desensitization of over tip leakage in an axial turbine rotor by tip ... performed in...

The Pennsylvania State University

The Graduate School

Department of Aerospace Engineering

DESENSITIZATION OF OVER TIP LEAKAGE IN AN AXIAL TURBINE

ROTOR BY TIP SURFACE COOLANT INJECTION

A Thesis in

Aerospace Engineering

by

Nikhil Molahally Rao

© 2005 Nikhil Molahally Rao

Submitted in Partial Fulfillment of the Requirements

for the Degree of

Doctor of Philosophy

August 2005

ii

The thesis of Nikhil Molahally Rao was reviewed and approved* by the following:

Cengiz Camci Professor of Aerospace Engineering Thesis Advisor Chair of Committee

Dennis K. McLaughlin Professor of Aerospace Engineering

Lyle N. Long Professor of Aerospace Engineering

Savas Yavuzkurt Professor of Mechanical Engineering

Timothy F. Miller Senior Research Associate

George S. Lesieture Professor of Aerospace Engineering Head of the Department of Aerospace Engineering

*Signatures are on file in the Graduate School

iii

ABSTRACT

Mechanical energy extraction in axial flow turbine rotors occurs through a change

in angular momentum of the working fluid. The gap between the turbine rotor and the

stationary casing is referred to as the tip gap. High pressure turbine blades are typically

un-shrouded and pressure driven flow through the tip gap is termed as over tip leakage.

Over tip leakage reduces efficiency of the turbine stage and also causes thermal distress

to blade tip surfaces. The gap height typically increases over the operational life of a

turbine, leading to increased efficiency drop. The thermal load on the tip surface also

increases with increasing gap height and is exacerbated by the radial transport of high

temperature fluid found in the core of the combustor exit flow. Thus over tip leakage not

only decreases stage efficiency, but also constrains it by limiting the maximum cycle

temperature.

Reducing the sensitivity of turbine performance to the effects of the tip gap is

termed Tip Desensitization. An experimental investigation of tip desensitization through

coolant injection from a tip surface trench was conducted in a large scale, low speed,

rotating research turbine facility. Five out of twenty nine rotor blades, referred to as

cooled blades, are provided with coolant injection at four locations, at 61%, 71%, 81%,

and 91% blade tip axial chord length. At each of the first three locations the coolant jets

are directed towards the blade pressure-side, while coolant is exhausted radially at the last

location.

Qualitative information of tip surface flow was obtained through the

implementation of surface flow visualization using an oil and pigment mixture.

iv

Quantitative measurements of absolute total pressure were conducted in a plane located at

30% axial chord downstream of the rotor exit plane using a fast response total pressure

probe aligned to the absolute tip velocity vector. Time accurate, phase locked total

pressure was measured and averaged over 200 rotations. Total pressure defects due to

over tip leakage vortex, passage secondary flow, and blade wakes are clearly observed in

total pressure maps over the entire circumference and a range of radial locations. The

effect of coolant mass flow rate and injection location was investigated by coolant

injection from a single cooled blade with a gap height of 1.40% blade height. Coolant

mass flow rates in the range of 0.4% - 0.7% of turbine mass flow rate were investigated.

Coolant injection from all five cooled blades was also investigated. The effect of casing

endwall surface roughness was also studied, without coolant injection.

The sensitivity of total pressure defect, due to over tip leakage, to tip gap height is

reduced by both coolant injection and roughening of the casing surface. The total

pressure defect due to the large gap height of 1.40% blade height is reduced to levels

comparable to the defect due to a gap height of 0.72% blade height. The strong total

pressure gradient that characterizes the leakage vortex due to the gap height of 1.40%

blade height is considerably diminished by both coolant injection and roughening of the

casing surface. Coolant injection from 81% chord location is most effective in reducing

both the total pressure defect and the total pressure gradient. Casing surface roughness

significantly shifts the leakage vortex towards blade suction surface reducing its

interaction with the upper passage vortex. The benefit of casing surface roughness is

greater at larger gap heights.

v

Surface flows on the rotor and casing endwalls, blade tip surface for gap heights

ranging from 0.72% - 1.4% blade height, and blade tip surface with coolant injection at a

gap height of 1.4% blade height were visualized. Highly overturned flows are observed

on both endwalls, caused by secondary flow exiting the stator. Important rotor endwall

features such as the horseshoe vortex system, cross-passage secondary flow, and the path

of the pressure-side leg of the horseshoe vortex are identified. A distinct reattachment

line occurs on the tip surface of un-cooled blades, at approximately twice the gap height

from the pressure-side corner. Surface patterns clearly indicate the presence of a gap

vortex and chord-wise flow on the tip surface of un-cooled blades.

Coolant jets are turned by the gap flow, towards suction-side of the tip gap.

Cooling films form on the tip surface due to coolant jets from the first two locations

while intense mixing of coolant and main gap flow is observed around the 81% chord

location. The coolant jets prevent reattachment on the tip surface and also affect pressure-

side corner separation of the gap flow. Suction surface traces with coolant injection

indicate that roll up of gap flow in the neighboring passage is not continuous.

Coolant injection directed towards the blade pressure-side corner and artificially

introduced surface roughness could, individually, be effective techniques to operate high

pressure axial flow turbines at large tip gap heights without the associated penalty on

efficiency. The combination of coolant injection and artificially introduced surface

roughness could lead to greater loss reduction. The specific coolant scheme used has been

shown, qualitatively, to have thermal benefits in addition to reducing the total pressure

defect associated with the roll up of gap flow into a vortex.

vi

TABLE OF CONTENTS

LIST OF FIGURES ..................................................................................................... ix

LIST OF TABLES.......................................................................................................xvi

NOMENCLATURE ....................................................................................................xvii

ACKNOWLEDGEMENTS.........................................................................................xxi

Chapter 1 Introduction ................................................................................................1

1.1 Axial Turbine Blade Terminology..................................................................4 1.2 Turbine Rotor Passage Flow...........................................................................6 1.3 Over Tip Leakage (OTL)................................................................................7

1.3.1 Aerodynamic Losses ............................................................................11 1.3.2 Thermal Effects ....................................................................................13

1.4 Tip Desensitization .........................................................................................14 1.5 Film Cooling of Turbine Blade Tips ..............................................................17 1.6 Objectives of Current Research ......................................................................18

Chapter 2 Facility Description ....................................................................................20

2.1 The Axial Flow Turbine Research Facility (AFTRF) ....................................20 2.2 The Turbine Stage...........................................................................................23

2.2.1 Stage Aerodynamic Design ..................................................................26 2.2.2 AFTRF Tip Clearance Distribution......................................................29

2.3 Tip Cooled Blades ..........................................................................................31 2.4 Air Transfer System (ATS) ............................................................................36 2.5 Flow Visualization..........................................................................................36 2.6 Instrumentation ...............................................................................................38

2.6.1 Monitoring Instrumentation..................................................................39 2.6.2 Performance Measurement ...................................................................40

2.6.2.1 High-Frequency Total Pressure Probe .......................................41 2.6.3 Coolant Mass Flow Meter ....................................................................42 2.6.4 Rotating Frame Pressure Transducer....................................................43

2.7 Data Acquisition .............................................................................................44 2.8 Data Processing ..............................................................................................47 2.8 Operation ........................................................................................................49

Chapter 3 Rotor Flow Visualization ...........................................................................51

3.1 Rotor Hub Endwall Flow Visualization .........................................................53 3.1.1 Visualization with the Oil-Dot Technique ...........................................53 3.1.2 Visualization with the Oil-Film Technique..........................................57

vii

3.2 Blade Tip Surface Flow Visualization............................................................65 3.3 Rotor Casing Surface Flow.............................................................................72

3.3.1 Oil-Film Visualization of Rotor Casing Surface Flow.........................73

Chapter 4 Effect of Tip Gap Height on Over Tip Leakage.........................................79

4.1 Oil Film Based Tip Surface Flow Visualization.............................................79 4.1.1 Operation ..............................................................................................80 4.1.2 Tip Surface Flow Visualization; Large Gap Height (t/h = 1.4%) ........81

4.1.2.1 Blade Pressure Surface...............................................................81 4.1.2.2 Blade Tip Surface.......................................................................82 4.1.2.3 Heat Transfer Implications.........................................................87

4.1.3 Tip Surface Flow Visualization; Small Gap Height (t/h = 0.71%) .....88 4.1.4 Tip Surface Flow Visualization of Other Gap Heights ........................90 4.1.5 Suction Surface Traces from Oil Film Based Tip Surface Flow

Visualization...........................................................................................93 4.2 Total Pressure Measurements .........................................................................96

4.2.1 Baseline, No Injection ..........................................................................97 4.2.1.1 Region above 85% Blade Height ...............................................99 4.2.1.2 Region 75% - 85% Blade Height ...............................................100 4.2.1.3 Passage Core ..............................................................................100

4.2.2 Repeatability.........................................................................................102 4.2.3 Effect of the Tip Gap Height ................................................................104

Chapter 5 Effect of Coolant Mass Flow Rate on Over Tip Leakage ..........................112

5.1 Visualizing the Effect of Coolant Injection ....................................................113 5.1.1 Effect of Tip Trench .............................................................................114 5.1.2 Injection at Minj = 0.4% at Gap Height of t/h = 1.40% ........................116

5.1.2.1 Injection Prior to Start-up...........................................................116 5.1.2.2 Injection at Operating Speed ......................................................119

5.1.3 Visualization at Other Injection Rates..................................................122 5.1.4 Suction Surface Traces .........................................................................125 5.1.5 Heat Transfer Implication.....................................................................127

5.2 Total Pressure Measurement...........................................................................128 5.2.1 Comparison of Averaged Values..........................................................135

Chapter 6 Effect of Injection Location on Over Tip Leakage ....................................138

6.1 Injection from Individual Holes......................................................................139 6.2 Injection from Combination of Holes.............................................................146

Chapter 7 Multiple Cooled Blades and the Effect of Casing Surface Roughness ......154

7.1 Baseline...........................................................................................................155 7.2 Variation of Coolant Mass Flow Rate ............................................................157

viii

7.3 Effect of Casing Surface Roughness ..............................................................163 7.3.1 Smooth Casing Surface ........................................................................163 7.3.2 Fine Surface Roughness (220 Grit) ......................................................166 7.3.3 Coarse Surface Roughness (100 Grit) ..................................................168

7.4 Comparison of the Averaged Total Pressure Coefficient ...............................169

Chapter 8 Summary and Conclusions.........................................................................173

8.1 Summary.........................................................................................................174 8.1.1 Surface Flow Visualization ..................................................................174 8.1.2 Total Pressure Measurement ................................................................177

8.2 Conclusions.....................................................................................................181 8.3 Recommendations for Future Work ...............................................................183

Bibliography ................................................................................................................185

Appendix A Total Pressure Probe Characteristics......................................................190

A.1 Angular Sensitivity ........................................................................................190 A.2 Frequency Spectrum of Flow Field ...............................................................191 A.3 Uncertainty Analysis......................................................................................193

Appendix B AFTRF Tip Clearance Distribution........................................................196

Appendix C Total Pressure Coefficient Contour Map................................................200

ix

LIST OF FIGURES

Figure 1.1: Axial Turbine Rotor Passage.....................................................................5

Figure 1.2: Design Pressure Distribution on Turbine Blade........................................5

Figure 1.3: Turbine Rotor Passage Flow, From Yamamoto [4]. .................................9

Figure 1.4: Tip Gap Flow Conceptual Model From Bindon [7]..................................9

Figure 2.1: Schematic of the Axial Flow Turbine Research Facility. .........................21

Figure 2.2: Turbine Casing Frame For Measurement Windows. ................................21

Figure 2.3: AFTRF Instrumentation Window. ............................................................24

Figure 2.4: Blade Velocity Triangles...........................................................................29

Figure 2.5: AFTRF Tip Clearance Distribution...........................................................31

Figure 2.6: Cooled Blade Tip Surface. ........................................................................33

Figure 2.7: Cooled Blade Tip Coolant Injection Arrangement....................................34

Figure 2.8: Cross-Sectional View of Cooled Blade.....................................................35

Figure 2.9: Air Transfer System (ATS). ......................................................................37

Figure 2.10: Schematic of Monitoring Instrumentation. .............................................38

Figure 2.11: Schematic of Performance Instrumentation. ...........................................39

Figure 2.12: Total Pressure Probe Static Calibration...................................................42

Figure 2.13: Coolant Discharge Measurement Orifice Calibration. ............................44

Figure 2.14: Coolant Mass Flow (Orifice) Meter Calibration Minj vs. VDC...............45

Figure 3.1: Rotor Endwall Surface Flow Visualization Using the Oil-Dot Technique. ............................................................................................................55

Figure 3.2: Rotor Endwall Static Pressure Distribution, From Xiao [3]. ....................56

Figure 3.3: Oil Film On Rotor Endwall Before Test. ..................................................57

Figure 3.4: Near Leading Edge Surface Flow Features on the Rotor Endwall Visualized Using the Oil-Film Technique............................................................59

x

Figure 3.5: Blade Passage Surface Flow Patterns on the Rotor Endwall Visualized Using the Oil-Film Technique. .............................................................................61

Figure 3.6: Near Trailing Edge Surface Flow Patterns on the Rotor Endwall Visualized Using the Oil-Film Technique............................................................63

Figure 3.7: Blade Suction Surface Trace Formed During Rotor Endwall Surface Flow Visualization................................................................................................64

Figure 3.8: Oil Dots on Blade (B21) Pressure Surface Before and After Test Run. ...66

Figure 3.9: Tip Surface Flow Visualization (t/h = 1.40%) by Oil Dots Applied Near Blade Tip......................................................................................................67

Figure 3.10: Near Leading Edge Detail of Tip Surface Flow Visualization Using the Oil-Dot Technique (t/h = 1.40%)....................................................................69

Figure 3.11: Tip Surface Flow Visualization (t/h = 1.40%) Using the Oil-Dot Technique; Oil Applied Near Blade Root. ...........................................................70

Figure 3.12: Tip Surface Flow Visualization (t/h = 0.71%) Using the Oil-Dot Technique. ............................................................................................................71

Figure 3.13: Casing Surface Flow Visualization During Turbine Operation Using the Oil-Dots Technique.........................................................................................75

Figure 3.14: Rotor Casing Surface Flow Visualization Using the Oil-Dot Technique. ............................................................................................................76

Figure 3.15: Rotor Casing Surface Flow Visualization Using the Oil-Film Technique; Before and After Test. .......................................................................77

Figure 3.16: Rotor Casing Endwall Surface Flow Visualization Using the Oil-Film Technique.....................................................................................................78

Figure 4.1: Oil Film on Blade Pressure Surface Before and After Test. .....................82

Figure 4.2: Surface Flow Patterns on Tip Surface of Blade (B21) With a Gap Height of t/h = 1.40%, Visualized Using the Oil-Film Technique.......................86

Figure 4.3: Surface Flow Patterns on Front Half of Tip Surface of Blade B21, Using the Oil-Film Technique. .............................................................................86

Figure 4.4: Surface Flow Patterns on Rear Half of Tip Surface of Blade B21............87

Figure 4.5: Surface Flow Patterns on Tip Surface of Blade (B7) With a Gap Height of t/h = 0.71%, Visualized Using the Oil-Film Technique.......................90

xi

Figure 4.6: Surface Flow Patterns on the Tip Surface of Blade (B2) With a Gap Height of t/h = 0.81%, Visualized Using the Oil-Film Technique.......................92

Figure 4.7: Surface Flow Patterns on Tip Surface of Blade (B21 at reduced gap height) With a Gap Height of t/h = 1.2%, Visualized Using the Oil-Film Technique. ............................................................................................................92

Figure 4.8: Influence of Gap Height on the Location of the Visualized Reattachment Line on the Blade Tip Surface. ......................................................93

Figure 4.9: Suction Surface Traces Formed During Oil-Film Based Tip Surface Flow Visualization; Gap Heights, t/h = 1.4% and t/h = 1.2%. .............................95

Figure 4.10: Suction Surface Traces Formed During Oil-Film Based Tip Surface Flow Visualization; Gap Heights, t/h = 0.81% and t/h = 0.72%. .........................95

Figure 4.11: Effect of the Tip Gap Height on Measurements From Suction Surface Trace Formed During Oil-Film Based Visualization of Tip Surface Flow. .....................................................................................................................96

Figure 4.12: Total Pressure Coefficient Contours With No Coolant Injection (Base1)..................................................................................................................98

Figure 4.13: Secondary Flow Vectors At Rotor Exit From LDA Measurements by Ristic et al [54]......................................................................................................101

Figure 4.14: Total Pressure Coefficient Contours With No Coolant Injection (Base3)..................................................................................................................103

Figure 4.15: Radial Distribution of the Rotor Averaged Total Pressure Coefficient; Baseline Repeatability. .....................................................................104

Figure 4.16: Repeatability of Wake Profiles at r = 0.96h With No Coolant Injection. ...............................................................................................................107

Figure 4.17: Repeatability of Wake Profiles at r = 0.57h With No Coolant Injection. ...............................................................................................................107

Figure 4.18: Repeatability of the Passage Averaged Coefficient For Cooled Blade B21........................................................................................................................108

Figure 4.19: Total Pressure Coefficient Contours With Tip Gap Height of Cooled Blade B21 Reduced to t/h = 0.72%. .....................................................................108

Figure 4.20: Effect of Reducing the Tip Gap Height of Blade B21 On the Wake Profile at r = 0.96h. ...............................................................................................109

xii

Figure 4.21: Effect of Reducing the Tip Gap Height of Blade B21 On the Wake Profile at r = 0.57h. ...............................................................................................109

Figure 4.22: Effect of the Tip Gap Height On the Passage Averaged Coefficient of Cooled Blade B21.................................................................................................110

Figure 4.23: A Comparison of the Passage Averaged Coefficient Distribution For Blade B7 and Blade B21.......................................................................................110

Figure 4.24: Variation in the Area Averaged Total Pressure Coefficient with Tip Gap Height. (Area = 20% span*1 passage). .........................................................111

Figure 5.1: Surface Flow Visualization of the Effect of Tip Trench on Cooled Blade B21. ............................................................................................................115

Figure 5.2: Surface Flow Visualization of the Effect of Coolant Injection at Minj = 0.4% From Cooled Blade B21; Injection While Turbine at Rest. ........................117

Figure 5.3: Surface Flow Visualization of the Effect of Coolant Injection at Minj = 0.4% From Cooled Blade B21; Injection While Turbine at Operating Speed.....................................................................................................................120

Figure 5.4: Surface Flow Visualization of the Effect of Coolant Injection at Minj = 0.5% From Cooled Blade B21; Injection While Turbine at Rest......................123



Figure 5.7: Suction Surface Traces for Minj = 0.4% and Minj = 0.5%. ........................126

Figure 5.8: Suction Surface Traces for Minj = 0.6% and Minj = 0.7%. ........................127

Figure 5.9: Total Pressure Coefficient Contours with Coolant Injection at Minj = 0.41%. ...................................................................................................................129

Figure 5.10: Wake Profile at r = 0.96h, Without and With Coolant Injection at Minj = 0.41% and Minj = 0.52%.............................................................................130

Figure 5.11: Total Pressure Coefficient Contours With Coolant Injection at Minj = 0.52%. ...................................................................................................................131


xiii

Figure 5.13: Wake Profile at r = 0.96h, Without and With Coolant Injection at Minj = 0.63% and Minj = 0.72%.............................................................................133


Figure 5.15: Wake Profile at r = 0.57h, Without and With Coolant Injection.............134

Figure 5.16: Effect of Coolant Injection On the Passage Averaged Coefficient of Cooled Blade B21.................................................................................................136

Figure 5.17: Area Averaged Coefficient For Blade B21 With Coolant Injection at a Tip Gap Height of t/h = 1.40%. .........................................................................137

Figure 6.1: Total Pressure Coefficient Contours for Coolant Injection From Blade B21 and Location H1 at 61% Cax. ........................................................................140




Figure 6.5: Effect of Injection Hole Location on the Passage Averaged Coefficient of Cooled Blade B21. ...........................................................................................145

Figure 6.6: Effect of Injection Location on the Wake Profile at r = 0.96h..................145

Figure 6.7: Total Pressure Coefficient Contours for Coolant Injection From Blade B21 and Locations H1+H2. ..................................................................................147



Figure 6.10: Total Pressure Coefficient Contours for Coolant Injection From Blade B21 and Locations H1+H2+H3..................................................................149

Figure 6.11: Total Pressure Coefficient Contours for Coolant Injection From Blade B21; Full Injection......................................................................................150

xiv

Figure 6.12: Effect of Injection Location Combinations on the Passage Averaged Coefficient of Cooled Blade B21. ........................................................................151

Figure 6.13: Rotor Averaged Coefficient With Combined Injection. .........................152

Figure 6.14: Effect of Injection Location Combinations on the Wake Profile at r = 0.96h. ....................................................................................................................153

Figure 7.1: Total Pressure Coefficient With Multiple Cooled Blades; No Coolant Injection (Baseline), Minj = 0.0%..........................................................................156

Figure 7.2: Wake Profile at r = 0.96h Comparing Baseline Distributions of Multiple Cooled Blade and Single Cooled Blade (B21). .....................................157

Figure 7.3: Total Pressure Coefficient With Multiple Cooled Blades; Minj = 0.43%..158

Figure 7.4: Total Pressure Coefficient With Multiple Cooled Blades; Minj = 0.62%. ...................................................................................................................160

Figure 7.5: Total Pressure Coefficient With Multiple Cooled Blades; Minj = 0.72%. ...................................................................................................................160

Figure 7.6: Wake Profiles at r = 0.96h for Multiple Blade Coolant Injection. ............162

Figure 7.7: Passage Averaged Coefficient Comparison for Multiple Blade Coolant Injection. ...............................................................................................................162

Figure 7.8: Total Pressure Coefficient With a Smooth Plastic Layer On the Casing Inner Surface.........................................................................................................164

Figure 7.9: Wake Profiles at r = 0.96h Comparing the Influence of Casing Surface Roughness.............................................................................................................165

Figure 7.10: Wake Profiles at r = 0.57h Comparing the Influence of Casing Surface Roughness................................................................................................166

Figure 7.11: Total Pressure Coefficient With Fine Sandpaper (220 Grit) On the Casing Inner Surface.............................................................................................168

Figure 7.12: Total Pressure Coefficient With Coarse Sandpaper (100 Grit) On the Casing Inner Surface.............................................................................................170

Figure 7.13: Passage Averaged Coefficient For Blade B21 With Different Casing Roughness Treatments..........................................................................................171

Figure 7.14: Rotor Averaged Coefficient With Different Casing Roughness Treatments. ...........................................................................................................172

xv

Figure A.1: Probe Response to Incidence. (Squares denote rotor averaged Cpt and circles denote passage averaged Cpt.). .................................................................192

Figure A.2: Frequency Spectrum of Rotor Exit Flow Near Rotor Tip. .......................193

Figure B.1: Clearance Gap Variation Along Blade Axial Chord Length For TCL1...199

Figure C.1: Total Pressure Contour Map Of the Entire Rotor Exit Flow Field In the Measurement Plane.........................................................................................201

xvi

LIST OF TABLES

Table 2-1: AFTRF Overall Stage Design Characteristics............................................25

Table 2-2: Design Thermodynamic Parameters. .........................................................26

Table 2-3: AFTRF Design Coefficients.......................................................................26

Table 2-4: AFTRF Nozzle Design...............................................................................27

Table 2-5: AFTRF Rotor Design. ................................................................................28

Table 2-6: Radial Variation of Turbine Stage Design Parameters From 1-D Mean-line Analysis. ........................................................................................................28

Table 2-7: Estimate of Coolant Discharge Mass Flow Rate, From Pudupatty [43]. ..34

Table 2-8: Some Calculated Parameters From Low Speed DAS. ...............................46

Table 5-1: Test Matrix of Flow Visualization with Coolant Injection. .......................114

Table 6-1: Test Matrix for Effect of Injection Location. .............................................139

Table A-1: Uncertainty and Nominal Values in Measured Parameters.......................195

Table A-2: Uncertainty in Derived Parameters ...........................................................195

Table B-1: Gap Height Variation Along Blade Axial Chord For Clearance Distribution TCL1 ................................................................................................197

Table B-2: Gap Height Variation Along Blade Axial Chord For Clearance Distribution TCL3 ................................................................................................198

xvii

NOMENCLATURE

Cax Blade tip axial chord length, m.

Cp Specific heat at constant pressure, kJ/kg. K

Cps Static pressure coefficient.

Cpt Total pressure coefficient.

Cpt, R Rotor averaged total pressure coefficient.

Cpt,, A Area averaged total pressure coefficient.

Cpt,P Passage averaged total pressure coefficient.

d Differential operator.

h Blade height, m.

h0 Total or stagnation enthalpy, kJ/kg.

h1 Distance of upper boundary of suction surface oil trace from tip surface,

m.

h2 Width of suction surface oil trace, m.

i Circumferential index.

j Radial index.

lc Distance of reattachment line measured from pressure-side corner,

perpendicular to camber-line.

lx Distance of reattachment line measured from pressure-side corner,

perpendicular to blade axial chord.

Minj Coolant injection mass flow rate, [-]

N Rotor speed, rpm.

xviii

p local static pressure, Pa

P0, p0 Total or stagnation pressure, Pa.

qm Mean wheel speed based dynamic pressure, Pa.

r Radius, m.

r,θ,x Cylindrical coordinates.

T Temperature, K.

t Tip gap height, m.

T0 Total or stagnation temperature, K.

U Blade speed, m/sec.

V Absolute velocity vector, m/sec

v Specific volume, m3/kg.

Vax Axial component of absolute velocity vector, m/sec.

Vt Tangential component of absolute velocity vector, m/sec.

W Relative velocity vector, m/sec

x,y,z Cartesian coordinates.

GREEK SYMBOLS

∅ Diameter, m.

ψ Loading coefficient.

α Absolute flow angle.

β Relative flow angle.

φ Flow coefficient.

xix

γ Ratio of specific heats.

ηtt Total-Total efficiency.

ρ Density, kg/m3.

SUBSCRIPTS

1 Stage inlet.

2 Nozzle exit / rotor inlet

3 Rotor exit.

amb Ambient (pressure or temperature).

hub Rotor hub.

m mid-span

max Maximum.

tip Rotor tip.

ACRONYMS

B(#) Blade number.

C.L. Centerline

EES Engine equivalent speed.

EGV Exit guide vane.

H(#) Injection hole number.

HP High pressure.

Hp Pressure-side leg of horseshoe vortex.

Hs Suction-side leg of horseshoe vortex.

xx

L.E. Blade leading edge.

LP Low pressure.

N Nozzle.

NGV Nozzle guide vane.

OTL Over tip leakage.

PS Blade pressure surface or pressure-side

R Rotor.

SS Blade suction surface or suction-side.

T.E. Blade trailing edge.

TCL Tip clearance.

xxi

ACKNOWLEDGEMENTS

I would like to thank Dr. Cengiz Camci for his guidance and encouragement in

helping me achieve my academic goals. I would also like to thank Dr. McLaughlin, Dr.

Long, Dr. Yavuzkurt, and Dr. Miller for serving on my doctoral committee and for

providing me with valuable suggestions.

This achievement was also made possible by the financial support of the

Department of Aerospace Engineering, through teaching assistantships. I would also like

to thank the staff of the Department of Aerospace Engineering for their cheerful

disposition, patience in answering my many questions, and encouragement.

I would particularly like to thank Mark Catalano and Rick Auhl for their

assistance in overcoming the technical difficulties. The maintenance and modifications of

the research facility were promptly addressed by Harry Houtz. In addition to this, I

appreciate his interest in my research and his willingness to share his experience with me.

I would also like to acknowledge J. D. Miller for providing me with the necessary

resources and for assisting me in overcoming technical difficulties.

My family has been a constant source of inspiration and determination, both of

which were needed in attaining this goal. Thank you for your faith in my abilities. I am

also extremely grateful to be able to share this with numerous friends.

1

Chapter 1

Introduction

Transformation of the energy potential of fuels into useful forms of energy has

been an important aspect of civilization ever since the Industrial Revolution. The

invention of the steam engine provided great impetus to the Industrial Revolution as it

enabled the harnessing of thermal energy for mechanical drive applications. Similarly,

the gas turbine engine greatly accelerated developments in aviation. In addition to

propulsion of large and medium aircrafts, gas turbine engines are also used extensively in

electrical power generation and in marine applications. Thus improving the performance

of the gas turbine engine has great economic and environmental value.

Gas turbine engines belong to a class of rotating, energy conversion systems

called turbomachines. The energy transfer in turbomachines is achieved by a change in

the angular momentum of the working fluid, in a rotating blade row termed the rotor.

Energy is extracted from the working fluid in turbines, while energy is transferred to the

fluid in compressors and pumps. Turbine engine flows are contained within a duct or

engine casing. A gap is therefore necessary between the rotor and the stationary casing.

This gap is responsible for the phenomenon of Over Tip Leakage (OTL), which in turn

deteriorates the aerodynamic and thermal performance of the axial flow turbine

The necessary gap between the turbine rotor and stationary casing is referred to as

the tip gap or the tip clearance. In multi-stage turbines it is typical to have the blade tips

of low pressure (LP) stages covered by a shroud and the tip gap is formed between the

2

shroud outer surface and the casing. The shroud constrains the blades against large

displacements due to vibration, as the LP turbine blades tend to be longer than those in

high pressure (HP) stages. It is possible to shroud the LP turbine due to the relatively

lower gas temperature and rotational speed. On the other hand, HP turbines form the front

end of multi-stage turbines and extract energy from hot gas exiting the combustor, where

gas temperatures can be higher than metal melting point temperatures. Effective cooling

is necessary for proper functioning of the HP turbine stages and a shrouded turbine rotor

would necessitate cooling of the shroud. Cooling air, which is extracted from the engine

compressor section, is accounted for as a loss of cycle work. The addition of metal at the

maximum radius might also constrain the design performance due to increased blade root

stress and differential thermal growth during engine transients. Hence, shrouded HP

turbine designs are uncommon. The space between the blade tip profile and stationary

casing in HP turbine rotors is referred to as the tip gap and flow through this gap is

termed over tip leakage (OTL). In multi-stage turbines the annulus area also changes

significantly along the length of the turbine section. The annulus height of HP turbines is

small and a gap of the same physical dimension accounts for a larger part of the annulus

area. Thus, the influence of over tip leakage on the passage core flow is relatively

stronger in HP stages.

The total-total stage efficiency (ηtt) is defined in Equation 1-1. The stage

temperature and pressure are denoted by T and p respectively, while the subscripts “01”

and “03” refer to the stage inlet and exit conditions respectively. Efficiency increases

with either an increase in inlet temperature or a decrease in exit temperature. Efficiency

drop due to over tip leakage results from the exit total temperature (T03) of the fluid that

3

participates in over tip leakage remaining elevated, since energy is not fully extracted

from it. Another source of energy loss due to over tip leakage in transonic axial flow

turbines has been recently reported by Thorpe, et al [1]. They show, both analytically and

through measurement of casing temperature, that work is done on the fluid within the tip

gap of a transonic rotor, leading to increase in total temperature of the working fluid.

The stage inlet temperature (T01) may be limited by a number of factors including

thermal considerations and the net efficiency gain due to cooling. Over tip leakage causes

thermal deterioration of the blade tip surface and according to Bunker [2] might

necessitate a decrease in operating temperature (T01) over time, causing a further

reduction in efficiency. The effect of lowering the inlet temperature may also be

considered from the perspective of entropy generation. Gibbs’ equation as given by

Equation 1-2 shows that entropy generation is inversely proportional to the initial

temperature. Losses may be minimized by increasing the cycle maximum temperature.

Thus, improvements in efficiency are limited by the maximum permissible operating

temperature.

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛ −

⎟⎟⎠

⎞⎜⎜⎝

⎛−

⎟⎟⎠

⎞⎜⎜⎝

⎛−

=

γ

γη

1

01

03

01

03

pp1

TT1

tt (1-1)

pTCsT vddd p −= (1-2)

4

The research reported in this thesis is specific to un-shrouded, high pressure

turbine rotors. This chapter summarizes the current understanding of the over tip leakage

process, the aerodynamic and thermal effects of over tip leakage, previous studies aimed

at minimizing the effects of over tip leakage, and the objective and organization of this

thesis.

1.1 Axial Turbine Blade Terminology

A typical turbine blade passage is shown in Figure 1.1. The blade profiles shown

form the tip section of the turbine blades in the rotating turbine research rig at The

Pennsylvania State University. The direction of rotation is from bottom to top, as shown

by the blade speed vector U and W denotes inlet velocity in the relative frame of

reference. The blade passage, highlighted by a hatched pattern, is bounded by pressure

surface (PS) on the top and suction surface (SS) at the bottom. A pressure distribution is

generated on each blade surface, as shown in Figure 1.2, as the flow is turned in the blade

passage. The design blade surface pressure distribution shown was obtained from Xiao

[3]. The pressure is shown in terms of a pressure coefficient, as defined by Equation 1-3,

where 1p is the passage averaged local inlet pressure and Um is the wheel speed at mean

radius. Pressure driven flow, from blade pressure-side to blade suction-side due to the

pressure difference across the gap is referred to as over tip leakage or tip leakage flow.

21

5.0 mps U

ppCρ−

= (1-3)

5

Figure 1.1: Axial Turbine Rotor Passage.

-2.00

-1.50

-1.00

-0.50

0.00

0.50

0.00 0.20 0.40 0.60 0.80 1.00 1.20

x/Cax

Cps

SSPS

LE TE

Figure 1.2: Design Pressure Distribution on Turbine Blade.

6

1.2 Turbine Rotor Passage Flow

Turbine rotor passage flows are highly three dimensional and a brief description is

given with reference to Figure 1.3. The schematic of passage flow, from Yamamoto [4],

summarizes the observations of several researchers. Since the function of the passage is

to guide and turn the fluid, the primary flow follows streamlines defined by the blade

surfaces bounding the passage. However, viscous effects, streamline curvature, and

presence of strong pressure gradients generate significant recirculatory flows that are

generally termed as secondary flows.

The inlet boundary layer (1) stagnates as it approaches the leading edge and rolls

up into a horseshoe vortex (2). The two legs of the horseshoe vortex, pressure-side leg (3)

and suction-side leg (4) eventually merge near the blade suction surface and may become

part of a larger secondary flow structure, the rotor endwall or hub passage vortex (9). The

passage vortex is a result of pressure-driven cross-passage flow in the rotor endwall

boundary layer (8). Similarly, there exists a casing endwall or tip passage vortex. Over tip

leakage is also shown in Figure 1.3. The tip leakage vortex (12) and the tip passage

vortex rotate in opposite directions. The interaction between these vortices generates

limiting streamlines on the blade suction surface. In general, the vortices transport fluid

with high momentum and temperature into the boundary layers and deposit the low

momentum boundary layer fluid into the mainstream. The rotational kinetic energy is

dissipated without the extraction of useful work. The interaction between vortices also

leads to increased turbulence and mixing.

7

1.3 Over Tip Leakage (OTL)

The process of over tip leakage in actual engines is very complex and is affected

by compressibility, shocks, rotational effects, and the influence of the stationary casing.

The physics and effects of OTL flow have been studied primarily in cascades and for

incompressible flows. Results from rotating rig experiments are limited and as such no

data exists for flow within the tip gap or near tip surface flow.

Tip leakage flow methodology was investigated by Booth, et al [5] in different

water flow rigs. This study experimentally validated Rains’ [6] hypothesis that leakage

flow rate may be predicted based on an inertial balance. Booth, et al [5] confirmed that

pressure driven flow occurred normal to the gap while transverse velocity was preserved

across the gap. Gap discharge coefficient was found to be independent of gap height, inlet

boundary layer, and cross-flow velocity when the gap discharge coefficient was defined

using the static pressure across the gap. The gap mass flow rate was quantified by a

velocity calculated from Bernoulli’s equation used in conjunction with a discharge

coefficient.

The first detailed measurement of OTL flow within the tip gap, reported by

Bindon [7], conducted in a linear cascade of turbine blades showed that the gap flow

constitutes of fluid originating from two sources, as shown in Figure 1.4. The inlet

boundary layer passes through the tip gap in the front part of the blade, while pressure

driven flow commences at approximately 30% axial chord length. The pressure driven

flow separates at the pressure-side corner of the blade, reattaches on the tip surface and

flows towards suction-side of the tip gap. A gap vortex, caused by the separation, exists

8

in the tip gap. A favorable pressure gradient within the separated region generates a

chord-wise flow on the tip surface, towards the blade trailing edge. Low momentum fluid

within the gap vortex is entrained by the main gap flow around the mid-chord region of

the blade tip surface, where the driving pressure difference reaches a maximum. In

addition to the separation vortex the wall static pressure on the tip surface, near the

pressure-side corner was shown to be significantly lower than both the suction-side

pressure and the cascade exit pressure. This low pressure was attributed to the contraction

effect of the separation bubble in the tip gap. The gap flow diffuses in the tip gap,

generating a wake near the tip surface. Thus, the gap flow is characterized by a wake

region (d) near the blade tip surface and a low loss, jet (c) in the remainder of the gap.

Velocity measurements at different heights within the tip gap were conducted by

Yaras et al [8] in a linear turbine cascade. Bulk of the gap flow was found to occur above

the gap separation vortex. The change in velocity from gap inlet to gap exit was measured

to be small, suggesting that flow was fully accelerated at gap inlet. Close to the tip

surface the velocity vectors were observed to change rapidly and a chord-wise flow was

detected within the separation vortex. The centerline of the gap separation vortex was

located at about 20% gap height. The presence of a chord-wise flow and gap vortex are

also supported by surface flow visualization by Sjolander and Cao [9]. The visualization,

conducted in an idealized tip gap, showed a highly organized flow in the near pressure-

side corner of the tip surface, where flow separates due to the sharp corner of the tip gap.

9

Figure 1.3: Turbine Rotor Passage Flow, From Yamamoto [4].

Figure 1.4: Tip Gap Flow Conceptual Model From Bindon [7].

10

The linear cascade investigations while providing detailed understanding of the

gap flow do not capture the effect of casing relative motion. This motion is in a direction

opposite to the leakage flow and its effect would depend on the gap height. The effect of

casing motion was investigated in a linear cascade by Yaras and Sjolander [10]. Casing

motion at 100% engine equivalent speed (EES) was found to achieve a global reduction

in gap velocity. This was attributed to a reduced pressure effect, caused by an enhanced

passage vortex. Increasing casing relative speed also confined the gap vortex closer to the

tip, primarily due to the reduction in gap velocity. Correspondingly, blade loading near

the tip was shown to be reduced by Yaras, et al [11]. Flow downstream of the blades

displayed greater vorticity due to the passage vortex and lower vorticity due to the tip

leakage vortex. The increased vorticity of the passage vortex was attributed to the relative

casing motion, while decrease in leakage flow vorticity was attributed to the reduction in

gap mass flow rate. Water model cascade tests with a moving belt by Graham [12]

confirmed that the gap mass flow rate decreased with both a decrease in gap height and

an increase in casing relative speed. A cut-off speed was also determined, at which speed

the gap flow was completely eliminated at gap height of 1%. Morphis and Bindon [13]

investigated the effect of relative casing motion in an annular cascade and showed that

the low pressure at the PS corner was not affected by the influence of the casing.

However, higher pressure was measured in the reattachment region behind the bubble

and was attributed to a reduction in leakage flow.

While the effect of the relative casing motion may be simulated in cascades by

incorporating moving belts, the effect of centrifugal and coriolis forces on the passage

flow is not accounted for in such studies. Yamamoto [14] compared the leakage vortex

11

location in a stationary cascade with that in a rotor and found that the leakage vortex was

confined to the tip region of the passage, possibly due to centrifugal effects in rotors. In

stationary cascades the vortex responded to the span-wise pressure gradient and moved

radially inwards towards the hub. Kaiser and Bindon [15] noted that the energy

associated with the over tip leakage fluid in rotating rigs is higher than that found in

cascades, since energy extraction from over tip leakage flow is incomplete. Velocity

measurements in the rotating frame by McCarter [16], in the large scale, low speed axial

turbine at Penn State showed the path of the leakage vortex along the blade passage was

not as steep as that measured in cascades. The losses due to the tip leakage vortex also

increased further downstream of the rotor. Thus, it is important for tip desensitization

investigations to be conducted in a rotating environment.

1.3.1 Aerodynamic Losses

Losses in a turbine stage result from different mechanisms and those occurring in

vane and blade passages may be broadly split into profile losses, secondary flow and tip

leakage losses, and losses due to shocks. Over tip leakage flow is a significant contributor

to losses in turbine stages. Waterman [17] and Booth [18] suggest that a third of the stage

losses may be attributed to tip leakage flows. Considering losses in the rotor alone,

Schaub, et al [19] indicate that OTL may be responsible for as much as 45% of loss

within the rotor passage. Most HP rotors are designed to operate with gap heights of 1% -

2% blade height, which makes the losses associated with over tip leakage

disproportionately large.

12

Loss measurements by Bindon [7] in a linear cascade indicate that total pressure

loss due to the clearance gap vary linearly with gap size, with the lowest gap size

investigated being 0.3%. The total loss due to the tip gap was also separated into internal

gap loss (39%), suction corner mixing loss (48%), and endwall/secondary loss (13%).

The internal gap loss is attributed to the low momentum fluid within the separation

bubble. Mixing losses were seen to arise only over the last 20% of axial chord. McCarter

et al [20] measured the total pressure loss in a large scale, low-speed rotating rig

(AFTRF) at Penn State. They showed that the leakage vortex strengthens dramatically at

around 80% axial chord, causing mixing losses due to interaction between leakage and

passage flows, to rise. Losses due to leakage vortex were found to be about 25% higher

than those caused by the passage vortex.

The loss generation also depends on the profile thickness. Booth et al [5] found

that performance dropped as blade tip profile thickness was reduced while maintaining

loading distribution. A similar result was obtained by Sjolander & Cao [9] who showed

that the mass averaged losses increased as the blade tip profile thickness decreased in

relation to the gap height. Graham [12] found that flow in the gap over thin tipped blades

showed little effect of the moving endwall, thereby concluding that the pressure-driven

flow was essentially unaffected by surface friction.

The interaction between the passage vortex and the leakage vortex in linear

cascades was found to increase, with increasing blade turning and increasing gap height

by Yamamoto [4]. The vorticity from the passage vortex was observed to decrease by

McCarter, in the region that the leakage vortex vorticity increased. The interaction

between the two vortices was found to occur at and beyond the rotor exit plane. An

13

important conclusion was that secondary flows in the rotating rig dissipated slower than

the secondary flows in cascades. Additionally, the tip leakage vortex dissipated slower

than the upper passage vortex.

1.3.2 Thermal Effects

A discussion on high-pressure turbine blade over tip leakage flow is incomplete

without the inclusion of the thermal effects of OTL. Even though this thesis does not

report any quantitative results from heat transfer measurements on turbine blade tips a

brief description of some of the public domain knowledge is provided to complete the

picture. As noted by Bunker [2] blade tips are exposed to hot gases on all sides and the

tip surface experiences some of the highest and lowest heat transfer coefficients leading

to severe thermal gradients. Tip clearance increases with operational hours due to casing

rub and oxidation of the tip surface. This leads to decreased efficiency. One of the most

common effects of OTL is the burn-out of the pressure-side corner that has been linked to

the chord-wise flow on the tip surface by Bindon [21], and Sjolander and Cao [9]. This is

one of the reasons limiting cycle maximum temperature and in turn the efficiency of the

work extraction process in turbine blade passages. Heat transfer coefficients measured by

Bunker et al [22] on the tip surface of a blade with sharp edges show the extreme

variation in heat transfer coefficient over the tip surface. Bunker [23] also notes that the

leakage vortex that forms near the suction surface and its interaction with the tip passage

vortex is responsible for increased oxidation of the suction surface.

14

1.4 Tip Desensitization

The clearance between the blade tips of a high pressure, un-shrouded, axial gas

turbine rotor and the stationary casing has been shown, in the previous sections, to be a

source of efficiency loss and a limiting factor to improving efficiency through increased

cycle maximum temperature. The gap height is also subject to growth due to both tip

surface material oxidation and tip surface rubbing into the casing during engine thermal

transients. The efficiency drop as a function of tip gap height is linear, as shown in

Bunker. More importantly the slope of the line is greater for smaller engines and high

pressure turbine stages due to the small annulus areas. Thus, tip desensitization may be

defined as reducing the sensitivity of turbine performance to the effects of the tip gap

height.

Most of the tip desensitization techniques reported involve modification of the tip

surface geometry in an effort to decrease the leakage mass flow rate, by decreasing the

discharge coefficient. A comprehensive study of various tip geometries by Booth et al [5]

included, among other geometries, tip surface extensions (winglets), squealer tips, and

double squealer or grooved tips. The single and double squealer configurations were

found to give the best reduction in discharge coefficient. However, their performance was

found to be very sensitive to geometry and gap Reynolds number. Winglets were found

to generate greater improvement due to both a reduction in discharge coefficient and a

reduced pressure drop across the tip gap surface. Bindon [24] visualized flow over

squealer tips using smoke in a linear cascade and found that at large gap heights the

leakage flow essentially passed over the tip surface as if the surface were flat. Linear

15

cascade results of Heyes et al [25] show that suction-side squealers are better than

pressure-side squealers.

The heat transfer on tip surfaces with squealer cavities has also been studied both

experimentally, in cascades, and through numerical simulation. Bunker and Bailey [26]

concluded that squealer tips decrease the heat load on the tip surface. Cavity surface heat

transfer coefficients decreased with increase in cavity depth and the distribution of heat

transfer coefficients was also more uniform. The overall heat load on the cavity wall also

decreased with increase in cavity depth. Numerical simulations by Acharya et al [27] also

indicate that the aerodynamic benefits of double squealer tips decrease with increase in

gap height. The reductions in leakage mass flow rate and tip heat transfer coefficients

obtained over flat tip geometry grew smaller as the gap height of double squealer tips was

increased. The presence of a squealer cavity decreased the suction surface heat transfer

coefficients due to modification in the leakage vortex strength. Papa et al [28] found that

for a blade tip surface with squealers the starting point of the leakage flow moved

rearwards with increasing gap height.

Morphis and Bindon [13] experimented with pressure-side corner rounding in an

effort to reduce separation related losses in the tip gap. In the 1% - 2% gap height range

the total-total stage efficiency of a single stage was found to improve with rounding-off.

The efficiency of the second stator row was found to improve with rounding-off of the

pressure-side corner of the first stage blades. Numerical simulation by Ameri [29]

showed that leakage mass flow rate increased by about 25% when the pressure-side

corner was rounded-off, due to an increase in the gap discharge coefficient. The addition

of a camber-line sealing strip decreased the leakage mass flow rate, however the mass

16

flow rate was still greater than that through the gap over a sharp edged, flat tip blade. The

reduction in total pressure loss caused by a camber-line strip on the tip surface with

pressure-side corner rounding-off was not proportional to the reduction observed in gap

mass flow rate. The tip surface heat transfer was found to increase due to rounding-off of

the pressure-side edge, by Bunker et al [22].

Research at Penn State has consisted of both experimental and computational

investigations of tip desensitization. The effect of various tip geometries was studied in

the low-speed rotating rig at Penn State by Dey [30]. The tip geometries studied included

pressure-side and suction-side extensions (winglets), and single and double squealers.

The axial coverage of these tip modifications was also investigated. The height of the tip

gap was found to be an important parameter and the effectiveness of squealers degraded

quickly as gap height was increased. The first investigation of coolant injection from a

rotating rig was done by Dey and Camci [31]. No appreciable change in rotor exit total

pressure was measured and this was attributed to the coolant injection holes being too

small. Computational investigation of tip desensitization through tip surface chamfering

by Tallman and Lakshminarayana [32] indicate that OTL losses may be reduced due to

turning of the leakage flow towards the camber-line.

Of the desensitization techniques reviewed only the double squealer tip finds

actual application in service. Double squealer rims offer the advantages of minimizing

contact surface area and partly shielding cavity surface from thermal degradation. The

other techniques, except tip coolant injection from flat tips, present added cooling

requirements that may not be easily addressed. In a comprehensive review of various

desensitization methods Harvey [33] notes that attempts to reduce the discharge

17

coefficient have proven contradictory. In fact, the use of shrouds is pointed out as

advantageous from the point of view of efficiency retention during service. However

turbine shrouds, in addition to requiring cooling, increase tip mass, which can introduce

mechanical complications. Furthermore, the desire to increase firing temperatures might

at some stage make shrouds untenable in turbine designs, particularly HP turbines.

1.5 Film Cooling of Turbine Blade Tips

It is typical for modern high pressure turbine blades to eject coolant into the tip

gap from radial bores on the tip surface. This may be done either for cooling of the blade

tip surface, or for purging dirt to maintain the effectiveness of the internal cooling

system. More recently the effect of radial coolant injection on blade tip surface heat

transfer and tip gap aerodynamics has been studied.

Experimental investigation of a modeled blade tip is reported in Kim et al [34].

Kwak and Han [35] investigated film cooling of gas turbine blade tips in a linear cascade.

Blade tip heat transfer coefficients were found to increase with gap height. Film cooling

from radial holes along the camber line was shown to be more effective at larger

clearances and at higher blowing ratios. Near tip coolant injection from the pressure

surface was found to increase film effectiveness over the tip surface. Hohlfeld et al [36]

numerically simulated the film cooling effect of dirt purge holes on turbine blade tips.

Coolant ejected from the purge holes served to block the leakage flow at a tip clearance

of 0.54% blade height. However, as gap height was increased to 1.64% span the blocking

effect decreased. At the large clearance, shroud cooling effectiveness increased with

18

blowing ratio, while tip surface cooling effectiveness first decreased and then increased.

Acharya et al [27] also computationally simulated film cooling of turbine blade tips. Film

cooling at three different gap heights was simulated. Coolant injection was found to alter

the leakage vortex and also decrease the heat transfer coefficient along the coolant

trajectory. Film cooling effectiveness was found to increase slightly with gap size. Two-

dimensional computations of over tip leakage with coolant injection by Koschel et al [37]

show that the coolant discharge has a minor effect on the gap discharge mass flow rate.

The influence of wall motion appeared to be more pronounced with coolant injection.

The effect of coolant injection into the tip gap was simulated by Chen et al [38] using

two-dimensional Navier-Stokes equations. A single radial jet from a tip surface slot was

found to decrease the gap inlet mass flow rate. The discharge mass flow rate remained

more or less the same and was independent of the coolant mass flow rate. The effect of

the secondary jet was found to be sensitive to the position and width of the tip surface

slot.

1.6 Objectives of Current Research

The design of modern high pressure turbine blades is complex and as noted by

Bunker [2] is governed by various, diverse parameters, including aerodynamic and

thermal variables. The most common blade tip designs include un-shrouded blade tips,

with and without squealer rims, and a few shrouded designs. The effectiveness of

squealer rims with cavity cooling is to some extent controlled by the necessary hading of

turbine casing. Furthermore, cooling of the blade tip region is almost always necessary.

19

The primary objective of this experimental investigation is to study the effect of

tip coolant injection on OTL in a rotating environment. The configuration used in Dey

and Camci [31] is modified to allow for greater coolant mass flow rates. The study

consisted of both quantitative and qualitative measurements. Time accurate total pressure

measurements were conducted downstream of rotor exit using a high-frequency

transducer aligned with the absolute flow. The measurements without and with coolant

injection are analyzed for performance benefits. Surface flow visualization was also used

to qualitatively understand the interaction between injected coolant and gap flow, not

only from an aerodynamic view point but also from the thermal benefits that may be

present. Some of the results presented and discussed in this thesis may also be found in

Rao and Camci [39], [40], [41], and [42].

The organization of this thesis is as follows. Chapter 2 describes the facility,

aerodynamic design of the turbine stage, instrumentation, and operation of the facility.

Flow visualization on the rotor endwall, the casing endwall, and the blade tip surface is

reported in Chapter 3. The effect of tip gap height is the subject of Chapter 4, where the

results presented include total pressure measurements and surface flow patterns. Chapter

5 reports the effect of coolant injection from a single rotor blade, where the effect of

coolant mass flow rate on OTL is discussed. The effect of coolant injection location on

the measured total pressure defect due to over tip leakage is discussed in Chapter 6.

Chapter 7 discusses coolant injection from multiple blades and the effect of casing

surface roughness. A summary of the results, conclusions and future work

recommendations are presented in Chapter 8.

20

Chapter 2

Facility Description

The facility used in this investigation is the Axial Flow Turbine Research Facility

(AFTRF) at the Pennsylvania State University. A description of the facility, turbine stage

design, and instruments used is included in this chapter.

2.1 The Axial Flow Turbine Research Facility (AFTRF)

The AFTRF is a large scale, low speed, single stage, axial flow turbine facility. A

schematic of the facility is shown in Figure 2.1. The shaded regions denote rotating

components. The main components in the flow path of this facility are listed below in

order,

a) Inlet section,

b) Turbine stage,

c) Transition ducting with acoustic muffler,

d) Two-stage axial fan, not shown in the figure,

e) Exhaust diffuser, not shown in figure

The facility inlet is housed in a large stagnation chamber covered with foam cloth.

Air flows through a smooth bell-mouth section with an axi-symmetric center-body that

terminates at the stator. The centre-body houses slip-ring units for data transfer from

rotating frame to stationary frame.

21

Figure 2.1: Schematic of the Axial Flow Turbine Research Facility.

Figure 2.2: Turbine Casing Frame For Measurement Windows.

Rotating Hub

Stub Shaft

Bell-mouth Inlet

Centre-body

0.5715 m

0.1524 m

0.2985 m0.2794 m

0.1334 m

Total pressure probe

Split line between rotating & stationary hub Nozzle ring

C.L.

Rotating hub

22

The turbine stage consists of a stator row (N) that accelerates the flow into the

rotor (R). The turbine outer casing is provided with a rectangular access frame that

measures (0.5715 m x 0.2794 m), as shown in Figure 2.2. The frame is symmetric about

the horizontal plane and starts at 0.1524 m upstream of stator row. Windows for probe

based measurements, shown in Figure 2.3, or optical access are mounted in this frame.

The windows are fabricated so that the inner surface of the windows are curved to the

turbine casing outer diameter (∅max = 0.9166 m) and fits in flush. A row of exit guide

vanes (EGV’s), three chord lengths downstream of the rotor remove the swirl induced in

the flow by the rotor blades. Air then passes through the transition ducting, which is lined

with acoustic damping liner to reduce noise generation. Airflow through the facility is

induced by the suction provided by a four stage axial fan. The combined flow capacity of

the fans is 10 m3/sec and they generate a combined pressure rise of 74.7 mm Hg

(approximately 40” water column). The fan duct is covered externally by thermal lagging

material to ensure adiabatic boundary conditions in the fan duct. Air energized by the

fans is then exhausted to atmosphere through the diffuser, which has an external moving

endplate that provides a means to control mass flow rate.

A rotating instrumentation drum is mounted on to the turbine disk. Instruments,

such as pressure transducers, hot-wire anemometers, probe traverse, mass flow measuring

devices, etc, are mounted in the instrumentation drum. All instruments are wired to a slip-

ring unit on the main shaft. An outer drum that forms the rotating hub, as marked in

Figure 2.1, extends 133 mm downstream of the rotor exit plane and forms the inner

diameter (∅hub = 0.6706 m) of the rotor annulus. The rotor endwall (hub) surface extends

23

beyond this in the stationary frame, with a small gap provided at the split line with the

rotating hub as marked in Figure 2.2.

A stub shaft, with a pulley, is flange mounted to the downstream end of the

rotating instrumentation drum. Torque is transmitted from stub-shaft to a brake shaft by

the belt & pulley system. A water-cooled, eddy current brake absorbs power generated by

the turbine. The brake is capable of maintaining the turbine speed constant to within ±1

rpm. The brake shaft is equipped with an inline torque meter for torque measurement. A

BEI systems optical shaft encoder is mounted in the stationary frame aft of the pulley.

The encoder shaft is coupled to a stub shaft mounted to the turbine pulley.

2.2 The Turbine Stage

Overall characteristics of the turbine stage are shown in Table 2-1. The rotor tip

radius is 0.4582 m and the inner radius is 0.3353 m, giving a hub-tip ratio of 0.7317. This

ratio is representative of high-pressure turbine stages, where the stage inlet density is high

and hence the annulus height tends to be small. The turbine generates about 60 kW of

power while operating at a nominal speed of 1300 rpm and a nominal through flow rate

of 11.05 kg/s. The design total-total isentropic efficiency of the turbine stage is 0.893.

Design performance data of the turbine stage is shown in Table 2-2. The design

inlet conditions to the turbine stage are that of standard atmosphere at mean sea level.

Low inlet to exit pressure and temperature ratios implies that compressibility effects are

insignificant. Additionally, there is minimal density change through the stage allowing

for a constant mean radius throughout the stage.

24

Figure 2.3: AFTRF Instrumentation Window.

Nozzle ring area Casing inner face

∅ = 0.9166 m Over tip region

Total pressure probe

Window View From OUTSIDE

Stationary frame probe traverse

Probe stem

25

For a specific turbine design, where the geometry and gas are fixed, the pressure

ratio, temperature ratio, and efficiency depend on the flow function and the speed

function as given by Equation 2-1. The definition and value of the flow function and the

speed function are listed in Table 2-3, along with other design coefficients. Typical inlet

conditions experienced were P01 = 98.0 kPa (980 mbar) and T01 = 301.15 K (28° C). The

departure from design inlet operating conditions can be corrected for by changing the

turbine mass flow rate and rotor speed such that the ratios on the RHS of Equation 2-1

are kept constant. Design performance is achieved by operating the turbine stage at a

mass flow rate of 10.466 kg/sec and a rotor speed of 1327 rpm. These two parameters

may also be represented in non-dimensional form by φ, the flow coefficient.

⎟⎟⎠

⎞⎜⎜⎝

⎛=

0101

01

01

03

01

03

TN,

PTm

,TT,

PP &

fη (2-1)

Table 2-1: AFTRF Overall Stage Design Characteristics.

Power; P (kW) 60.6

Rotor Tip Radius; rtip (m) 0.4582

Rotor Hub Radius; rhub (m) 0.3353

Rotor hub-tip ratio 0.7317

Mass Flow Rate; m& (kg/sec) 11.05

Rotational Speed; N (rpm) 1300

Total-Total Isentropic Efficiency; ηtt 0.893

26

2.2.1 Stage Aerodynamic Design

Turbine stage nozzle guide vane (NGV) design parameters are listed in Table 2-4.

The stator ring has 23 nozzle vanes that turn the inlet axial air stream by about 70°. The

specific vane exit angle is chosen in order to maximize vane efficiency. Design Reynolds

numbers at nozzle inlet and exit are representative of modern high pressure stages. The

spacing between the stator row and rotor is variable between 20% and 50% blade tip

Table 2-2: Design Thermodynamic Parameters.

Total Inlet Pressure;Po1 (kPa) 101.36

Total Inlet Temperature;To1 (K) 289

Total Pressure Ratio; Po1/Po3 1.0778

Total Temperature Ratio; To3/To1 0.981

Pressure Drop; Po1-Po3 (mm Hg) 56.04

Table 2-3: AFTRF Design Coefficients.

Flow Function; m& √T / P (kg√oK m2 / kN sec) 1.85

Speed Function; N/√T (rpm/√oK) 76.47

Flow Coefficient; φm = (V2ax/U)m 0.568

Work Coefficient; ψm = ∆ho/Um2 1.85

Specific Work Output; ∆ho/ m& (kJ/kg) 5.49

27

axial chord length. The spacing used in the current investigation is 20% blade tip axial

chord length.

Turbine rotor design data is listed in Table 2-5. The rotor blade count is 29. The

maximum relative Mach number in the rotor is 0.24. The blades were designed to have a

nominal tip clearance of 0.9 mm or 0.76% blade height (h = 0.1229 m), on a rotor

average basis. Blade velocity triangles at three sections are shown in Figure 2.4. Blade

inlet angles change considerably from hub to tip, giving the blade a twisted, three-

dimensional shape at the leading edge. The rotor inlet geometric parameters indicate a

small departure from free-vortex conditions (r * Vθ = Constant). The design absolute

velocity vector at rotor exit is aligned at 25.16° to axial at the tip and at 35.13° to axial at

the hub. The AFTRF turbine stage is a reaction stage, which means that part of the static

pressure drop occurs across the turbine rotor, as shown in Table 2-6. Thus, more than half

the static pressure drop at the tip radius takes place in the turbine rotor, while very little

static pressure drop occurs within the rotor at the hub.

Table 2-4: AFTRF Nozzle Design.

Number of Vanes 23

Stator Zweifel Coefficient 0.7247

Chord; (m) 0.1768

Spacing; (m) 0.1308

Maximum Thickness; (mm) 38.81

Turning Angle; 70o

Reynolds Number ( ÷105) inlet / exit (3~4) / (9~10)

Stator Efficiency; ηs 0.994

28

Table 2-5: AFTRF Rotor Design.

Number of Blades 29

Rotor Zweifel Coefficient 0.9759

Relative Mach Number 0.24

Blade Height; hb (m) 0.1229

Tip Clearance; (mm) 0.9

Turning Angle; Tip / Hub 95.42o / 125.69o

Chord; (m) 0.1287

Axial Tip Chord; (m) 0.084

Spacing; (m) 0.1028

Maximum Thickness; (mm) 22

Reynolds Number ( ÷105) inlet / exit (2.5~4.5) / (5~7)

Rotor Efficieny; ηR 0.8815

Table 2-6: Radial Variation of Turbine Stage Design Parameters From 1-D Mean-line Analysis.

Parameter HUB MID TIP

Loading, ψ 2.156 1.854 1.635

Reaction, R 0.184 0.382 0.507

Flow Coefficient, φ 0.754 0.568 0.429

Turning, ∆β 125.09 110.54 86.20

29

2.2.2 AFTRF Tip Clearance Distribution

The distribution of tip gap height in the turbine rotor is shown in Figure 2.5. The

gap height was measured at the split line of the measurement window using feeler gages.

The values reported here are within ±0.0254 mm (0.001”) or t/h = 0.02% of the actual

value. Measurements were made in three equal regions along blade axial chord (near

leading edge, mid-chord, near trailing edge) and an average was computed for each

Figure 2.4: Blade Velocity Triangles.

U

W2

V2

V3

W3

30

region. The values reported in Figure 2.5 are average values for the tip surface. The

variation of gap height of each blade, with axial distance from the leading edge, is shown

in Appendix B. Four distributions were active over the course of the project. The design

tip clearance for the AFTRF is 0.9 mm or t/h = 0.76%. Five blades in the rotor, referred

to as, “cooled blades,” were modified to have a slightly larger tip gap. TCL1, the dashed

blue line, shows a single blade (B21) with a large gap height of t/h = 1.40%, whereas the

other blades have a gap height closer to the design gap height. The gap height of the other

cooled blades was reduced to the levels shown by applying precision plastic layers to the

tip surface with the help of double-sided tape of thickness 0.102 mm (0.004”). The

distribution TCL2, solid green line, was obtained by reducing the gap height of blade B21

to t/h = 0.72%, using precision plastic layers and double-sided tape. These two

distributions are relevant to results in Chapter 4, and results from “isolated injection,”

presented in Chapter 5 and Chapter 6. TCL3 represents gap height as measured from bare

metal tip surface of each blade. Thus, this represents the maximum tip clearance levels

achievable in the rig in this configuration and was used for coolant injection from all

cooled blades, referred to as “multiple injection,” results of which are presented in

Chapter 7. The last distribution, TCL4, was obtained by applying precision plastic layer

to the casing surface in the measurement window region and is relevant to casing surface

roughness results presented in Chapter 7.

31

2.3 Tip Cooled Blades

As mentioned in the previous section five blades, referred to as cooled blades,

were modified to a slightly larger tip clearance gap of 1.65 mm or t/h = 1.34%. These

blades were also modified to include features for injecting coolant air from the tip surface

of the blades. The tip surface of one of the cooled blades is shown in Figure 2.6. The tip

trench has four injection holes, three of which connect to the radial plenum chambers

shown in the inset of Figure 2.6. The fourth injection hole, marked H4, is located at the

end of the trench, near the trailing edge and serves as its own plenum. The radial plenums

0.30%

0.50%

0.70%

0.90%

1.10%

1.30%

1.50%

1.70%

0 5 10 15 20 25 30

Blade Number

Gap

Hei

ght (

t/h) %

TCL1

TCL2

TCL3

TCL4

COOLED BLADES

Figure 2.5: AFTRF Tip Clearance Distribution.

32

are blocked by a plug, inserted and permanently fixed after all modifications were

completed. All radial plenum chambers open into a common blade root plenum. A

standard ¾” Aluminum tube is fitted into the base of the blade root plenum to allow for

air lines to be connected to the chamber.

Geometric details of the injection scheme are shown in Figure 2.7 & Figure 2.8.

The tip trench is 60 mm long and 2 mm wide, as shown in Figure 2.7. The trench extends

from blade mid-chord to 0.91 Cax and is at a 60° angle with respect to the axial direction.

The three injection locations (H1 – H3) within the trench are made up of two holes, each

of ∅ = 0.762 mm, while the last (H4) is a single hole of ∅ = 1.8 mm. The number of

holes at each location and their diameters are specified in the table in Figure 2.7.Based on

a network analysis by Pudupatty [43], the discharge mass flow rates from each injection

location are as shown in Table 2-7. The trailing edge location H4 accounts for about half

the total discharge area and a third of the total coolant mass flow rate. While H3 has the

same discharge area as H1 and H2, the radial plenum diameter to this location is smaller,

as shown in Figure 2.7, thereby resulting in a small reduction in the mass flow rate.

The four injection sets are located 0.10 Cax apart, starting at 0.61 Cax. The distance

is measured from blade leading edge to the mid point of the line joining the centers of the

two injection holes. As shown in the cross-sectional details in Figure 2.8, the two

injection holes at H1 are inclined at 10° to each other and at 45° to the tip surface,

directed towards the blade pressure-side corner. It must be noted that the jets are not

blocked by the edge of the tip trench. The two holes open into the corresponding radial

plenum. This detail is repeated at H2 and H3.

33

Figure 2.6: Cooled Blade Tip Surface.

34

Figure 2.7: Cooled Blade Tip Coolant Injection Arrangement.

Table 2-7: Estimate of Coolant Discharge Mass Flow Rate, From Pudupatty [43].

Injection location Area (mm2) / Fraction of total area

Radial plenum area

(mm2)

Fraction of total mass flow rate

H1 0.912073 / 17.24% 12.4898 22.56%

H2 0.912073 / 17.24% 12.4898 22.56%

H3 0.912073 / 17.24% 7.05539 22.32%

H4 2.55431 / 48.28% 2.55431 32.55%

35

Figure 2.8: Cross-Sectional View of Cooled Blade.

36

2.4 Air Transfer System (ATS)

Figure 2.9 describes the air transfer system used for bringing coolant air from the

stationary frame to the plenum chamber of the cooled blades, in the rotor frame of

reference. The shaded regions in Figure 2.9 denote rotating components, while the un-

shaded regions denote stationary components. A custom designed air chamber is located

near the downstream bearing of the turbine rotor assembly. The rotating face of the

chamber is connected to the rotating instrumentation drum, while the stationary part of

the chamber is fixed to the bearing housing. Two precision seal systems minimize

leakage between the rotating and stationary surfaces of the air transfer system. The two

stationary seals work against plasma coated surfaces on the turbine shaft and rotating

drum of the AFTRF. Five barbed pipe fittings are threaded into the rotating part of the

ATS. High-pressure nylon tubes are connected between the fittings and metallic straight

connectors mounted in the rotating drum. One of these connectors is modified by

inserting an orifice for mass flow rate measurement. High-pressure tubes then supply the

air to the blade plenum chambers.

2.5 Flow Visualization

Surface flow visualization experiments were conducted using a mixture of oil and

pigment. The oil used was ash less dispersant SAE 40 Aviation oil (Aeroshell oil W80).

Initial experiments conducted resulted in the conclusion that oil by itself was not a viable

medium for surface flow visualization due to the high wall shear stresses encountered in

37

the rotor blade passages. Two types of pigments, Titanium White artists’ oil color

containing titanium oxide and zinc oxide, and Zinc Yellow Hue artists’ oil color

containing Arylide yellow and Zinc Oxide were used for surface flow visualization. The

visualization material used contained about 500 mm3 of pigment mixed in with oil and

weighed approximately 2 gm in a volume of 10 ml. The mixture with Zinc Yellow Hue

was observably more viscous than the mixture with Titanium White Artists’ oil color.

Figure 2.9: Air Transfer System (ATS).

Air-Flow

38

2.6 Instrumentation

A comprehensive instrumentation and measurement system, schematics of which

are shown in Figure 2.10 and Figure 2.11, was set-up as part of the current research

effort. The monitoring system shown in Figure 2.10 is used to control operation of the

facility, while the performance measurement system shown in Figure 2.11 is used to

measure the turbine rotor exit flow field to assess the effect of desensitization.

Encoder Signal Unit

Thermocouple Signal Conditioning Unit

ANALOG OUTPUT

DIGITAL OUTPUT

LOW SPEED DASTIMING I/O CONNECTIONS

ANALOG INPUTS

P03

P01

Inlet Pitot-Static Probe

V2ax / Um

6-Channel Pressure

Transducer

ρV12 / 2

1 / rev. Trigger Pulse

6000 / rev. Clock P

ulse

To High Speed DAS

Exit Pitot-Static Probe

Inlet Thermocouple

Exit Thermocouple

Room Thermocouple

Bearing Thermocouple

OMEGAThermocouple Display

Units

ρV32 / 2

T01

Troom Tbrg

T03

Himmelstein Torque Meter + Display Unit

Power + Torque

Figure 2.10: Schematic of Monitoring Instrumentation.

39

2.6.1 Monitoring Instrumentation

The measurements for monitoring test conditions consist of pressures and

temperatures at stage inlet and exit, bearing temperature, ambient conditions, and shaft

output, as shown in Figure 2.10. A low speed data acquisition system described in

Section 2.7 acquires and processes the data.

Turbine inlet flow conditions are measured using a single Pitot-static probe and a

single K-type thermocouple, both located at turbine mean radius and about 1.5 vane tip

KRONHITE Filter

Kulite Power & Signal Unit

Total Pressure Probe with Kulite Sensor

VALIDYNE Pressure

Transducer

Encoder Signal Unit

Analog Signal From Low-Speed

DAQ

DIGITAL OUTPUT

HIGH SPEED DAS

TIMING I/O CONNECTIONS

ANALOG INPUTSP03

P01

Inlet Pitot-Static Probe

V2ax / U

m

HONEYWELL Pressure

Transducer

ρV1 2 / 2

1 / rev. Trigger Pulse

6000 / rev. Clock Pulse

Stepper Motor Controller

Figure 2.11: Schematic of Performance Instrumentation.

40

axial chord lengths upstream of the nozzle. The Pitot-static probe is connected to pressure

transducers to measure inlet total pressure and inlet kinetic energy per unit volume, as

shown schematically in Figure 2.10. Inlet total temperature, measured using the

thermocouple, is used to control turbine operating speed. Similarly, the rotor exit total

pressure, kinetic energy per unit volume, and total temperature are measured using a

single Pitot-static probe and K-type thermocouple located at mean radius and 0.6 rotor tip

axial chord length downstream of the rotor. The pressure transducers used in conjunction

with the Pitot-static probes are a Validyne transducer and Honeywell transducers.

Bearing temperatures are monitored using K-type thermocouples on the bearing

housings. Shaft power and torque are measured by an inline torque meter connected to a

Himmelstein display. The power and torque readings are not connected to the data

acquisition system and are noted down.

2.6.2 Performance Measurement

The effect of the implemented desensitization methods on over tip leakage flow is

assessed by measuring the total pressure downstream of the rotor exit plane. The

instruments used consist of a high-frequency total pressure probe and inlet pressure

probe. The signals from these instruments are connected to a high speed data acquisition

system, which is described in Section 2.7. An analog voltage signal generated by the

monitoring system is also connected to this system.

41

2.6.2.1 High-Frequency Total Pressure Probe

A high frequency response pressure sensor, XCS-062-5D, manufactured by Kulite

Semiconductors is used to measure the exit total pressure field. The sensor is a sealed

cylindrical tube of diameter 1.6 mm, rated at 5 psid, and operates in a differential mode.

A protective B-screen mounted in front of the sensor diaphragm reduces the diaphragm

frequency response from 150 kHz to approximately 20 kHz, Kulite [44]. The sensor is

powered by a regulated 10 VDC supply and the transducer signal is amplified through an

instrumentation amplifier. Static calibration of the probe using a manometer yields a

linear calibration curve, as shown in Figure 2.12. The zero offset of the transducer is

subtracted from measured voltage when the transducer is pressurized.

The transducer is housed in a 3.5 mm outside diameter tube with a square cut

face, such that it is flush with the square face of the tube and all gaps between the sensor

and tube are sealed to prevent cavities. The differential pressure tube is left open to the

room, which means that one end of the diaphragm is at Pamb. The probe is mounted in a

single axis, stepper motor driven, linear traverse system on the casing window. The

traverse with probe is shown in Figure 2.3. This allows the probe to be traversed with

increments of 1/16th inch and greater. The probe face is positioned 30% chord

downstream of the rotor exit, at mid-pitch with respect to the upstream nozzle passage,

and aligned to the absolute tip flow vector at the rotor exit, as shown in Figure 2.2.

Characteristics of the total pressure probe, such as angular sensitivity and frequency

response are included in Appendix A. Through measurements at rotor exit the probe is

shown to be insensitive to incidence angles in a range ±15.

42

2.6.3 Coolant Mass Flow Meter

Coolant mass flow rate to cooled blade B21 is measured by an ∅ 8.0 mm orifice

installed on the supply line in the rotating frame. The orifice is installed in a ∅ 9.525 mm

(3/8”) pipe with two static pressure taps located 2D upstream and 5D downstream of the

orifice plate. The orifice meter is calibrated to relate the pressure difference (∆P) between

the static pressure taps to the standard volume flow rate through the orifice. The

calibration chart shown in Figure 2.13 relates the pressure difference to the non-

dimensional mass flow rate Minj. Minj as defined in Equation 2-2 is the ratio of orifice

R2 = 1.00

-3.000

-2.500

-2.000

-1.500

-1.000

-0.500

0.000

0.00 0.50 1.00 1.50 2.00 2.50

Applied Gage Pressure, dP (kPa)

Mea

sure

d V

olta

ge C

hang

e Fr

om V

0, dV

(VD

C)

dP (Pa) = 925.352 dV (VDC)V0 = Voltage at (dP = 0).

Figure 2.12: Total Pressure Probe Static Calibration.

43

mass flow rate to turbine design mass flow rate, assuming coolant is being injected from

all 29 blades in the blade row. The orifice meter pressure taps are connected to a PSI

systems pressure transducer in the rotating frame, as detailed in the Section 2.6.4. The

calibration equation used to set the coolant mass flow rate during the test is obtained from

a modified curve-fit relating dependence of Minj on the transducer signal due to applied

pressure difference, as shown in Figure 2.14. The actual turbine mass flow rate varies for

each test run and may also be different from the design mass flow rate. Thus, the value of

Minj reported with the results has been corrected for the actual operating conditions.

2.6.4 Rotating Frame Pressure Transducer

A PSI Systems ESP-48 transducer mounted in the rotating drum is used for

measuring the static pressures from the orifice meter. The transducer has 48 sensing ports

and one reference port and can measure a pressure difference of ±1 psi between any of

the sensing ports and the reference port. The static pressure taps from the orifice meter

are connected to the PSI transducer in the differential mode, to directly measure the

pressure difference. The upstream static pressure tap is connected to the reference port,

while the downstream static pressure tap is connected to a sensing port. The transducer is

powered by a 10VDC regulated supply. Channel indexing is achieved by an 8-bit digital

signal.

T

c inj m

29*m M&

&=

(2-2)

44

2.7 Data Acquisition

Data acquisition is accomplished by using two National Instruments DAQ boards

and associated LabView programs. A low speed, 8-channel, board PCI-6024E in a PC-

AT computer is used for acquiring data for test monitoring. Analog inputs to this board

consist of inlet and exit total pressure, inlet and exit kinetic energy per unit volume, total

temperatures at stage inlet and exit, inlet chamber temperature, and bearing temperature.

Digital input consists of a 6000 per rev pulse generated by the shaft encoder. The output

functions consist of an analog output voltage that is proportional to the flow coefficient

0.00

0.50

1.00

1.50

2.00

2.50

3.00

0.00 0.50 1.00 1.50 2.00 2.50

Wall Static Pressure Difference Across Orifice Plate, (psid)

Min

j, (%

)

Figure 2.13: Coolant Discharge Measurement Orifice Calibration.

45

and digital lines for indexing channels on the rotating frame pressure transducer. The

input data is acquired and averaged over 60 seconds. A virtual instrument written using

LabView processes the inputs and displays various parameters, some of which are shown

in Table 2-8, along with the use of each parameter.

y = 1.094x0.4257

R2 = 0.9949

0.00

0.50

1.00

1.50

2.00

2.50

3.00

0.00 1.00 2.00 3.00 4.00 5.00 6.00

Transducer Signal, (VDC)

Min

j, (%

)

0.332

Figure 2.14: Coolant Mass Flow (Orifice) Meter Calibration Minj vs. VDC.

46

A high speed, 4-channel, simultaneous sampling DAQ board, the PCI-6110E is

used for performance improvement measurements. The analog inputs include signal from

the high-frequency total pressure probe, inlet kinetic energy per unit volume, inlet total

pressure, and a voltage signal from the low-speed data acquisition system. Timing and

I/O operations are controlled by a one per rev., trigger pulse and a 6000 per rev. clock

pulse. Digital output lines are used to control the stepper motor.

The associated Labview instrument initiates data acquisition at each radial

position of the total pressure probe upon being triggered and acquires 6000 points per

revolution of the rotor. This is an important feature implemented into the current data

acquisition process. Typically, during a single run the turbine speed is varied with

temperature, to hold the speed function constant. Additionally, the turbine speed for

different runs tends to be different. The use of the encoder pulse to control the scan clock

ensures that each data point falls in the same circumferential bin, independent of the rotor

speed. The unsteady total pressure downstream of the rotor is acquired 6000 times per

Table 2-8: Some Calculated Parameters From Low Speed DAS.

Parameter Operational Importance

Pressure Ratio, Po3/Po1 -

Temperature Ratio, To3/To1 -

Flow Coefficient at mid-span, φ Controls acquisition of data on high speed DAS

Rotor Speed, N Rotor speed control

Actual Mass Flow Rate, m& Controls acquisition of data on high speed DAS

47

revolution, which corresponds to a nominal sampling rate of 132 kHz. The data is

ensemble averaged over 200 ensembles. Stepper motor control also rests with this virtual

instrument.

2.8 Data Processing

The unsteady total pressure downstream of rotor exit is ensemble averaged and

converted to a non-dimensional total pressure coefficient as given by Equation 2-3. The

pressure drop across the stage is a measure of both the energy extracted from the air as

well as the energy loss in the stage. The advantage of the specific non-dimensional form

is that this value is invariant with operating conditions. Um may be expressed in terms of

the speed function as shown in Equation 2-4. The air density term in the numerator of

Equation 2-3 may be canceled out by taking inlet total pressure out in the numerator and

multiplying and dividing the RHS of Equation 2-3 by R. Then the total pressure

coefficient is given by Equation 2-5. The pressure ratio term in the numerator is a

function of the flow function and the speed function, which if maintained constant should

result in the same pressure ratio, irrespective of operating conditions. Since all terms in

the denominator are constant, maintaining the flow coefficient and hence the stage total

pressure ratio is important to ensure repeatability in the measured total pressure

coefficient. The uncertainty in total pressure coefficient, calculated by the uncertainty

propagation method of Kline and McClintock [45], is δCpt = ±0.024 or ±0.58%.

48

The measured total pressure coefficient is circumferentially averaged over each

passage to obtain a passage averaged coefficient as defined in Equation 2-6, at each radial

position. The average is essentially the mean of 207 total pressure coefficient values. The

number of circumferential points per passage is obtained by dividing the total number of

points per rotation (6000) by the number of blades (29) and truncating the fraction. The

radial distribution of the passage averaged coefficient isolates the effect of OTL in each

passage and allows for comparison of tip gap behavior with and without coolant

injection.

2

0103

21

),(),(m

pt

U

PjiPjiCρ

−= (2-3)

des

operdes

mm T

TN

60r2U ∗=π (2-4)

60r2A where

TN

A21

1P

),(P

),( m2

des

des

01

03

π=

⎟⎟⎠

⎞⎜⎜⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛−

=

jiR

jiC pt (2-5)

∑+

=207

, ),()(i

iptPpt jiCjC (2-6)

49

For each test the total pressure coefficient is circumferentially averaged over the

entire rotor (6000 points) to obtain the rotor averaged coefficient as defined in

Equation 2-7. The radial distribution of the rotor averaged coefficient is assumed to be

invariant if the flow within the rotor passages is locally modified. Thus the effect of

changing the gap height of cooled blade B21 and the effect of coolant injection from B21

should produce no effect on the rotor averaged coefficient. This allows each data set to be

compared for repeatability and is used as a criterion for acceptable data.

An area averaged coefficient is also calculated for individual passages as defined

in Equation 2-8. The average is computed over an area of one passage (207 points) and a

height of 20% span starting at the tip. This coefficient gives a single value than captures

the effect over the entire passage.

2.8 Operation

Operation for total pressure measurements consists of starting up one of the axial

flow fans and allowing the turbine rotor to spin at part speed and load for 5 minutes. The

second fan is then started and the rotor is allowed to spin for an additional 5 minutes

∑=

=6000

1, ),()(

iptRpt jiCjC (2-7)

∑ ∑=passage heightblade

ptApt jiCC%20

, ),( (2-8)

50

before increasing speed to value indicated by inlet total temperature. The facility is then

allowed to run at constant speed for about 30 minutes in order to let the ambient air, the

facility, and the total pressure probe to achieve thermal equilibrium. Turbine speed is

corrected if necessary and data acquisition is commenced when the measured flow

coefficient is close to the required value. Data acquisition is triggered by a one per rev

trigger pulse from the shaft encoder and 6000 data points are acquired per revolution.

Upon completion of data set at a given radius the probe is automatically traversed radially

to the next radial position. After each movement a 5 second delay occurs before data

acquisition. Two radial positions are measured in a minute of run time. During the test,

speed changes are effected depending on the change in temperature. For tests involving

coolant injection the coolant flow is started at least 5 minutes before data acquisition.

This allows the coolant flow to stabilize.

Operation procedure for flow visualization experiments is detailed in subsequent

chapters. The start-up process is much quicker and the turbine rotor is brought to

operating speed within 2 minutes. Visual observations were conducted during the

operation to help terminate the test.

51

Chapter 3

Rotor Flow Visualization

Surface flow visualization is a qualitative flow measurement technique that is

used frequently in the study of internal and external flows. Although surface flow

visualization in the study of turbine passage flows is common in cascades, results from

rotating turbomachinery flows are rare. Langston [46] notes that surface flow

visualization has been extensively used in cascades to explain complex cascade passage

flows and also to guide the alignment of probes for measurements near surfaces.

Merzkirch [47] defines mechanical interaction techniques for surface flow visualization

as those that consist of application of dots or film of oil and pigment mixture to the

surface being studied. The slope of the surface shear stress lines indicates the ratio of

two-dimensional shear stresses, as given by Equation 3-1, from Langston [46]. The flow

is assumed to be in the x-z plane, with y-axis normal to the plane. The velocities are u in

the x-direction and w in the z-direction.

0

0)(

=

=

∂∂

∂∂

=

y

y

yu

yw

dxxdz (3-1)

52

An ink-dot-film technique of surface flow visualization was developed by Eckerle

and Langston [48] and used to study the formation of horseshoe vortex due to the

interaction of endwall boundary layer and a vertical cylinder. Aunapu et al [49], [50] used

a similar technique in their investigation of cascade endwall flow and the effect of an

endwall fence on cascade secondary flows. Allen and Kofskey [51] employed smoke

flow visualization to study the development and interaction of secondary and tip leakage

flows in a rotating rig. Dring and Joslyn [52] studied blade surface and tip surface flows

by injecting ammonia from surface pressure taps. The ammonia interacted with diazo

paper applied to the endwall surfaces and indicated the direction of surface streamlines.

Results from the application of mechanical interaction techniques to visualize flows on

rotating surfaces are rare in the public domain.

In this chapter the visualization of rotor endwall flow, blade tip surface flow, and

turbine casing endwall flow is presented. It is believed that the surface flow visualization

results discussed in this and subsequent chapters are the first such surface flow

visualization results from a rotating rig. The visualization material used consists of a

mixture of oil and pigment as detailed in Section 2.5 and following Merzkirch’s [47]

terminology the techniques used will be referred to as oil-dot technique or oil-film

technique. The surface shear stress lines formed by the visualization material are referred

typically referred to as surface or limiting streamlines. They will be referred to as streaks

in this document.

53

3.1 Rotor Hub Endwall Flow Visualization

Rotor endwall (hub) surface flow visualization was conducted using oil dots and

oil film, applied to the rotating endwall surface. Oil dots provide discrete surface markers

that enable unambiguous representation of surface flow. The origin of the visualization

material and surface streamline is easily identified when using dots. The information

gained from this is useful in analyzing the patterns left in the oil-film, which provides a

better global view of the surface flows. The origin of the visualization material is not

clear when using the oil-film technique and as suggested by Langston [46] the results

must be interpreted carefully since the accumulation of visualization material might be

due to absence of wall shear, as in separated flow, or due to application of transported

visualization material. In all tests the timing was determined by visual observations and

ranged from 10 minutes of steady operation when using oil dots to 20 minutes when

using oil film.

3.1.1 Visualization with the Oil-Dot Technique

The rotor endwall surface was cleaned and the visualization mixture was applied

to the endwall surface in the form of a grid of dots. The visualized streaks on the rotor

endwall are shown in Figure 3.1. The negative of the original image is presented for

improved clarity and hence the blade surfaces appear to be in shadow.

Figure 3.2, from Xiao [3], shows the static pressure distribution on the rotor

endwall. Points A and B denotes local maxima in static pressure, while point C denotes

54

the minimum static pressure measured on the rotor endwall. Streaks near the rotor inlet

plane, shown in Figure 3.1, illustrate the complex boundary layer development in this

region. The oil streaks at the leading edge of blade B23 are tilted in the z-direction, up to

about mid-pitch (Point 1), as the inlet flow moves around the suction surface. The flow

from mid-pitch to about ¾ pitch (Point 2) is directed axially and towards the suction

surface of B23 in the last quarter of the passage width. This behavior is consistent with

the static pressure distribution in Figure 3.2 where strong pitch-wise gradients occur near

the blade leading edge and a large, mostly uniform pressure field occurs centered around

mid-pitch. The lowest static pressure in the front half of the passage is measured in an

area approximately symmetric about the point of maximum blade curvature. Figure 3.1

shows that all visualization material in the passage up to the dotted white line is swept

into this area, marked by the dashed white curve.

Visualization material along the pressure surface – endwall corner is swept

towards the suction surface of blade B23. The orientation of the streaks changes rapidly

from the leading edge to a maximum angle of 75° near the point of maximum curvature

on the pressure surface. This may be explained by the strong, cross-passage pressure

gradients seen in Figure 3.2. The highest static pressure is measured at B and the isobars

up to mid-pitch have a strong axial orientation, indicating that the boundary layer flow is

mostly in the negative z-direction. The pressure field changes character around mid-chord

with stronger chord-wise gradients appearing. This is reflected in the oil streaks in

Figure 3.1. Streaks near the suction surface – endwall corner are mostly parallel to the

suction surface. Oil streaks in the passage are directed towards rotor exit, rather than the

suction surface of blade B23.

55

Figure 3.1: Rotor Endwall Surface Flow Visualization Using the Oil-Dot Technique.

1

2

B23

B22

x

z

56

Figure 3.2: Rotor Endwall Static Pressure Distribution, From Xiao [3].

57

3.1.2 Visualization with the Oil-Film Technique

Figure 3.3 shows the rotor endwall with an initial oil film layer applied over the

surfaces of two blade passages. Oil upstream of the leading edge is applied with a

tangential orientation, while oil within the passage follows approximately the contour of

the passage. This reduces the ambiguity that the observed patterns are due to paint brush

strokes during application. While the oil film is thicker in Passage 1, the thickness of oil

application is reasonably uniform in each passage.

Figure 3.3: Oil Film On Rotor Endwall Before Test.

Tangential brush strokes

Brush strokes aligned with passage curvature

58

Figure 3.4 shows the surface streaks formed in the oil film, in the near leading

edge region. Streaks upstream of rotor inlet and on the rotating hub are oriented mostly in

the axial direction, indicating that inlet boundary layer development on the rotating hub is

predominantly in the axial direction. As with visualization using oil dots, the orientation

of the oil streaks varies in the pitch-wise direction. In region 1, the streaks are inclined

towards the pressure surface, while in region 2 the streaks are pointing away from the

blade pressure surface. In between these regions and immediately upstream of the blade

leading edge the inflow is axial. A detailed description of inlet boundary layer behavior

as it approaches a bluff body is given by Eckerle and Langston [48]. The inlet boundary

layer stagnates as it approaches the blade leading edge (LE) and eventually separates

from the rotor endwall. This leads to the accumulation seen within the oval region. The

separated flow then impinges on the leading edge and is turned back towards the

stagnating flow. As these opposing flows meet they create a saddle point and fluid is

turned to flow around the blade leading edge. This leads to the formation of the two legs

of the horseshoe vortex.

The pressure-side leg (Hp) is driven across the passage by transverse pressure

gradient and is clearly identified in the patterns formed on the rotor endwall surface. The

vortex impinges closer to the pressure surface of the blade. Two dividing lines are

formed, one between Hp and the inlet flow, and the other between Hp and cross-passage

flow. The vortex moves in a straight line except for a small shift to the right, probably

caused by the strong cross-passage flow. The suction-side leg (Hs) of the horseshoe

vortex is not as distinct. However, close to the leading edge the dividing line separating

inlet boundary layer and Hs is identified by noting that the inlet boundary layer and the

59

re-circulating flow approach the line from opposing directions, as shown by the dotted

lines. The vortex impinges on the blade suction surface and is subject to stretching due to

acceleration. A large amount of visualization material is washed away from the rotor

endwall, indicating increased wall shear. This is consistent with the presence of thin

boundary layers due to acceleration around the suction surface curvature.

Figure 3.4: Near Leading Edge Surface Flow Features on the Rotor Endwall Visualized

Using the Oil-Film Technique.

2

1

Saddle Point

Hp

Hs

L.E.

x

z

60

Figure 3.5 shows an overall picture of the rotor endwall surface flow. The

pressure-side leg of the horseshoe vortex is marked by continuous curves. The curves are

drawn in regions where the visualization material is completely washed out, as in the

region close to the leading edge marked by a dotted circle. The pressure-side leg of the

horseshoe vortex divides the surface flow in to two distinct regions. On the impingement

side of the vortex the passage boundary layer is highly skewed by cross-flow from

pressure surface to suction surface. Similar boundary layer behavior is also observed in

cascade tests as reported by, among others, Gregory-Smith [53].

Flow on the other side of the pressure-side leg is blocked from following the

passage flow and instead flows towards the suction surface. The lower dividing line

becomes a little indistinct as the vortex approaches the blade suction surface, while the

upper dividing line is seen to continue downstream, as observed by the lack of

visualization material parallel to the blade split line. The rotor endwall surface static

pressure distribution in Figure 3.2 shows that strong chord-wise gradients occur in this

region, attributed to the passage vortex. Thus it is expected that increased wall shear due

to the entrainment of cross-flow by the pressure leg of the horseshoe vortex causes

greater oil to be removed in this region. Three dashed, white curves also indicate the

streak curvature, as the cross-flow approaches the split line. The last, dotted curve drawn

turns towards rotor exit after the split line. Indeed, downstream of this curve it can be

observed that visualization material crosses and also accumulates in the split line.

Therefore, the dotted curve probably represents the location where the passage vortex

lifts off the rotor endwall. The dividing line of the suction leg of the horseshoe vortex is

indistinguishable beyond point A. The continuous curve drawn leading up to this point

61

indicates a bifurcation pattern observed close to the leading edge. Thus it is suggested

that the suction leg of the horseshoe vortex lifts off the endwall in this region.

Accumulation of visualization material in this region of the rotor endwall also indicates a

reduction in wall shear.

Figure 3.6 displays surface flow features in the near trailing edge region of the

blade passage. The secondary flow is seen to encompass the entire passage. Some oil

Figure 3.5: Blade Passage Surface Flow Patterns on the Rotor Endwall Visualized Using

the Oil-Film Technique.

A

x

z

62

accumulation is evident on the endwall-blade corner, indicating reduced wall shear. The

green, dashed curve drawn separates the surface flow into two regions. Wall shear stress

above this curve appears to be lower than that below the curve. This could be a result of

higher velocity away from the pressure surface due to the significantly lower endwall

static pressures shown in Figure 3.2. At rotor exit, flow from pressure surface to mid-

pitch appears to exit the rotor at an angle of 64°, close to the design relative flow angle of

65°, as marked out. The blade wake is identified by the region of low removal rate of

visualization material. Flow closer to the suction surface is overturned, due to the passage

cross flow, and not much visualization material reaches beyond the rotor exit plane.

Visualization material is also swept up the blade suction surface, as shown in

Figure 3.7. The tip gap height for this blade is (t/h = 0.72%). The most significant

observation is the large white streak, enclosed between the two continuous curves drawn,

just below the blade tip, starting at about mid-chord and extending the length of the blade

surface. The mostly radial lines are drawn out by the visualization material as it is forced

up the blade suction surface by centrifugal action on the oil reaching the rotor endwall –

suction surface corner. It is clear that in the enclosed region the visualization material is

subject to intense, probably turbulent, shear that prevents the formation of the radial

streamlines. Due to its proximity to the blade tip the streak is believed to be caused by the

interaction between the tip leakage vortex and the tip passage vortex. The streak starts off

from the tip, in a radially inward trajectory before turning to move in the stream-wise

direction. The size of the streak also increases with chord-wise distance and at the trailing

it extends from about 13% blade height to about 25% blade height.

63

Figure 3.6: Near Trailing Edge Surface Flow Patterns on the Rotor Endwall Visualized

Using the Oil-Film Technique.

Blade Wake

64°

x

z

64

Figure 3.7: Blade Suction Surface Trace Formed During Rotor Endwall Surface Flow

Visualization.

Lines caused by visualization material from rotor endwall moving up the blade surface due to centrifugal action

Streak of visualization material due to interaction between tip leakage and tip passage vortices

65

3.2 Blade Tip Surface Flow Visualization

Blade tip surface flow visualization was conducted by applying oil dots on the

pressure surface of certain rotor blades. The visualization material moves up the blade

pressure surface under the action of centrifugal forces and is eventually thrown off the

blade surface at the pressure-side corner. The effect of centrifugal force on the

visualization material is greatest at the start of the experiment due to its finite volume.

During operation the thickness of the visualization material decreases and the effect of

aerodynamic forces is greater than that at start-up. Some of the material is carried on to

the tip surface by leakage flow entering the gap. The visualization material deposited is

then subject to the action of aerodynamic shear forces on the tip surface. The use of oil

dots gives discrete markers on the surface and information gained from this is then

applied towards understanding patterns that are caused by the application of oil film on

the pressure surface. These results will be discussed in subsequent chapters.

Figure 3.8 shows the pressure surface of blade B21 (t/h = 1.40%), before and after

the test run. The dots were applied about 1” below the tip surface, in two staggered rows.

The motion of the dots up the blade surface is governed by the ratio of the centrifugal

force to aerodynamic shear. It is clear from the second picture that the centrifugal force is

the dominant force on the visualization material over most of the blade surface, as the

dots move in discrete lines, directed radially outward. Towards the trailing edge however

the aerodynamic forces increase due to acceleration of flow over the rear half of the

blade. The lines are seen to tilt towards the trailing edge and visualization material in the

last 3-4 columns does not reach the tip surface.

66

The tip surface patterns for this test are shown in Figure 3.9. All dot traces, except

those close to the trailing edge, deposit material on to the tip surface. A continuous

limiting streamline is visible on the tip surface along the blade length indicating chord-

wise flow in the near pressure-side corner of the tip surface. Oil streaks directed towards

the suction-side corner of the tip gap are seen to originate from this limiting streamline.

As will be shown in the next chapter a strong re-circulatory pattern is observed on the tip

surface and the limiting streamline is likely formed at the edge of the re-circulating flow

and the corner separated flow. This chord-wise flow was also measured by Bindon [7] in

a linear cascade setup and observed in surface flow visualization of an idealized tip gap

by Sjolander and Cao [9].

Figure 3.8: Oil Dots on Blade (B21) Pressure Surface Before and After Test Run.

BEFORE RUN

AFTER RUN

67

Figure 3.9: Tip Surface Flow Visualization (t/h = 1.40%) by Oil Dots Applied Near

Blade Tip.

0.2 Cax.

0.5 Cax.

Limiting streamline moves towards suction-side of tip surface

H2 at 0.71 Cax.

H1 at 0.61 Cax.

Low momentum region

Minimal transfer of oil on to tip

surface

68

The limiting streamline is first visible, as shown in the enlarged detail in

Figure 3.10, close to 0.2 Cax. It stays very close to the pressure-side corner until about

mid-chord and then moves observably towards the blade suction surface. The location of

the limiting streamline in relation to the trench shows that it passes closest to the injection

hole at 71% chord. The limiting streamline is not observed over the last 5% of the blade

chord length. In this region, visualization material from the pressure-surface does not

reach the tip surface as most of it is turned around the trailing edge. Additionally, as will

be shown in the next chapter, gap flow fails to reattach on to the tip surface in this region.

A second test was conducted by applying the oil dots near the blade root. The

blade tip surface was also coated with smooth, flat black paint. The tip surface pattern

from this is shown in Figure 3.11. The repeatability between the two tests is very good.

The limiting streamline is seen to follow the same path as shown in Figure 3.9. There is a

noticeable shift towards the blade suction-side corner between 50% and 60% chord and

then the limiting streamline runs fairly parallel to the pressure-side corner. The lack of oil

up to mid-chord is believed to be due to insufficient visualization material reaching the

tip surface.

The tip surface flow patterns were also investigated on a blade with half the gap

height, t/h = 0.71%. The surface flow pattern is shown in Figure 3.12. A curved streak is

observed on the tip surface at about 0.3% Cax. The limiting streamline beginning from

this point terminates at about 0.4 Cax in a long straight streak across the tip surface. The

lack of curvature of this streak suggests that the near surface gap flow is mainly

responding to the gap pressure differential and is not affected by the chord-wise pressure

differential across the gap. The break in the limiting streamline might be due to a sudden

69

increase in gap velocity in this region. Beyond this point the limiting streamline begins

again at 0.45 Cax and is continuous up to the trailing edge. It is closer to the blade

pressure-side corner, in comparison to the large gap height previously discussed. The

streaks that cross the tip surface after mid-chord show greater inclination towards the

trailing edge, when compared to the single streak at 0.4 Cax.

Figure 3.10: Near Leading Edge Detail of Tip Surface Flow Visualization Using the Oil-

Dot Technique (t/h = 1.40%).

≅ 0.2 Cax.

70

Figure 3.11: Tip Surface Flow Visualization (t/h = 1.40%) Using the Oil-Dot Technique;

Oil Applied Near Blade Root.

0.5 Cax.

≅ 0.55 Cax.

Limiting streamline moves towards suction-side of tip surface

71

Figure 3.12: Tip Surface Flow Visualization (t/h = 0.71%) Using the Oil-Dot Technique.

0.3 Cax.

0.4 Cax.

0.45 Cax.

72

3.3 Rotor Casing Surface Flow

Rotor casing surface flows were visualized using both the oil-dot technique and

the oil-film technique. The oil dots were placed in three staggered rows along the rotor

axis and covered the rotor footprint. Figure 3.13 is an image taken during while the

turbine was running. Position of a few oil dots is marked with circles. The slope of the

surface streaks depends on the ratio of the shear stress in the x-direction and in the z-

direction. The characteristic velocity in the x-direction is the axial component of velocity

and in the z-direction is the tangential component added vectorially to the blade speed.

Visualization material displaces almost linearly in the first 30% blade chord and

then turns smoothly towards the axial direction. Most of the oil dots coalesce in this

region to generate a single streak. The orientation, with respect to the axial direction, of

the linear portion of the streaks is marked at various positions. The first two streaks, S1

and S2, are inclined at 87° degrees and 84° degrees respectively, which is greater than the

70° degree design angle. The proximity of these dots to the vane suction surface suggests

that this orientation is a result of overturning caused by vane passage secondary flow.

Similar pattern of overturned flow was observed on the rotor endwall with oil film

visualization. Indeed, these streaks merge immediately downstream of the vane trailing

edge and follow a path that is parallel to the vane trailing edge camber-line.

The third and fourth oil streaks, S3 and S4, which are immediately upstream of the

rotor inlet plane, are measured at 75° and 66°, respectively. Near blade mid-chord the

streak (S5) is measured at 46°, indicating almost equal shear stresses in the xy and zy

planes. The tangent to S4, measured at mid-chord is also inclined at 47° and as observed

73

the shear stress in the xy plane becomes more dominant with chord-wise distance. Dots in

the last 1/3 of the blade passage follow a very shallow path and are quickly turned

towards the axial, and then in the –z-direction beyond the rotor exit plane. The angle

measured for the streak S6 is -26°, which is close to the design exit angle at blade tip. It is

observed that not all the streaks are turned away from the direction of rotation. This

behavior is emphasized in Figure 3.14 acquired after the rotor was stopped. The two

streaks SA and SB indicate that the surface flow is as per design. The streaks close to SA

show similar orientation, while the streaks just below SB are directed axially. This axial

orientation changes as we approach streak SA. It is believed that the axial orientation of

the streaks is caused by the highly overturned flow from the nozzle guide vane exit flow

field being dominated by strong endwall cross flows.

3.3.1 Oil-Film Visualization of Rotor Casing Surface Flow

Rotor casing flow visualization was also conducted by applying an oil film to the

transparent casing window, as shown in the image on the left in Figure 3.15. The initial

oil film coverage is a little over one nozzle (N) pitch circumferentially, and about 1.5

rotor tip axial chord length, in the axial direction. Application of the visualization

material (brush strokes) is axially oriented. The image on the right in Figure 3.15 was

acquired after the test was completed.

74

An image of the outer casing flow patterns taken after the test was completed and

rotor was stopped is shown in Figure 3.16. The orientation of the streaks observed is very

similar to that observed with the oil-dot visualization. A region of decreased shear stress

is identified by a pair of white dotted lines. The fluid in this region appears to originate

from near the suction surface of nozzle guide vane N1 and the visualization material ends

up at a location exactly downstream of nozzle guide vane N2. This confirms the previous

hypothesis that the axial orientation of the streaks downstream of the rotor is caused by

the highly overturned flow exiting the nozzle passage. The reason for the greater

visualization material within this feature suggests that it is the path of the vane tip

passage vortex. Downstream of rotor exit there are regions of differing oil accumulation.

Oil streaks leading into region A appear to cover the greatest distance, carrying more

visualization material. It is clear from this picture that the wall shear imposed on the

visualization material is significantly larger in the footprint of the rotor than downstream

of the rotor.

75

Figure 3.13: Casing Surface Flow Visualization During Turbine Operation Using the Oil-

Dots Technique.

S1 (87°)

S2 (84°)

S3 (75°)

S4 (66°)

Streamlines converge

S5 (48°)

S6 (-26°)

Vane T.E.

Vane T.E.x

z

76

Figure 3.14: Rotor Casing Surface Flow Visualization Using the Oil-Dot Technique.

SA (-26°)

Vane (N2) T.E.

Vane (N1) T.E.

SB

Axially directed streamline(s)

x

z

77

Figure 3.15: Rotor Casing Surface Flow Visualization Using the Oil-Film Technique;

Before and After Test.

Blade Trailing

Edge

Vane Trailing

Edge

78

Figure 3.16: Rotor Casing Endwall Surface Flow Visualization Using the Oil-Film

Technique.

N1

N2

A

79

Chapter 4

Effect of Tip Gap Height on Over Tip Leakage

The results presented in this chapter discuss the effect of tip gap height on over tip

leakage (OTL) flow and the observed total pressure downstream of the rotor. Changes in

the tip surface flow due to variation in gap height were investigated qualitatively, by

using the oil-film technique of surface flow visualization. As introduced previously, OTL

flow is ejected into the passage from the blade suction-side and forms a vortex in the

blade passage thereby affecting the passage flow. The variations in leakage vortex

footprint are studied in the stationary frame by measuring total pressure downstream of

the rotor. A high resolution total pressure map at stage exit in the cold research turbine

could be used as a measurement of aerodynamic efficiency. One of the objectives of this

chapter is to set-up baseline data to compare to the effect of coolant injection.

4.1 Oil Film Based Tip Surface Flow Visualization

The oil film technique of surface flow visualization is implemented to study over-

tip leakage. The oil dot technique discussed in the previous chapter was able to

distinguish the chord-wise flow on the tip surface. However identification of micro-flow

features, as observed in the case of rotor endwall flow visualization, is difficult when

using discrete flow markers.

80

4.1.1 Operation

The blade surfaces are first coated with smooth flat black paint to allow for better

contrast with the white pigment used and to avoid reflections during image acquisition.

Visualization material is then applied on the pressure surface, with a soft brush as evenly

as possible. Images of the blade pressure surface with the un-disturbed oil film were

recorded before each test. The optical window was then fastened and the first fan stage

was started, which brought the turbine rotor to a stable speed of about 1240 rpm in

approximately 30 seconds. The second fan stage was then started and both the turbine

rotor speed and flow rate stabilized sufficiently in 30 seconds. Thereafter, the rotor speed

was increased to the corrected operating speed. Thus, the time required for the rotor to

reach operating speed was less than 90 seconds in a total run time of about 20 minutes. A

strobe light controlled by the AFTRF encoder one per rev., pulse was used to conduct

visual observations of the blades during the tests.

During the startup, the thickness of the visualization material on the blade

pressure surface is reduced, as the centrifugal action forces oil to move up the blade

surface and eventually the oil is ejected off the surface. It was observed that some of the

oil started splattering on the casing when the speed reached about 700 rpm. Some of the

oil thrown off the pressure surface is carried into the tip gap by the leakage flow. Thus,

by the time the rotor reached operating speed some amount of visualization material is

carried on to the tip surface. The thickness of the visualization material deposited on the

tip surface is extremely small and was observed to move in well defined dots across the

tip surface, under the influence of the aerodynamic shear generated by the gap flow. The

81

test was terminated when no motion was observed for over a minute, which on average

was 20 minutes after startup. The extremely thin streak patterns observed are thus due to

steady or time averaged flow in the tip gap. It must be noted that the mixture never dried

hard and could be smeared even after the test. Tests were also conducted to ensure that

the oil deposited on the tip surface did not dry by the time the rotor reached operating

conditions.

4.1.2 Tip Surface Flow Visualization; Large Gap Height (t/h = 1.4%)

Flow visualization was done on blades with different gap heights from a

maximum of 1.4% to a minimum of 0.72%. The surface patterns are most distinct at the

largest gap height and are discussed first in detail. At the smaller gap heights many of the

same features are present and the discussion focuses more on drawing out the differences.

4.1.2.1 Blade Pressure Surface

Figure 4.1 shows images of the blade pressure surface with oil film before (on the

left) and after (on the right) the visualization experiment. The initial oil film is brush

painted so that it has no preferred direction. Care is taken to ensure reasonably uniform

film thickness. A definite radial orientation of the oil film is visible in the image acquired

after the test. The oil layer on the surface of the rotating concave surface is forced

radially outwards due to centrifugal forces on the oil film. During startup the

aerodynamic shear, imposed by the passage flow in the relative frame, on the oil layer is

82

less significant in comparison to the centrifugal forces due to the initial thickness of the

oil film. The patterns on the pressure surface are not indicative of the passage flow.

However, the curvature of the patterns is indicative of the relative strengths of

aerodynamic and centrifugal forces imposed on the oil film. The blade surface is visible

in some areas where all the paint has washed out.

4.1.2.2 Blade Tip Surface

Figure 4.2 shows the tip surface patterns on the test blade (B21) with a sharp

corner, flat tip, and a tip gap height of t/h = 1.40%. The trench is blocked by applying a

smooth and thin layer of tape on the tip platform and cutting it to the tip shape. An

Figure 4.1: Oil Film on Blade Pressure Surface Before and After Test.

BEFORE TEST

AFTER TEST

83

extremely thin layer of smooth flat black paint is then sprayed on the tape. A grid of lines

is drawn at intervals of 0.05 Cax to allow positioning of various features.

One of the general observations that can be made is the oil accumulation on the

tip surface all along the PS edge of the blade. As compared to the rest of the tip surface,

only a few streaks are seen to originate from the accumulated oil up to 0.30 Cax along the

pressure-side corner. Additionally, these streaks do not extend to the suction-side corner

of the tip surface. As discussed in the previous chapter, the oil-dot technique also showed

a limiting streamline originating at about 0.20 Cax. Gap flow in this region comprises

mainly of the inlet flow, thereby generating lower wall shear on the tip surface. The

streaks that do form indicate turning of gap flow towards the camber-line as it progresses

to the suction-side of the tip gap. Blade tip and casing surface pressure measurements by

Xiao et al [54] in the AFTRF indicate minimal gap pressure differential in this region.

Similar behavior is also observed in linear cascade, shroud surface pressure

measurements by Bunker et al [22] and numerical simulation of flow in the Bunker

cascade by Ameri [29]. The numerical simulation also shows that gap flow streamlines in

the front quarter of the tip surface originate primarily from the rotor inlet flow. A definite

reattachment and recirculation pattern is observed on the tip surface from about 0.3 Cax,

extending the length of the tip surface.

Enlarged images of the surface patterns are shown in Figure 4.3 and Figure 4.4 for

clarity. A dashed curve near the leading edge of the tip surface, in Figure 4.3, indicates

the oil streak due to inlet flow passing through the tip gap. Close to the suction surface of

the gap the streak curves towards the exit and this is likely due to the gap flow

responding to the chord-wise pressure gradient set up on the suction surface. The

84

continuous curve traced on the tip surface, in Figure 4.3, is a limiting streamline between

flow towards the blade suction-side corner and re-circulating flow towards the pressure-

side corner. Oil carried by the leakage flow entering the tip gap, around the pressure-side

corner is deposited on to the tip surface. Subsequently, wall shear forces the oil towards

either the suction-side corner or the pressure-side corner. Wall shear near reattachment is

low and hence, according to Yang [55], is expected to yield a wide zone of oil

accumulation. In the present study however, the position of the reattachment is believed

to be fairly accurate due to the bifurcation pattern observed. Chord-wise orientation of

streaks at reattachment may indicate a three-dimensional reattachment. The direction of

streaks in the recirculation zone indicates that flow within the recirculation bubble is

towards the trailing edge. The recirculation patterns between 0.4 Cax and 0.5 Cax are not

as well defined. The smearing could be a result of the separation bubble lifting off the tip

surface and being ejected into the passage, as conceptualized by Bindon [7].

Measurements by Xiao [3] in the AFTRF show tightly spaced pressure contours in this

region of the tip gap. The rapid acceleration of fluid entering the gap is accompanied by

increasing wall shear, causing more of the visualization material to be washed away.

Immediately behind this smeared region and in the direction of the leakage flow is an

area that contains very little visualization material. Since visual observations indicated no

wetness, due to oil, it is possible that most of the oil carried into the gap does not reach

the tip surface in this region.

The features from blade mid-chord to the trailing edge are very similar, oil

accumulation at the pressure-side corner, a reattachment line on the tip surface, and over

tip leakage flow towards the suction-side corner are observed in Figure 4.4. The near

85

pressure-side corner oil accumulation is caused by both the re-circulating flow and the

separated flow in this region. The location of the reattachment line normal to the

pressure-side corner changes very little along the blade length. Similar behavior was

observed with the line denoting chord-wise flow in the previous chapter. It is believed

that the limiting streamline obtained through the oil-dot visualization occurs between the

recirculation bubble and flow within the separation bubble near the pressure-side corner.

Reattachment occurs at about 4% Cax (2*gap height) from the pressure-side corner

between 0.3 Cax and 0.65 Cax. The distance decreases at about 0.75 Cax along PS edge,

where the reattachment occurs at about 3% Cax. The reattachment occupies half of the

gap at about 0.8 Cax and the gap flow is fully separated in the last 5% of the blade chord.

The oil streaks leading towards the suction-side corner are more or less normal to

the camber-line up to about 0.8 Cax along camber-line, after which the streaks are turned

more towards the camber-line. An interesting observation is the path followed by the oil

after impinging on the tip platform. The turning of the streaks is seen to be more gradual

for flow towards the suction-side corner than in the direction of the bubble. Therefore, the

re-circulating flow experiences greater acceleration towards the pressure-side corner.

This behavior is consistent with the significantly low wall static pressure that occurs in

the near pressure-side corner of the tip surface, as measured by Xiao [3] in the AFTRF.

After about 0.6 Cax the streaks turn more sharply towards the suction-side corner, due to a

drop in chord-wise pressure gradient in the region of separated flow on the tip surface.

The surface streamlines observed are similar to the streak patterns, on a plane parallel to

and very close to the tip surface, obtained from a steady computational simulation by

Prasad and Wagner [56].

86

Figure 4.2: Surface Flow Patterns on Tip Surface of Blade (B21) With a Gap Height of t/h = 1.40%, Visualized Using the Oil-Film Technique.

Figure 4.3: Surface Flow Patterns on Front Half of Tip Surface of Blade B21, Using the Oil-Film Technique.

Recirculation

Normal Leakage Direction

Reattachment Inlet Boundary Layer Fluid

Accumulation Due to Separation

Normalized distance markers

Chord-wise Flow

87

4.1.2.3 Heat Transfer Implications

The tip surface flow patterns observed also have important heat transfer

implications. Low heat transfer areas may be expected in about the first 1/3rd of the tip

surface due to inlet boundary layer flow through the gap. Impingement of over tip

leakage flow causes higher heat transfer to the region near the pressure-side corner of the

blade. Acceleration of fluid into the recirculation bubble results in higher wall shear,

enhancing heat transfer to the tip surface around the reattachment line. Tip heat transfer

coefficients measured by Kwak et al [57] in a linear cascade show a region of maximum

heat transfer coefficient beginning at the pressure-side corner and extending towards the

blade suction surface. This region also spans most of the profile length. Additionally,

chord-wise flow in the recirculation bubble could trap high temperature fluid entering the

blade row, thereby exposing the near pressure-side corner of the tip surface to high total

temperatures. Previous hot streak studies, for example Prasad and Hendricks [58],

indicate that the maximum temperature in the core of the passage flow entering the blade

Figure 4.4: Surface Flow Patterns on Rear Half of Tip Surface of Blade B21.

Reattachment lineFully separated flow region0.95Cax

Reattachment lineFully separated flow region0.95Cax

Gap flow turning towards T.E.

Normal gap flow direction.

88

row usually migrates to the PS corner of the blade tip. Fully separated gap flow, as

observed in the last 5% blade chord, would lead to decreased heat transfer coefficients

due to low momentum activity in this region.

4.1.3 Tip Surface Flow Visualization; Small Gap Height (t/h = 0.71%)

The tip surface flow patterns on blade B7 (t/h = 0.71%), with half the clearance

gap as that of the test blade (B21) are shown in Figure 4.5. The patterns on blades B21

(t/h = 1.4%), B7 (t/h = 0.71%), and B2 (t/h = 0.81%) were obtained in a single test.

Minimal streaks are formed in the region up to about 0.25 Cax along the pressure-side

edge. The existing streaks are curved, indicating acceleration towards the suction-side of

the gap. Normal to 0.35 Cax along the pressure-side, a circle identifies a region of

smeared surface lines, similar to that shown in Figure 4.3 and attributed to the possible

ejection of the separation bubble. Thus, there seems to be a shift towards the leading edge

of some of the consistent flow features when the gap height is reduced.

The region near the pressure-side corner indicating reattachment and recirculation

is reduced at the smaller clearance gap, which causes greater oil accumulation along the

pressure-side edge. Results shown in Chapter 3, where the chord-wise, limiting

streamline formed closer to the pressure-side edge, support this observation.

Reattachment occurs at about 2% Cax (2.1*gap height) from the pressure-side edge. In a

direction normal to the pressure-side corner and in the region between 0.65 Cax and 0.7

Cax along the pressure-side there is a region of higher shear, as evidenced from the

observation that more of the tip surface is visible. This is not observed for large tip

89

clearance (t/h = 1.4%) and hence would seem to be characteristic of smaller clearance

gaps. In the oil-dot visualization shown previously it was observed that the only streaks

seen on the tip surface were in this region. As will be shown later in the paper, suction

surface traces show that significant leakage flow appears to enter the passage just

upstream of this region. Hence, the higher velocity and shear, in this region could be due

to proximity of the leakage vortex to the suction surface. Measured blade pressure

coefficient in Xiao, et al [54] shows a significant dip in the suction surface pressure

around 0.6 Cax. The return of dense streak patterns to the left of this region would

indicate the movement of the leakage vortex away from blade suction surface. The

leakage vortex forms farther away from the suction surface when the gap height is large

(t/h = 1.4%) and hence has a reduced influence on the gap flow.

There is no indication of a fully separated flow over the tip surface at the small

gap height, as even at the trailing edge there is a distinct direction to the streaks

indicating flow post reattachment towards the suction-side edge. Oil Streaks closer to the

trailing edge are however not oriented normal to the camber-line. This is attributed to the

increased tangential momentum of the leakage flow as it enters the tip gap. It was shown

earlier that pressure surface patterns qualitatively indicate increased wall shear due to

passage flow. The gap flow starts to turn away from normal to camber-line at about 0.8

Cax along the camber line. Bindon [7] indicates that in the near TE region of the tip, flow

in the separation bubble may be moving away from the trailing edge. The orientation of

streaks in the recirculation zone indicates however that the flow is indeed towards the

trailing edge. While it is not clear from the oil patterns, it is expected that the bubble is

ejected into the blade wake.

90

4.1.4 Tip Surface Flow Visualization of Other Gap Heights

Visualization was also carried out for two other gap heights, of t/h = 0.81% and

t/h = 1.2%. The tip surface patterns are shown in Figure 4.6 and Figure 4.7 respectively.

A contoured, precision plastic layer was applied to the tip surface of B21 with double-

sided adhesive tape, to obtain the gap height of t/h = 1.2%. Blade B2 was used without

modification for t/h = 0.81%.

At both gap heights the oil accumulation at pressure-side corner is observed,

along with the re-circulation. The differences between gap height 0.71% and 0.81% are

minimal. It does appear that the reattachment line is a little farther away from the

pressure-side corner, as the recirculation patterns are better defined at this gap height.

There is a region of greater oil removal from 0.65 Cax to 0.8 Cax, similar to that observed

at the smallest gap height. Gap flow is still fully attached near the trailing edge.

Figure 4.5: Surface Flow Patterns on Tip Surface of Blade (B7) With a Gap Height of t/h = 0.71%, Visualized Using the Oil-Film Technique.

0.1

0.2 0.3

0.4 0.5 0.6

0.7

0.8

0.9

1.0

Region of higher wall shear stress

91

The tip surface patterns in Figure 4.7 describe a flow field similar to that

discussed up to now. With increase in gap height the recirculation patterns tend to be

more distinct, as the reattachment line moves farther away from the pressure-side corner

leading to lesser accumulation on the tip surface. The high shear area identified at the

smaller gap heights is not evident on the tip surface. There are indications of separated

flow in the last 5% blade chord.

The distance of the reattachment line from the pressure-side corner was measured

on the images at 0.6 Cax and is presented in Figure 4.8. The distance is measured normal

to the axial chord (lx) and also normal to the camber-line (lc). The measurement is

rendered non-dimensional with respect to the blade tip axial chord length and the gap

height. The variation of the measurement with gap height is plotted. The variation with

gap height of the distance normalized by blade tip axial chord is linear. The exception to

this behavior is the measurement at t/h = 0.81%, which is believed to be due to

measurement accuracy. The linear fit is not unexpected as previous research has indicated

that characteristic measures of OTL, such as gap mass flow rate, total pressure losses and

efficiency drop, vary linearly with gap height. The location of the reattachment line

measured normal to camber-line and normalized with gap height does not change much

with gap height. Indeed the outlier in this case is again the measurement at t/h = 0.81%.

Thus it can be concluded that the reattachment occurs about 2*t from the pressure-side

corner. Of course it must be noted that this might only be valid for a range of gap

pressure difference.

92

Figure 4.6: Surface Flow Patterns on the Tip Surface of Blade (B2) With a Gap Height of t/h = 0.81%, Visualized Using the Oil-Film Technique.

Figure 4.7: Surface Flow Patterns on Tip Surface of Blade (B21 at reduced gap height) With a Gap Height of t/h = 1.2%, Visualized Using the Oil-Film Technique.

0.1

0.2 0.3

0.4 0.50.6

0.7

0.8

0.9

1.0

Region of higher wall shear stress

Re-attachment line

Accumulation due to corner separation

0.1

0.2 0.3

0.40.5 0.6

0.7

0.8

0.9

1.0

More distinct re-attachment line

Accumulation due to corner separation

93

4.1.5 Suction Surface Traces from Oil Film Based Tip Surface Flow Visualization

Visualization material carried across the tip surface is deposited on the blade

suction surface in the form of a trace, similar to that observed in the case of rotor endwall

flow visualization. Images of the suction surfaces of the blades at all gap heights were

also acquired and are shown in Figure 4.9 and Figure 4.10. The oil that is deposited on

the suction surface is carried over by the leakage flow as it exits the gap suction-side and

enters the passage. It was shown in Chapter 3 that visualization material from oil film on

0.0%

2.0%

4.0%

6.0%

8.0%

10.0%

12.0%

0.5% 0.7% 0.9% 1.1% 1.3% 1.5%

Tip Gap Height, t/h [-]

Loc

atio

n of

Rea

ttac

hmen

t, [-

]

0.00

0.50

1.00

1.50

2.00

2.50

3.00

Loc

atio

n of

Rea

ttac

hmen

t, l c

/ t [-

]Measured normal to axial chord, lx/t

Measured normal to camber-line, lc/t

lc / t

Figure 4.8: Influence of Gap Height on the Location of the Visualized Reattachment Line on the Blade Tip Surface.

94

the rotor endwall was forced up the blade suction surface and resulted in a similar trace.

This was attributed to visualization material being transported in the interaction zone

between the tip leakage vortex and the tip passage vortex. It must be reiterated that the

oil-pigment mixture used for gap heights of 0.71%, 0.81%, and 1.40% was from the same

batch. It appears that as the gap height is decreased the start point of the trace is better

defined. The traces appear to originate on the suction surface in the region 0.65 Cax to

0.75 Cax. McCarter [16] showed through velocity measurements in the AFTRF that the

leakage flow at 0.7 Cax is very close to the suction surface and develops rapidly at 0.8

Cax.

The distance from the tip surface to the upper boundary of the trace (h1) and width

of the trace (h2) were measured off the blade surfaces and are presented in Figure 4.11 as

a function of the gap height. The measurements are normalized by the blade height (h)

and both measurements vary linearly with gap height. As the gap height is reduced the

lower boundary of the trace moves up towards the blade tip and the width of the trace

grows smaller. Thus, it might be concluded that as the gap height is decreased the leakage

vortex is closer to the blade surface, as well as closer to the blade tip. The measurements

for blade B7 (t/h = 0.71%) are smaller than that observed in Figure 3.7 and this is

attributed to the origin of the visualization material. In the results presented here the oil is

carried by the gap flow, while in Figure 3.7, visualization material is present on the

suction surface, in addition to OTL capturing oil thrown off the blade suction surface.

95

Figure 4.9: Suction Surface Traces Formed During Oil-Film Based Tip Surface Flow

Visualization; Gap Heights, t/h = 1.4% and t/h = 1.2%.

Figure 4.10: Suction Surface Traces Formed During Oil-Film Based Tip Surface Flow

Visualization; Gap Heights, t/h = 0.81% and t/h = 0.72%.

h2

h1

96

4.2 Total Pressure Measurements

Quantitative results from total pressure measurement are presented in this section.

Total pressure measurements downstream of the rotor are recorded for all 29 blade

passages using a phase-locked measurement technique. The results are presented in form

of contour plots that show the total pressure distribution in the r-θ plane, radial

distributions of passage averaged and rotor averaged coefficients, circumferential

distribution of total pressure coefficient at individual radial locations, and area averaged

total pressure coefficient for individual passages. The relevant equations are in Section 2..

The term “Base#” is used to refer to tests conducted without coolant injection and with

0.02.04.06.08.0

10.012.014.016.0

0 0.5 1 1.5Gap Height, (t/h %)

SS T

race

Mea

sure

men

t, (h

1 /h %

; h2

/h %

)

h1/h

h2/h

Figure 4.11: Effect of the Tip Gap Height on Measurements From Suction Surface Trace

Formed During Oil-Film Based Visualization of Tip Surface Flow.

97

the cooled blade B21 gap height of t/h = 1.40%. Initial data sets cover a span from 0.051h

to 0.981h. This gives an overall picture of passage flows, including near rotor endwall

flow. Since variations below 0.438h were found to be unaffected by gap height, within

uncertainty limits, most of the data sets cover a span from 0.438h to 0.981h.

The radial distributions of the passage averaged and rotor averaged total pressure

coefficients are plotted as continuous curves, rather than points. The curve itself is

generated by a three point moving average of the data with the end points unchanged.

The passage averaged coefficient is computed for the passage that is bounded on the

suction surface by the blade number referenced. So, a passage averaged coefficient for

B21 is the measured total pressure coefficient averaged across the passage that is

bounded by the suction surface of B21 and pressure surface of B20. Thus, the passage

defined contains the leakage vortex of B21. The passage is defined to start at the edge of

blade wake and core passage flow and contain 207 points. The circumferential

distributions obtained at a fixed radial position are termed as “wake plots” throughout the

manuscript.

4.2.1 Baseline, No Injection

Baseline tests were conducted with a test blade (B21) tip gap height of t/h =

1.40%. The rotor tip clearance distribution is TCL1, as shown in Figure 2.5. Thus, the

baseline is relevant to only the isolated coolant injection study. The measurements were

conducted using the high response total pressure probe described in Section 2.6.2.1. The

probe is in the stationary frame of reference and samples the total pressure field 6000

98

times for each revolution of the rotor. Thus a complete map of the rotor exit total pressure

field is obtained, an example of which is briefly discussed in Appendix C. The contour

plots presented show only a sector containing the cooled blade.

Figure 4.12 is a contour plot of the total pressure coefficient, as defined in

Equation 2-3. Radial locations corresponding to the hub, the casing, and 50%, 75%, and

85% blade span are shown in solid curves. The number (#) and gap height (t/h) of each

blade is shown above the casing boundary. The direction of rotation is from right to left,

as represented by the blade speed vector below the hub boundary. The total pressure

distribution over five passages, with the cooled blade B21 at the centre, shown is

designated Base1. Solid boundaries in the tip region and dashed curves in the mid-span

region are drawn to visually track changes in the total pressure field, which is possible

due to the phase-locked measurement technique employed. These details are consistently

reproduced in all subsequent contour plots.

Figure 4.12: Total Pressure Coefficient Contours With No Coolant Injection (Base1).

99

4.2.1.1 Region above 85% Blade Height

This region is dominated by two distinct flow regimes. The tip leakage vortex,

seen in the wake region of each blade, is characterized by low total pressure and a near

elliptical shape. The shape results from the total pressure data being projected onto a

plane normal to the rotor axis. The leakage vortices in each blade passage are seen as the

lowest total pressure regions in the passage. The size of the leakage vortices and the total

pressure defect increase with increasing gap height. The entrainment of low momentum

fluid by the leakage flow and diffusion beyond the rotor exit plane has caused the blade

wake to be completely overshadowed by the leakage flow.

The leakage vortex of B21 occupies about 15% of the blade span and extends well

into the blade passage. The vortex has a well defined core and strong gradients in both

radial and circumferential directions. The total pressure at the core is approximately 1*qm

lower than the maximum total pressure coefficient measured. The minimum test blade

total pressure, measured in the core of the tip vortex, is also lower than that of the

neighboring blades shown because of the much larger gap height of this blade.

The second distinct flow feature in the last 15% blade height is the intermediate

total pressure zone between the two adjacent tip vortices. This zone (Cpt=-3.85, green) is

bounded by the outer casing that is in relative motion with respect to the rotating passage,

the core flow in the middle of the passage and the two subsequent tip vortices. The

secondary flow pattern near the outer casing is more distorted in comparison to its

counterpart near the hub surface because of the leakage flow. Velocity measurements by

McCarter et al. [20] show higher axial and reduced tangential velocities in this region.

100

This is consistent with under-turning of the passage flow due to the blockage presented

by the tip leakage vortex. Such under-turning would be the greatest in the passage

containing the test blade leakage vortex, and hence the measured total pressure would be

lower than that of neighboring passages.

4.2.1.2 Region 75% - 85% Blade Height

The main flow feature present in this region is the tip-side passage vortex, as

marked in the figure. The identity of this structure is also confirmed by results in

McCarter et al [20]. It exists just below the tip leakage vortex and has a smaller total

pressure defect than that of the leakage vortex. The interaction of the passage vortex with

the leakage vortex is greatest for the test blade, as seen by the poor definition of the

passage vortex in the wake. Another interesting characteristic of this region is a band of

lower total pressure that runs across the passage, culminating in the passage vortex. This

suggests that the secondary flow effect is greatest in this region, with the flow probably

being over-turned as it migrates towards the suction side of the passage.

4.2.1.3 Passage Core

The passage core occupies about 50% of the passage. It has a well defined core of

high total pressure at around 50% blade height. Just below 75% span, close to the blade

suction-side, a dip in the passage core is observed. This is possibly due to a vortical

structure. The extent of the dip suggests that this structure is formed early in the blade

101

passage and is convected downstream. Below 50% span the passage core is strongly

affected by the hub-side passage vortex, which rotates in the same direction as the

leakage vortex. Although the hub-side passage vortex is not as clearly discernible as the

tip-side passage vortex, its effect on the passage core is clear. A better definition of the

hub-side passage vortex is clearly observed in the past, phase-locked LDA measurements

shown in Figure 4.13, obtained by Ristic et al. [59] in the same facility. It must be noted

that the nominal absolute flow angle at the hub is 35°, which is a 10° incidence on the

probe. Additionally, the passage flow tends to be over-turned closer to the hub, due to the

rotor endwall passage vortex. In the passage containing the test blade, the passage core is

shifted radially downwards, probably by the radially inward migration of fluid due to the

blockage of the tip leakage vortex. The core in the passage formed by B21 & B22 is at a

higher radius, suggesting a strong radial flow towards the tip due to the enlarged

clearance of the test blade.

Figure 4.13: Secondary Flow Vectors At Rotor Exit From LDA Measurements by Ristic

et al [54].

Hub-side passage vortex

102

4.2.2 Repeatability

Repeatability of baseline data is presented through contour plot in Figure 4.14,

radial distribution of rotor averaged pressure coefficient in Figure 4.15, wake plots at two

radial locations in Figure 4.16 and Figure 4.17, and radial distribution of passage

averaged coefficient in Figure 4.18. A total of three baseline data sets were acquired over

a six month period. Figure 4.14 is a contour plot from the last data set, designated Base3,

shows good comparison of the total pressure coefficient distribution with Base1. While

the radial extent of the data set is shortened, the magnitude of total pressure coefficient

and the geometric definition of flow structures match well with that shown in

Figure 4.12. The radial distribution of circumferentially averaged total pressure

coefficient, along all 29 passages in the rotor, also shows good repeatability. In the tip

vortex dominated zone and core of the passages the repeatability is sufficiently

maintained for three baseline experiments performed at different dates. While general

repeatability is good, greater deviation occurs for Base1 in the region between 70% and

80% blade height, which in some instances is at the limit of the uncertainty band. Overall

repeatability for the measurements is deemed to be good within a 95% confidence level.

The circumferential distribution of total pressure coefficient at selected radial

locations of r/h = 0.96 and r/h = 0.57 is shown in Figure 4.16 and Figure 4.17,

respectively. These locations correspond respectively to the location of the leakage vortex

dominated zone and the core flow, where minimal influence of the leakage vortex is

expected. Data shown extends from blade B19 to blade B23. Overall, the repeatability is

very good at both radial locations. The influence of the test blade, with a relatively large

103

tip clearance of t/h = 1.40%, is noticeable in Figure 4.16 as the largest total pressure

defect induced by the large tip leakage vortex. However, when the same distribution at a

lower radius (r/h=0.57) is examined, the tip vortex influence diminishes.

Radial distributions of circumferentially averaged total pressure coefficient for the

passage defined by blade B20 and blade B21 (test blade) are shown in Figure 4.18 . The

distribution is very similar to the results from all 29 passages (rotor averaged) except that

the tip region results are influenced by the large tip clearance of blade B21 (t/h=1.40%).

The effect of the increased clearance of the test blade is smoothed out in the rotor

averaged coefficient, and hence the peak in the passage averaged coefficient distribution

for r/h>0.85 may be interpreted as a total pressure defect zone. As in the case with rotor

averaged coefficient, the repeatability between the three baseline sets is very good, except

in the region between 70% and 80% blade height where Base1 displays a greater

difference, which in some instances is near the limit of the uncertainty band.

Figure 4.14: Total Pressure Coefficient Contours With No Coolant Injection (Base3).

104

4.2.3 Effect of the Tip Gap Height

Figure 4.19 shows a contour plot of total pressure coefficient obtained with the

test blade tip clearance is reduced to t/h=0.72%, from t/h=1.40%. The clearance was

reduced by applying precision plastic layer on to the tip surface with double-sided

adhesive tape. The effect of reducing the tip clearance is evident in the distribution and

level of measured total pressure at stage exit. The solid lines marked around the tip vortex

zones are from the baseline case and used for comparison purposes. The tip leakage

vortex of cooled blade B21 is much smaller in size and closer to the suction-side of the

Figure 4.15: Radial Distribution of the Rotor Averaged Total Pressure Coefficient;

Baseline Repeatability.

Uncertainty band

105

passage. The total pressure measured in the vortex is higher than that of the baseline by

about 0.2 qm. The passage flow above 85% span should experience lesser under-turning

and hence register as a higher absolute total pressure, in comparison to the baseline. The

tip-side passage vortex of blade B21 is better defined due to reduced interaction with the

leakage vortex. The passage core itself has rotated in a counterclockwise direction due to

the reduced blockage of the leakage flow. Importantly, the tip leakage vortices in the

neighboring passages are unaffected and hence the influence of changing the gap height

appears to be localized to the passage bounded by blades B20-B21.

The wake plot at r/h = 0.96, presented in Figure 4.20 clearly shows the movement

of the wake of blade B21 towards the suction surface of the blade. The blue triangles

indicate a significant total pressure defect in the tip leakage vortex when the clearance is

large, at t/h=1.40%. The total pressure defect is reduced (orange squares) and the

magnitude is similar to that of the neighboring blades. The wake plot at r/h=0.57,

Figure 4.21, shows a slight drop in the total pressure at the mid-span. This is probably

due to the aforementioned rotation in the passage flow due to a reduction in the influence

of the leakage flow on the passage flow. Tip vortex impact at this lower radial position is

not measurable in the passage of the test blade.

The radial distribution of the passage averaged pressure coefficient in Figure 4.22

shows that there is a definite shift in the flow field towards the blade tip, in addition to

the increase in total pressure. This confirms that the lower total pressure seen in the wake

plots is due to a change in the passage flow structure. The defect due to the tip-side

passage vortex is also better defined than in the baseline sets because the tip

vortex/passage vortex interaction is reduced when the tip gap is reduced to t/h=0.72% .

106

Figure 4.23 compares the radial distribution of the passage averaged coefficient

from two tests. The curves in RED were obtained with tip gap height of blade B21 at t/h

= 1.40%, while the curves in GREEN were obtained with tip gap height of blade B21 at

t/h = 0.72%. The passage averaged coefficient of passage containing the tip leakage

vortex of blade B7 (t/h = 0.71% in dash-dot curves) is compared with that of passage

containing the tip leakage vortex of blade B21 (t/h = 1.40%, continuous and dashed

curves). Blade B7 is diametrically opposite to blade B21 in the 29 blade rotor. The

repeatability of the radial distributions for blade B7 is good, as expected, since the effect

of changing the gap height of blade B21 is expected to create changes locally. The radial

distribution of blade B7 is almost identical to that of blade B21 with a tip gap height of

t/h = 0.72%. The tip gap heights of neighboring blades B6 is t/h = 0.73%, B8 is t/h =

0.71% , B20 is t/h = 0.77%, and B22 is t/h = 0.83%. Thus, it appears that the gap height

of neighboring blades has little influence on over tip leakage flow, as long as the

variation in gap heights is within 15%.

A comparison of the area averaged coefficient as a function of tip gap height is

shown in Figure 4.24. The area average is computed using Equation 2-8, over one

passage containing the tip leakage vortex of blades with varying tip gap heights and a

height of 20% span, from 0.8h – 1h. The variation of the area averaged coefficient is

quite linear. The maximum deviation of the area averaged coefficient based on the curve-

fit, from the actual measured data is 0.2% of the measurement at the tip gap height of t/h

= 0.92%. This trend agrees well with that observed in the measurements from flow

visualization.

107

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.6

-4.5

-4.4

-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

Base1: t/h = 1.40%, Minj = 0Base2: t/h = 1.40%, Minj = 0Base3: t/h = 1.40%, Minj = 0

Figure 4.16: Repeatability of Wake Profiles at r = 0.96h With No Coolant Injection.

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

-3.6

-3.5

-3.4

-3.3

Base1: t/h = 1.40%, Minj = 0Base2: t/h = 1.40%, Minj = 0Base3: t/h = 1.40%, Minj = 0

Figure 4.17: Repeatability of Wake Profiles at r = 0.57h With No Coolant Injection.

108

Figure 4.18: Repeatability of the Passage Averaged Coefficient For Cooled Blade B21.

Figure 4.19: Total Pressure Coefficient Contours With Tip Gap Height of Cooled Blade B21 Reduced to t/h = 0.72%.

Uncertainty band

109

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.6

-4.5

-4.4

-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

Base3: t/h = 1.40%, Minj = 0t/h = 0.72%, Minj = 0

Figure 4.20: Effect of Reducing the Tip Gap Height of Blade B21 On the Wake Profile at

r = 0.96h.

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

-3.6

-3.5

-3.4

-3.3

Base3: t/h = 1.40%, Minj = 0t/h = 0.72%, Minj = 0

Figure 4.21: Effect of Reducing the Tip Gap Height of Blade B21 On the Wake Profile at

r = 0.57h.

110

Figure 4.22: Effect of the Tip Gap Height On the Passage Averaged Coefficient of

Cooled Blade B21.

Figure 4.23: A Comparison of the Passage Averaged Coefficient Distribution For Blade

B7 and Blade B21.

-4.16 -4.04

111

-4.11-4.10-4.09-4.08-4.07-4.06-4.05-4.04-4.03

0 0.5 1 1.5 2Tip Gap Height, (t/h %)

Are

a A

vera

ged

Coe

ffic

ient

, Cpt

,A

Figure 4.24: Variation in the Area Averaged Total Pressure Coefficient with Tip Gap Height. (Area = 20% span*1 passage).

112

Chapter 5

Effect of Coolant Mass Flow Rate on Over Tip Leakage

The effect of coolant injection on over tip leakage is studied in two parts. This

chapter presents the effect of coolant mass flow rate on over tip leakage flow. Coolant is

injected from a single blade, referred to as the test blade (B21), and the series of tests are

referred to as isolated injection. Precision plastic layers were attached with double-sided

adhesive tape to the tip surfaces of the four other blades with injection capability. This

makes the four blades inactive from the coolant injection point of view and also changes

the rotor clearance distribution. The clearance distribution with these modifications is

TCL1, as specified in Figure 2.5. It was shown in Chapter 4 that the total pressure

signature of the tip leakage vortex of cooled blade B21 was not affected by the gap size

of blades preceding it. A single large disturbance was also preferable to maintain flow

repeatability of the rotor and allows for the measurement of the effect of an isolated blade

injecting coolant into the tip gap.

The effect of coolant mass flow rate is investigated both qualitatively, using flow

visualization, as well as quantitatively by measuring total pressure downstream of the

rotor exit. The mass flow rate of coolant injection into the tip gap is an important

parameter. Air used for turbine hot-section cooling is high-pressure air bled from the

engine compressor section. Thus, a part of total cycle work is trapped in the coolant air.

Since coolant does not participate in the work generation, cooling air represents a loss in

113

efficiency or in available work. This loss must however be compared to the benefits that

are accrued from coolant injection, for example, air used in internal blade cooling and

film cooling of turbine blades allows turbine inlet temperatures to be higher than metal

melting point, which in turn increases the thermal efficiency of the engine. In the case of

tip injection the efficiency lost due to injecting coolant into the tip gap must be balanced

against loss reduction in the high-pressure turbine stages due to its effect on the OTL

flow, as well as the heat transfer benefits. After the leading edge, the tip surface is the

hottest part of a turbine blade and heat transfer benefits might indicate that higher inlet

temperatures are possible, thereby increasing the overall thermal efficiency of the gas

turbine engine.

5.1 Visualizing the Effect of Coolant Injection

The flow visualization technique employed to study OTL flow was described in

the previous chapter. The advantage of using flow visualization with coolant injection is

that in addition to obtaining information on the fluid dynamic interaction between the

coolant jets and gap flow, potential heat transfer benefits may be identified, in the

rotating frame. As before the blade surfaces were coated with smooth flat black paint

prior to application of the visualization material. The optical window was then fastened

and the first fan was started, followed by the second fan in 30 seconds. The rotor speed

was increased to operating speed after flow through the facility had stabilized.

Coolant mass flow rate injected is stated as a percent of turbine mass flow rate,

assuming coolant is injected from all blades as shown in Equation 2-2. Four injection

114

mass flow rates, 0.4%, 0.5%, 0.6%, and 0.7% were tested. In all cases, except one,

coolant injection was initiated and stabilized prior to start-up. For the case of Minj = 0.4%

however, a test was conducted where injection was commenced after the rotor had

reached operating speed. The various test cases are summarized in Table 5-1 .

5.1.1 Effect of Tip Trench

Tip surface flow over a flat tip at a gap height of t/h = 1.40% was discussed in

Chapter 4. The effect of the tip trench was investigated using flow visualization, as

shown in Figure 5.1, at the gap height of t/h = 1.40%. As compared to the flat tip surface

flow patterns Figure 4.2, there is, as expected, little change in surface flow in the front

half of the blade. A distinct reattachment line starts to form near 0.35 Cax and the re-

circulation pattern is clearly visible. The reattachment line intersects with the near

pressure-side edge of the tip trench at about 0.6 Cax. In the region from 0.5Cax, where the

Table 5-1: Test Matrix of Flow Visualization with Coolant Injection.

Test Case Blade # t/h% Minj% Description

T1 21 1.40 0 Tip with trench

T2 21 1.40 0.4 Injection at N = 0

T3 21 1.40 0.4 Injection at operating speed

T4 21 1.40 0.5 Injection at N = 0

T5 21 1.40 0.6 Injection at N = 0

T6 21 1.40 0.7 Injection at N = 0

115

trench starts, to 0.55Cax the streaks in the leakage flow direction deposit visualization

material in the trench. It appears that gap flow does not reattach on the tip surface from

0.55 Cax to the trailing edge.

The streak pattern between the trench and the pressure-side corner indicates

recirculation up to H1, suggesting that the gap vortex is still present and is separated from

the tip surface. Recirculation is also observed beyond H2 at 71% chord. Between H1 and

H2 the region between the trench and the pressure-side corner shows more accumulation

than on the flat tip. It is likely that the trench weakens the recirculation and hence reduces

wall shear stress in this region. There is no accumulation in the trench, except close to 0.5

Cax, another indication that the gap vortex is lifted off the tip surface. One of the

interesting changes from the flat tip is the region of low oil concentration to the left of

H1. The reason for this is unknown and it is not observed at the other injection locations.

A limiting streamline, similar to that seen with oil dot visualization, is observed on the tip

Figure 5.1: Surface Flow Visualization of the Effect of Tip Trench on Cooled Blade B21.

H1H2H3H4

Oil in trench RecirculationChord-wise flow

0.1

0.2 0.3

0.40.50.6

0.9

1.0

116

surface injection hole H2. In visualization without the trench, recirculation due to the gap

vortex accumulates visualization material near the pressure-side corner of the blade and

likely prevents the formation of a clear chord-wise limiting streamline. Hence, this

observation confirms that the gap vortex influence on the tip surface is reduced.

5.1.2 Injection at Minj = 0.4% at Gap Height of t/h = 1.40%

This particular injection case was tested in two different ways, as described in

Table 5-1. Figure 5.2 shows the surface flow patterns when injection was commenced

before startup, while Figure 5.3 was obtained for injection after the turbine rotor reached

operating speed.

5.1.2.1 Injection Prior to Start-up

Figure 5.2 shows the surface patterns on the blade tip with injection before

startup. Surface flow patterns up to about 0.6 Cax along the PS edge are similar to those

observed for tip surface with trench and no injection, indicating no effect of coolant

injection in this region. A re-circulatory pattern appears around 0.35 Cax and extends up

to 0.6 Cax. Beyond this point the observed patterns are distinctly different due to the

effect of the coolant jets. The pressure-side corner accumulation is intermittent. Jet

penetration into the leakage flow, due to coolant injection from H1, leaves an arc-like

imprint in the accumulated oil, on the tip surface, between the trench and the pressure-

side corner. Separation is not completely eliminated due to the limited coverage

117

available. Additionally, divergence of the jets from each other allows oil to accumulate in

between the jets.

A large area with little to no streaks is seen between the trench and blade suction-

side corner, in the region surrounding H1. Thus the coolant jets spread out as they turn

and flow towards the suction-side corner of the blade. The leading jet does not lean

towards the trailing edge and appears to be turned back sharply. The trailing jet on the

other hand does have a 10° angle towards the trailing edge and is turned more gradually.

The lack of accumulation at 0.65 Cax along the pressure-side corner is due to the trailing

jet in H1 blocking the passage flow as it enters the gap. Once this jet is turned back it

covers a larger area on the tip platform than the leading jet. Immediately behind the

injection holes there are very few streaks, when compared to the no injection cases. The

streaks formed are located in between the injection holes and indicate low momentum

activity. The leakage flow that passes between the jets diffuses out on the tip surface and

Figure 5.2: Surface Flow Visualization of the Effect of Coolant Injection at Minj = 0.4% From Cooled Blade B21; Injection While Turbine at Rest.

Gap flow turning around coolant jet

Oil accumulates between jets

Little to no streaks

Trench accumulation

0.1

0.3 0.4

0.6

0.70.8

0.9

1.0

118

also gives up some energy in overcoming the shear imposed by the coolant jets. The

chord-wise flow on the tip surface would also be blocked by the coolant jets. No oil-dot

based visualization was conducted and hence it is difficult to say with any certainty the

path taken by this fluid in the separated flow region near the pressure-side corner of the

tip surface.

Accumulation of visualization material, due to pressure-side corner separation, is

observed between H1 and H2. The streaks in this region however are not as sharp as

those observed in the forward half of the blade and might be a result of visualization

material accumulated in the trench. It was shown earlier that with no injection there was

no accumulation in the trench and hence path of leakage flow is altered by the coolant

jets. Visualization material accumulates on the tip surface up to the trench and is

probably driven by gap flow forced in to this region by jet blockage. In the previous

chapter it was discussed that visualization material accumulates near the pressure-side

corner of the tip surface partly due to the separation effect and partly due to paint being

deposited by the re-circulating flow. The extent of deposition indicates that recirculation

is augmenting the material deposited by the separated flow.

The path of leakage flow around the leading jet from H2 is discernible by the

curvature of the streaks, particularly around the leading jet. The curvature suggests flow

of leakage fluid around the blockage presented by the coolant jets. The accumulation at

the pressure-side corner is lesser than that observed near H1. The jets are again turned

back, leaving a large clear region between the trench and the suction-side corner. There is

greater accumulation in the trench between H2 and H3 and consequently, the streak

patterns between trench and suction-side corner are indistinct. Tip separation is reduced

119

along the pressure-side corner from about 0.72 Cax to 0.85 Cax, the region of influence of

H2 and H3.

Injection from H3 is not much different in the effect it has on the gap flow. The

coolant jets start out towards the pressure-side and are turned to flow towards the suction-

side. The streak patterns between the trench and the suction-side corner, in the region

around H3, are considerably smeared and could indicate considerable mixing between

gap flow and coolant fluid. The location of this injection hole is close to the impingement

location of normal gap flow. The last injection hole (H4) is radial and hence allows for

separation to occur between 0.85 Cax and 0.9 Cax. It appears however to block the leakage

flow beyond 0.9 Cax, forcing the leakage to turn towards the blade trailing edge.

5.1.2.2 Injection at Operating Speed

Figure 5.3 shows tip surface flow patterns for the case when the injection is

initiated after the rotor reached operating speed. The patterns are qualitatively different.

In this case, by the time the injection is initiated, there is some deposition and shearing of

visualization material on the tip platform. Injection is fully stabilized 30 seconds after

rotor speed is set to operating speed. Subsequently, with injection it is expected that areas

of low flow momentum activity will display greater concentration of paint, while a lower

concentration of paint will indicate flow with greater momentum.

Interrogating the images with this perspective it is clear that both Figure 5.2 and

Figure 5.3 essentially represent the same flow patterns. Observations are confined to the

area beyond mid-chord. The first set of injection jets are turned towards the suction-side

120

by the pressure driven gap flow. Accumulation behind H1 is dense with no clear streak

pattern indicating low flow velocities in this region. It might be recalled that in the

previous discussion the patterns in this region were interpreted to show low wall shear

due to wake like flow after leakage fluid passes between the coolant jets. The

accumulation is particularly high behind the trailing jet of H1, which is expected since it

is along the path the leakage flow is expected to take. Penetration of the jets in to the

separation zone is indicated by the removal of visualization material close to the

pressure-side corner. To the left of H1 and between the trench and suction-side corner the

patterns are lighter than those observed in the forward part of the blade, which can only

happen if due to the greater wall shear generated by the coolant jets being turned around

to form a film over the tip surface. In Figure 5.2 this appears as a region clear of streaks.

Figure 5.3: Surface Flow Visualization of the Effect of Coolant Injection at Minj = 0.4% From Cooled Blade B21; Injection While Turbine at Operating Speed.

Accumulation due to low momentum activity

Coverage of coolant jets

121

There is very little accumulation in the trench, to the left of H1 this is expected as

injection was started after rotor reached operating speed. The accumulation that does

exist indicates that some visualization material is still being carried over from the

pressure surface even two minutes into the test. The effect of lower momentum flow is

also seen in the pattern between H2 and the SS edge. The accumulation of paint due to

lower momentum flow occurs to the left of the trailing jet, instead of behind it. The

patterns between H2 and H3, behind the trench are more distinct streak patterns than

those observed in the previous injection test. There is also some accumulation within the

trench. The mixing effect of the coolant jets is in clear evidence near the location H3. The

streaks that would have formed prior to injection, between the trench and the SS edge, are

completely smeared indicating much greater shear than without injection and hence

attributable to coolant injection. Accumulation along the pressure-side corner is

intermittent, indicating jet penetration into the leakage flow. To the left of the last

injection hole, at 0.95 Cax, a dense line of accumulation extends across the tip surface. In

Figure 5.2 this same feature is seen as a dark line due to lack of paint accumulation. This

line is probably caused by a change in the dominant shear direction due to the coolant jet

at H4. The dominant shear direction is expected to be normal to the pressure-side corner

and blockage due to the coolant jet at H4 causes the leakage flow to turn towards the

trailing edge.

One of the important pieces of information that this test yields is that the

visualization material is not dry and is able to trace out the path of the tip surface flow.

Since injection was commenced after the rotor reached the operating speed, if the

visualization material had dried out then the pattern should resemble that shown in

122

Figure 5.1 . However, the pattern is distinctly different and the physical reasoning behind

the appearance of the pattern is supported by Figure 5.2. Hence, it is concluded that the

visualization material is not dry, two minutes into the test and is influenced by wall shear.

5.1.3 Visualization at Other Injection Rates

The other injection rates tested were, Minj = 0.5%, 0.6%, and 0.7%. These results

are shown in Figure 5.4, Figure 5.5, and Figure 5.6, respectively. In each case, leakage

flow up to about 0.6 Cax is unaffected by the injection. While there is some change in the

density of streaks and their definition, the important features like reattachment and

recirculation are identifiable. The variations are believed to be caused partly by variations

in the mixture and partly by the operating temperature. These effects were not studied, as

it is believed that the available information is sufficient to make consistent observations.

Focusing on the flow patterns seen near injection location H1, higher jet

momentum has almost eliminated accumulation due to pressure-side corner separation.

Immediately behind the injection holes there is greater oil concentration than that seen for

the lowest injection rate, and this increases with increase in the injection rate. At the

highest injection rate of 0.7%, the streaks behind the injection holes show a converging

pattern. On either side of the jets, there is evidence of film formation, with the leading jet

displaying better coverage than the trailing jet. This is opposite to that observed for Minj =

0.4%. Additionally, it appears that the influence of the trailing jet begins farther

downstream along the blade chord than for Minj = 0.4%. This is again consistent in that

123

greater jet momentum will better resist turning due to leakage flow. There is also

significant accumulation in the trench.

Streaks formed due to injection from H2 and H3 are similar in nature to that near

H1. There is again very little accumulation near the pressure-side corner. The wake like

behavior of the streamlines between the trench and the suction-side corner is more

pronounced, especially near H2, as observed by the greater accumulation very near the

suction-side corner of the tip surface. In this region the leakage flow is expected to be

lower in momentum than that near H1 due to the natural reduction in driving pressure

differential across the tip surface. The most interesting change is exhibited by injection

from H3. The jet trajectories are defined better at the higher injection rates of Minj = 0.6%

and Minj = 0.7%. The leading jet appears to be more effective in covering the tip surface

than the trailing jet. The influence of the trailing jet is farther to the left of the jet.

Injection from H4 appears to stagnate around the hole and flow over the SS edge into the

passage.


PS corner accumulation eliminated

Trench accumulation

124



0.1

0.2 0.3

0.40.5 0.6

0.7

0.8

0.9

1.0

0.1

0.3 0.40.5

0.6 0.7

0.8

0.9

1.0

PS corner accumulation and streamlines near T.E.

125

5.1.4 Suction Surface Traces

The suction surface traces of the tip leakage vortex are shown in Figure 5.7 for

Minj = 0.4%, 0.5%, and Figure 5.8 for Minj = 0.6%, 0.7%. For Minj = 0.4% the streak

pattern is similar to that obtained without tip injection and is seen from approximately 0.6

Cax. This is actually closer to the leading edge than in the case of t/h = 0.71%. It appears

that blockage due to coolant injection from H1 is causing the leakage flow to turn away

from the trailing edge. Visual observations showed that there was more oil deposition.

This could result from the leakage vortex moving closer to the suction surface. The width

of the trace, at the trailing edge, is about 10% blade height and the clear region measures

5% blade height from the tip. In comparison to the flat tip case there is some movement

of the vortex closer to the blade tip. The trace appears to start closer to the trailing edge

for Minj = 0.5%. Indeed, the trace is not continuous, rather shows multiple leakage entry

paths, at 0.8 Cax and 0.88 Cax along the SS edge, into the passage. There is a oil free

region between the two entry points and the streak extends 8% blade height from the tip.

The greatest oil deposition occurs after 0.95 Cax.

Injection at Minj = 0.6% also displays multiple entry points along the suction

surface, the first one located at 0.75 Cax along SS edge. Maximum oil deposition occurs

beyond 0.95 Cax. The trace measures 10% blade height from the tip. At Minj = 0.7 % the

first clear entry point is at 0.8 Cax. There is a very light trace that appears to start at 0.73

Cax. It was noticed that even when using the same mixture, the sharpness of the patterns

varied, probably due to initial room temperature. This might be a reason for the traces to

show up much later along the suction surface for two higher injection rates. The trace

126

extends 12% blade height from the tip for the highest injection rate. The maximum paint

deposition occurs after 0.95 Cax along SS edge. It is clear that other than at the lowest

injection rate, the injection jet near the trailing edge does cause leakage flow near the

trailing edge to be channeled into the passage. From a heat transfer perspective these

results show that the heat transfer coefficient on the suction surface may be greater than

that for t/h=0.71%. It also identifies a potential problem area, close to the trailing edge

region where the heat transfer may actually be increased beyond that experienced by a

blade with t/h=0.71%.

Figure 5.7: Suction Surface Traces for Minj = 0.4% and Minj = 0.5%.

Minj = 0.4% Minj = 0.5%

127

5.1.5 Heat Transfer Implication

The tip trench is located ideally, to prevent strong reattachment on the tip surface.

Additionally, coolant injection reduces the effects of flow separation, albeit

intermittently, and reattachment on the tip surface. The recirculation of high temperature

fluid in the separation bubble is also affected. In the previous chapter, the reattachment

and associated recirculation were identified as the possible reasons for the enhanced heat

transfer observed on the near pressure-side corner of the tip surface. Thus, the location of

the trench and coolant injection should decrease the heat transfer coefficient on the tip

surface influenced by the coolant jets. The surface flow patterns also indicate that in the

region between the tip trench and the suction-side corner of the tip surface a reduction in

tip heat transfer is possible. Blockage of the leakage flow leads to low momentum

Figure 5.8: Suction Surface Traces for Minj = 0.6% and Minj = 0.7%.

Minj = 0.6% Minj = 0.7%

128

activity between the injection holes and the suction-side corner. Additionally, the coolant

jets turn towards the suction-side corner and form a film over the tip surface. Both these

effects would also reduce heat transfer to the tip surface.

The picture is not all positive however. Clearly the presence of reattachment near

the trailing edge is of concern as this will lead to greater heat transfer coefficients in this

region of the tip surface. However, with the temperature of the fluid considerably lower

than inlet temperatures, it is possible that this modification will not cause undue heating

of the tip surface. Radial holes are probably not the most effective way to cool the tip

surface, while jets angled towards the pressure-side and also towards or away from the

trailing edge appear to allow better area coverage.

5.2 Total Pressure Measurement

As in the previous chapter the results are presented in form of contour plots, radial

distributions of passage averaged coefficient, wake plots, and area averaged total pressure

coefficient. The comparisons made in this section are with respect to the baseline data

shown in Figure 4.14. Injection was initiated from the test blade only.

Contour plot for injection at Minj = 0.41% is shown in Figure 5.9. There is

considerable reduction in the tip leakage vortex of the test blade, as measured by the total

pressure downstream of the rotor exit. The minimum total pressure in the wake increases

by about 0.2 qm. This effect is similar in magnitude to that obtained with reduced tip

clearance of t/h = 0.72% and no injection. However, the flow structure is less like a

vortex with a smaller pitch-wise footprint. Additionally, the severe gradient observed in

129

the tip leakage vortex due to the large gap height of t/h = 1.40% is also eliminated. The

vortex appears to hold its position, as there is no observable movement towards the blade

suction surface, as seen when the gap height was reduced to t/h = 0.72%. Injection has no

effect on either the passage core flow or the interaction between the leakage vortex and

the tip side passage vortex. The wake plot in Figure 5.10 shows the reduction in wake

depth with tip injection as compared to the baseline. While there is some thinning of the

wake, the wake does not move to the right.

Figure 5.9: Total Pressure Coefficient Contours with Coolant Injection at Minj = 0.41%.

130

When the coolant mass flow rate is increased to Minj = 0.52%, Figure 5.11, the

leakage vortex core shows a slightly increased total pressure drop, as compared to Minj =

0.41%. The measured total pressure is however greater than that measured for the large

tip clearance. The shape of the vortex is also better defined than for Minj = 0.41%. The

size of the leakage vortex is smaller than that observed at the large gap height however

the vortex shows no appreciable movement towards the blade suction surface. The

interaction between the tip side passage vortex and the leakage vortex is still strong and

no effect is observed on the passage core. The wake plot, Figure 5.10, only serves to

confirm that the energy defect, while not as great as that observed for the large tip gap is

slightly greater than that obtained at the lower injection rate.

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.6

-4.5

-4.4

-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

Base3: t/h = 1.40%, Minj = 0t/h = 1.40%, Minj = 0.41%t/h = 1.40%, Minj = 0.52%

Figure 5.10: Wake Profile at r = 0.96h, Without and With Coolant Injection at Minj =

0.41% and Minj = 0.52%.

131

Injection at higher coolant mass flow rates begins to change the nature of the

leakage vortex from the test blade. As shown in Figure 5.12, at a coolant mass flow rate

of Minj = 0.63%, the tip leakage vortex from the test blade is not only reduced, but also

shifted more towards the casing. This observation may be expected when the gap height

is reduced. The tip-side passage vortex of the test blade is better defined due to reduced

interaction between the two flow structures. Injection also affects the tip leakage vortex

due to blade B22. The tip vortex of blade B22 is shifted towards the blade suction surface

and has a higher value of total pressure associated with it. Figure 5.13 compares the wake

profile with injection Minj = 0.63% () and the baseline and it is clear that the total pressure

defect of blade B21 is reduced. The reduction is however not as great as compared to

Figure 5.11: Total Pressure Coefficient Contours With Coolant Injection at Minj = 0.52%.

132

injection at 0.41%. The wake shift and defect reduction in the wake of blade B22 is also

seen in the wake profile.

At the highest coolant mass flow rate of Minj = 0.72%, shown in Figure 5.14, the

tip leakage vortex from blade B21 is shifted towards the casing and also appears to have

moved out into the passage. The increase in total pressure coefficient is similar to that

observed for coolant mass flow rate of Minj = 0.63%. The higher jet momentum however

affects the tip leakage vortex of blade B22 even more. This leakage vortex is shifted even

more, this is more apparent when the position of the leakage vortex is related to that of

the tip-side passage vortex. The structure also shows signs of complete mixing and a

greater total pressure coefficient. It appears that the jet momentum is also affecting the tip


133

leakage vortex of blade B23, as a slight shift is observed. In the baseline it was shown to

be inside the tip-side passage vortex, while injection at Minj = 0.72% has actually moved

it to the outside, probably due to jet penetration.

In all cases, the passage core flow is not affected much, as seen by the boundaries

representing the baseline in the contour plots. Indeed when comparing the wake plots

with and without injection at a radial location of 0.57h, shown in Figure 5.15, it is seen

that there is practically no change in the wake structure close to mid-span due to injection

from the blade tip.

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.6

-4.5

-4.4

-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

Base3: t/h = 1.40%, Minj = 0t/h = 1.40%, Minj = 0.63%t/h = 1.40%, Minj = 0.72%

Figure 5.13: Wake Profile at r = 0.96h, Without and With Coolant Injection at Minj =

0.63% and Minj = 0.72%.

134


Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

-3.6

-3.5

-3.4

-3.3

Base3: t/h = 1.40%, Minj = 0t/h = 1.40%, Minj = 0.41%t/h = 1.40%, Minj = 0.52%t/h = 1.40%, Minj = 0.63%t/h = 1.40%, Minj = 0.72%

Figure 5.15: Wake Profile at r = 0.57h, Without and With Coolant Injection.

135

5.2.1 Comparison of Averaged Values

The radial distribution of passage averaged values of total pressure coefficient for

passage containing tip leakage vortex of blade B21, shown in Figure 5.16 compares the

distributions for gap heights of t/h = 1.40%, t/h = 0.72%, and coolant injection at a gap

height of t/h = 1.40%. One of the general observations that may be made is that coolant

injection reduces the distinct peak due to the large leakage vortex of blade B21. The

radial distributions with coolant injection are very close to that obtained at gap height of

t/h = 0.72%. In addition, the averaged total pressure coefficient in the leakage vortex

affected zone is similar to that in the tip-side passage vortex zone. The radial distributions

of passage averaged coefficient with coolant injection at gap height of t/h = 1.40% and

that due to gap height of t/h = 0.72% lie within the uncertainty band in the region from

0.75h – 1h. Thus, it appears that in the range of coolant mass flow rates studied there is

about equal effect and this effect is similar to that of reducing the gap height of the

cooled blade B21 to t/h = 0.72%.

Figure 5.17 compares the area averaged total pressure coefficient in the passage

containing the tip leakage vortex of cooled blade B21. The area averaged coefficient is

shown for tests without and with injection at the large gap height of t/h = 1.40%, and also

at the small gap height of t/h = 0.72%. The extent of the area average covers one passage

and 20% blade height. The small gap height (in blue) is used as a reference value against

which the effectiveness of coolant injection may be compared qualitatively. All area

averaged values due to coolant injection at gap height of t/h = 1.40% lie on the line

representing the averaged value of gap height t/h = 0.72%. Thus, it is possible to

136

conclude that the aerodynamic effect of the large gap height of t/h = 1.40% has been

modified by coolant injection to that due to a gap height of t/h = 0.72%.

Figure 5.16: Effect of Coolant Injection On the Passage Averaged Coefficient of Cooled Blade B21.

Uncertainty band

137

-4.15

-4.10

-4.05

-4.00

-3.95

-3.90

0 1 2 3 4 5

[-]

Are

a A

vera

ged

Tot

al P

ress

ure

Coe

ffic

ient

Minj = 0.41%Minj = 0.52%

Minj = 0.63%

Minj = 0.72%

t/h= 1.40%

t/h= 0.72%

Figure 5.17: Area Averaged Coefficient For Blade B21 With Coolant Injection at a Tip Gap Height of t/h = 1.40%.

138

Chapter 6

Effect of Injection Location on Over Tip Leakage

The chord-wise location of coolant jets is also an important parameter in

optimizing coolant injection. The focus of this chapter is the effect of injection from

individual locations and combinations of locations on OTL flow. As before, coolant is

injected from a single cooled blade, referred to as the test blade (B21) and the rotor

clearance distribution is TCL1, as shown in Figure 2.5. Test cases are referred to by the

number of the injection locations active in each case. Thus, H1 refers to injection from

the location 61% chord (H1) and H1+H3 refers to combined injection from the locations

61% chord (H1) and 81% chord (H3). Table 6-1 summarizes the various tests conducted.

Injections holes were rendered inactive by blocking the holes with silicone sealant. The

sealant was allowed at least 24 hours to set and the seal was pressure tested at 40 psig,

both before and after the test. Thus it was ensured that coolant was injected from only the

desired locations. Only total pressure measurements were done to study the effect of

injection location. In all cases the ATS pressure was maintained constant at 10 psig. This

was done since in actual engines the coolant stream total pressure is maintained constant.

The choice of pressure was made to conform to the lowest injection mass flow rate tested

and also to maintain the total pressure ratio to below critical pressure ratio. The injection

rate was stabilized for at least 5 minutes before data acquisition was commenced.

139

6.1 Injection from Individual Holes

Figure 6.1 shows the total pressure contours when injecting coolant from H1 only,

located at 61% chord. Comparing the pitch-wise extent of the vortex to the boundary

marking the vortex with no injection, it is apparent that coolant injection has reduced the

size of the vortex. The minimum total pressure measured in the leakage vortex has not

changed appreciably, however the core of the vortex appears to have shifted closer to the

test blade suction surface. The strong gradient characterizing the tip leakage vortex due to

a large tip gap height is also observed.

Table 6-1: Test Matrix for Effect of Injection Location.

Test Case Active Injection Hole

Minj%

H1 H1 0.2

H2 H2 0.2

H3 H3 0.2

H4 H4 0.2

H1+H2 H1 and H2 0.3

H1+H3 H1 and H3 0.3

H2+H3 H2 and H3 0.3

H1+H2+H3 H1, H2, and H3 0.35

Full Injection All locations 0.42

140

The pitch-wise extent of the leakage vortex is also reduced when injecting from

H2 (at 71% chord) only as shown in Figure 6.2. The reduction in total pressure defect is

greater than when injecting from H1 only and the gradient across the tip leakage vortex is

smoother. Figure 6.3 shows the effect of injecting from H3 (at 81% chord). The reduction

in pitch-wise extent is similar to that observed for H2 injection. The total pressure defect

however has been significantly reduced, by about 20% qm. This difference is similar to

that observed, as shown in the previous chapter, both when the gap height was reduced to

t/h = 0.72% and also with coolant injection at Minj = 0.42% into a gap height of t/h =

1.40%. The vortex core is more uniform with smoother gradients across the vortex. The

passage flow above 85% blade height and between blades B20 and B21 appears to be less

energetic. Coolant injection from H4 has very little effect on the leakage vortex, as shown

Figure 6.1: Total Pressure Coefficient Contours for Coolant Injection From Blade B21 and Location H1 at 61% Cax.

141

in Figure 6.4. The size of the vortex is about the same as that in the baseline and the low

total pressure core has moved back to a location comparably similar to that observed in

the baseline. There does appear to be some reduction in the total pressure defect. The

more significant effect of coolant injected from H4 is the reduction in total pressure

measured in the passage above 85% blade height. It is not possible to confirm that this is

due to additional total pressure loss.


142



143

The reduction in the passage area occupied by the leakage vortex and the

movement the tip leakage vortex towards the blade suction surface can occur if the gap

normal momentum is reduced, similar to when the gap height is reduced. Thus coolant

injection from locations H1, H2, and H3 does cause reduction in gap normal velocity,

either by forcing the gap flow to turn towards axial or by reducing gap mass flow rate and

producing a wake like behavior. The gap thickness to gap height ratio and the gap

pressure differential is greatest at H1. The gap flow may be expected to contain a wake

region due to diffusion beyond the vena-contracta, as reviewed earlier. Coolant jets at H1

originate very close to the location of reattachment and may cause a more intense

diffusion of the gap flow around the blockage presented by the coolant jets. This

reduction in normal gap momentum manifests itself as a reduction in the pitch-wise

extent of the tip leakage vortex.

The greatest defect reduction is observed to occur due to coolant injection at the

81% blade axial chord location. Flow visualization results indicate that injection hole H3

is located close to the reattachment line. Additionally, the tip gap surface flow is seen to

turn towards the trailing edge, due to a drop in pressure difference across the gap. Flow

visualization also indicates intense mixing of gap flow and injected coolant. It would

appear then that the total pressure defect observed in the measurement plane is caused

due to the gap flow diffusing close to the suction-side of the gap, where coolant injection

prevents a strong recirculation and subsequent diffusion. As reviewed earlier the leakage

vortex was seen to increase dramatically in size between 80% and 90% chord and the

area mass averaged shed vorticity also more than doubled in this region. The trailing edge

jet injects coolant radially into the gap where the gap flow is fully separated and the

144

negligible effect may be attributable to either entrainment of coolant by the leakage flow

or the fact that there isn’t much diffusion within the gap occurring in the first place.

The radial distribution of the passage averaged total pressure coefficient for the

passage bounded by the suction-side of the test blade is shown in Figure 6.5. The effect

of coolant injection from individual locations is large enough to affect the passage

averaged total pressure for individual injection from the locations H1, H2, and H3. Also

shown in Figure 6.5 is the radial distribution of passage averaged total pressure

coefficient at a tip gap height of t/h = 0.72%. The reduction in total pressure defect due to

coolant injection from individual locations is not quite as large as that obtained by

reducing the gap height of blade B21 from t/h = 1.40% to t/h = 0.72%. Below 90% span

the values follow that of the baseline pretty closely. The total pressure defect reductions

are also localized to about the last 10% of blade height. In the case of injection from H4

there is little change from the baseline, with the near casing total pressure actually lower

by about 2%. This is caused partly by the reduced total pressure observed in between the

two tip leakage vortices of blades B20 and B21, as shown in Figure 6.4.

The wake structure downstream of the blades is shown in Figure 6.6. The change

in wake depth follows closely what has been observed in the contour plots. The greatest

reduction in wake depth is observed for coolant injection from H3, while the other three

cases show similar reduction in depth. The similarity in the wake depth of injection cases

H1, H2, and H4 is more likely due to movement of the vortex, as clearly the effect of H4

injection is negligible.

145

Figure 6.5: Effect of Injection Hole Location on the Passage Averaged Coefficient of

Cooled Blade B21.

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.6

-4.5

-4.4

-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

-3.6

Base3: t/h = 1.40%, Minj = 0t/h = 1.40%, H1t/h = 1.40%, H2t/h = 1.40%, H3t/h = 1.40%, H4

Figure 6.6: Effect of Injection Location on the Wake Profile at r = 0.96h.

Uncertainty band

146

6.2 Injection from Combination of Holes

Isolated injection studies indicated that coolant injection from the near trailing

edge, radial hole did not affect the tip leakage vortex significantly. Hence in studying the

combinations, H4 was not considered.

Figure 6.7 shows the effect of combined injection from H1 & H2. The pitch-wise

extent of the tip leakage vortex is reduced more than that when individually injecting

from H1. In addition, the vortex has shrunk in its span-wise extent. The decrease in

measured total pressure drop, in the leakage vortex region, is also greater than the

individual injection cases. Higher total pressure from the passage flow is seen within the

continuous curve enclosing the baseline tip leakage vortex. Coolant injection from the

combination of H1+H3 also decreases the area occupied by the tip leakage vortex of

cooled blade B21, as shown in Figure 6.8. The reduction in the total pressure defect

associated with the leakage vortex of blade B21 is greater than that achieved by the

combination of H1+H2. The tip leakage vortices of blades B21 (t/h = 1.40%) and blade

B19 (t/h = 0.92%) resemble each other closely, indicating an effective reduction in gap

height is achievable through coolant injection from the combination of H1+H3.

The total pressure map with coolant injection from the combination H2+H3 is

shown in Figure 6.9. The area occupied by the tip leakage vortex of cooled blade B21 is

smaller in comparison to that with no injection. The area of low total pressure is however

greater than that observed in the two previous combinations, where H1 was active.

Additionally, higher total pressure (GREEN Zone) is not observed within the boundary

denoting the tip leakage vortex of blade B21 without injection. This compares well with

147

the observations made for isolated injection, where coolant injection from H1 caused the

greatest reduction in pitch-wise extent of the tip leakage vortex of blade B21. It is not

unexpected that the tip leakage vortex of cooled blade B21 is significantly reduced in size

when coolant is injected from the combination of H1+H2+H3, as shown in Figure 6.10.

The associated total pressure defect is also reduced, along with a drop in the total

pressure gradient across the tip leakage vortex. Higher total pressure (GREEN Zone) is

also observed within the boundary of the tip leakage vortex with no coolant injection.

Figure 6.7: Total Pressure Coefficient Contours for Coolant Injection From Blade B21 and Locations H1+H2.

148



149

The effect of coolant injection from all locations is shown in Figure 6.11. The

measured coolant mass flow rate is Minj = 0.42%. The effect of coolant injection at a

mass flow rate of Minj = 0.41% was previously discussed with reference to Figure 5.9.

The total pressure distribution in the tip leakage vortex of cooled blade B21 in

Figure 6.11 is remarkably similar to that observed for the tip leakage vortex of blade B21

in Figure 5.9. This further supports the good repeatability between tests as discussed in

Section 4.2.2. The area of influence of the tip leakage vortex of blade B21 is significantly

reduced. The minimum total pressure measured in the tip leakage vortex is about 0.2 qm

greater than that measured with no coolant injection.

Figure 6.10: Total Pressure Coefficient Contours for Coolant Injection From Blade B21 and Locations H1+H2+H3.

150

The variation in passage averaged total pressure distribution, due to coolant

injection from various combinations of injection locations is shown in Figure 6.12. The

passage average is computed for the passage containing the tip leakage vortex of cooled

blade B21. All combinations tested increase the averaged total pressure in the span-wise

region from 0.8h – 1h. The total pressure in this region is affected by the tip leakage

vortex, the tip-side passage vortex, and the interaction between these secondary flow

structures. Thus the combined effect of the injection locations tested affects the passage

flow more than individual injection. The improvement due to all combinations tested is

about the same, since all the curves are contained within the uncertainty band. The effect

of the tested combinations is also similar to the effect of reducing the gap height of the

test blade from t/h = 1.40% to t/h = 0.72%.

Figure 6.11: Total Pressure Coefficient Contours for Coolant Injection From Blade B21; Full Injection.

151

In the case of H1+H3 it is seen that below 84% blade height there appears to be

significant total pressure recovery. This however is not due to injection alone. Figure 6.13

shows the rotor averaged total pressure coefficient for the various combinations. In

general there is good repeatability, except in the case of H1+H3, where the curve below

88% blade height is shifted to the right. The shift is just within the uncertainty band and

hence the data is treated as having greater uncertainty. The same behavior was observed

Figure 6.12: Effect of Injection Location Combinations on the Passage Averaged

Coefficient of Cooled Blade B21.

Uncertainty band

152

in the passage averaged results for this case and hence it is not possible to separate the

effect of coolant injection, if any, on the flow in this region.

Wake plots for the injection combinations are shown in Figure 6.14. At the near-

tip radial location of 0.96h the wake depth is reduced for all combinations tested. The

greatest reduction occurs for the full blowing case and the least for the combination of

H1+H2. However, there is not much difference between the cases. The most interesting

feature, in comparing Figure 6.6 and Figure 6.14, is that the wake profile of B21 is

observably shifted towards the blade suction-side. This behavior is not observed for the

individual injection cases. This means that combined injection is successful in moving

the tip leakage vortex closer to the blade suction surface.

Figure 6.13: Rotor Averaged Coefficient With Combined Injection.

Uncertainty band

153

Blade Number

Tota

lPre

ssur

eC

oeff

icie

nt,C

pt

19 20 21 22 23-4.6

-4.5

-4.4

-4.3

-4.2

-4.1

-4

-3.9

-3.8

-3.7

-3.6

Base3: t/h = 1.40%, Minj = 0t/h = 1.40%, H1+H2t/h = 1.40%, H1+H3t/h = 1.40%, H2+H3t/h = 1.40%, H1+H2+H3t/h = 1.40%, H1+H2+H3+H4

Figure 6.14: Effect of Injection Location Combinations on the Wake Profile at r = 0.96h.

154

Chapter 7

Multiple Cooled Blades and the Effect of Casing Surface Roughness

The results presented so far dealt with coolant injection from a single cooled

blade (B21) and the tests were referred to as isolated injection tests. It was shown that

higher injection rates had some effect on the leakage flow in neighboring passages, due to

increased momentum of the coolant jets. As noted in Chapter 2, five blades (B17-B21)

were modified for coolant injection. The tip gap on four of the five cooled blades was

reduced by applying precision plastic shims on the blade tip surfaces. In order to study

the effect of coolant injection from multiple blades these shims were removed and the

effect of multiple cooled blades on the tip leakage flow over cooled blade B21 is

presented in this chapter. Coolant mass flow rate measurement was done only on the

supply line to blade B21, as described in Chapter 2. The removal of the plastic shims on

four of the five cooled blades resulted in the clearance distribution TCL3, shown in

Figure 2.5.

A preliminary investigation of the effect of casing surface roughness was also

conducted by artificially roughening the casing inner surface. The roughness was

introduced by applying various grades of sandpaper using double-sided adhesive tape.

The nominal clearance distribution for the surface roughness study is TCL4 in Figure 2.5.

155

7.1 Baseline

The total pressure coefficient contour map with five cooled blades and no coolant

injection is shown in Figure 7.1. The discussion in this section is centered on the

differences between Figure 7.1 and Figure 4.14, where the large tip gap was present on

cooled blade B21 only. Note that the gap height for blades B19 and B20 is greater than

that in the results presented in previous chapters. The effect of enlarging the tip gap

height of these blades increases the area occupied by their respective tip leakage vortices.

The tip leakage vortices of blade B19 and B20 also shift to the left and away from the

blade suction surface, as shown by the boundary of the baseline tip leakage vortices from

Figure 4.14. It was shown in Chapter 4 that the tip leakage vortex of cooled blade B21

moved significantly towards the blade suction surface when the tip gap height was

reduced from t/h = 1.40% to t/h = 0.72%. Thus, the behavior observed here is consistent

with the change in the tip gap height.

The position of the tip leakage vortex of cooled blade B21 has not changed much

and the slight shift towards the left may be attributed to the increased tip gap height of

blade B20. Increasing the gap height of blade B20 leads to lower momentum in the tip-

side passage vortex, allowing the leakage vortex of blade B21 to affect more of the

passage. The total pressure associated with the leakage vortex of blade B21 is unchanged

in magnitude, while the total pressure coefficient of blades B19 and B20 has dropped by

0.2qm, in comparison to Figure 4.14. For the blade sector shown, the largest total pressure

defect is seen for blade B19, which has the largest tip gap. The GREEN zone between

leakage flow structures of blades B20 and B21 has also increased considerably.

156

The wake structure of blades B19 and B20 is also different, in comparison to the

results presented in Figure 4.14. The distinct tip-side passage vortex has vanished and is

replaced by a zone of uniformly low total pressure. This occurs due to the increased

interaction of the tip leakage vortex with the tip-side passage vortex. The passage core

between blades B20 and B21 indicates a radial outward shift, due to a reduction in the

influence of the tip-side passage vortex on the core flow.

The change in the wake profile at the span-wise location of 0.96h is shown in

Figure 7.2. The comparison of the isolated injection baseline and multiple injection

baseline shows that blade B21 is mostly unaffected by the increased tip gap height of

blades B17-B20. However the wake profiles for blade B17-B20 show considerable

Figure 7.1: Total Pressure Coefficient With Multiple Cooled Blades; No Coolant Injection (Baseline), Minj = 0.0%.

157

increase in depth, particularly for blade B19. The tip gap height of blade B19 was

increased from t/h = 0.92% to t/h = 1.54%. The shift in the wake profile to the left is also

observed and is on average about 20% pitch.

7.2 Variation of Coolant Mass Flow Rate

Coolant mass flow rates of 0.43%, 0.62%, and 0.72% were investigated. From

previous coolant injection results it was seen that coolant mass flow rate of Minj = 0.41%

had the most beneficial effect on the leakage vortex. Injection at a coolant mass flow rate

Figure 7.2: Wake Profile at r = 0.96h Comparing Baseline Distributions of Multiple

Cooled Blade and Single Cooled Blade (B21).

158

of Minj = 0.63% appeared to be the most effective overall, while the coolant mass flow

rate of Minj = 0.72% had the greatest effect on the tip leakage flow structure of blade B22.

The total pressure contour map with multiple blade coolant injection at Minj =

0.43% is shown in Figure 7.3. The tip leakage vortices of the cooled blades (B17 – B21),

with enlarged tip gap heights, are seen to have moved to the right and closer to the blade

suction surfaces. The reduction in total pressure defect is 0.15qm, on average. Multiple

blade coolant injection appears to have an effect on the tip leakage flow structures of

blades B22 and B23, this is not seen when coolant was injected from cooled blade B21

only at Minj = 0.41%. Thus, multiple blade coolant injection appears to be generating a

strong flow in the near casing region, opposite in direction to the leakage flow, thereby

affecting the tip leakage flow over blades with no coolant injection. A comparison with

Figure 5.9, where coolant was injected from blade B21 only at Minj = 0.41%, shows that

coolant injection at the lowest coolant mass flow rate has an almost identical effect.

Figure 7.3: Total Pressure Coefficient With Multiple Cooled Blades; Minj = 0.43%

159

An increase in the coolant mass flow rate to Minj = 0.62% produces a greater

reduction in the area occupied by the tip leakage vortex of cooled blade B21, as shown in

Figure 7.4. This was not observed when injecting from a single cooled blade (B21), as

discussed in Chapter 4. The tip leakage vortex of blade B21 displays a greater movement

to the right and is almost a part of the blade wake. It may be recalled that with isolated

injection the tip leakage vortex retained a compact structure and was displaced closer to

the casing. This would also suggest that multiple blade injection is setting up a near

casing flow that is opposing the leakage flow. Thus, when injecting at Minj = 0.62%, the

tip leakage vortex of blade B21 shows greater reduction in the total pressure defect. A

visual comparison shows that the total pressure defect associated with the tip leakage

vortex of blade B21 (t/h = 1.40%) with coolant injection at Minj = 0.62%, in Figure 7.4, is

much smaller than that associated with the total pressure defect of blade B23 (t/h =

0.77%) in Figure 7.1. The higher kinetic energy (GREEN) zone in between adjacent tip

leakage vortices of blades B20 and B21 occupies a larger area, due to a reduction in the

influence of the tip leakage vortex of blade B21 on the passage flow.

The characteristics of the tip leakage vortex of blade B21 with coolant injection at

a mass flow rate of Minj = 0.72% are very similar to that observed at the coolant mass

flow rate of Minj = 0.62%, as shown in Figure 7.5. The area occupied by the vortex, total

pressure defect, and location of the vortex are almost identical in both cases. The

increased momentum of the coolant streams appears to affect the tip-side passage vortex.

The tip-side passage vortex of blade B22 occurs at a lower radial location, closer to 75%

span, thereby reducing the interaction between the tip leakage and tip-side passage

vortices.

160

Figure 7.4: Total Pressure Coefficient With Multiple Cooled Blades; Minj = 0.62%.

Figure 7.5: Total Pressure Coefficient With Multiple Cooled Blades; Minj = 0.72%.

161

The wake profiles of coolant injection from five cooled blades and the

corresponding baseline is shown in Figure 7.6. The five cooled blades, with large tip gap

heights, are clearly distinguished in the baseline by the much deeper wakes observed for

blades B17-B21. The wake depth is reduced at all coolant mass flow rates and the

movement of the wake profile to the right is also observed. The change in wake depth of

blade B21 indicates a total pressure defect reduction of approximately 0.3 qm, greater

than the 0.2 qm reduction observed with isolated injection. The circumferential shift in the

wake profile also increases with increase in the coolant mass flow rate. The effect of

coolant injection of the tip leakage flow over blade B22 is also observed at the selected

radius. The shift in the wake profile of blade B22 is more pronounced at the highest

coolant mass flow rate. The reduction in total pressure defect, while greater than the

uncertainty band, is small.

The radial distribution of the passage averaged coefficients shown in Figure 7.7

compares the effect of multiple blade coolant injection to that of no coolant injection. The

effect of multiple cooled blades on blade B21 is presented. Total pressure gains are

observed in the region from 0.85h to 1h at all coolant mass flow rates. As with coolant

injection from blade B21 only, variation in coolant mass flow rate appears to have

negligible effect on the averaged total pressure. There is little effect of coolant injection

over the rest of the passage height over which the data is available. Consistently higher

total pressure is measured in the near casing region for all coolant mass flow rates due to

energizing of the passage flow by the higher momentum present in the coolant jets.

162

Figure 7.6: Wake Profiles at r = 0.96h for Multiple Blade Coolant Injection.

Figure 7.7: Passage Averaged Coefficient Comparison for Multiple Blade Coolant

Injection.

Uncertainty band

163

7.3 Effect of Casing Surface Roughness

The effect of casing surface roughness on over tip leakage flow was investigated

by applying coarse and fine sand paper to the inner surface of the casing window, with

the aid of double-sided adhesive tape. This method of artificially roughening the casing

inner surface changes the tip clearance distribution. A smooth surface of equivalent tip

clearance was also tested by applying a precision plastic shim of the same thickness. This

was done to isolate the effect of artificially roughening the casing surface from the effect

of tip gap height reduction due to treatment thickness. The nominal thickness of the

applied casing treatment is 0.3175 mm or 0.26% blade height and the resulting tip

clearance distribution is TCL4, as shown in Figure 2.5. Only total pressure measurements

with no coolant injection were conducted.

7.3.1 Smooth Casing Surface

The total pressure coefficient distribution in the measurement plane with a smooth

casing surface at a uniformly reduced tip gap height is shown in Figure 7.8. The

treatment thickness is 0.26% of blade height. The tip gap height of each blade is noted

above the casing boundary and the clearance of cooled blade B21 is t/h = 1.14%. The tip

gap height reduction leads to the expected drop in total pressure defect for all blades, as

compared to the total pressure coefficient observed in Figure 7.1. The smallest tip gap

height obtained is t/h = 0.51% for blade B23 and consequently the corresponding tip

leakage vortex is very weak and contained within the wake. It must be noted that the total

164

pressure defect due to leakage vortex of blade B21 (t/h = 1.14%) is greater than that of

blade B21 with a tip gap height of t/h = 0.72%, which was discussed in reference to

Figure 4.14. The expected movement of the tip leakage vortices to the right is also

observed. The higher total pressure zones between subsequent tip leakage vortices are

more energized. The tip-side passage vortices in the wakes of blades B22 and B23 are

quite distinct and occur below 0.85h.

The wake profiles shown in Figure 7.9 compare the baseline for multiple cooled

blades with the wake obtained by uniformly reducing the tip gap height through

application of the smooth surface treatment. Also shown are the wake profiles due to

Figure 7.8: Total Pressure Coefficient With a Smooth Plastic Layer On the Casing Inner Surface.

165

artificially introduced surface roughness, which will be discussed later. The reduction in

tip gap height causes the expected reduction in wake depth and circumferential

movement of the wake profile to the right. The reduction in wake depth for the cooled

blade B21 is 0.14 qm. The wake profile near mid-span is shown in Figure 7.10 and

indicates no effect at mid-span due to the uniform reduction in tip gap height.

Figure 7.9: Wake Profiles at r = 0.96h Comparing the Influence of Casing Surface

Roughness.

166

7.3.2 Fine Surface Roughness (220 Grit)

The total pressure coefficient distribution shown Figure 7.11 was obtained after

applying 220 Grit sandpaper to the casing inner surface. The thickness of the applied

treatment is 0.26% blade height. It is clear that the artificially introduced surface

roughness has considerable effect on both the leakage flow and the tip-side passage

vortex. The tip leakage vortices of the cooled blades shown are greatly reduced in size.

The total pressure defect is also substantially reduced, along with the gradients. The tip

leakage vortex is contained within the wake and is located to the right of the tip-side

passage vortex. The tip leakage vortices of blades B22 and B23 are completely mixed in

Figure 7.10: Wake Profiles at r = 0.57h Comparing the Influence of Casing Surface

Roughness.

167

with the wake fluid. It is of course unclear whether the leakage flow is eliminated or if it

is completely mixed in with the wake. The increased surface roughness is expected to

locally increase turbulent kinetic energy within the tip gap, reducing the normal

momentum exiting the tip gap suction-side corner. Thus the secondary kinetic energy

associated with the tip leakage vortex in the blade passage is reduced and the over tip

leakage fluid is found within the blade wake.

The tip-side passage vortices appear to be stronger and especially so at the larger

tip gap heights. This is partly due to the reduced interaction between the tip leakage flow

and the tip-side passage secondary flow. The passage cores are shifted considerably to the

left and this is probably due to the effect of the tip-side passage vortex on the passage

core.

The wake profiles in Figure 7.9 show that the artificially introduced surface

roughness leads to considerable reduction in wake depth, in the range of 0.3 qm to 0.4 qm.

The wake depth of cooled blade B21 is reduced by 0.38 qm, which is greater than the 0.14

qm reduction obtained by the effect of reducing the gap height only, as discussed earlier.

Thus, the effect of artificially roughening the casing inner surface is about 1.7 times the

effect of reducing the tip gap height. The total pressure defect reduction is also greater at

the larger tip gap heights, indeed for the wake profiles shown there is small difference

between the smooth wall and fine roughness wake depths at the smaller tip gap heights.

There is however a global shift in the wake profiles to the right. The effect on the passage

core is also observed in the wake profiles at 0.57h, shown in Figure 7.10. The wake

profile may be observed to have shifted to the left, a behavior also seen in the contour

plot. The wake depth however remains almost the same.

168

7.3.3 Coarse Surface Roughness (100 Grit)

Figure 7.12 is a contour plot of the total pressure field downstream of the rotor

with coarse sand paper applied on to the casing. While the treatment thickness (0.3% h) is

slightly greater than the previous two treatments, the difference (0.04% h) is believed to

be too small to have an effect on the tip gap flow. Hence, any differences seen may be

attributed to the increased surface roughness.

The effect of coarse roughness on the tip leakage flow appears to be similar to

that of fine sandpaper. The tip leakage vortices are greatly reduced in area and the total

pressure field within the vortices is uniform. The tip leakage vortices at the small gap

Figure 7.11: Total Pressure Coefficient With Fine Sandpaper (220 Grit) On the Casing Inner Surface.

169

heights are again mixed in with the blade wakes. The tip-side passage flow also appears

to be more uniform across the passage. The tip-side passage vortices are stronger due to

the reduced interaction with the tip leakage flow. The tip-side passage vortices have also

moved to a lower radial location. The reduced interaction between the secondary flows

near the blade tip is a result of reduction in leakage flow momentum exiting the tip gap,

caused by greater turbulent kinetic energy within the tip gap. There is considerable

spacing between the two tip secondary flow structures, indicating that the increase in

surface roughness is responsible for this shift. The movement of the passage core to the

left is also observed and hence it can be concluded that this is due to the increased surface

roughness.

The wake profiles at r = 0.96 h, due to the increased surface roughness are quite

similar to those obtained with the fine surface roughness as shown in Figure 7.9. This

implies that the effect of surface roughness is not enhanced significantly by increasing

the roughness quality of the flow path. The wake profiles close to mid-span, shown in

Figure 7.10, also indicate that the effect of surface roughness is unchanged by the

increase in the roughness quality of the casing inner surface.

7.4 Comparison of the Averaged Total Pressure Coefficient

The passage averaged total pressure coefficient, computed for the passage

containing the tip leakage vortex of cooled blade B21 (t/h = 1.40%, without casing

treatment) with different casing roughness treatments is shown in Figure 7.13. The

170

baseline with five cooled blades (solid RED) and coolant injection at Minj = 0.62%

(dotted GREEN) are also shown for comparison. The decrease in tip gap height by

0.26%h without artificial roughness (smooth casing) reduces the total pressure defect

observed in the region from 0.85h to 1h. The reduction is however at the limit of the

uncertainty band. The introduction of surface roughness doubles the total pressure defect

reduction over the entire passage, from 0.05qm to 0.1qm at the location 0.93h. The

benefits of artificial roughness are comparable to that obtained through multiple blade

coolant injection in the region influenced by the tip leakage vortex.

Figure 7.12: Total Pressure Coefficient With Coarse Sandpaper (100 Grit) On the Casing Inner Surface.

171

The rotor averaged total pressure coefficient for the multiple cooled blade

baseline and the three casing treatments is shown in Figure 7.14. The total pressure defect

due to the five large tip gap heights is visible in the rotor averaged coefficient. This is in

contrast to the rotor averaged coefficient in Figure 4.15. Thus, the performance of the

rotor may be expected to decrease when the rotor tip clearance distribution is changed

from TCL1 to TCL3. The artificial introduction of surface roughness appears to have no

observable effect in the region of the tip leakage vortex, over that of reducing the tip gap

height by 0.26%h. This supports the earlier conclusion that artificial roughening of the

casing surface is most beneficial at the larger gap heights.

Figure 7.13: Passage Averaged Coefficient For Blade B21 With Different Casing

Roughness Treatments.

Uncertainty band

172

Figure 7.14: Rotor Averaged Coefficient With Different Casing Roughness Treatments.

Uncertainty band

173

Chapter 8

Summary and Conclusions

Turbomachines are widely used in the transfer of thermal energy to mechanical

energy in power generation and aircraft engines. Axial flow turbines extract energy from

the working fluid by effecting a change in the angular momentum of the flow. The

efficiency of this ideally isentropic process depends on how closely the blade passage

flow occurs to the design streamlines. Secondary flow in blade passages, such as passage

vortices, horseshoe vortices and over tip leakage cause the fluid to deviate considerably

from design streamlines thereby reducing the efficiency. The gap between rotating

turbine blades and stationary casing, in high pressure, un-shrouded, axial flow turbines is

called the tip gap or tip clearance. Flow in this gap from blade pressure-side to blade

suction-side, termed over tip leakage, is pressure driven, generates considerable losses,

and increases the thermal load in regions of the tip surface. The tip gap height also

increases with service, thereby causing further decrease in engine efficiency and also the

exhaust gas temperature. Reducing the effects of this necessary gap on turbine

performance is termed tip desensitization and an experimental investigation of tip surface

coolant injection as a method for desensitizing turbine blade tips was reported in this

thesis. Many of the extensively researched desensitization techniques serve to block over

tip leakage mass flow rate by reducing gap discharge coefficient. These methods

typically present increased cooling requirements. Even the most commonly used tip

174

surface geometry that of double squealer tips is subject to burn-out and thermal

degradation. Tip surface coolant injection directed towards blade pressure-side corner,

reported in this thesis, aims at modifying existing, radial coolant injection schemes and

thereby does not incur the penalty of additional cooling requirements.

8.1 Summary

Two different measurement approaches were used to study tip desensitization in

this study. Qualitative investigation of the effects of tip gap height and tip surface coolant

injection was studied using surface flow visualization. Surface flows on the rotor endwall

and turbine casing were also visualized. The visualization techniques used consisted of

oil-dot technique and the oil-film technique. Quantitative measurements consisted of

obtaining high resolution, total pressure distributions at 30% chord downstream of the

rotor exit using a fast response total pressure probe aligned with the absolute tip velocity

vector. The influence of tip gap height, tip surface coolant injection from rotor blades,

and the effect of casing surface roughness were investigated in detail.

8.1.1 Surface Flow Visualization

Surface flows on the rotor endwall, turbine casing, and rotor blade tip surfaces

were visualized. Rotor endwall surface flow visualization showed the formation of the

horseshoe vortex upstream of the blade leading edge. Variations in boundary layer

growth on the rotating hub, upstream of the rotor inlet plane were also captured. The path

175

of the pressure-side leg of the horseshoe vortex is well defined. The cross-passage

boundary layer flow that leads to the passage vortex was discussed with reference to

previously obtained rotor endwall static pressure distributions. At rotor exit the flow in

the rotating frame was shown to exit the blade passage near design angle from pressure

surface up to mid-pitch and then the flow was highly overturned due to the passage

secondary flow. Visualization of the interaction zone between tip leakage vortex and tip

passage vortex was also possible. A streak of visualization material was observed to

extend from about 60% blade tip axial chord length all the way to the blade trailing edge.

The streak diverged along its length and its lower boundary was measured at 75% span at

the trailing edge.

Rotor tip surface visualization was performed by applying the oil and pigment

mixture to the blade pressure surface and allowing rotational effects and leakage flow

entering the gap to transport oil on to the tip surface. Visualization material carried on to

the tip surface when using the oil dot technique is discrete. Chord-wise flow within the

separation zone on the tip surface transported the oil in a well defined streamline. The

streamline moves further away from the pressure-side corner, close to 0.5 blade tip axial

chord and subsequently runs parallel to the blade profile, maintaining its distance from

the pressure-side corner. The distance of the streamline from the pressure-side corner also

varies with gap height, being closer to the blade pressure-side corner at smaller gap

heights. Two gap heights were investigated. The distance of the chord-wise streamline

from the pressure-side corner was measured to be 1.4*t for a gap height of t/h = 1.40%

and 1.1* t for a gap height of t/h = 0.72%. In terms of blade tip axial chord length the

distances are 3% Cax for t/h = 1.40% and 1.2% Cax for t/h = 0.72%.

176

Oil film applied to blade pressure surface clearly identifies a separation region

near the pressure-side corner, followed by a well defined reattachment line on the tip

surface. Re-circulating flow is also clearly identified by this technique. The regions

where reattachment and re-circulation are observed are closely coupled with areas of high

heat transfer rates on the blade tip surface.

The position of the reattachment line from blade pressure-side corner was

measured at 60% Cax for gap heights (t/h) of 0.72%, 0.81%, 1.2%, and 1.40%. The

distance measured normal to blade tip axial chord length, when normalized by the blade

tip axial chord length, was found to increase linearly with gap height. The distance almost

doubled when the gap height was increased by almost twice. The distance measured

normal to the camber-line and normalized by gap height was found to stay reasonably

constant at about 1.95*t – 2.0*t.

Turbine casing surface flow visualization was also conducted using the two

techniques. Oil dots indicate the highly overturned nozzle exit flow in the near suction

surface region at vane exit. Flow angles at rotor exit are near design angles for flow

originating away from the nozzle suction surface. The highly over-turned nozzle exit

flow is seen at rotor exit as being under-turned, with a predominantly axial direction. This

underscores the importance of conducting probe based measurements with the probe

positioned at nozzle mid-pitch. A region of low momentum fluid originating from the

suction-side of nozzle vane was tracked up to blade mid-chord.

Surface flow visualization with coolant injection showed that pressure-side corner

separation was substantially reduced in the region influenced by the coolant jets, at all

coolant mass flow rates tested. Recirculation was completely eliminated in the last 40%

177

of blade tip axial chord length. This should reduce not only the gap mixing losses, but

also decrease the heat transfer rate to the tip surface in this region. Flow in the last 1/3rd

of the blade is substantially changed, including the elimination of fully separated tip

region.

Coolant jets form localized films on the tip platform, at all injection rates. The

leading jet of each injection set forms a film to the right of the hole, while the trailing jet

forms a film to the left, due to its inclination towards the trailing edge.

At 0.4% coolant mass flow rate the trailing jets cover the largest area on the tip

surface. At the other three coolant mass flow rates the best surface area coverage is

obtained from the leading jets. Thus, orientation of cooling jets appears to depend upon

the amount of coolant injected. Increasing the coolant mass flow rate also appears to

cause more mixing between the coolant and the gap flow. Leakage flow is blocked by the

coolant jets, leading to low momentum activity between the injection holes and the SS

corner.

8.1.2 Total Pressure Measurement

Total pressure measurements of the flow field 30% blade tip axial chord lengths

downstream of the rotor exit plane were conducted using a high-frequency, total pressure

probe aligned with the angle of the tip velocity vector in the stationary frame of

reference. Time accurate, phase-locked, total pressure measurements are averaged over

200 rotor revolutions and non-dimensionalized with a mean wheel speed based dynamic

pressure (qm). The total pressure probe attached to the outer casing of the turbine stage is

178

traversed in the radial direction. Thus, a complete two-dimensional mapping of the rotor

exit total pressure field for all 29 passages is possible. Distinct flow structures such as tip

leakage vortices, passage vortices, blade wake, and core flow were detected by

employing a phase-locked ensemble averaging technique. Circumferential averaging of

the total pressure coefficient was done for individual passages and for the entire rotor.

The rotor averaged coefficient was used to ascertain repeatability of data, while the

passage averaged coefficient enabled distinguishing the effect on individual passages.

At a gap height of t/h = 1.40% the tip leakage vortex is relatively large, occupying

about 15% blade span. The tip leakage vortex for this case presents a significant blockage

to the passage flow. The total pressure defect due to the large vortex is about 20% qm

greater than that of blades with nominal blade clearance.

Reduction of tip clearance effectively reduced the blockage presented by the tip

clearance flow to the passage flow. The total pressure drop measured in the leakage

vortex was reduced by about 20% qm. The size of the leakage vortex was greatly reduced,

by almost 50%, at the smaller gap height. The reduced interaction between the leakage

vortex and the passage vortex was evident from the improved definition of the passage

vortex. A reduction in tip gap height caused the leakage vortex to move towards the blade

suction-side.

Coolant injection from the tip trench was successful in filling in the total pressure

defect originally resulting from the leakage vortex without injection. Coolant mass flow

rates of 0.41%, 0.52%, 0.63%, and 0.72% were investigated for coolant injection from a

single blade with a gap height of t/h = 1.40%.

179

At all injection rates, the reduction in total pressure defect is very similar to that

when the gap height t/h = 1.40% is reduced by about half to 0.72%. This result shows

that directed tip coolant jets can be effective in creating beneficial “sealing” effects

previously observed only from small clearances.

Injection at Minj = 0.41% core flow was the most effective in reducing the total

pressure loss in the leakage flow of the test blade. This was observed at a radius near the

core of the tip vortex. However, it appears that 0.63% injection is the most effective from

a global point of view, as shown by the passage averaged pressure coefficient obtained in

the last 25 % of the blade height.

The tip leakage vortex moves closer to the tip in a radially outward direction,

especially at the higher injection rates. The cross section of the new tip leakage vortex,

with coolant injection, is smaller and some of the total pressure defect is eliminated by

the injection process. The upper passage vortex is better defined when the tip leakage

vortex cross section is smaller and located nearer the casing.

The relatively high radial position of the leakage vortex resulting from Minj values

0.63% and 0.72% implies that the interaction with the passage vortex is less pronounced.

Thus, tip injection is capable of reducing losses, especially those due to leakage

flow/passage vortex/blade wake interaction.

At the two highest injection rates the leakage vortices of adjacent blades (to the

right of the test blade) are affected by the injection. The high momentum associated with

these jets moves the tip vortices in adjacent channels against the direction of rotation.

This might be due to the alteration of the near casing flow between the pressure side

corner of the test blade and the adjacent blade. Since the tip jets are faced towards the

180

pressure side corner of the test blade, especially at high blowing rates, the static pressure

distribution near the outer casing in the adjacent passage is expected to be altered.

The location of the coolant injection holes was also studied, by individual

injection from each hole and combinations thereof. The coolant supply pressure was

maintained constant for all tests in this series. Individual injection results show that

coolant injection from 61%, 71%, and 81% chord locations reduce the leakage vortex

size at the measurement location. This is attributed to a reduction in normal momentum

exiting the tip gap.

Injection from 81% chord is the most successful in filling the total pressure defect

in the vortex core. Thus it appears that leakage flow responsible for the greatest total

pressure deficit occurs around 80%.

The injection location at 91%, with the largest hole diameter of 1.8 mm, does not

have a significant effect on the leakage flow. This hole is very close to the trailing edge

of the blade. However, the tip-side passage flow shows an increase in the total pressure

drop coefficient.

Combined injection in general shows better desensitization. There is observable

movement of the leakage vortex towards the suction-side of the test blade. Combined

injection from H1-H3 is found to be almost as effective as injection from all locations.

Coolant injection from all five cooled blades was studied at coolant mass flow

rates of 0.43%, 0.62%, and 0.72%. In all cases the effect of injecting from more than one

blade caused a greater total pressure recovery. Increasing coolant mass flow rate leads to

greater defect reduction, although it appears that 0.62% might be an inflection point in

the performance benefit curve. Multiple blade injection is believed to cause a strong near

181

casing flow in a direction opposing over tip leakage and might be expected to change the

static pressure distribution on the casing wall. Tip passage vortex appears to be more

energized.

Casing surface roughness was investigated by applying smooth plastic shim and

two grades of sandpaper to the casing surface with double-sided adhesive tape. Casing

surface roughness greatly reduced the leakage vortex defect measured downstream of the

rotor exit. The area occupied by the tip leakage vortex was also substantially smaller. The

increase in turbulent kinetic energy within the tip gap is believed to reduce the normal

momentum exiting the tip gap, leading to a smaller leakage vortex. The leakage vortex is

also contained within the blade wake and hence it can be concluded that the leakage flow

is turned more efficiently. The tip-side passage vortex was found to be more energized

especially with the coarse grade of sandpaper causing the tip-side passage vortex to move

towards the hub and away from the blade wake. This is attributed to the reduced

interaction with the tip leakage vortex, which also means that the mixing losses due to

this interaction must be lower.

8.2 Conclusions

The following conclusions are drawn form the results,

1. Surface flow visualization is an extremely effective tool in understanding the flow

physics in rotating turbine blade passages.

2. Rotor endwall flow features observed in the rotating frame of reference compare

well with those observed in previous stationary cascade facilities.

182

3. Low momentum fluid from the nozzle suction surface propagates through the

rotor passage causing under-turning of flow at rotor exit.

4. Chord-wise flow on the tip surface occurs in a region between the separated flow

at blade pressure-side corner and re-circulating flow within the tip gap.

5. The location of the reattachment line, measured from the pressure-side corner

increases linearly with gap height.

6. The distance of the reattachment line measured from the pressure-side corner is

about 2*gap height, at all gap heights.

7. Tip gap flow is fully separated over the last 5% axial chord length of the blade,

where the driving pressure differential across the tip gap is minimal.

8. The tip trench causes the gap vortex to be weakened, by preventing it from

remaining attached to the tip surface.

9. Coolant injection from the test blade causes the leakage vortex to decrease in size

and total pressure defect is reduced to levels observed for blades with half the gap

height. Tip surface coolant injection could serve as a highly effective tip leakage

sealing system.

10. Coolant injection closer to the trailing edge, where the blade tip profile is thin,

caused a much greater reduction in total pressure defect. Coolant injection closer

to the leading edge on the other hand is more effective in reducing the area

occupied by the tip leakage vortex.

11. Heat transfer benefits might arise from the elimination of reattachment and re-

circulation of gap flow on the tip surface and the blockage presented by the

coolant jets.

183

12. Coolant injection from multiple blades increases the reduction in total pressure

defect, by about 50%. Leakage flow of un-cooled blades also appears to be

affected beneficially.

13. Increasing the surface roughness of the turbine casing leads to the tip-side passage

vortex to appear at a lower radius and the tip leakage vortex to be greatly reduced

in both size and total pressure defect.

14. Tip leakage reduction methods studied in this thesis are highly applicable to many

axial flow turbomachinery systems. The tip injection and casing roughness based

OTL control are candidates for use in modern counter-rotating ducted fan systems

built into many unmanned flight vehicles.

8.3 Recommendations for Future Work

This experimental study has shown that tip surface coolant injection, with coolant

jets directed towards the pressure-side corner, can serve as an effective tip leakage

sealing strategy and can also improve the thermal performance of turbine blade tips.

There are however certain parameters that cannot be addressed economically in a rotating

rig. These may however be studied either in stationary cascades or through numerical

investigations. The influence of tip gap height on the exit flow field may be used to

“calibrate” such investigations. Parametric studies of the location of the coolant injection,

from the pressure-side corner can be conducted relatively inexpensively. Similarly, the

potential thermal benefits of tip surface coolant injection may be easier to quantify,

184

particularly in stationary cascades. The effect of casing surface roughness may also be

simulated numerically and compared to available experimental results.

185

Bibliography

1. Thorpe, S. J., Yoshino, S., Ainsworth, R. W., and Harvey, N. W., 2004, “An Investigation of the Heat Transfer and Static Pressure On the Over-Tip Casing Wall of an Axial Turbine Operating at Engine Representative Flow Conditions. (2). Time Resolved Results,” International Journal of Heat and Fluid Flow, 25, 2004, pp. 945 – 960.

2. Bunker, R. S., “A Review of Turbine Blade Tip Heat Transfer,” Presented at International Symposium “Turbine 2000” on Heat Transfer in Gas Turbine Systems, August 13-18, Cesme, Turkey.

3. Xiao, X., 2001 “Investigation of Tip Clearance Flow Physics in Axial Flow Turbine Rotors,” Doctoral Thesis, Pennsylvania State University, May 2001.

4. Yamamoto, A., 1988, “Interaction Mechanisms between Tip Leakage Flow and the Passage Vortex in a Linear Turbine Rotor Cascade,” ASME J. of Turbomachinery, 110, pp. 329-338.

5. Booth, T. C., Dodge, P. R., Hepworth, H. K., “Rotor-Tip Leakage: Part 1- Basic Methodology,” ASME Paper No. 81-GT-71.

6. Rains, D. A., “Tip Clearance Flows in Axial Flow Compressors and Pumps,” Rept. 5, Hydrodynamics and Mechanical Engineering Laboratories, California Institute of Technology.

7. Bindon, J. P., “The Measurement and Formation of Tip Clearance Loss,” ASME Paper No. 88-GT-203.

8. Yaras, M., Zhu, Y., Sjolander, S. A., 1989, “Flow Field in the Tip Gap of a Planar Cascade of Turbine Blades,” ASME J. of Turbomachinery, 111, pp. 276-283., Paper No. 88-GT-29.

9. Sjolander, S. A., and Cao, D., 1995, “Measurements of the Flow in an Idealized Turbine Tip Gap,” ASME J. of Turbomachinery, 117, pp. 578-584.

10. Yaras, M. I., Sjolander, S. A., 1992, “Effects of Simulated Rotation on Tip Leakage in a Planar Cascade of Turbine Blades. Part 1: Tip Gap Flow,” ASME J. of Turbomachinery, 114, pp. 652-659.

186

11. Yaras, M. I., Sjolander, S. A., Kind, R. J., 1992, “Effects of Simulated Rotation on Tip Leakage in a Planar Cascade of Turbine Blades. Part 2: Downstream Flow Field and Blade Loading,” ASME J. of Turbomachinery, 114, pp. 660-667.

12. Graham, J. A. H., 1986, “Investigation of a Tip Clearance Cascade in a Water Analogy Rig,” ASME J. Engineering for Gas Turbines and Power, 108, pp. 38-46.

13. Morphis, G., Bindon, J. P., “The Performance of a Low Speed One and A Half Stage Axial Turbine With Varying Rotor Tip Clearance and Tip Gap Geometry,” ASME Paper No. 94-GT-481.

14. Yamamoto, A., Tominaga, J., Matsunuma, T., Outa, E., 1994, “Detailed Measurements of Three-Dimensional Flows and Losses Inside an Axial Flow Turbine Rotor,” ASME Paper 94-GT-348.

15. Kaiser, I., Bindon, J. P., “The Effect of Tip Clearance on the Development of Loss Behind A Rotor and A Subsequent Nozzle,” ASME Paper No. 97-GT-53.

16. McCarter, A., “Investigation of Tip Clearance Flow Fields in a Turbine Rotor Passage,” Masters Thesis, Pennsylvania State University, May 2000.

17. Waterman, W. F., “Turbine Loss Correlations and Analysis,” VKI/Pennsylvania State University Lecture Course, “Tip Clearance Effects in Axial Turbomcahines,” Apr.

18. Booth, T. C., 1985, “Importance of Tip Clearance Flows in Turbine Design,” VKI Lecture Series 1985-05, “Tip Clearance Effects in Turbomachines,” 1985.

19. Schaub, U. W., Vlasic, E., and Moustapha, S. H., 1993, “Effect of Tip Clearance on the Performance of a Highly Loaded Turbine Stage,” AGARD-CP-537, “Technology Requirements for Small Gas Turbines,” Paper 29.

20. McCarter, A., Xiao, X., Lakshminarayana, B., “Tip Clearance Effects in a Turbine Rotor: Part 2- Velocity Field and Flow Physics,” ASME Paper No. 2000-GT-477.

21. Bindon, J. P., 1987, “Pressure Distributions in the Tip Clearance Region of an Unshrouded Axial Turbine as Affecting the Problem of Tip Burn-out,” ASME Paper No. 87-GT-230.

22. Bunker, R. S., Bailey, J. C., and Ameri, A. A., 2000, “Heat Transfer and Flow on the First-Stage Blade Tip of a Power Generation Gas Turbine: Part 1-Experimental Results,” ASME J. Turbomachinery, 122, pp. 263-271, Paper No. 99-GT-189.

187

23. Bunker, R. S., 2004, Blade Tip Heat Transfer and Cooling Techniques. VKI Lecture Series 2004-02, “Turbine Blade Tip Design and Tip Clearance Treatment,” 2004.

24. Bindon, J. P., 1986, “Visualization of Axial Turbine Tip Clearance Flow using a Linear Cascade,” CUED/A-Turbo TR122, Whittle Laboratory, Cambridge University.

25. Heyes, F. J. G., Hodson, H. P., and Dailey, G. M., 1991 “The Effect of Blade Tip Geometry on the Tip Leakage Flow in Axial Turbine Cascades,” ASME Paper No. 91-GT-135.

26. Bunker, R.S., Bailey, J.C., 2000, “Blade Tip Heat Transfer and Flow with Chord-wise Sealing Strips,” International Symposium on Transport Phenomena and Dynamics of Rotating Machinery (ISROMAC), Honolulu, Hawaii, pp.548-553.

27. Acharya, S., Yang, H., Prakash, C., and Bunker. R., 2003, “Numerical Study of Flow and Heat Transfer on a Blade Tip with Different Leakage Reduction Strategies,” ASME Paper No. GT-2003-38617.

28. Papa, M., Goldstein, R. J., Gori, F., 2002,”Effects of Tip Geometry and Tip Clearance on the Mass/Heat Transfer From a Large Scale Gas Turbine Blade,” ASME Paper GT-2002-30192.

29. Ameri, A.A., 2001, “Heat Transfer and Flow on the Blade Tip of a Gas Turbine Equipped with a Mean Camberline Strip,” ASME J. of Turbomachinery, 123, pp.704-708.

30. Dey, D., 2001d, “Aerodynamic Tip Desensitization in axial Flow Turbines,” Doctoral Thesis, Pennsylvania State University, December 2001.

31. Dey, D., Camci, C., 2000, “Development of Tip Clearance Flow Downstream of a Rotor Blade With Coolant Injection From a Tip Trench,” In Proceedings of the 8th ISROMAC Conference, Honolulu, Hawaii pp. 572-579.

32. Tallman, J., and Lakshminarayana, B. L., 2001, “Methods for Desensitizing Tip Clearance Effects in Turbines,” ASME Paper No. 2001-GT-0486.

33. Harvey, N. W., 2004, Aerothermal Implications of Shroudless and Shroude Blades. VKI Lecture Series 2004-02, “Turbine Blade Tip Design and Tip Clearance Treatment.”

34. Kim, Y. W., Downs, J. P., Soechting, F. O., Abdel-Messeh, W., Steuber, G., and Tanrikut, S., 1995, “A Summary of the Cooled Turbine Blade Tip Heat Transfer and Film Effectiveness Investigations Performed by Dr. D. E. Metzger,” ASME J. Turbomachinery, 117, pp. 1-11.

188

35. Kwak, S. J., and Han, J. C., 2002, “Heat Transfer Coefficient and Film-Cooling Effectiveness on a Gas Turbine Blade Tip,” ASME Paper No. GT-2002-30194.

36. Hohlfeld, E. M., Christophel, J.R., Couch, E. L., and Thole, K. A., “Predictions of Cooling from Dirt Purge Holes Along the Tip of a Turbine Blade,” ASME Paper No. GT-2003-38251.

37. Koschel, W., Schmidt, H., and Vornberger, A., 1990, “Analysis of The Rotor Tip Leakage Flow With Tip Cooling Air Ejection,” AGARD-CP-469, “Secondary Flows in Turbmachines.”

38. Chen, G., Dawes, W. N., Hodson. H. P., 1993, “A Numerical Investigation of Turbine Tip Gap Flow,” AIAA 93-2253, AIAA/SAE/ASME/ASEE, 29th Joint Propulsion Conference & Exhibit, June 28-30, 1993, Monterey, CA.

39. Rao, N., and Camci, C., 2004, “Axial Flow Turbine Tip Desensitization by Injection from a Tip Trench. Part 1- Effect of Injection Mass Flow Rate,” in Proceedings of ASME Turbo Expo 2004, Vienna, Austria, June 2004, ASME Paper GT2004-53256.

40. Rao, N., and Camci, C., 2004, “Axial Flow Turbine Tip Desensitization by Injection from a Tip Trench. Part 2- Leakage Flow Sensitivity to Injection Location,” in Proceedings of ASME Turbo Expo 2004, Vienna, Austria, June 2004, ASME Paper GT2004-53258.

41. Rao, N., and Camci, C., 2004, “A Flow Visualization Study of Axial Turbine Tip Desensitization by Coolant Injection From a Tip Platform Trench,” in Proceedings of 2004 ASME International Mechanical Engineering Conf., Anaheim, California, November 2004, ASME Paper IMECE2004-60943.

42. Rao, N., and Camci, C., 2005, “Visualization of Rotor Endwall, Tip Gap, and Outer Casing Surface Flows In a Rotating Axial Turbine,” in Proceedings of ASME Turbo Expo 2005, Reno, Nevada, June 2005, ASME Paper GT2005-68264.

43. Pudupatty, R., 1998, Personal communication to Camci, C., Solar Turbine Inc.

44. Kulite Semiconductor Products, Inc., Kulite Miniature IS Silicon Diaphragm Pressure Transducers, Bulletin: KS-1000F.

45. Kline, S. J., McClintock, F. A., 1953, “Describing Uncertainties in Single-Sample Experiments,” Mechanical Engineering, 75, pp.3-8.

46. Langston, L. S., 1990, “Research on Cascade Secondary and Tip Leakage Flows – Periodicity and Surface Flow Visualization,” AGARD-CP-469, “Secondary Flows in Turbomachines,”

189

47. Merzkirch W., 1987, Flow Visualization 2nd edition London Academic

48. Eckerle, W. A., and Langston, L. S., 1986, “Horseshoe Vortex Formation around a Cylinder,” ASME Paper No. 86-GT-246.

49. Aunapu N V., Volino R J., and, Flack K A., 2000, “Surface Flow Visualization of a Scaled-up Turbine Blade Passage,” J. Meas. Sci. Technol., 11, pp. 987-991.

50. Aunapu N V., Volino R J., Flack K A., and Stoddard R M., 2000, “Secondary Flow Measurements in a Turbine Passage With Endwall Flow Modification,” ASME J. Turbomachinery, 122, pp. 651-658.

51. Allen, H. W., and Kosfskey, M. G., Visualization Study of Secondary Flows in Turbine Rotor Tip Regions. NACA TN-3519

52. Dring, R. P., and Joslyn, H. D., 1981, “Measurement of Turbine Rotor Blade Flows,” ASME J. Engineering for Power, 103, pp. 400-405.

53. Gregory-Smith, D. G., Graves, C. P., Walsh, J. A., 1988, “Growth of Secondary Losses and Vorticity in an Axial Turbine Cascade,” ASME J. Turbomachinery, 110, pp. 1-8.

54. Xiao, X., McCarter, A., Lakshminarayana, B., “Tip Clearance Effects in a Turbine Rotor: Part 1- Pressure Field and Loss,” ASME Paper No. 2000-GT-476.

55. Yang W., 1989, Handbook of Flow Visualization, Hemisphere Publishing Corporation.

56. Prasad, A., and Wagner, J. H., 2000, “Unsteady Effects in Turbine Tip Clearance Flows,” ASME Paper No. 2000-GT-0444.

57. Kwak, J. S., Ahn, J., Han, J. C., Lee, C. P., Bunker, R. S., Boyle, R., and Gaugler, R., 2003, “Heat Transfer Coefficients on The Squealer Tip and Near Tip Regions of a Gas Turbine Blade With Single or Double Squealer,” ASME Paper No. GT-2003-38907.

58. Prasad, D., and Hendricks, G. J., 2000, “A Numerical Study of Secondary Flow in Axial Turbines With Application to Radial Transport of Hot Streaks,” ASME Paper No. 2000-GT-0448.

59. Ristic, D., Lakshminarayana, B., Chu, S., 1998, “Three-Dimensional Flow Field Downstream of an Axial Flow Turbine Rotor,” AIAA paper 98-3572, presented at the 34th AIAA/ASME/SAE/ASEE Joint Propulsion Conference and Exhibit, Cleveland, Ohio.

190

Appendix A

Total Pressure Probe Characteristics

The sensitivity of the total pressure coefficient to probe incidence was tested at

two radial locations in the AFTRF and the results are presented Section A.1. The

frequency content of the total pressure field in the measurement plane and in the region

of influence of the tip leakage vortex was measured to decide on the low pass filter cut-

off frequency. The frequency spectrum is discussed in Section A.2.

A.1 Angular Sensitivity

Measurement sensitivity to exit flow angle was measured in the AFTRF by

rotating the probe in increments of 5°, on either side of α3 = 25.4°. This was done at two

radial locations, 0.93h (near tip region) and 0.49h (mid span). Counterclockwise (CCW)

rotation of the probe is positive and makes the probe more tangential with every

increment. The rotor averaged, and passage averaged total pressure coefficient for blade

B21, are shown in Figure A.1. The absolute velocity vector at mid span is at an angle of

29.19° from axial, corresponding to an incidence angle of 3.79°. The mean value of Cpt,P

at mid-span in a ±15° range is -3.916 and all values in this range lie within the

uncertainty band of δCpt = ± 0.024. Furthermore, the passage averaged coefficient for

blade B21 indicates that the blade passage behaves almost identically as the rest of the

191

rotor. At 0.93h however, the difference between the rotor averaged and passage averaged

values is considerable, due to the large leakage vortex. Passage averaged coefficient in

the range -25° to 10° are within uncertainty limits of the mean value (Cpt,P = -4.22),

computed for the range ±15°. The value at 15° is just outside the uncertainty band.

McCarter, et al. [16] measured lower relative tangential velocity in the region dominated

by the leakage vortex. This means that the leakage vortex approaches the probe at a

negative incidence. Thus, from measurements in the range of -25° to 0° it is possible to

conclude that error in the measurement of total pressure associated with the tip leakage

vortex of the test blade is within uncertainty bounds.

A.2 Frequency Spectrum of Flow Field

The frequency content of the rotor exit flow field, 30% chord length downstream

of the rotor, was obtained by feeding the signal to a frequency analyzer and computing

the spectrum shown in Figure A.2. The peaks identified are those at the blade passing

frequency (BPF) and its harmonics. The 2nd harmonic of the BPF is not as distinct as the

fundamental and 1st harmonic. Furthermore there is at least a 10 dB drop in signal power

between the fundamental and 2nd harmonic. The drop off is around 30 dB at a frequency

of 12 kHz and the spectrum appears to be leveling out. Based on this it was decided that

using a low-pass filter cut-off frequency of 20 kHz would not remove any content from

the signal. Furthermore, the full range of the sensor frequency response would be utilized.

192

Figure A.1: Probe Response to Incidence. (Squares denote rotor averaged Cpt and circles

denote passage averaged Cpt.).

193

A.3 Uncertainty Analysis

A sample calculation of the uncertainty analysis is shown in this section. The

propagation of uncertainty is calculated by the method derived by Kline and McClintock

[45]. The expression for the total pressure coefficient is as shown in Equation A-1. The

uncertainty in the total pressure coefficient is obtained from Equation A-2, which is

-6.00E+01

-5.00E+01

-4.00E+01

-3.00E+01

-2.00E+01

-1.00E+01

0.00E+00

1.00E+01

2.00E+01

3.00E+01

0.00E+00 2.00E+03 4.00E+03 6.00E+03 8.00E+03 1.00E+04 1.20E+04 1.40E+04

Frequency, (kHz)

Mag

nitu

de, (

dB)

BPF Harmonics

Figure A.2: Frequency Spectrum of Rotor Exit Flow Near Rotor Tip.

194

obtained by differentiating Equation A-1 and dividing the resulting expression by

Equation A-1.

where W[] is the uncertainty associated with the parameter in the square brackets.

The individual uncertainty and nominal value of each measured parameter (used

in the denominator) is shown in Table A-1. These in turn are obtained from manufacturer

specifications of the precision associated with the measurement. Uncertainty values for

air density (ρ) and blade mean wheel speed (Um) are calculated in a similar fashion. The

values are listed in Table A-2, without presenting the sample calculations.

2

0

21

),(m

ptU

PjiC

ρ

∆= , where 01030 ),( PjiPP −=∆

(A-1)

222

0

20

⎟⎟⎠

⎞⎜⎜⎝

⎛+⎟

⎟⎠

⎞⎜⎜⎝

⎛+⎟

⎟⎠

⎞⎜⎜⎝

⎛

∆=

∆

m

UP

pt

C

U

WW

P

W

C

Wmpt

ρρ

, (A-2)

195

Table A-1: Uncertainty and Nominal Values in Measured Parameters

Parameter Precision Error Uncertainty Nominal Value

∆P0 0.1% of 34.474*103

(kPa) ±34.474 (Pa) 7471.387 (Pa)

Tamb ±0.5 (K) ±0.5 (K) 300 (K)

Pamb 0.1% ±100 (Pa) 98700 (Pa)

N ±1 (rpm) ±1 (rpm) 1328 (rpm)

Table A-2: Uncertainty in Derived Parameters

Parameter Uncertainty

pt

C

C

Wpt ±0.0058

0

0

PW P

∆∆ ±0.00461

ρρW

±0.0035

m

U

U

Wm ±7.53*10-4

196

Appendix B

AFTRF Tip Clearance Distribution

Tip clearance measurements in three regions of the blade chord are presented in

tabulated form and graphically. Data is presented for two of the four clearance

distributions reported in this manuscript. The clearance distributions TCL1 and TCL3 as

defined in Chapter 2 are presented in Table B-1 and Table B-2, respectively. The

graphical representation of the variation of measured gap height along blade axial chord

length is shown in Figure B.1. Only the variation in distribution TCL1 is shown.

197

Table B-1: Gap Height Variation Along Blade Axial Chord For Clearance Distribution TCL1

Blade #

Clearance Average Clearance

LE-33% 33-66% 66-TE mm t/h% mm t/h% mm t/h% mm t/h%

1 1.02 0.83% 0.99 0.81% 1.02 0.83% 1.01 0.82% 2 0.97 0.79% 0.99 0.81% 1.02 0.83% 0.99 0.81% 3 0.99 0.81% 0.99 0.81% 0.97 0.79% 0.98 0.80% 4 0.97 0.79% 1.02 0.83% 0.97 0.79% 0.98 0.80% 5 0.94 0.76% 0.91 0.74% 0.94 0.76% 0.93 0.76% 6 0.91 0.74% 0.91 0.74% 0.86 0.70% 0.90 0.73% 7 0.89 0.72% 0.86 0.70% 0.86 0.70% 0.87 0.71% 8 0.91 0.74% 0.86 0.70% 0.84 0.68% 0.87 0.71% 9 0.91 0.74% 0.86 0.70% 0.86 0.70% 0.88 0.72%

10 0.89 0.72% 0.89 0.72% 0.89 0.72% 0.89 0.72% 11 0.94 0.76% 0.94 0.76% 0.94 0.76% 0.94 0.76% 12 0.94 0.76% 0.94 0.76% 0.91 0.74% 0.93 0.76% 13 0.94 0.76% 0.91 0.74% 0.91 0.74% 0.92 0.75% 14 0.94 0.76% 0.86 0.70% 0.86 0.70% 0.89 0.72% 15 0.89 0.72% 0.86 0.70% 0.84 0.68% 0.86 0.70% 16 0.91 0.74% 0.89 0.72% 0.86 0.70% 0.89 0.72% 17 0.99 0.81% 1.04 0.85% 1.09 0.89% 1.04 0.85% 18 0.76 0.62% 1.14 0.93% 1.09 0.89% 1.00 0.81% 19 1.07 0.87% 1.12 0.91% 1.22 0.99% 1.13 0.92% 20 0.89 0.72% 0.97 0.79% 0.99 0.81% 0.95 0.77% 21 1.63 1.32% 1.70 1.38% 1.83 1.49% 1.72 1.40% 22 1.02 0.83% 1.02 0.83% 1.02 0.83% 1.02 0.83% 23 0.91 0.74% 0.97 0.79% 0.97 0.79% 0.95 0.77% 24 0.97 0.79% 0.97 0.79% 1.04 0.85% 0.99 0.81% 25 1.02 0.83% 1.02 0.83% 1.04 0.85% 1.02 0.83% 26 1.02 0.83% 1.07 0.87% 1.09 0.89% 1.06 0.86% 27 1.09 0.89% 1.02 0.83% 1.02 0.83% 1.04 0.85% 28 0.97 0.79% 0.99 0.81% 1.02 0.83% 0.99 0.81% 29 0.97 0.79% 0.99 0.81% 0.97 0.79% 0.98 0.79%

198

Table B-2: Gap Height Variation Along Blade Axial Chord For Clearance Distribution TCL3

Blade # Clearance Average Clearance

LE-33% 33-66% 66-TE

mm t/h% mm t/h% mm t/h% mm t/h% 1 1.02 0.83% 0.99 0.81% 1.02 0.83% 1.01 0.82% 2 0.97 0.79% 0.99 0.81% 1.02 0.83% 0.99 0.81% 3 0.99 0.81% 0.99 0.81% 0.97 0.79% 0.98 0.80% 4 0.97 0.79% 1.02 0.83% 0.97 0.79% 0.98 0.80% 5 0.94 0.76% 0.91 0.74% 0.94 0.76% 0.93 0.76% 6 0.91 0.74% 0.91 0.74% 0.86 0.70% 0.90 0.73% 7 0.89 0.72% 0.86 0.70% 0.86 0.70% 0.87 0.71% 8 0.91 0.74% 0.86 0.70% 0.84 0.68% 0.87 0.71% 9 0.91 0.74% 0.86 0.70% 0.86 0.70% 0.88 0.72%

10 0.89 0.72% 0.89 0.72% 0.89 0.72% 0.89 0.72% 11 0.94 0.76% 0.94 0.76% 0.94 0.76% 0.94 0.76% 12 0.94 0.76% 0.94 0.76% 0.91 0.74% 0.93 0.76% 13 0.94 0.76% 0.91 0.74% 0.91 0.74% 0.92 0.75% 14 0.94 0.76% 0.86 0.70% 0.86 0.70% 0.89 0.72% 15 0.89 0.72% 0.86 0.70% 0.84 0.68% 0.86 0.70% 16 0.91 0.74% 0.89 0.72% 0.86 0.70% 0.89 0.72% 17 1.78 1.45% 1.78 1.45% 1.78 1.45% 1.78 1.45% 18 1.68 1.36% 1.70 1.38% 1.73 1.41% 1.70 1.38% 19 1.80 1.47% 1.98 1.61% 1.91 1.55% 1.90 1.54% 20 1.70 1.38% 1.70 1.38% 1.73 1.41% 1.71 1.39% 21 1.63 1.32% 1.70 1.38% 1.83 1.49% 1.72 1.40% 22 1.02 0.83% 1.02 0.83% 1.02 0.83% 1.02 0.83% 23 0.91 0.74% 0.97 0.79% 0.97 0.79% 0.95 0.77% 24 0.97 0.79% 0.97 0.79% 1.04 0.85% 0.99 0.81% 25 1.02 0.83% 1.02 0.83% 1.04 0.85% 1.02 0.83% 26 1.02 0.83% 1.07 0.87% 1.09 0.89% 1.06 0.86% 27 1.09 0.89% 1.02 0.83% 1.02 0.83% 1.04 0.85% 28 0.97 0.79% 0.99 0.81% 1.02 0.83% 0.99 0.81% 29 0.97 0.79% 0.99 0.81% 0.97 0.79% 0.98 0.79%

199

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

1.60%

0 5 10 15 20 25 30

Blade Number

Cle

aran

ce G

ap H

eigh

t, t/h

(%)

LE-33%33%-66%66%-TEAverage

Figure B.1: Clearance Gap Variation Along Blade Axial Chord Length For TCL1.

200

Appendix C

Total Pressure Coefficient Contour Map

The total pressure contours presented and discussed previously showed a sector of

five rotor blades, including the cooled, test blade B21. A complete map of the total

pressure field is shown in Figure C.1. As noted in Section 4.2.1, the total pressure probe

is in the stationary frame and the data acquisition system acquires 6000 data points per

revolution of the rotor. Thus, in the total pressure map of the entire rotor 29 blade wakes,

passage cores, and tip leakage vortices are seen as shown in Figure C.1. The variations in

the tip leakage vortex structure with tip gap height are clearly observed.

201

Figure C.1: Total Pressure Contour Map Of the Entire Rotor Exit Flow Field In the

Measurement Plane.

ΩBlade Wake

Passage Core

Tip Leakage Vortex

VITA

Nikhil Molahally Rao

Nikhil Rao received his Bachelor of Engineering degree in Mechanical

Engineering, from the University of Mysore in February 1994. His career in

turbomachinery engineering started in May 1994, at Turbotech Precision Engineering (P)

Ltd., in Bangalore, India. The author left his position in May 1997 to commence his

graduate study at The Virginia Polytechnic Institute and State University and received his

M.S. degree in Mechanical Engineering in May 1999. Subsequently, he entered the

graduate program in Aerospace Engineering at The Pennsylvania State University and

successfully defended his doctoral thesis on May 02, 2005. The author will be graduating

with a Ph.D. in Aerospace Engineering in August 2005 and has accepted the position of

Engineer in the Turbine Engineering Group, at The Elliott Company in Jeannette,

Pennsylvania, starting in June 2005.

Refereed Papers

Rao, N., and Camci, C., 2005, “Visualization of Rotor Endwall, Tip Gap, and Outer Casing Surface Flows In a Rotating Axial Turbine,” ASME Paper GT2005-68264, ASME Turbo Expo 2005, Reno, Nevada, June 2005.

Rao, N., and Camci, C., 2004, “A Flow Visualization Study of Axial Turbine Tip Desensitization by Coolant Injection From a Tip Platform Trench,” ASME Paper IMECE2004-60943, 2004 ASME International Mechanical Engineering Conf., Anaheim, California, November 2004.

Publications

Rao, N. M., Feng, J., Burdisso, R. A., and Ng, W. F., 2001, “Experimental Demonstration of Active Flow Control to Reduce Unsteady Stator-Rotor Interaction,” AIAA Journal, Vol.39, No.3, March 2001, pp.458-464.

Presentations

Rao, N., “Axial Flow Turbine Tip Desensitization by Injection From a Tip Trench. Part 1- Effect of Injection Mass Flow Rate,” Presented at ASME Turbo Expo 2004, Vienna, Austria, June 14, 2004.

Rao, N., “Axial Flow Turbine Tip Desensitization by Injection From a Tip Trench. Part 2- Leakage Flow Sensitivity to Injection Location,” Presented at ASME Turbo Expo 2004, Vienna, Austria, June 14, 2004.

desensitization of over tip leakage in an axial turbine rotor by tip ... performed in...

Documents