Measurements with Front Window: Second Series (Planes 4 and 5)

5.2.1 Plane 4: 0.8 Speed Ratio

Plane 4 was measured at the impeller exit. The flow pattern
showed general flow direction from the suction side of the impeller
passage to the pressure side (Figure 29). The average velocity of the fluid
flow was 3.0m/s. The flow magnitudes increased (+2 and x direction)
from the suction to the pressure side, with the exception of along the

bottom of the plane where the flow moves in the —z and +x direction. For

53

the upper half of the plane the fluid flow moved across the impeller
passage, where the vector angle was generally increasing (horizontally
and vertically) as the flow continued to the pressure side. For the bottom
half of the plane the flow moved straight across (+x) or even downward (-
z) as it approached the pressure side. There were steeper velocity

gradients also observed along the bottom of the plane.

5.2.2 Plane 4: 0.4 Speed Ratio

The flow pattern is shown in Figure 30. This flow magnitudes
varied more than in the 0.8 speed ratio case. The average velocity of the
fluid flow was 5.5m/s. As the fluid flow moved from the suction side to
the pressure side (+x direction), the flow does not have the same general
increasing angle (+2 and x direction) as with the upper half of the 0.8
speed ratio case. Instead, for the majority of the plane the flow direction
moved downward (-2) as it continued toward the pressure side. A larger
portion along the bottom of the impeller passage also included steeper

velocity gradients.

5.2.3 Plane 4: Comparison of Both Speed Ratios

Plane 4 was measured at the impeller exit (Figures 29 and 30).

Blade-passing effects of the turbine were found in both the 0.4 and 0.8

54

speed ratios. Again, McCarrick found these effects in the gap region, but
not at the impeller exit [21]. Here the flow is moving out of the impeller
exit, in the +2 and x direction, from the suction side to the pressure side
of the passage. The time delay from the wave that propagates from the
turbine previously mentioned (Plane 2) is also evident in the animations.
The fluid flow characteristics from the suction to the pressure side of the
passage showed a distinct s-curve pattern, including higher velocities on
the pressure side overall. Steeper velocity gradients were observed at the

bottom of the plane than at the top.

5.2.4 Plane 5: 0.8 Speed Ratio

Plane 5 was measured 1mm from the impeller exit in the gap region.
The flow pattern moved in a general trend in the gap region toward the
turbine (Figure 31). The average velocity of the fluid flow was 3.5m/s.
The flow magnitudes increased (+2 and x direction) from the suction to
the pressure side. The vector angle was generally increasing (form
horizontal) as the fluid flow moved across the gap region. In the middle
of the plane (approximately in the center of the passage), a band of
steeper velocity gradients were observed. This is due to the turbine
blade-passing effects. A section of flow (purple) smaller in magnitude on

the left-hand side of the plane appear to follow the band of steeper

55

velocity gradients. There also were steeper velocity gradients observed

along the bottom of the plane.

5.2.5 Plane 5: 0.4 Speed Ratio

The flow pattern moved in a general direction in the gap region
toward the turbine, however this flow contained steeper velocity gradients
than in the 0.8 speed ratio case (Figure 32). The average velocity of the
fluid flow was 7.0m/s. As the fluid flow moved from the suction side to
the pressure side (+x direction), the flow did not exhibit increasing angle
with respect to the x-axis, as with the 0.8 speed ratio case. Instead, the
flow moved up (+z), then down (-z) as it continued toward the pressure
side. As in the 0.8 speed ratio case the middle of the plane
(approximately in the center of the passage), steeper velocity gradients
were observed. This also is due to the turbine blade-passing effects. In
the upper left-hand corner of the plane (left of the band of blade-passing
effects) vectors lower in magnitude were observed. These showed a flow
that was circular in nature. There were also steeper velocity gradients

observed along the lower portion of the plane.

56

5.2.6 Plane 5: Comparison of Both Speed Ratios

Plane 5 was measured 1mm from the impeller exit in the gap region
(Figures 31 and 32). Blade-passing effects from the turbine were quite
prominent. McCarrick found this condition also [21]. Although the
ripple effect was present, it did not take as long for the wave to propagate
back across the gap toward the impeller, as in plane 4. This would be
expected since the distance to the turbine blade is less at this measured
point. As with the previous cases the animation generated for this series

shows these effects clearly.

57

Pressure

I
11{3913
. 10.7102

/ ,1 i 10.0292
p I 1 . 9.31313

/
i 8.66709

gr I IMM / i 7.90501
_‘ / _. 7.305
' I 3 6.62395 '
5.91291
_w 5.26186

 

Green Arrow:
Measured Perspective
White Arrow:
View Perspective

 

Figure 23. Plane 1 (mid-chord of impeller) at 0.8 speed ratio

58

Shroud
Side

Pressure
Side

1259025
11.9010
11.0011
10.1101
9.21971

”.29902

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

Figure 24. Plane 1 (mid-chord of impeller) at 0.4 speed ratio

59

/.37832
6.45763
5.53691
4.61624

n1/s

 

 

 

Suction

\ *1

'131111\gx‘f

\

i
\

1
‘1 'l ‘\ :| , 1 11. . 1‘ W1
‘ ‘1 -. ‘ 1 ' ‘

1
I

\11113

\11\\\\\\x\1331

'5 I
r‘ 6’
r’ i 4-
/’ x
('7 r‘
/' _ /
("7 f‘
. / n’
K r
(If— .-" r’
, ,J ‘

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

Figure 25. Plane 2 (impeller exit) at 0.8 speed ratio

 

(I /’
foressure
Side

10.8317

. 9.98519

9.13864
8.29209
7.44554
6.59899

».S.75241

4.90589
1.05934

_ 3.21279

 

 

i” .7. 1.99392
’/ _ 1.43142
// ‘ 3.06392

1.29611
2.72891
2.16111
1.59391
1.02611
L___ 0.458912

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 26. Plane 2 (impeller exit) at 0.4 speed ratio

6]

Turbine *— ’

Suction

Side \

7 Turbine
Pressure

l/
- / 9.51497
/ . 8.74662
/ _ 7.97828
1 7.20991

' ' .2
I'Iil/
xi/
/ 6.4416

___7_ , ' 5.67325

—-“__‘_ 1 .90191

. ,. ,/ . ‘ 1.13657

. l " ///

3.36823

LCD, 2.59988

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 27. Plane 3 (1mm inside turbine) at 0.8 speed ratio

62

Turbine
Sucfion

Side \

Turbine
l’ressure

a] 10.6731
.9.92116
9.16891

’ 0.11665
7.66439
6.91213
6.15907

5 . 40761
1.65535
.11.90309

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 28. Plane 3 (lm inside turbine) at 0.4 speed ratio

63

Suction

Side 7 p _ , . /f

/
/’ Pressure
,7! Side
a“

,. 5.00201
.- 1.5811
4.07936
3.57762
3.07530
_____.__—-—-fi , ’ 2.57115
rid—“eff“r—“F , 7:1 ~ ,, 7 , 2.07241
- " 1.57067

_“___.._ I

4" fl x 1.06893

L (1.567186

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 29. Plane 4 (impeller exit) at 0.8 speed ratio

74/
;;71’ressure
/./‘)' Side

/ 10. 4141

1/77/ . 9.31963
// 0.22510
. 1 7.13072

. 6.03627
/' ' 1.94102
.'~ -15,,,1,_,_5_5 . . 3.04736

, 2.75291
1.65046
L_1_ 0.564004

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 30. Plane 4 (impeller exit) at 0.4 speed ratio

65

Suction ,. '
Side

éy/T’ressure
/ Side
.7 .

6.06331
' :4 5.41107
4.75002
' 4.10650
'1. 45434
’ 2.0021
2.14906
1 . 49762
0.045374
L._ (I. 193132

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 31. Plane 5 (1mm from impeller exit in gap region) at 0.8 speed ratio

66

Side

A955“ re
lll/

15.0679

. 14.2113
12.5548
10.0902
9.24171

' 7.50517
.5.92863
4.27209
2.61555
L— 0.95901

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 32. Plane 5 (1mm from impeller exit in gap region) at 0.4 speed ratio

67

5.3 Measurements with Side Window: Third Series (Planes 6-9)

5.3.1 Plane 6: 0.8 Speed Ratio

This plane was measured 2mm inside the impeller passage from the
impeller exit The average velocity of the fluid flow measured in plane 6
was 8.0m/s (Figure 33). The flow characteristics showed the main flow
moving toward the exit of the impeller (+y direction), with the highest
velocities occurring near the center of the passage. Velocities as low as

4.0m/s were observed in the upper left-hand corner of the plane.

5.3.2 Plane 6: 0.4 Speed Ratio

The average velocity of the fluid flow for the 0.4 speed ratio was
8.8m/s (Figure 34). The main flow also was moving toward the exit of the
impeller (+y direction), with the highest velocities occurring near the
center of the passage. Velocities as low as 6.3m/s were observed in the

upper left-hand corner of the plane.

5.3.3 Plane 6: Comparison of Both Speed Ratios

This plane was measured 2mm inside the impeller passage from the

impeller exit (Figures 33 and 34). In the impeller passage the data rate

68

was lower in part due to the limited optical access within the passage
along with the high reflectivity from the impeller walls. A filter was
incorporated in the TQANLZR code to remove the reflective data. The
remaining data allowed for some flow interpretation. There was no
noticeable turbine blade-passing effects found at this location. The
direction of the flow measured was generally uniform for both speed
ratios and followed the impeller passage (+y direction) toward the exit
and turbine as expected. This toriodal flow measurement had the highest

velocities occurring near the center of the impeller passage.

5.3.4 Plane 7: 0.8 Speed Ratio

Plane 7 was measured 1mm inside of the impeller passage from the
exit. The average velocity of the fluid flow was 7.5m/s (Figure 35). The
main flow was observed moving toward the exit of the impeller (+y
direction), with the highest velocities occurring near the center of the
passage. Velocities as low as 2.5m/s were observed at the top of the

plane.

5.3.5 Plane 7: 0.4 Speed Ratio

The average velocity of the fluid flow for the 0.4 speed ratio was

9.0m/s (Figure 36). The main flow was observed moving toward the exit

69

of the impeller (+y direction), with the highest velocities occurring near
the center of the passage. Velocities as low as 6.6m/s were observed

along the upper left-hand corner of the plane.

5.3.6 Plane 7: Comparison of Both Speed Ratios

For plane 7, the data rate was also lower in part due to the geometry
of the passage along with the high reflectivity from the impeller walls
(Figures 35 and 36). As with plane 6 a filter was incorporated in the
TQANLZR code to remove the reflective data. The features of the flow
pattern measured were generally uniform for both speed ratios, and moved
(+y direction) toward the impeller exit and the turbine as expected. The
toriodal flow showed the highest velocities occurring near the center of

the impeller passage with no turbine blade-passing effects at this location.

5.3.7 Plane 8: 0.8 Speed Ratio

Plane 8 was measured at the impeller exit. The average velocity for
the 0.8 speed ratio case was 7.0m/s (Figure 37). A band of higher
velocity vectors showing blade-passing effects can be observed diagonally
along the plane. This ripple effect has a time delay that follows the
turbine blade as it passes, causing the wave to propagate back across the

gap region toward the impeller. A series of vectors smaller in magnitude

70

(3.6m/s) were observed along the top of the plane above the impeller exit.
This flow is believed to be leaving the gap region, bypassing the turbine
and flowing above and behind the turbine and into the lock-up clutch

region. This causes a negative affect of the overall converter efficiency.

5.3.8 Plane 8: 0.4 Speed Ratio

For this plane the average velocity was 7.0m/s (Figure 38). A band
of higher velocity vectors showing blade-passing effects also was
observed diagonally along the plane. The ripple effect that follows the
turbine blade is also evident. A series of vectors was observed along the
top of the plane above the impeller exit, which were smaller in magnitude
(1.1m/s) than that of the rest of the plane. This flow also is believed to
be leaving the gap region and flow above and behind the turbine and into
the lock-up clutch region, having a negative effect on the converter

efficiency.

5.3.9 Plane 8: Comparison of Both Speed Ratios

Plane 8 was measured at the impeller exit (Figures 37 and 38).

There is a generally uniform flow from the impeller exit out into the gap

(+y direction). Blade-passing effects were observed for both speed ratios.

The time delay (ripple effect) was also observed. The blade-passing

71

effects were more conspicuous in this plane than for planes 2 and 4. It
may be more apparent in this case since the flow measured here is the
primary flow component, whereas for planes 2 and 4 the components
measured are the secondary flow. Also, for this plane the data rate was
higher than that of planes 6 and 7, making the particular flow
characteristics, such as blade-passing effects, more evident. This is due
in part to the geometry of the impeller passage. The curvature of the
passage is not as severe at the impeller exit as it is farther inside the
passage, allowing for higher data rate. For the 0.4 speed ratio similar
flow characteristics were observed, with the blade-passing effects a little
more prominent. Also, the geometry of the converter allowed for flow to
be measured above the top of the impeller passage at the exit. This can be
seen in both speed ratios. The flow leaves the impeller passage and
appears to flow above and behind the turbine blade into the lock-up clutch
region. This flow loss will have a negative effect on the overall torque

converter efficiency.

5.3.10 Plane 9: 0.8 Speed Ratio

Plane 9 was measured 1mm from the impeller exit out into the gap

region. For this plane the average velocity of the vectors was 5.0m/s

(Figure 39). Even with the lower data rate, a band of higher velocity flow

showing blade-passing effects was observed diagonally along the plane.

72

The blade-passing effect appears to verify the same effect observed in the
plane 5 (front window) series. A series of vectors smaller in magnitude
(1.8m/s) was observed along the top of the plane above the impeller exit.
This flow is believed to be leaving the gap region and flow above and
behind the turbine and into the lock-up clutch region, which negatively

affects overall converter efficiency.

5.3.11 Plane 9: 0.4 Speed Ratio

In plane 9 the average velocity was 5.4m/s (Figure 40). A band of
higher velocity vectors showing blade-passing effects was observed
diagonally along the plane. The ripple effect that follows the turbine
blade is also evident and does not appear to be as long a time delay as
with the plane 8 case. A series of vectors was observed along the top of
the plane above the impeller exit, which were smaller in magnitude
(1.8m/s). This flow also is believed to be leaving the gap region flowing
above and behind the turbine and into the lock-up clutch region, having a

negative effect on converter efficiency.

5.3.12 Plane 9: Comparison of Both Speed Ratios

Plane 9 was measured 1mm from the impeller exit out into the gap

region (Figures 39 and 40). In this case the 0.8 speed ratio resulted in

73

low data rate. The location being 1mm out into the gap region did not
create any problem with reflectivity from the impeller walls, however the
problem is believed to result from the particle distribution in the fluid
flow. At the 0.8 speed ratio the impeller and turbine are turning at nearly
a 1:1 ratio. The fluid flow characteristics are mainly toroidal in nature at
that speed ratio, therefore there would be little, if any mixing of the
particles in the flow. The fluid flow does follow the passage straight out
into the gap region (+y direction), as expected. The vector magnitude of
the data does change in a diagonal pattern showing the blade-passing
effects along the center of the passage.

In the 0.4 speed ratio case there is much more data than in the 0.8
case. This is due to the better mixing of the particles in the fluid flow.
The turbine blade-passing effects are more apparent in the 0.4 case as
well. The blade-passing effects are clearly evident showing the ripple
effect from the turbine blade as it passes. As with the previous side
window planes the flow is generally uniform across the measured area

with toroidal flow (+y direction) out into the gap region.

74

Sucfion .5

/

Side

11.39

k—~1057u
I 9.75009
‘ 0.94333

1 0.12770
I 7. 31222
6.19667

‘ 560ml

1.86556
L— 4 4.05

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 33. Plane 6 (2mm inside impeller) at 0.8 speed ratio

Suction
Side

/ Pressure
Side

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 34. Plane 6 (2mm inside impeller) at 0.4 speed ratio

76

Suction '
Side

Pressure
Side '

-_ 9.97

v 9.14333
I 8. 31667
7.49

6 . 66333
' ' 5. 03667
7 5. 01

4.18333 .
3.35667
L— 2.53

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 35. Plane 7 (1mm inside impeller) at 0.8 speed ratio

77

Pressure
Side

A
J
v.
a»
J-.'

‘L’

 

Green Arrow:

Measured Perspective
White Arrow:

View Perspective

 

Figure 36. Plane 7 (1mm inside impeller) at 0.4 speed ratio

78

Suction ~ 2 1

Side .. .
- .. 1 $43,. ”3" Pressure
‘ Side

/ - 10. 09

/ 9. 35222

-: 68.1444
7.87667
7.13889
6. 40111
b 5. 66333
1 4. 92556

1.18778
1.; 3.45

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 37. Plane 8 (impeller exit) at 0.8 speed ratio

79

Pressure
Side

_ 11.92

‘. 10.7122
_ 9.50444
' 0.29667

. 7.00009
5.001111 “1/5
1.67333
1.46556
2.25770

L... 1.05

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 38. Plane 8 (impeller exit) at 0.4 speed ratio

80

Pressure
Side .

2.63770
1.02009‘
___.1.02

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 39. Plane 9 (1mm from impeller exit in gap region) at 0.8 speed ratio

Pressure
Side

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 40. Plane 9 (1mm from impeller exit in gap region) at 0.4 speed ratio

82

5.4 Data Comparison

5.4.1 Front Window: Comparison Planes 2 and 4 at Both Speed Ratios

The front window data measured at the impeller exit, plane 2 was
similar to the data measured at the same location for plane 4 (Figures 41
and 42). The animation showed blade-passing effects, which were
noticeable in both series. It is noted that blade-passing effects were much
more noticeable in the plane 1mm outside the impeller passage (plane 5)
than in either plane 2 or 4. This would be expected with the time delay
and viscous losses taken into account.

The overall flow featured shown in planes 2 and 4 were similar.
More specifically, the comparisons of the results in plane 4 contained
steeper velocity gradients (particularly at the bottom of the passage) than
in plane 2 for both speed ratios. For the 0.4 speed ratio case, flow
magnitudes were quite similar for both planes. However, for the 0.8
speed ratio case the flow magnitudes were larger in the plane 2, and were
much more of a general uniform flow. This may be due, in part, to
position differences between the two planes when measured.

The location of plane 2 may be slightly outside the impeller passage
while plane 4 may be located slightly inside. That position difference
may be large enough to show these differences in the results. The exit of

the impeller passage is a very location sensitive area, in that the velocity

83

magnitudes change rapidly with spatial location. There are uncertainties
that must be included, such as the time difference between the measured
planes, which included disassembly of the converter for periodic
maintenance. Also the refractive index, temperature effect, vibration, and
measurement volume should be taken into account. Planes should be
measured on both sides of plane 2 and 4 (inside and outside of the
impeller exit) to gain further understanding on the characteristics of the
fluid flow in these areas. This would provide data in a more volumetric
fashion.

Another area of data to be analyzed further is planes 1 and 4,
Which are located mid-chord of the impeller passage and impeller exit,
reSpectively. The flow trends found on the shroud side of planes 1 and 4
Show that flow differences other than the primary flow exists in that area
(Figure 43). This flow characteristic may be connected all the way
thl‘Ough the impeller passage, possibly influenced by the reintroduction of
fluid in the slots in the shroud where the tabs secure the blade. This

region, as well as other regions need further investigation.

Pressure
. Side

Plane 2 223‘?“
1st Series

 

Pressure

Plane 4 ‘1 Suction : I. . . ... " Side

Side
2nd Series

I L'l-U‘JJ
I) 567186

 

Figure 41 . Comparison of planes 2 and 4 (both located at impeller exit) at 0.8 speed ratio

85

 

Plane 2
1st Series 222,?“

L_ _ (I 458917.

 

Suction ; i ‘ . l' - : ' ' ‘ Pressure
Plane 4 Side 1 7 : ; ‘ 7 Side
2nd Series ‘

1) 561004

 

Figure 42. Comparison of planes 2 and 4 (both located at impeller exit) at 0.4 speed ratio

86

 

Shroud
Side

' i I I I V 7 ' n
Plane 1 » 0335"”
, Suction ”a" "

0.4 Side

Plane 4 —)

Plane 1

 

Suction
Side

Pressure
Side

Plane 4 _ : g . . .

 

Figure 43. Comparison of planes 1 and 4 at 0.4 speed ratio

87

5.5 Steady State Condition Velocity Analysis

Another important task of this study was to provide useful
information for CFD code development and to compare with CFD results.
CFD codes have been developed to predict the internal flow
characteristics in the automotive torque converter [8-12]. In general,
CFD codes are calculating the flow velocities regardless of the turbine
blade locations. The flow velocities from the CFD code are calculated
when the impeller and turbine rotation speeds are fixed. This type of
velocity profile can be defined as a steady state condition velocity
analysis of the internal flow of a torque converter.

In the animated series the measured velocity clearly shows the
velocity profile at any point in the converter, which varies with respect to
the location of the turbine blades. The velocity has a periodic oscillation
that depends on the relative locations between the impeller and turbine
blades. Because of this periodic oscillation nature of the internal flow,
the measured velocity cannot be compared with that of the CFD results.
For this purpose the steady state condition velocities were calculated by
means of averaging the measured velocities over the complete range of the
turbine rotation. Therefore, the steady state condition velocities do not
depend on the relative locations between the impeller and turbine blades,
which is similar to velocity profiles from the CFD simulation. The

advantage is the average flow rate in meters per second can be estimated

88

and used to gain a perspective of the overall fluid flow measured. The
disadvantage is distinct flow characteristics, such as blade-passing effects
at and beyond the impeller planes are not clearly discernable. The
averaged steady state condition velocity profiles are shown in Figures 44-

61.

5.5.1 Plane 1: Comparison of Both Speed Ratios

For plane 1, which was measured mid-chord of the impeller
passage, two different flow fields were observed for both speed ratio
cases (Figures 44 and 45). The primary flow in the middle of the impeller
passage moves in one general direction (+2 and x) from the suction to the
pressure side and near the shroud show the flow direction changes to the —
z direction. The magnitude of the velocity vectors was smaller in the
shroud area as well, as compared to that of the rest of the plane. In the
0.8 case the average velocities were 8-lOm/s for the majority of the plane,
whereas the shroud side showed average velocities of 5-6m/s. For the 0.4
case the average velocities were 8-lOm/s with the vertical component (+z)
larger in magnitude than in the 0.8 speed ratio case. For the shroud

region the average velocities were 5-6m/s.

89

5.5.2 Plane 2: Comparison of Both Speed Ratios

For plane 2, which was measured at the exit region of the impeller
passage shows flow fields were generally uniform for both speed ratio
cases (Figures 46 and 47). The general the direction of the flow
magnitudes (+x direction) moved from the suction side to the pressure
side of the impeller. The magnitudes of the flow were smaller on the
suction side of the impeller passage than that of the pressure side for both
speed ratios. For the 0.8 case there are steeper velocity gradients along
the bottom of the plane (+x direction) with velocities approximately 7-
8m/s. The average velocities for the rest of the plane were 4-6m/s, with
the smaller velocity vector magnitudes on the suction side of the passage.
For the 0.4 speed ratio as the flow moved from the suction side to the
Pressure side of the impeller passage (+x direction), the flow magnitudes
Were not at a generally increasing angle (horizontal to vertical) as with
the 0.8 speed ratio case. Instead, the overall flow pattern moved up (+z
direction), then down (-z direction), in a type of s-curve, as it continued

toward the pressure side. Average velocities were 3-6m/s with the

s\owest velocities near the bottom of the plane.

5.5.3 Plane 3: Comparison of Both Speed Ratios

Plane 3 was measured 1mm into the turbine passage (Figures 48 and
49). The general the direction of the flow magnitudes moved in the +2
and x direction. For the 0.8 speed ratio the flow was generally uniform
with average velocities of 4-6m/s. For the 0.4 case there are steeper
velocity gradients along the bottom of the plane (+x direction) with
velocities approximately 7-9m/s. The average velocities for the rest of
the plane were 4-6m/s, with the smaller velocity vector magnitudes on the

suction side of the passage.

5.5.4 Plane 4: Comparison of Both Speed Ratios

Plane 4 was measured at the impeller exit region (Figures 50 and
51). The flow pattern showed general flow direction from the suction
side of the impeller passage to the pressure side. For the 0.8 speed ratio
the average velocity of the fluid flow was 2-5m/s with slower velocities
(purple) on the left side of the plane of l-2m/s. The flow magnitudes
increased (+2 and x direction) from the suction to the pressure side, with
the exception of along the bottom of the plane where the flow moved in
the +x direction. In the 0.4 speed ratio case the flow contained steeper
velocity gradients. Average velocities were 3-6m/s with higher velocities

(orange) of 9-lOm/s. For the upper half of the plane the fluid flow moved

91

across the impeller passage, where the vector angle was generally
decreasing (-z and x direction) as the flow continued to the pressure side.
For the bottom half of the plane the steeper velocity gradients moved
straight across (+x) or even downward (-z) as it approached the pressure

side.

5.5.5 Plane 5: Comparison of Both Speed Ratios

Plane 5 was measured 1mm from the impeller exit in the gap region
(Figures 52 and 53). Similar flow pattern, as in the impeller exit region
were also observed showing the general flow direction (+2 and x) from the
suction side of the impeller passage to the pressure side. For the 0.8
speed ratio the average velocity of the fluid flow was 2-4m/s with slower
velocities (purple) on the left side of the plane of 1-2m/s. The flow
magnitudes increased (+2 and x direction) from the suction to the pressure
side, with the exception of along the bottom of the plane where the flow
moved in the +x direction. Along the bottom of the plane there were
steeper velocity gradients (green) approximately 5-6m/s. In the 0.4 speed
ratio case the flow contained steeper velocity gradients in the bottom half
of the plane. Average velocities were 4-6m/s with higher velocities
(yellow and orange) of 7-lOm/s. For the upper half of the plane the fluid
flow moved across the impeller passage, where the vector angle was

generally straight or slightly decreasing (-z and x direction) as the flow

92

continued to the pressure side. The upper left-hand side of the plane

contained slower velocity gradients of approximately l-2m/s.

5.5.6 Plane 6: Comparison of Both Speed Ratios

This plane was measured 2mm inside the impeller passage from the
impeller exit (Figures 54 and 55). The direction of the flow measured was
generally uniform for both speed ratios and followed the impeller passage
(+y direction) toward the exit and turbine as expected. This toriodal flow
measurement had the highest velocities occurring near the center of the
impeller passage. There were slower than average velocities (green) of 5-
6m/s at the left and right upper portions of the plane for both speed ratios.
For the 0.8 speed ratio the average velocities were 7-8m/s, with some
steeper velocity gradients (orange) of 9-10m/s. The 0.4 speed ratio case
average velocities were 9-10m/s, with the highest velocities occurring at

the center of the passage.

5.5.7 Plane 7: Comparison of Both Speed Ratios

For plane 7, the measured location was 1mm inside of the impeller

passage from the exit (Figures 56 and 57). The features of the flow

pattern measured were generally uniform for both speed ratios, and moved

(+y direction) toward the impeller exit and the turbine as expected. The

93

toriodal flow showed the highest velocities occurring near the center of
the impeller passage. For the 0.8 speed ratio average velocities were 7-
lOm/s, with a line of slower velocities (green) along the top of the plane
of 5-6m/s. The 0.4 speed ratio also had velocities of 7-10m/s and only a
couple of slower velocities (green) occurring at the upper left-hand corner

of the plane at 5-6m/s.

5.5.8 Plane 8: Comparison of Both Speed Ratios

Plane 8 was measured at the impeller exit (Figures 58 and 59).
There is a general uniform flow from the impeller exit out into the gap
(+y direction) for both speed ratios. The toriodal flow showed the highest
velocities occurring near the center of the impeller passage. For the 0.8
speed ratio average velocities were 7-10m/s, including a line of slower
velocities (green) along the top of the plane of 4-6m/s. The 0.4 speed
ratio also had velocities of 7-10m/s and slower velocities (green)
occurring at top of the plane at l-3m/s. The slower velocities at the top
of the plane for both speed ratios shows fluid flow leaving the impeller
and continuing above the turbine, which negatively affects converter

efficiency.

94

5.5.9 Plane 9: Comparison of Both Speed Ratios

Plane 9 was measured 1mm from the impeller exit out into the gap
region (Figures 60 and 61). As with the previous side window planes the
flow is generally uniform across the measured area with toroidal flow (+y
direction) out into the gap region for both speed ratios. For the 0.8 speed
ratio the average flow velocity was 5-6m/s. The 0.4 speed ratio has
velocities of 5-6m/s, including higher velocity flow (yellow and orange)
in the middle of the plane of 7-9m/s. There was a band of slower
velocities along the top of the plane of l-3m/s, for both speed ratios.

This has a negative affect of overall converter efficiency.

95

Shroud

f; Sucfion
fide

\ .

It.

OF—Nw-‘U‘O‘NWWH

 

Green Arrow:
Measured Perspective
White Arrow:
View Perspective

 

Figure 44. Steady state condition velocities plane 1 at 0.8 speed ratio

96

Shroud

/
/
,/ Pressure
Side

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 45. Steady state condition velocities plane 1 at 0.4 speed ratio

97

 

t. 1.

é»..........p....

ovrearwoe1w..:¢
refriudrraspadi
rrrrtrrgiraaara
rrrirrrrrirrrw.‘
rf‘rrrrrraaw.

 

Green Arrow

6
.w
t
c
e
p
s
r
e
P
w
r
u
S
a
W

e
.w
a

6
mp
am
MP
t
mm
wv

 

Figure 46. Steady state condition velocities plane 2 at 0.8 speed ratio

98

 

S
/

/, 0
// 19975513210
l/fi/ ,.
1
i i.
ll”?
52...... /
'/

«...—...... ....J

x

.......:..._

...-6466.65;

r~ngmmm~¢m..

..mccmmm§.§mm.d.ea
pug-.mmmkarL-rr
..._...mmes.6...~:._.
mememx‘Umnm6tesege
rummu..._s6._.6~..\.6... r
coach... 26.16.45.
6....hs.. Rises:

_. s
w c.

.c-au

mm.
a...

 

Green Arrow
Measured Pers

 

e
...V.
t

m

01

White Arrow
View Perspective

Figure 47. Steady state condition velocities plane 2 at 0.4 Speed ratio

99

 

 

Pressure

{art {.74. r f
n.‘ ... f...’.4s r A. 7 1
3.? gal/12.} ..r r 1.1.
I r 1 5.9.1.! If. a ..

.i..a/.I...r././,4. ...). c

if 4. 0.1/1.1! II 1. .., ..

.... a 0.1.11.1. ell/.1 p n .1

. . ... .
f as... If 6.61. III ... r .0.

.A .1 2.4.;

..rnufwflé.‘.fié(on

2.15.94 d P .r

.7! 0491/; 9.451,.

.. .... r a .p ...,

 

.1 1.7.6....» 91 ....

l I ..1 .

Green Arrow

8
.w
t
C
e
p
m
e
P
w
m
S
a
e
M

W
W
W.

 

e
.w
t
c
e
p
S
r.
e
P
w
.w
V

Figure 48. Steady state condition velocities plane 3 at 0.8 speed ratio

 

Suction
Side ' /

Side

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 49. Steady state condition velocities plane 3 at 0.4 speed ratio

101

 

Suction : . . - i . .i Z
Side . ' ~ /

/
//
7/

Pressure
Side

/,/
’/

(‘7

O

o—mws—mmwmor—t

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 50. Steady state condition velocities plane 4 at 0.8 speed ratio

102

 

 

/

7917711?!

Pressure
Side

,..,,,,.,..
Ialrg'rn

fi
1,.
Q.
:5
D
c .
12.,
t-
‘5,
\
Q.
‘ .

inset/g7?

O

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 51. Steady state condition velocities plane 4 at 0.4 speed ratio

103

 

// Pressure
/ Side

O

. .
a—mws—wmwmecu-fi

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 52. Steady state condition velocities plane 5 at 0.8 speed ratio

 

I Q .
or Q
5 '1
_ 6.
r c
D. C"
. C‘w
6. .

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 53. Steady state condition velocities plane 5 at 0.4 speed ratio

105

 

Pressure
Side

map-0
D

ar-‘reu—hmmu

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 54. Steady state condition velocities plane 6 at 0.8 speed ratio

Pressure
Side

0

.7
D—‘NU—L‘Ofiflfihfil—l

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 55. Steady state condition velocities plane 6 at 0.4 speed ratio

107

Pressure
fide

Sucfion

fide

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 56. Steady state condition velocities plane 7 at 0.8 speed ratio

108

Pressure
Side

Suction
Side

I I
a

OPNW~U1¢NODwH

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 57. Steady state condition velocities plane 7 at 0.4 speed ratio

Pressure
Side

9

Db—‘Mw-E-U‘O‘NGDGH

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

   

Figure 58. Steady state condition velocities plane 8 at 0.8 speed ratio

110

Pressure
Side

6

D—‘Nu-i-U‘O‘Nmufit-l

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 59. Steady state condition velocities plane 8 at 0.4 speed ratio

111

Pressure
Side

O

1

GHmu—U‘JNNIGJQH

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 60. Steady state condition velocities plane 9 at 0.8 speed ratio

112

Pressure
Side

D

D—Nui-U‘ONCQQH

 

Green Arrow:
Measured Perspective
White Arrow:

View Perspective

 

Figure 61. Steady state condition velocities plane 9 at 0.4 speed ratio

113

5.6 Mass Flow Rates

The mass flow rates at certain measured planes, which can also be
used for CFD analysis was calculated with measured velocities. The flow
through a passage of the impeller, turbine and stator was preferred to
determine mass flow rates in the primary flow direction. Not all of the
planes measured fit these criteria. The vertical component (+z) for plane
1 was useful for the primary flow mid-chord of the impeller passage
measured in the front window. Planes 6-9 also provided primary flow
characteristics measured in the side window. Planes 2-5 cannot be
utilized since the components measured in these locations were not the
flow direction component needed. The code Massflow was used to
produce an output file, which are then plotted in turbine angle versus
mass flow rate format (Figures 62-67).

The mass flow rates were used to understand the total flow
variations with respect to the relative turbine locations, the fluid flow
through the slots in the shroud and the region between the turbine and
housing. The region in plane 9 where fluid flow loss was observed
travelling out of the gap region was calculated based on the measurement
velocities in plane 1 and the gap region. For the 0.8 speed ratio the flow
region between the turbine and housing was estimated at 4.7%, while the
0.4 speed ratio was found to be 2.8%. Further studies need to be
performed in order to understand the detailed flow characteristics in this

region. Since the location of the flow was calculated 1mm outside of the

114

impeller exit in the gap region. Since the location of the flow was
calculated 1mm outside of the impeller exit in the gap region, a plane
should be measured closer to the turbine and calculations performed to
see if the flow losses estimated are actually leaving the region, or

reentering the turbine.

5.6.1 Mass Flow Rates: Plane 1

The mass flow rates versus turbine angle at plane 1 (measured at
mid-chord of impeller passage) is shown in Figure 62. For the 0.8 speed
ratio the flow ranges from 0.255-O.3kg/s. The oscillations are fairly
constant per turbine angle. The 0.4 speed ratio case shows similar overall
flow rate oscillations with slightly lower amplitudes of 0.475-0.523kg/s.
For both speed ratio cases the flow rate increases until the turbine
position is approximately 10 degrees. The flow rates are fairly uniform
during the turbine positions between 10 and 18 degrees, then the flow rate

decreases.

5.6.2 Mass Flow Rates: Plane 6

Plane 6 (measured 2mm inside of impeller passage from the

exit) had mass flow rates of 0.255-0.3kg/s for the 0.8 speed ratio and

0.474-0.523kg/s. The oscillations of the 0.4 case showed lower

115

amplitudes than with the 0.8 speed ratio (Figure 63). For both speed ratio
cases the flow rate increases until the turbine position is approximately 10
degrees. The flow rates are fairly uniform during the turbine positions
between 12 and 20 degrees, then the flow rate decreases. For the 0.4
speed ratio the flow rate decreases between 2 and 8 degrees, then

increasing, while the 0.8 speed ratio is more uniform (less sinusoidal).

5.6.3 Mass Flow Rates: Plane 7

For plane 7 (measured 1mm inside of impeller passage from
the exit) the overall flow rate amplitudes were quite similar for both
speed ratios (Figure 64). Flow rates were 0.12-0.158kg/s for the 0.8
speed ratio and 0.194-O.215kg/s for the 0.4 speed ratio. The flow rates
decrease between 2 and 8 degrees, are fairly uniform during the turbine
positions between 8 and 12 degrees, then increases to a peak between 17

and 22 degrees, then the flow rate decreases.

5.6.4 Mass Flow Rates: Plane 8

The blade-passing effects by the turbine are evident in plane 8
(measured at the impeller exit). This shows that the turbine blade-passing
effects are influencing the fluid flow back through the entire gap region to

the impeller exit (Figure 65). The overall flow rate amplitudes were

116

lower for both speed ratios as compared to Figures 63 and 64. Flow rates
were 0.372-0.425kg/s for 0.8 speed ratio and 0.365-0.45kg/s for the 0.4
speed ratio. The flow rates decrease between 2 and 10 degrees, are fairly
uniform during the turbine positions between 10 and 15 degrees, then
increases to a peak between 17 and 22 degrees, then the flow rate

decreases.

5.6.5 Mass Flow Rates: Plane 9

The blade-passing effects by the turbine are also evident in
plane 9 (measured 1mm from the impeller exit into the gap region).
Amplitudes were even lower here than for the other planes (Figure 66).
Flow rates were 0.166-O.213kg/s for 0.8 speed ratio and 0.33-0.378kg/s
for the 0.4 speed ratio. The flow rates for the 0.8 speed ratio decrease
between 2 and 5 degrees, are fairly uniform during the turbine positions
between 5 and 10 degrees, then increases to a peak between 20 and 25
degrees, then the flow rate decreases. For the 0.4 speed ratio the flow
rate decreases between 2 and 10 degrees, then fairly uniform between 10
and 15 degrees, then increasing.

Since the area measured from plane to plane is different the mass
flow rate results are particular to each plane (Figure 67). This is true
even when comparing data in the same plane at both speed ratios the

results are subject to different areas due to different data rates. In most

117

cases the planes can be compared. Plane 9 had significantly different data
rates from one speed ratio to another (considerably more for the 0.4 case
than the 0.8 case). Plane 1 has similar areas at both speed ratios due to
close data rates and therefore can be compared to each other. Plane 8 also
can be compared to both speed ratios for the same reason. Both speed
ratios compared gives an indication of the different flow rates in the same

plane.

118

 

 

m1flowparllflnaanflalllqnadrllo

 

 

 

Figure 62. Mass flow rate plane 1 atboth speed ratios

119

 

plunelluwparlmmuapaadrllo

 

 

 

 

Figure 63. Massflowrateplane6atbothspeedratios

120

pl-IaTMparWn-Idauapaadnlo

 

 

Figure 64. Mass flowrate plane 7 atboth speedratios

121

 

M (U31
9

 

 

 

99.09
0808

MM
o

 

 

 

'° .0 '° .0 a
8 0 8 21 .7.
O

 

Figure 65. Mass flow rate plane 8 at both speed ratios

122

meflowparmmanflauapaadrdo

021. ~ 7 - , r (7 .. 577.457‘~# f — , --. - .—

02

 

«A»

i 0.18

 

 

3 A /J
\Wf

 

 

 

0.10
0.15 . f . fl .
o s 10 is 20 25
pollen
planafllowpartuhlmarflaOAapaadratlo
0.30.. ,_ -- -W- 1-- _ - , . _.__ +_.-..____.-______. -1

 

'\ /\/
\ /

... \ N 1

0.&

/
K

 

 

Figure 66. Mass flow rate plane 9 at both speed ratios

123

 

 

 

 

“Ammo
manor-Io

 

 

 

Figure 66. Mass flow rate compan'son planes 1, 6-9 both speed ratios

124

CHAPTER 6

SUMMARY AND CONCLUSIONS

In this study the characteristics of the fluid flow were examined in
an automotive torque converter. The goal was to gain an improved
understanding of the complexity of the fluid flow phenomena inside a
torque converter, to provide database for numerical simulation and to
identify the important region locations that will benefit from additional
studies. Measurements were made in a total of 9 planes. Three planes
were examined in the impeller passage, three at the impeller exit, two in
the gap region, and one in the turbine passage. Experiments were
conducted at two different speed ratios. Animations, steady state
condition velocities, and mass flow rates were the results showing that
LDV system mapping can assist in providing the design information

needed for developing improved torque converters.

The following conclusions can be made from this study.

1 LDV is capable for measuring internal flows in production torque
converter. The MSU torque converter test stand is capable of
providing constant temperatures, pressures, torques and flow rates

needed to simulate realistic torque converter conditions.

125

2 The software animation developed by MSUERL is a powerful tool,
which is used to visualize and analyze the velocity profiles. The
animation results validate the post-processing algorithm (TQANLIZR)
by showing that the measured velocity profile results in the correct
locations without manual placement.

3 Turbine blade-passing effects were clearly observed in the gap region
between the impeller and turbine from both the front and side window
data series. This included a time delay of the turbine blade-passing
effects as the wave generated during turbine blade passing propagated
back toward the impeller exit.

4 The impeller exit area was found to exhibit high spatial and temporal
velocity gradients for both speed ratios. This is evident from the data
results in planes 2 and 4 that were measured at the impeller exit area.

5 In plane 1, at mid-chord of the impeller passage, the vertical
component of flow velocities are generally uniform across the passage
for the 0.4 speed ratio as compared with the 0.8 speed ratio. Also in
the impeller passage, two distinct flow trends were observed, in the
passage, and near the shroud, in both the 0.4 and 0.8 speed ratios.

6 In the turbine passage, higher velocities were observed on the pressure
side then on the suction side of the turbine.

7 The steady state condition velocities generated showed overall flow

trends of entire series in one frame.

126

 

 

8 Fluid flow through the gap between the turbine and converter housing
was observed, which negatively affects overall converter efficiency.
Flow into the region between the turbine and shell was calculated
using mass flow rates to be 4.7% for the 0.8 speed ratio and 2.8% for
the 0.4 speed ratio. Fluid velocity decreases as it enters the gap region

from the impeller toward the turbine.

127

CHAPTER 7

RECOMMENDATIONS

One area that needs further fluid flow studies done is in the
impeller passage along the shroud side. Flow separation here may result
may result in significant loss in efficiency. There is evidence of flow
patterns in the impeller passage that indicate a region of flow different
than the rest of the plane measured (Planes 1 and 4). Additional planes
should be measured from below plane 1 to beyond plane 4. This would
provide volumetric results to help determine the full nature of these flow
phenomena. Other locations such as near the shroud surface and the
region around the slots that hold the blades to the shroud, need to be
studied. Small perturbations of flow in these areas can result in
significant turbulence generation downstream of these slots and thus,

overall converter efficiency.

The following are recommendations for future studies.
1 Further measuring and mapping needed for additional understanding of
fluid flow. An example being planes measured on both sides of

impeller exit.

128

2 Include flow visualization in finding optimum locations for future
LDV studies. This would help determine pertinent fluid flow locations
that can be studied using LDV.

3 Additional regions such as the gap region, turbine, and the region
above the turbine should be investigated. Through further studies
optimal distances between both the gap region (between the impeller
and turbine) and the region between the turbine and shell can be
determined, which can improve the overall converter efficiency.

4 Windows to be fabricated and installed in the turbine blade and on the
surface of the turbine side of the torque converter allowing for the

study of fluid flows more in depth in that area.

129

LIST OF REFERENCES

. Lee, K., Novak, M. and Schock, H., “3-D LDV Measurement of In—
Cylinder Air Flow in a 3.5L Four-Valve SI Engine,” SAE Paper No.
950648, 1995.

. Hwang, K., Lee, K., Mueller, 1., Stuecken, T. and Schock, H.,
“Dynamic Flow Study in a Catalytic Converter Using Laser Doppler
Velocimetry and High Speed Flow Visualization,” SAE Paper No.
950786, 1995.

. Hascher, H., Kasser, J., Novak, M., Lee, K. and Schock, H., “An
Evaluation of Turbulent Kinetic Energy for the In-Cylinder Flow of a
Four-Valve 3.5L SI Engine Using 3-D LDV Measurements,” SAE Paper
No. 970793, 1997.

. Kasser, J., Hascher, H., Novak, M., Lee, K. and Schock, H., “Tumble
and Swirl Quantification within a Motored Four-Valve SI Engine
Cylinder Based on 3-D LDV Measurements,” SAE Paper No. 970792,
1997.

. By, RR. and Maloney, J.E., “Technology Needs for the Automotive
Torque Converter-Part1: Internal Flow, Blade Design, and
performance,” SAE Paper No. 880482, 1988.

. Fister, W. and Adrian, F. W., “Experimental Researches of Flow in
Hydrodynamic Torque Converters,” Proceeding of the 7‘” Conference

of Fluid Machinery in Budapest, Hungary, Vol. 1,pp. 210-224, 1983.

. By, R.R. and Lakshminarayna, B., “Measurement and Analysis of
Static Pressure Field in a Torque Converter Pump,” Journal of Fluid
Engineering, Vol. 117, pp. 109-115, March 1995.

. Lakshminarayna, B. and von Backstrom, T.W., “ Perspective: Fluid
Dynamics and Performance of Automotive Torque Converters: An

Assessment,” ournal of Fluids Engineering, Vol. 118, pp. 665-676,
December 1996.

. Watanabe, H., Kurahashi, T. and Kojima, M., “Flow Visualization and
Measurements of Torque Converter Stator Blades Using a Laser Sheet
Lighting Method and a Laser Doppler Velocimeter,” SAE Paper No.
970680, 1997.

130

LIST OF REFERENCES (CONT.)

10.Bahr, H.M., Flack, R.D., By, R.R, and Zhang, J.J., “Laser Velocimeter
Measurements in the Stator of a Torque Converter,” SAE Paper No.
901769, 1990.

11.Fax received from G.R. Becraft, Ford Motor Company Torque
Converter Laboratory, Livonia, Michigan, 1992.

12.Kubo, M. and Ejiri, B., “A Loss Analysis Design Approach to
Improving Torque Converter Performance,” SAE Paper No. 981100,
1998.

13.Abe, K., Kondoh, T., Fukumura, K. and Kojima, M., “Three-
Dimensional Simulation of the Flow in a Torque Converter,” SAE
Paper No. 910800, 1991.

14.Ma, W., Luo, B., and Wu, 8., “The Research on Quasi-Three
Dimensional Flow Design of Hydrodynamic Torque Converter Blades,”
SAE Paper No. 912201, 1991.

15. Ejiri, E., “A New Approach to Developing a More Efficient Torque
Converter Stator,” SAE Paper No. 901765, 1990.

16. Schultz, H., Greim, R. and Volgmann, W., “Calculation of Three-
Dimensional Viscous Flow in Hydronamic Torque Converters,”
Transactions of the ASME, Vol. 118, pp. 578-579, July 1996.

17. Khalid, A., Legrand, J. and Robert, J.M., “Turbulent Flow Induced by
an Impeller in a Closed Toroidal Loop,” Journal of Fluids
Engineering, Vol. 118, pp. 677-684, December 1996.

18. Gruver, J.K., Flack, RD. and Brun, K., “Laser Velocimeter
Measurements on the Pump of an Automotive Torque Converter: Part
1- Average Measurements,” Transactions of the ASME, Vol. 118, pp.
562-569, July 1996.

19. Gruver, J.K., Flack, RD. and Brun, K., “Laser Velocimeter
Measurements on the Pump of an Automotive Torque Converter: Part
II- Unsteady Measurements,” Transactions of the ASME, Vol. 118, pp.
570-577, July 1996.

131

LIST OF REFERENCES (CONT.)

20. Numazawa, A., Ushijima, F., Fukumura, K. and Ishihara, T., “An

21.

Experimental Analysis of Fluid Flow in a Torque Converter,” SAE
Paper No. 830571, 1983.

McCarrick, D.W., “Experimental Investigation of the Flow Field in an
Automotive Torque Converter”, A Thesis, Michigan State University,
E. Lansing, Michigan, 1993.

132

MICHIGAN STATE UNIV. LIBRQRIES
lllllllllllWWI1Wllll111111111”1111111111
31293016880449