Cooled variable-area radial turbine technology program

NASA CR-166408 (11Sl-CB-165408) COOL£il VARllaLE-AB~A BADIAL TUBBIN£ l~CtiNCLO~Y fLCuuAd Final .• hepoct 'Garr~tt Tucbin~ ~~9ine Co.} 303 p Unclcts liC A 14/!lF Au1 ~SCL 21E G3/07 09.240 ..

COOLED, VARIABLE-AREA RADIAL

TURBINE PROGRAM

G.D. Large L.J. Meyer Garrett Turbine Engine Company ~ A Division of the Garrett Corporation }

a

, .J JANUARY 1982 Prepared for NATIONAL AERONAUTICS AND SPACE ADMINISTRATION Lewis Research Center Under Contract No. NAS3-22004

..., ... '"~ .. " ..I, , 1. Report No. 3. Recipient's Catalog No.

I 2. Government Accession No.

NASA CR-165408 4. Title and Subtitle 5. Report Date ..

January 1982 Final Report - Cooled Variable-Area Radial Turbine Technology Program 6. Performing Organization Code 7. Author(s) 8. Performing Organization Report No.

G. D. Large and L. J. Meyer 21-3914

t

i t----------------------------~--- 10. Work Unit No.

~ 9. Performing Orglnization Name and Address ., Garrett Turbine Engine Company ..

11. Contract or Grant No.

111 S. 34th Street, P.O. Box 5217

!

I Phoenix, Arizona 85010 J NAS3-22004 ..

1-- ____________________________ ---1 13. Type of Report and Period Covered 12. Sponsoring Agency Name and Address Conttactor Report U.S. Army Research & Technology Laboratories (AVRADCOM) 14. Sponsoring Agency Code Propulsion Laboratory ~wls.Resea~~~ ce~~er lL162209AH76 l..:leveland. Onl0 44135 15. Supplementary Notes Project Manager, Peter L. Meitner, U.S. Army Research and Technology Laboratories (AVRliOCOM), Propulsion Laboratory, Lewis Research Center, Cleveland, Ohio 44135.

16. Abstract The objective of this study was a conceptual evaluation and design analyses of a cooled variable-area radial turbine capable of maintaining nearly constant high- efficiency when operated at a constant speed and pressure ratio over. a range of flows corresponding to 50- to lOa-percent maximum engine power. The results showed that a 1589K (2400·F) turbine was feasible that would satisfy a 4000-hour duty cycle life • goal. The final design feasibility is based on 1988 material technology goals. A peak aerodynamic stage total efficiency of 0.88 was predicted at 100 percent power • Two candidate stators were identified: an articulated trailing-edge and a locally ,.I movable sidewall. Both concepts must be experimentally evaluated to determine the optimum configur.ation. A follow-on test program is proposed for this evaluation.

;11' 17. Key Words (Suggested by Author/s)) 18. Distribution Statement Radial Turbine Variable Geometry Cooled Turbine Cooled Vane

•

22. Price· 19. Security a'lSif. (of this report) 20. Securitv CllIlSif. (of this page) 21. No. of Pages UNCLASSIFIED UNCLASSIFIED • For sale by the National TechOicallntormatlon SerVICe, Spllngtie/d. V:rgllIlJ 22161 NASA-C-I68 (Rev. 10-75)

TABLE OF CONTENTS Page

1. a SUMMARY'

2.0 INTRODUCTION .

• 2.1 Background 2.2 Program Objective ...

\ , 2.3 Cooled Radial Rotor Technology i 2.4 Major Program Considerations 3.0 AERODYNAMIC PERFORMANCE CORRELATIONS AND SYSTEM OPTIMIZATION TECHNIQUE 3.1 Design-Point Performance Evaluation . 3.2 Off-Design Performance Evaluation 3.3 Effects of Cooling Flow and No:z1e Leakage 4. a TASK I - STATOR AND ROTOR CONCEP'l'S EVALUATION 4.1 Variable-Area Radial Turbine Require~ents 4.2 lOa-Percent Power Cycle Study 4.3 Preliminary Parametric Study (Stator/Rotor Concept Evaluation) 4.3.1 Design Point Study 4.3.2 Turbine Off-Design Performanc~ Characteristics 4.4 Stator Concepts Aero/Mechanical Evaluation 43' Vane Profile Design for Stator Concepts 4.4.1 Full-Vane Rotation Concept 4.4.2 4.4.3 Articulated Trailing-Edge Concept Insertable Minivane Concept 4.4.4 Locally Movable Sidewall Concept 4.4.5 Material Evalua'tion 4.5 Stator 4.5.1 Airfoil Rupture, 53 Stress 4.5.1.1 Resistance 4.5.1.2 Coated Oxidation Capability 4.5.1.3 Overtemperature Thermal Fatigue 4.5.1.4 Sidewall 4.5.2 • [ Recommended Variable Stator Concepts 4.6 I I Evaluation 58 Turbine'Rotor Materials 4.7 , • i

TABLE OF CONTENTS (Contd) Page 4.8 Cooled Rotor Concept AerO/Mechanical Evaluation 59 4.9 Recommend~d CoOled Rotor Concept 65 5.0 TASK II - DETAILED PARAMETRIC STUDY ., ~ ., 5.1 Selected Speed and Cycle Conditions 5.2 Effects of Stage Cooling Flows 5.3 Effects of Stator Leakage 5.4 Cooled Rotor Aerodynamic Analysis 5.5 Uncooled Rotor Aerodynamic Evaluation 5.6 Uncooled Rotor Mechanical Analysis 5.7 Cooled Rotor Mechanical Analysis 5.7.1 Influence of Radius Ratio on Rotor Stresses 5.7.2 Incorporation of Cooled, Dual-Alloy Configuration 5.7.3 Optimum Rotor Selection 5a8 Selected Stage Configuration 6.0 TASKS III AND IV - VARIABLE-AREA STATOR AERODYNAMIC AND MECHANICAL DESIGN 6.1 Detailed Aerodynamic D~sign Procedure 6.2 Articulated Trailing-Edge Stator - Detailed Aerodynamic Design 6.3 Detailed Aerodynamic Design of the Movable Sidewall Stator l2~ 6.4 Stator Flow Field - 3-Dimensional, Finite- Element Analysis 6.4.1 Background 6.4.2 Analysis and Program Description 6.4.3 Application of the Finite-Element Flow Analysis to the Variable Geometry Stator 6.5 Detailed Aerodynamic Design Conclusions 150 6.6 Articulated Trailing-Edge Stator - Mechanical Substantiation 159 6.6.1 Mechanical Description 6.6.2 Impact of Cycle Conditions on Stator Design 6.6.3 Stator Aerodynamic Boundary Conditions • 6.6.4 Vane-Cooling Circuit Design 6.6.5 Sidewall Cooling Design ii

TABLE OF CONTENTS (Contd) 6.7 Movable Sidewall Stator - Mechanical Substantiation - .

6.7.1 Mechanical Description '\ 7.0 TASKS III AND IV - COOLED ROTOR DETAILED AERODYNAl4IC DESIGN AND MECHANICAL SUBSTANTIATION 7.1 Cooled Rotor Detailed Aerodynamic DeSign 7.1.1 Rotor-Blade Geometry Definition 7.1.2 Rotor Aerodynamic Flow Analysis 7.2 Detailed Substantiation of Cooled Rotor DeSign 7.2.1 Mechanical Configuration 7.2.2 Rotor Thermal DeSign Analysis 7.2.2.1 Rotor Blade Aerodynamic Boundary Conditions 7.2.2.2 Geometric Assumptions 7.2.2.3 Blade Cooling Configuration 7.2.2.4 Thermal Analysis Method 7.2.2.5 Blade Cooling Passage Design 7.2.2.5.1 Hub Supply Passage 7.2.2.5.2 Inducer Passages 7.2.2.5.3 Inducer Tip 7.2.2.5.4 Exducer Region 7.2.2.6 Thermal-Analysis Results 7.2.3 Rotor Mechanical Design Analysis 7.2.3.1 Blade Configuration 7.2.3.2 Disk Configuration 7.2.3.3 Stress Analysis 7.2.3.4 Vibration Analysis 7.2.3.5 Life Analysis 7.2.3.6 Rotor DeSign - Conclusions and Recommendations 8.0 FINAL PERFORMANCE EVALUATION - SELECTED COOLED, VARIABLE-AREA RADIAL TURBINE DESIGN 9.0 FOLLOW-oN TEST PROGRAM PLAN 9.1 Program Objective , , .

9.2 Test Rig Description iii

TABLE OF CONTENTS (Contd) page 9.3 Overall Turbine Performance Measurement Instrumentation 9.4 Program Plan 10.0 CONCLUSIONS AND RECOMMENDATIONS " I

•

11.0 REFERENCES • I • iv

LIST OF ILLUSTRATIONS Title Figure NASA Specific Speed Correlation Corrected to Zero Clearance 2 Variation of Optimum Tip Speed with Turbine Stage Specific Work i

•

Cooled, Variable-Area Radial Turbine,

t System Definition

, ~ £ .

4 Effects of Swirl on Interstage DllCt Loss t Coefficient Results of System Optimization for TARADCOM

t 5

l. 12 Advanced Radial Turbine " i 6 Summary of Radial Turbine Rotor Clearance

[

Effects Predicted and Tested Off-Design Efficiency Characteristic Comparison 8 Predicted and Tested Off-Design Flow Characteristic Comparison 17 9 1-Dimensional Vector Diagram Comparison (100- Percent and SO-Percent Inlet Corrected Flows) 18

"

Cooled, Variable-Area Radial Turbine Study,

" 10

T4 • l533K (2300 F) at Maximum (lOa-Percent) Power 11 Cooled, Variable-Area Radial Turbine Study, T4 • l589K (2400 F) at Maximum (laO-Percent) Power Cooled, Variable-Area Radial Turbine Study, T4 • 16441 (2500 F) at Maximum (lOa-Percent) Power Uncooled, Variable-Area Radial Turbine Study, T4 • 14781 (2200 F) at Maximum (lOa-Percent) Power Preliminary Parametric Study for Cooled Rotor • t Concepts, 12:1 Cycle Pressure Ratio, T4 • 16441 (2500 F), N • 6178 radls (59,000 RPM) I , ! • v

\ ) LIST OF ILLUSTRATIONS (Contd) Title Figure Preliminary Parametric Study for Cooled Rotor Concepts at 16:1 Cycle Pressure Ratio, T4 • 1644 (2500 F) N - 6807 rad/s (65,000 RPM) Preliminary Parametric Study for Cooled Rotor

I'

Concepts at 20:1 Cycle Pressure Ratio, • T4 • 16441 (2500 F), N· 7435 rad/s (71,000 RPM) Parametric Study for Cooled Rotor Concepts, 16441 (2500 F) Case Effect of Cycle Pressure Ratio on Flow-Path Geometry Characteristic Rotors for Evaluation of Concepts Effect of Rotational Speed on Turbine Performance and Geometry at 16441 (2500 F) Effects of Stator Exit Angle on Stage and Nozzle Design at 16441 (2500 F), 7435 rad/s (71,000 RPM) and at 671 m/s (220G Ft/Sec) Selected Maximum Power Design Points for Off- Design Analysie. 16:1 Cycle Pressure Ratio, T4 • 16451 (2500 F) and N • 6807 rad/sec (65,000 RPM) 23 Off-Design Turbine Characteristics for Case No.1 at 16441 (2500 F), 6807 -rad/s (65,000 RPM), and 640 m/s (2100 Ft/Sec) 39 24 Off-Design Turbine Characteristics for Case No. 2 at 16441 (2500 F), 6807 rad/s (65,000 RPM) and 671 m/s (2200 Ft/Sec) 40 25 Off-Design Turbine Characteristics for Case No. 3 at 16441 (2500 F) 6807 rad/s (65,000 RPM) and 671 m/s (2200 Ft/Sec) 41 Case No. 3 Turbine I-Dimensional Vector Diagrams for 60- and lOO-Percent Power at 16441 (2500 F), 6807 rad/s (65,000 RPM) and RIR • 0.643

. I

Stator Profile Design for Variable Area Concepts I , Evaluation (19 Vanes) i i vi

I

J

\

~

LIST OF ILLUSTRATIONS (Contd)

l

~ Title Page ~ Figure l ( I Stator Ring at Maximum Power Setting 46 ~ k' Full Vane Rotation, Nozzle Area Variation t concept 48 Articulated Trailing-Edge Noz~le Area Variation

\' 30

Concept 49

•

Insertable Minivane, Nozzle Area Variation Concept 51

i

Locally Movable Sidewall, Nozzle Area Variation I I Concept 52

t

Larson-Miller Parameter Showing Stress Life t Capabilities of Candidate Superal~oys 54 t Stress-Rupture Capabilities of Candidate Turbine Super alloys 55 Tensile Properties of Hub Materials, Dual-Alloy Tur bine Rotor Comparison of 4000-Hour Mission Life and Density-Corrected Stresses at Inducer Tip at 640 mls (2100 Ft/Sec)

•

USARTL Cooled, Laminated Rotor Design Preliminary Cooled Radial Rotor Design Recommended Cooled Rotor Concept Cooled, Variable-Area Radial Turbine Compressor Speed Study Compressor Turbine Efficiency Product Characteristics Turbine Cooling Flow Model for the Detailed Parametric Study Articulated Trailing-Edge Nozzle, Sidewall .

Clearance Loss Models •

I

Effect of Stator Leakage on Off-Design Performance 78 •

j

vii f

t

!

LIST OF ILLUSTRATIONS (Contd) Figure Title 4S Detailed Parametric Study, ~B • 0 radian (0°), Maximum Power Condition 46 Detailed Parametric Study, ~B • 0.17 radian (10°), Maximum Power Condition 80

, I

Detailed Parametric Study, ~B • 0.35 radian ,

•

(20 0), Maximum Power Condition 81 Cooled Rotor Configurations from Detailed Parametric Study 83

49 Off-Design Turbine Characteristics With ~ = 1.04

1589K (2400 F), 5969 rad/s (57,000 RPt:), ind 6~0

m/s (2100 Ft/Sec), R/R = 0.60, 6 = O. rad (0°). 84

B Off-Design Turbine Characteristics with A4 • 1.0, l589K (2400 F), 5969 rad/s (57,000 RPM), and 640 mls (2100 Pt/Sec), R/R • 0.65, ~B • 0 rad (0°) 85 51 Off-Design Turbine Characteristics with A4 • 1.0, l589K (2400 F), 5969 rad/s (57,000 RPM) and 640 mls (2100 Ft/Sec), ..

R/R • 0.65, 8 • 0.17 rad (10°) 86 B 1. 0, 52 Off-Design Turbine Charabteristics With \4 = 1589 K (2400 F), 5969 rad/s (57,000 RPt!) 3nn (20° )

m/s (2100 Ft/Sec), R/R = 0.65, b = 0.35 Rad 87

B 53 Off-Design Turbine Characteristics with A4 • 1.0, 1589K (2400 P), 5969 radls (57,000 RPM), 640 mls (2100 Pt/Sec), RIR • 0.70, P • 0 rad (0°) 88 B 54 Detailed Parametric Study for uncooled rotor with 14 Blades at Maximum Power, 5969 rad/s (57,000 RPM), 1478K (2200 P), and 640 m/s (2100 Ft/Sec) 90 55 Off-Design Analysis for Uncooled Rotor at l478K (2200 F), {JB • 0.17 rad (10°), A4 = 0.975, R/R • 0.65, 640 mls (2100 Pt/Sec) and 5969 radls (57,000 RPM) 56 Uncooled Rotor Configuration 57 Mode Temperature Distribution, Uncooled Rotor 94 • viii

--..".4.0'.&. " .. " .... '1*_.14"~ ............... ____ ~ ..

t

,

•

'", 'l; LIST OF ILLUSTRATIONS (Contd) ; figure Title Page

i Normal Thickness Distribution, Uncooled Rotor 95

" t· ,\ Uncooled 14781 (2200·F) Rotor, 2-Dimensional ~ , Finite-Element Model 96 t Effect of Rotor Flow Path on Disk Stress 99

•

* ~, I Thickness Distribution for 0.67-Radlus-Ratio Configuration 100

t

Normal Slade Wall Thickness, Dual-Alloy i.

Geometry Study 102 t ~ I , i Single-Alloy Solid Rotor at Room Temperature 103 r , Dual-Alloy Rotor With and Without Full-Cooling Passage at Room Temperature 105 65 Dual-Alloy Solid Rotor with Full-Cooling Passage at Maximum Temperature 106 , Dual-Alloy Solid Rotor with Sondline at Largest Practical Radius, Room Temperature 107 67 Dual-Alloy Solid Rotor with Sondline at Largest , Radius and With Full-Cooling Passage, Room Temperature 108 Dual-Alloy Rotor, Highest Sondline Radius, Full- Cooling Passage, Maximum Power Temperature 110 Assumed Tempecatures and Equivalent Stresses 112 70 Effects of Radius Ratio and Inlet Blade Angle on System Performance at 60-Percent Power 115 Selected Cooled Rotor, Meridional Flow Path at 15891 (2400·F), UT4 of 640 m/s (2100 Ft/Sec) and 5969 radls (57,000 RPM) 116 72 I-Dimensional Vector Diagram for Selected Cooled- Stage Configuration, 100-Percent Power 117 Articulated Trailing-Edge Vane Concept 120 t' .

74 Articulated Trailing-Edge at 100-Percent Power r Design Point (Rotated Closed at 60-Percent Flow) I • I I I ix I I , •

LIST OF ILLUSTRATIONS (Contd) Title Figure Articulated Trailing-Edge Configuration, Vane Velocity Distributions for 100-Perc.nt Power Design Point (Stream-Function Solutiona) Articulated Trailing-Edge, 60-Percent Power Design Point (Rotated Open ,st 100-Percent Flow) Articulated Trailing-Edge Configuration, Vane i7 Velocity Distribution for 60-Percent Power Design point (Stream Function Solution) Final Articulated Trailing-Edge 17 Vane Profile Design for SO-Percent Power Design point Articulated Trailing-EJge Con~ept, Final Stator, Inlet Meridional Gec~etry Vane velocity Distribution for Final Articulated Trailing-Edge Stator Design (17 Vanes) Movable Sidewall Original Configuration ,.

Movable Sidewall Configuration (Purely Rotating Insert), Schematic Diagram • Movable Sidewall Configuration (Rotating- Translating Insert), Schematic Diagram Final Vane Profile, Movable Sidewall Concept MOVable Sidewall Stator Vane Velocities with 19 Vanes (stream-Function Solution) Stator Finite-Element Analysis Discretization Model Articulated Trailing-Edge Configuration, 60-Percent Pewer Setting, 80-percent Power Deaign Point Articulated Trailing-Edge Configuration, 100-percent Power Setting, at 80-percent Power Design Poir.t Finite-Blement ~odel for 100-Percent Plow ..

Geometry • x

t

I LIST OF ILLUSTRATlvNS (Contd) ! Title Figure t· !

Finite-Elem~nt Model for Movable Sidewall Configuration, (Purely Rotating Insert) at 60-percent Power Setting ..

Finite-Element Model for Movable Sidewall Configuration, (Rotating-Translating Insert) at 60-percent power S~tting

•

Contours of the Flow Passage Axial Width Within the purely Rotating Insert at 60-Percent Flow Setting Contours of the Flow-passage Axial Width within the Rotating-Translating Insert at 60-percent Flow setting stator Geometry and Station Designation Articulated Trailing-Edge Configuration Vane Velocity Distribution for lOa-percent power setting (Design point: 80-percent Power) • Articulate~ Trailing-Edge Configuration Vane Velocity Distribution for 60-~ercent Power Setting (Design Point at 80-Percent Power, 3-Dimensional Finite-Element AnalySiS) , Movable Sidewall Configuration Vane Velocity Distribution for laO-Percent Fower purely Rotating Sidewall Configuration Vane Velocity Distribution for 60-percent Power Rotating-Translating Movable Sidewall Configuration, Vane Velocity Distribution for 60-Percent Power Setting Movable Sidewall Configuration, Critical Mach No. Contours for 100-Percent Power setting . (3-Dimensional Finite-Element Analysis) Purely Rotating Sidewall Configuration, Critical Mach No. Contours for 60-Percent power Setting (3-Dimensional Finite-Element Analysis) Rotating-Translating Sidewall Configuration, Critical Mach No. Contours for 60-Percent power , setting (3-Dimensional Finite-Element Analysis) I • xi

..

LIST OF ILLUSTRATIONS (Contd) Title Figure Comparison of Flow properties at Stator Exit Station (Three-Dimensional Finite Element Analysis 157 Inlet Gas Temperature Versus Coolant-Added Variable-Area Stator Articulated Trailing-Edge stator Vane Velocities at lOa-Percent Power, Katsanis Solution Articulated Trailing-Edge Stator Vane Velocities at 60-Percent Power 163 Articulated Trailing-Edge Stator Vane Static Pressure Distribution 164 Articulated Trailing-Edge stator Vane, Suction- Side Surface Heat-Transfer Coefficients 166 Articulated Trailing-Edge Stator Vane, Pressure- Side Surface Heat-Transfer Coefficients 167 Thermal Loading on Articulated Trailing-Edge (No Film Cooling) Articulated Trailing-Edge Stator Vane Forward Section, Cooling Configuration .

~ Articulated Trailing-Edge Stator Vane Cooling Configuration Articulated Trailing-Edge Stator Vane 173 Articulated Trailing-Edge Stator Vane Cooling Flow Summary 174 Articulated Trailing-Edge Stator Sidewall Viewed from Flow-path Side 176 Movable Sidewall - Baseline Configuration 178 Variable Sidewall Concept with Purely Rotating Insert 179 Variable Side~all Concept with Rotating- Translating Insert 183

•

xii

'.

LIST OF ILLUSTRATIONS (Contd)

i

i-

Title Page Figure

t

Rotating-Translating Sidewall Concept, Surface * Static Pressure Distribution 185

t

Rotating-Translating Sidewall Concept Heat Transfer Coefficients 187 •

\

, • " A preliminary Concept for the Movable Sidewall Vane 188

t

•

~ Movable Sidewall Configuration Showing Vane J cooling 189 • i.

t

Movable Sidewall Configuration Vane, Flow Distribution 191

f

\ t Static Pressure Distribution Along the Midchannel • , Streamline of the Rotating-Translating Sidewall ~ , stator 192 , " Rotating-Translating Sidewall Configuration f Cooling, Forward sidewall at 60-Percent power 194 l " ~.

Rotating-Translating Movable Sidewall Stator, ~ , Heat Transfer Coefficients, Midchanne1 Streamlines 195 Radial Rotor Geometry program Capabilities 197 128 Final Variable-Area Radial Turbine Meridional Flow Path with Articulated Trailing Edge stator Shown at 100-Percent Power Final Rotor Blade Angle Distribution Rotor Blade Velocity Distribution at 60- and 100-percent Power Condition (14 Blades) Variable-Area Radial Turbine vector Diagram, 60-Percent Power Condition Variable-Area Radial Turbine Vector Diagram, 100-percent power Condition Airfoil Cooling, Final Configurat~on Cooling Passage Internal Pressure at Blade

i L

Periphery (Choke Discharge Orifice Requirement) , I I Hub supply ~assage for Blade Cooling , ..

Configuration xiii

LIST OF ILLUSTRATIONS (Contd) Title Page Figure Final Inducer Cooling Configuration Showing Cooling Flow Passages 213 Inducer Wall Target Temperatures Versus Achieved Temperatures 214 Inducer Tip Cooling Configuration with Impingement Cooling Techniques Exducer Cooling passage Configuration Final Rotor Cooling Flow Design (Percent of Core Flow at 60- and laO-percent power) Pressure-Side External Surface Temperature Distrib~tion at laO-Percent Power Pressure-Side Internal Surface, Temperature Distribution at laO-Percent Power Suction-Side Internal Surface Temperature Distribution at laO-Percent Power suction-Side External surface Temperature Distribution at lOa-Percent Power Disk Axisymmetric Temperature Distribution at lOa-percent Power Blade Temperature D!stribution for Cooled Rotor Configuration at laO-Percent Power Blade Thickness Distribution for Cooled Rotor Configuration Rake Angle and polar Angle Relationships suction-Side Surface Wall Thickness Distribution, Final Configuration 230 pressure-Side Surface Wall Thickness Distribution, Final ConfigurAtion 231 Total Wall Thickness Distribution (Pressure , .

Plus Suction Side Thickness) 232 t i

, Blade Total Surface-to-surface Distance

Distribution 233

!

r xiv

;

r

(Contd) LIST OF ILLUSTRATIONS

I

~ ..

~, Page Title Figure ~.

~ Stress Model Finite-Element 2-Dimensional

l

, Stress Model, 3-Dimensional Finite-Element ~ Axiometric View ~ (' Stress Model, 3-Dimensional Finite-Element End View Initial Turbine Design Equivalent Elastic Blade

• 156

Stress, Pressure-Side External Surface (Fore- shortened View). Uniform Temp - 294K (70°F) Initial Turbine Design Blade Pressure-Side Internal Surfac~ Equivalent Stress at 5969 rad/s (57,000 RPM) Uniform Temp - 294K (70°F) Initial Turbine Design Blade Suction-Side Internal . ; Surface Equivalent Stress at 5969 radls (57,000 RPM) • t Uniform Temp - 294K (70°F). 240 Initial Turbine Design Blade Suction-Side External Surface Equivalent Stress, lOO-Percent Power at 5969 radls (57,000 RPM). Uniform Temp - 294K (70°F) 241 t Blade Pressure-Side External Surface Equivalent Stress, 0ptimized Design at 5969 radls (57,000 RPM). Uniform Temp - 294K (70°F) Blade pressure-Side Internal Surface Equivalent Stress, Optimized Design at 5969 rad/s (57,000 RPM). Uniform Temp - 294K (70°F) Blade Suction-Side Internal Surface Equivalent Stress, Optimized Design at 5969 radls (57,000 RPM). Uniform Temp - 294K (70°F) Surface Equivalent Blade suction-Side External Optimized Design at 5969 radls Stress, (70°F) Uniform Temp - 294K (57,000 RPM) • rad/s Disk Radial Stress at 5969 Design Initial (70°F) Uniform Temp - 294K (57,000 RPM) • Disk Tangential Stress at 5969 radls Initial Design (70°F) Uniform Temp - 294K (57,000 RPM) • rad/s Disk Equivalent Stress at 5969 Design Initial Uniform Temp - 294K (70°F) RPM) • (57,000 167 Final Design Blade Pressure-Side External Surface Equivalent Stress at 100-Percent Power 249 xv

LIST OF ILLUSTRATIONS (Contd) Title Figure Final Design Blade suction-Side External Surface Equivalent Stress, Optimized Design at Maximum Power Final Design Disk Radial Stress at Maximum power 170 Final Design Disk Tangential stress at Maximum Power Final Design Disk Equivalent Stress, 100-Percent Power Condition Campbell Diagram Showing Interference Between Natural Blade Frequencies and Engine Excitation Orders at Room Temperature Campbell Diagram Showing Interference Between Natural Blade Frequencies and Engine Excitation Orders at Maximum Power Mode Shape No. 1 with Normalized Displacements at 5969 radls (57,000 RPM) Mode Shape No. 2 with Normalized Displacements

258 • I

at 5969 rad/s (57,000 RPM) Mode Shape No. 3 with Normalized Displacements I at 5969 radls (57,000 RPM) Mode Shape No. 4 with Normalized Displacements at 5969 radls (57,000 RPM) Mode Shape No. 5 with Normalized Displacements at 5969 radls (57,000 RPM) 261 Part-power Efficiency Characteristics for the Selected Cooled, Variable-Area Radial Turbine With the Articulated Trailing-Edge stator Concept 266 Cool~d, Variable-Area Radial Turbine Component Cold-Air Test Rig Cooled, variable-Area Radial Turbine, Program Schedule , xvi

---,---

LIST OF TABLES Title Table lOO-Percent Power Parameters Used for Cycle I ..

Analysls stress-Rupture Capabilities of Candidate II ~ S3 Turbine Super alloys Overtemperature Capabilities of Candidate III •

r

S6 stator Materials "

•

Elastic Moduli (Room Temperature) of Candidate ~ IV r S7 Super alloys ~' ~ Evaluation of Cooled rotor Life As a Function v of Rotor-Inlet Temperature, Duty Cycle, and Tip Speed Mechanical Analyses of Uncooled Rotor Disk VI Comparison of Mechanical properties for Cooled VII Rotor Models Selected Cooled, Variable-Area Radial Turbine VIII • Stage Characteristics Final Vane Design Par~meters for Selected IX Variable-Area Stator Concepts Estimated cooled, Variable-Area Radial Turbine

x

Efficiency for Selected Stage Configuration (lOO-Percent Power) Instrumentation for Overall Turbine Performance XI

I

xvii I

j

-.---.~--.--. ----------

ABBREVIATIONS AND SYMBOLS Stator and rotor passage width, em (in.)

b Force.'mass conversion factor 32.174 lb -ft/lb -sec g m f Mass flow, kg/s (lb/sec) W Stage inlet corrected flow, kg/s (lb/sec)

w/elo

Critical velocity, m/s (ft/sec) • Acr Variable geometry stator throat area ratio AN/AND 2 2 Stator throat area, cm (in. ) At 2 2 Rotor exit annular area, cm (in. ) AS tt

Trailing edge blockage, s-

B p Stator sidewall clearance, mm (in.)

C Inducer - shroud axial clearance, mm (in.)

C A Backface axial clearance, mm (i~.)

C • B • Exducer - shroud radial clearance, mm (in.)

C R .

High pressure compressor HPC High pressure turbine HPT Mechanical equivalence of heat, 778.0 ft-lbf/Btu u Rotor axial length, em (in.)

L Merijional distance, em (in.)

M Relative mach number MNR Rotational speed, rpm N Number of rotor blades NB Number of stator vanes N V Corrected speed, rpm N/JO Throat dimension, mm (in.)

' h ' 2 ( .)

Compressor lSC arge pressure, N/m pSla d

Pc

xviii

,

·

t

, ABBREVIATIONS AND SYMBOLS (Contd) f " ,: •

t

Compressor pressure ratio ~ ~ Pressure ratio r • l.

, Total pressure, N/m2 (psia) P'r

f

Total-to-total STAGE pressure ratio

{

PIP) T-'l' • £ p v 2 f· t. Dynamic pressure ~ N/m (psia) q I Radidl coordinate, cm (in.)

R

f

I Reynold's number RE t , p.

, ~xducer hub radius, cm (in.)

R ~ H5 t.

, Stage reaction t ;: RST(; Rotor tip radius, cm (in.)

RT Exducer tip radius, cm (in.)

RT • Rotor Exducer tip-to-inducer tip radius ratio R/R Spacing, cm (in.)

Tangential Thickness, cm(tnl Temperature, oK (OR) TN2 Stator vane leading edge normal thickness, mm (in.)

~tator vane trailing edge normal thickness, mm (in.)

Total temperature based on the relative velocity, oK (OR) Rotor tip speed, m/s (ft/sec) Absolute velocity, m/s (tt/sec) Radial velocity component, m/s (ft/sec) Tan~ential velocity component, m/s (ft/sec) Axial velocity component, m/s (ft/sec)

I

I Relative velocity, m/s (ft/sec)

t ~

Cooling flow rate, kg/s (lb/sec) ; xix

t

t

f

l

ABBREVIATIONS AND SYMBOLS (Contd)

x, Y, Z

Cartesian coordinates cm (in.)

Angle of .t1bsolute velocity vectoI with reference to radial direction, degrees Rotor exit absolute angle at mean radius rad. (degrees) • Stator vane inlet angle, degrees Flow angle at the pivot axis, degrees Angle of relative veloc~ty vector, degrees Rotor blade inlet angle, degrees Stage specific work (total-to-total) kJ/kg (Btu/lb) Average cooling effectiveness l1ave Total-to-total compressor efficiency l1 itical Cooling effectiveness at life critical section cr Total-to-total stage efficiency with the stator sidewall leakage included Total-to-total efficiency Work Coefficient Rotor inlet work coefficient Nean work coefficient based on the mean wheel speed at the rotor exit Static density kg/me (Slug/ft ) j.}s w Pressure loss coefficient Subscr ipts : BASE Refers to base total stage efficiency from specific speed curve BFS Refers to the rotor back face seal BORE Refers to the rotor bore xx

ABBREVIATIONS AND SYMBOLS (Contd) Subscripts (Contd) C Refers to compressor EW Refers to endwall EXT Refers to rotor exducer tip cooling flow discharge • Refers to rotor inducer cooling flow IND Refers to the rotor hub H Refers to individual cooling flow components I Refers to properties at the mean rotor exit radius M Refers to mixed flow properties MIX Refers to primary turbine mass flow P Refers to the pressure surface PS Refers to the rotor back face RBF Refers to rotor internal cooling RI Refers to the rotor shroud RS Refers to the shroud S Refers to rotor back face scallop SCALLOP Refers to parameters concerning the turbine stage STG SYS Refers to parameters concerning the entire system with the interturbine duct included SS Refers to the suction surface Refers to total-to-total conditions

!

T-T I ) Refers to the trailing edge , TE Refers to the stator vane

v

Refers to vane internal cooling 1

VI "

I

I

t

xxi .

~REVIATIONS AND SYMBOLS (Contd) ..

Turbine System Station Nomenclature: Combustor discharge Stator inlet • Stator exit Rotor inlet Rotor exit Interturbine duct exit !

. , xxii ---_._- -- ----

.... ..- .. --.--...

, i ~,

l

r , FINAL REPORT COOLED, VARIABLE-AREA ~)IAL TURBINE PROGRAM ~.

•

f

~.

~ , 1.0 SUMMARY ~ f This is the final report for the conceptual evaluation and ~ design analysis for the Cooled, Variable-Area Radial Turbine Technology PrO<jram. This program was conducted by the Garrett

i

Turbine Engine Company and Wp.s funded by the U. S. Army Research

( •

and TechnolO<jy Laboratory. The program was jcintly monitored by ~ the U.S. Army R~search and Technology Laboratory and the NASA- Lewis Research Center under NASA Contract No. NAS3-22004.

i., I-.

;

·

The objective of the program was to evaluate the aerodynamic

·

~ and .echanical p~tential of a variable-area radial turbine capable ~ of maintaining n nearly constant high efficiency when operated at i : a constant spe"':d and pressure ratio over a range of flows I" ; correspondi.ng tv 50- to 100-percent maximum engine power.

· , t I- The program consisted of four major tasks: "!

o Task I - Nozzle-area variation and rotor-cooling con- cepts evaluation.

o Task II - Detailed aero/mechanical parametric studies based on the concepts selected in Task I t~ deter.ine the optimum turbine geometry. Heat-transfer and stress analyses were conducted in sufficient depth to ensure meeting the 4000-hour turbine-life requirement.

o Task III - Detailed aerodynalftic designs of the vari- able-area stator and the selected rotor.

o Task IV - Heat-transfer and mechAnical-design analyses to substantiate the design.

After completion of Task IV, a follow-on test program was defined

I

, which consisted of completion of the detailed design, preparation i of fabrication drawings, actual rotor fabrication, and aerodynamic testing of the selected radial turbine rotor.

j

I

In general, the results of the program showed that a l589K ..

(2400·', cooled, var iable-area radial turbine was feasible that

~

would satisfy the 4000-hour duty cycle life goal. The parametric- study showed that although a relatively constant aerodynamic efficiency could be achieved at from 60- to 95-perc~nt power, the

I

predicted penalties for interturbine duct loss, cooling flows, stage reaction, and stator leakage resulted in severe part-power

•

performance degradation. These calculated penalties, however, were based on limited data, and in some instances extrapolations from axial turbine data. Therefore, actual part-power performance must be established experimentally. The final cooled rotor selection was based on 1988 materials technology, and features directionally solidified (OS) Mar-M 247 cooled laminated blades wi th 0.17-radian (lO-degree) rotor inlet angle, a 0.26 radian (15 degrees) inlet rake angle and a powder-metal disk that allowed a rotor tip speed 640 m/s (2100 ft/sec). A p~ak aerodynamic stage tlncooled total efficiency of 0.88 was predicted at 100 percent xwer i • • The prQgr~m·identified tWQ CAndidate variable-area stators.

These were designated the articulated trailing-edge and rotating- translating movable &idewall concepts. However, detailed aero- dynamic and mechanical analysis comparisons failed to identify an optimum configuration. To determine an optimum configuration, both concepts must be experimentally evaluated in ~he follow-on test program outlined in Section 9.0.

I

!

I ; ..

•

2.0 INTRODUCTION 2.1 Background The cu:r .tnt world-wide fuel shortage h.a placed increasfld emphaais on minimizing future engine fuel consumption. Even wich the development of aynthetic fuels, costa are a~ill expected to be relatively high when compared to pre-criaia oil coata. Under theae conditions, complex, higher-coat ~ngine concept a become more attractive if aignificant reductions in fuel uaage can be aChieved. Two such engine concepti are the variable-flow capacity engine and the recuperated-regenerated engine.

2.2 Program Objective The objective of thia program wal to evaluate the aerodynamic and mechanical potential of a high-temperature, variable-area radial turbine for use in a variable-flow capacity engine for rotorcraft application. For auch applicationa, the gaa-turbine engine ia required to operate over a wide range of power lettings.

With conventional fix~d-geometry enginel, part-power operation il achieved by reducing engine speed (preaaure ratio) and tempera- tute, thus reducing cy~le efficiency.

However, a variable-flow capacity turboshaft engine has the potential to significantly reduce Ipecific fuel consumption (SFC) at cruiae conditionl. A combination of a variable-diffu8er cen- trifugal compresaor and a variable-atator radial turt-ine would allow reduced engine power, while maintaining nearly constant com- pressor pressure ratio and turbine inlet temperature. The power f range over which this operating mode could be maintained is a function of both the variable-geometry components and the effect

l

, of variable-geometry operation on engine match And component efficiencies.

2.3 Cooled Radial Rotor Technology Recently, the radial turbine has received considerable atten- tion for both automotive an~ small turbosh~ft and turboprop appli- cations. It ia recognized that in smaller-flow class engines, the radial turbine has the ability to achieve higher efficiencies at higher stage work levels than its low-asp£ct-ratio axial turbine counterpart. Unfortunately, the inability to internally cool the radial turbine rotor has restricted its use to primarily lower temperature auxiliary power unit (APU) 3nd turbocharger applica- tiona. However, recent adva~ces in lamination techniques(l) have shown that a mechanically viable, high-temperAture cooled rotor i& now feasible. A 8IIall, high-temperature, cooled radial turbine rotor was designed and manufactured utilizi~g the Garrett laminate , . process developed under Azmy Contract No. DAAJ02-77-C-0032.

2.4 Major Program Considerations To realize the benefits associated with the variable-flow capacity engine, the variable-area radial tUltine must provide a nearly constant efficiency from cruise through maximum power set- tings. For a given turbine efficiency, minimulD SFC will occur at relatively high turoine inlet temperatures and cycl~ pressure ratios. Thus, high stage-work levels will be requtred. Given these considerations, the following issues were addressed in the program: • o The aerodynamic performance potential of a radial turbine as a function of cycle pressure ratio and inlet temperature using projected 1988 material properties; o The d~sign procedures necessary to optimize radial turbine performance over the entir~ duty cycle from 50- to lOO-percent power; .

The effects of turbine cooling flows, and the maximum o allowable inlet temperature for a cooled or uncooled rotor configuration with a duty-cycle-life of 4000 hours; o Defini tion of variable-geometry techniques that could , I minimize performance decrements from both a leakage and • !

i vane loading standpoint.

, i

\

f

" J I I I .. .

-,,""-.........--

t

3.0 AERODYNAMIC PERFORMANCE CORRELATIONS AND SYSTEM OPTIMIZATION TECHNIQUE t f 3.1 Design-Point Performance Evaluation ( ~ t The maximum attainable (base) efficiency for radial turbines !, ~ is currently correlated from either a specific speed relationship, ~, or from stator and rotor loss coefficients derived from experi- f ( mental data. The goals of each method are to evaluate the basic profile and secondary flow losses occurring in the radial turbine for a required combination of rotational speed, flow rate, and

I

• I , work levels. Although the evaluation of individual stator and ~ rotor loss levels on a more fundamental stator and rotor loss coefficient basis was desirable, the actual benefits associated , with this approach have not been achieved for arbitrary designs , ;, due to the complex 3-dimenaional flows that exist in the radial

r

turbine.

( For this program, the maximum attainable efficiency of the radial turbine was based on the specific speed correlation established by NASA in TND-660S(2) and on recent Garrett turbine designs. The NASA correlation corrected to zero clearance is presented in Figure 1. The reduction in peak efficiency at low specific speed is associated with low-aapect ratio blading, while the reduction in performance at high specific speed is associated with high Mach number. It should be emphasized that the NASA data was established for relatively low pressure-ratio stages, and that achievement of these p~rformance levels at higher pressure ratios is based on the assumption of equivalent loss coefficients.

However, correlation with recent Garrett radial turbine designs has indicated that this is a good assumption, as long as stator and rotor exit shock losses are not present.

Regardless of how a base efficiency is established, previous in-house stUdies have shown that the primary deterrent to achiev- ing peak performance is the inability to achieve optimum tip speed. The speed required for peak radial turbine efficiency can be established from the centrifugal compressor slip factor derived by Stani tz (3) and is a function of the overall work level, as shown in Figure 2. It was concluded that high-work radial tur- bines are generally tip-speed-limited, and peak efficiency would not be achieved. However, studies have ~hown(4,S) that, under these cond it ions, an opt imizat ion procedure based on a system

I

analysi s approach could be uti 11 zed to ar ri ve at max imum system performance. The system for this program was defined as a

I

var lable-area stator, a radial turbine stage, and included the downstream interturbine duct, as shown in Figure 3. The objective of the system analysis was to ~inimize the combined loss~s asso- ciated with nonoptimum tip speed (incidence) and interturbine duct 10S8. For an overall imposed stage work coeff icient (,\ STAGE)' the

I

i ~ I !

0\ 1.0 NASA TN 0-6605 DATA CORRECTED TQ

ZERO CLEARANCE TO RE S 3.0 X 10"

0.95 > u

z

w

~------.---

-

/" t ------

~ ~

/ "

~ 0.90 I EXPERIMENTAL DATA , I FROM NASA TN 0-6605 , -'

I \

~ ~

I

~

I

~ 0.85 If(

I

~

I

e

I

I

0.80 0.7 0.8 0.6 0.2 0.3 0.4 0.5 t I I I I , I I 1 (30 ) (40) (50) ( 60) (70) (80 ) (90 ) (100 ) (110 ) SPECIFIC SPEED, DIMENSIONLESS, [RPM (FT3/4/SEC1/2)] Figure 1. NASA Specific Speed Correlation Corrected to Zero Clearance • • .......... ..,1' ......... -.'-' "'~-''''''..........,.--- .

, ..

.

" ~ j: ~ ., , ~" t' ~ STAGE WORK LEVEL RANGE REQUIRED FOR COOLED. VARIABLE

f

., (3000)

~ AREA RADIAL :......t

,.

TURBINE PROGRAM ~, .'

'i'

"

I-; (2800) ~O~------~----~--~--~----~~--~ ~, OPTIMUM TIP SPEED , BASED ON

•

it (2600) i.

- r (,)

w

i

~

f. ~

! (2400)

-

J; .

~ ~ ~ .~ ;::

;!

(2200) ~ :J t ,

o·

w w (2000)

as

Q.

t- ~ (1800) ~~------~~--~------~----~----~ ::) ~

~

(1600) (1400) (1200) 100 300 400 600 , , i I i I (300) (200) (260 (SO) (100) (150)

f

~ STAGE SPECIFIC WORK, kJ/kg (BTU/LBM)

variation of Optimum Tip Speed with

I

Figure 2.

Turbine Stage Specific Work.

..,

• SYSTEM FOIt PARAMETRIC STUDY IS DEFINED 17\ COMBUSTO" EXIT FROM STATOR INLET-TO-INTERTURBINE \.!.II PLANE DUCT EXIT

~~I- ____ -~-j--l

.... -COOLED VARIABLE AREA STAr\lR I WITH CONTOURED SIDEWALLs I -'IDooo<:r-"-~ I DOWNSTREAM VARIABLE POWER TURBINE STATOR

I

t

j

z z. 0 em (0 IN.)

I

, Figure 3. Cooled~ Variable-Area Radial Turbine, System Definition.

....... w. Jd!'ll P'!AI ....

I

: .

, , magnitude of the incid~nce loss is a direct function of the !

inducer-to-exducer work split selected. This relationship is expressed in the following manner:

t

i'

gJ~HT_T ( Rr4S/ )2 A t (1) .'

. ----:ii2~ 1:1: A4 - R.r MS 1'1 ASTAGE

U 4 T Therefore, the imposed rotor inlet work coefficient (A4) depends on the selected rotor exit work coefficient AMS as well as the exducer-to-inducer radius ratio RM~/RT4. This, combined with the continuity equation, estab1ished~the rotor-exit swirl and Mach number level.

When the incidence loss is calculated from the kinetic-energy difference between nonoptimum stage rotor-inlet work coefficient (A4,ACT) and an ideal work coefficient (A4, IDEAL) based on the Stanitz slip factor, the following relationship was derived for the ratio of efficiency with incidence to the efficiency using optimum rotor-inlet cond~tions:

= ----------T.----~------------------

(2)

=-=-_l]=B,ASE (A ) 2

1 + 2 >-STAGE 4, ACT - A4, IDEAL

where: = Imposed stage work coefficient due to a

ASTAGE specified work and allowable tip speed • Imposed rotor-inlet work coefficient = 1 - ~ (slip factor with radial blades).

t This relationship shows that there is a strong incentive to reduce th~ rotor-inlet work coefficient (A4,ACT)' However, reduc- ing A 4 ,ACT by increasing A SM increases the losses of the down- stream interstage duct. This is due to increases in rotor-exit kinetic energy and swirl. The variation of inter stage duct loss r (as a function of average rotor-exit swirl) is presented in r , , Figure 4. The data was normalized by the minimum loss coefficient measured for two separate tests and shows good agreement for the range of exit swirl angles investigated. However, the minimum loss coefficient varied significantly for each duct configuration.

This indic~tes that, in addition to the level of swirl, the duct configuration also influenced the magnitude of the duct loss.

Inlet duct exper iments made by Dovzhik et al. (6) were based on uniform temperature and pressure, with straight duct sidewalls.

The Garrett Model GTP30S-2 (7) duct data shown in Figure 4 was based on actual rotor-exit conditions (inlet-duct conditions), using moderately curved sidewalls and radius ratios between tur- bines. The minimum duct-loss coefficient of 0.148 was derived ••

7.0 o GTP306-2 DATA (NO STRUTS): WMIN • 0.148,

/

z I

-

.I

6.0

31 ~

-'

/

ex: DOVSZHIK. S.A .• KARTAVENKO, V.M.

....

I

"MEASUREMENTS OF THE EFFECT OF z

i

I w FLOW SWIRL ON THE EFFICIENCY OF :z:: 5.0 § ~ ANNULAR DUCTS AND EXHAUST ....

~ NOZZLES OF AXIAL TUR80MACHINES"

/

~ i w WITH OPTIMUM GEOMETRY I = 0.075/ WMIN t- z w

/

4.0

u

I ~ P P

~

~ ". T5 - T6 'I w w= t- (J q5 ::l

I

c 3.0 w

§

• I < I t;

Iii

• I

::l a::

/

C w 2.0 . f ....

V w Z

-

~ :)

~

V

a::

~ 10_.0

,..

w 1.0

-

z ~

-

-----

-

z ~

-

o , , , I I I I I 0.8 0.9 0.1 0.2 0.3 0.4 0.5 0.6 0.7 I , , I I I I (0) ,20) (30) (40) (50) \ 10) ...

TURBINE EXIT SWIRL (DUCT INLET), RADIANS (DEGREES) Figure 4~ Bffects of Swirl on lnterstage Duct Loss CI:>eff icient.

from the GTP30S-2. However, the var iable flow-capaci ty eng ine concept will require a variable geometry power turbine with conventional axial rotating vanes. Interturbine duct struts will be needed to support the duct hub contour. Under these condi- tions, the duct minimum loss coefficient will increase. Previous in-house experimental programs have shown that, when struts are required, a minimum loss coefficient of 0.200 is representative.

When the effects of rotor-inlet incidence and inter turbine duct loss are combined and examined over a range of rotor-inlet work coefficit!nts, peak system performance is achieved for a range of rotor tip speeds (Figure 5). This determination was based on a study conducted by Garrett for TARADCOM for an Advanced Armored Vehicle Gas Turbine Engine (under Army Contract No. OAAK30-C- 0093). Two turbine rotor blade configurations were studied; a

conventional radial blade (,88 = 0 radian (0 degrees)], and a

3-dimensional, nonradial rotor blade (~8· 0.35 radian ( 20 degrees)] at the rotor inlet. The potential benefits of both are shown in Figure 5.

Another recent study(5) using 3-dimensional rotor blades was conducted by Garrett for l644K (25000~) automotive radial turbine.

This study was based on projected ceramic matp.rial technology cer- amic stress capability of 621 MPa (90 ksi) and ~ Weibull modulus of 15.

It was concluded from this study that: # o Exceptionally high rotor-tip speeds 701 mls (2300 ftl sec) were feasible for radial-bladed rotors.

o Nonradial rotors could achieve equivalent performance at lower tip speeds. However, the increased mechanical design complexity did not justify nonradial blades for this application.

In 1978, a Garrett company-sponsored research program was conducted to investigate the aerodynamic and mechanical feasi- bility of metallic, nonradial-bladed rotors. An existing radial rotor was modified to a 3-dimensional design for application in an APU with a l3l1K (1900 F) turbine-inlet temperature, a tip speed of 549 mls (1800 ft/sec), and a O. 35-radian (20-degree) blade angle. Rig test results showed that the increased performance achieved was in good agreement with that predicted from a decrease in inducer incidence loss. However, it must be determined whether the favorable aerodynamic characteristics offset the mechanical- design complexity involved. This would depend on material

Jevelopment and application. If dual-alloy rotors with os blades

allowed optimum tip speed Lo be approached, then the increased

...-' ...... ----------

HIGHER INTERTURBINE DU:r LOSS HIGHER ,NC;IDENCE LOSS 0.90 10.36 RAD ~D - (200)

A--

" 0"" ....

/' l3 (00) "

a•

.I' _ ) 17T - T 2-6 CYCLE • 0.878

0.88

-

~"".

,

-~

\-~-=- -IT-

N ..,'''' _ 0.35 7 0.86 Ut.t .., 610 mil

;,' .... ,..;13 (200)

l; (2000 ft/sec) >'

"

CJ z 0.84 w u

13n -(0°)

~

I.L I.L Ut.t '"' 549 mi.

X

w 0.82 (1800 ft/sec) 0.35

'\

..

~

/{J B -(20°)

w ~ , ,--- .........

en ., ....

> 0.80 en

,,'

" •

, N .. 4084 RAD/S (39,000 RPM) .

-' I « NO. OF BLADES - 14 t- I O I I 0.78

! J

, 1-- -- NONRADIAL BLADES • (0°) VOS

7 "

..... ..............

RADIAL BLADES / .......

0.78 ~

e

Ut.t • 488 mi. :~

(1600 h/lICi

/

0.74 1,4 1.7 1.8 1.9 1.3 1.5 1.6 1.0 1.1 1.2 • ROTOR-INLET WORK COEFFICIENT, ).4 Figure 5. Results of System Optimization for TARADCOM Advanced Radial Turbine.

""' benefits and mechanical complexity of nonradi3l blades would not be warranted. However, in variable-flow capacity engine applica- tion, rotor-inducer loading increases with reduced power. Under these conditions, the added complexity of nonradial blading may be justified. Therefore, the effects of rotor inlet-blade angles of from 0 to 0.35 radian (0 to 20 degrees) were evaluated for this program.

In addition to rotor incidence and duct loss, additional .

losses are incurred that must be accounted for in the parametric study. The correlations are described in the following para- graphs: t' o Reynolds Number Effect: Based on results from NASA(8) data.

o Rotor Clearance Effects: Based on correlations derived from several sources{9,10) and on Garrett test results.

The performance penalties are a function of both axial (Ca) and radial (Cr) clearance. The experimental data from these sources is presented in Figures 6A and 6B.

Additional rotor clearance effects are present with rotor scallops. The performance effects as a function of back face clearance recently have been evaluated for the Model GTP305-2 turbine(7) and are presented in Figure 6C. In additior., the results of a 1978 Garrett- sponsored radial turbine research program showed the effects ot scallop depth on radial turbine performance.

These data will be used to supplement the correlation derived from the Model GTP305-2 turbine.

o Rotor Blade Number Effects: Rased on the Models GTCP305-l, GTCP36-4, and published Pratt and Whitney data (11) • The resultant change in turbine efficiency represents the effect of increased blade loading as blade number is reduced for a given work requirement.

Rotor Backface Disk Friction Effects: Estimated based o on the results of Referonces{12, 13). It should be noted that tab loss is based on a full rotor back face disk. The rotor scallop effect on rotor disk friction ' is currently not available.

Experience gained from axial turbine designs has shown that the effects of rotor-exit hub blockage and stage reaction should not be ignored during the program turbine design process. At present, only preliminary correlations are available for these two effects (5) • However, recent investigations conducted by Meitner(l4) and McLallin(l5) will allOW further refinement for the effects of reaction and blockage.

1.oo~ce--...,...- ..... ~-~- ........... - ...

>" AXIAL CLEARANCE, PERCENT u I z 0.98 J-!~~....p"'c-+--t--"""-·""'-1.71 ~.. 7.1 ~ u. 0.98 t--T ....... d-"II~~~~-+~-I-J=12.5

~w 1n

W~

..J i7 0.14 t--t--t---t---"lIIfiod~"""-l~-I--+--I

~l; ~ ~ 0.92 t---+--+--+--+--f---f:llt"l!~--I I I ot;

~ 0.90 0 1 2 3 4 & 8 7 8

~ RADIAL CLEARANCE.

~ PERCENT OF EXDUCER PASSAGE HEIGHT (A) Summary of Radial and Axial Clearance Effects on Total Stage Efficiency (9), 1. 02 >" RADIAL' CLEARANCE. PERCENT u z ~ w 00 1.

o 0.26

~ :l..

-

5:! I:l 0.92 -no.

~

D 3.0 -----...

o 7.0

~ l~ ~.

--.

94 o.

~ ~ 0.92 0 10 20 30 AXIAL CLEARANCE.

PERCENT OF INDUCER PASSAGE HEIGHT (B) Summary of Radial and Axial Clearance Effects on Total S~ Efficiency It Design Equivalent Values of Speed and Pressure Ratio(10), 0.98 SCALLOP RADIUSI - INDUCER TIP RADIUS • 0.781 0.98 ..... _ ..... - ......_ ...... - ....... _ .... _ ..

o 10 20 30 •

BACKFACE CLEARANCE, PERCENT OF INDUCER PASSAGE HEIGH1 ; (C) Effect of Roto. ~ Clearance on Stage Total Effic.;iency at Design Conditions . .

Figure 6. Summary of Radial Turbine Rotor Clearance E~fects.

3.2 Off-Design Performance Evaluation ~.

l The accurate prediction of off-design performance character- ; istics of the cooled, var iable-area radial tUJ":bine was a major ~ :r. aspect of this program. A design point, as such, 1s not meaning- ~.

ful from an aerodynamic standpoint, aince the normal mode of oper- f ation was between 50- and lOO-percent maximum flow. The radial turbine off-design performance predicticn technique used for the i ; program was similar to that used by NASA (16), except that the , incidence model was based on a slip factor. An intermediate loss coefficient was included between the stator and rotor t.o account

t

r.

for stator trailing-edge mixing and secondary-flow loss, in addi- tion to the vane less-space loss. Radial stator flow calibrations , indicated that the flow coefficient up to ~he stator throat WAS ..

between 0.985 and 0.995. If losses higher than tllis are assigned, ".

I then the predicted flow characteristics (as a function of speed ~ and pressure ratio) would be in error. Input to the program ~ consists of the optimized vector-diagram quantities. Based on f .'

previous test results, an initial loss split between stator and .

rotor was specified. LoSS coelf icients used in the off-design .1 ~ mode were then calculated. The computer program ability to accur- ~.

ately predict the off-design radial-turbine efficiency and flow characteristics is illustrated in Figures 7 and 8. The compar- isons shown are relative to the tested performance of the advanced Model GTPJ05-2 radial turbine(7). For the cooled, variable-area radial turhine, the objective is to establish a design point

I

between 50- and 100-percEont flow that would result in maximum ..

.

efficiency. The off-design computer model was modified to account I , for both 3-dim~nsional rotor blading and downstream inter turbine , '0. duct loss. The duct-loss correlation used in the off-design calculation was identical to the design-point system-optimization correlation. This allowed the off-design system performance to be

I

evaluated in a consistent manner. The selection of several design points allowed optimization of system performance over the entire 50- to 100-percent tlow range. Figure 9 illustrates the extremes in turbine operatillg conditions that will exist with the var iable- area radial turbine.

t I; 3.3 Effects of Cooling Flow and Nozzle Leakag!

f

t • The inability to internally cool radial turbine rotors has

r been a signif icant deterrent to their use and to the amount of

i- research devoted to this component. The advent of laminated con-

f

struction, however, not only eliminated this cor.straint, but also increased radial turbine applications. Unfortunately, little data on the effects of rotor internal-cooling flow on the performance

•

of radial turbines is available. Three cooling approaches appeared feasible;

I

....

1.00 r-.,.-.------r"------r----~~..._:~---_r_-----------

--- .~

- ...

0.191 1~ __ ~~ POINT ~ 0.98 ," ~ ~ N~u-l~

~ ... 0.97 --1-

f

'7 0.96 t--~~ t;

~.. 0.96 I I I

~ 0--0---6 GTP305-2 RADIAL STAGE TEST DATA

~ 0.94 - - - - PREDICTED OFF-DESIGN PERFORMANCE+- ~ BASED ON DESIGN POINT MATCH ~ 0.93 ---

~

e 0.92

• ...

c( I- 0.91 --.- I " ~ 0.90 ' -L ' 2.0 2.5 3.0 3.5 4.0 45 5.0 PRESSURE RATIO. PIP'T-T ur Fi9 e 7. Predicted and Tested Off-Desi9n Efficiency Characteristic Ca.parison.

.......... ,.. '.-., .. , ".,' ..... ''''...

(0.84) 0.29

-

- ~

G w

IP ~~-

Ie (0.62)

0.28 - J DESIGN POINT

CD ~ M (0.60) I 0.27

"'

~

~ (O.M) ..

0.26 -

~

1-+

-' 0-0--6 GTP306-2 RADIAL STAGE TEST DATA u.. (0.56) 1------#1-----+----+ - - - - PREDICTED OFF-OESIGN w PERFCRMANCE BASED ON 0.25 -.

b

DEStcN POINT MATCH w t---.,

~ 10.&4) 1

0.24 - (0.52) I • 2.0 2.5 3.0 3.5 4.0 5.0 4.5 0.23 - PRESSURE RAT~. PIP'T-T Figure 8~ Predicted and Tested Off-Design Plo~ Characteristic ~ .....

COliparison.

100' CORR. FLOW (w..4l/6) C • 1.28 RAIl (72.0·)

8 - 0.16 RAIl (9.1758·)

uil' cr' • 0.8772 6 - 1.124 RAIl (1:>4.4]") M5 ~5 • 0.363 RAIl (20.80·)

'\.s. 5.772 eM (2.2725 IN)

U/A~r' • 0.~251 (ROTOR BlUT MEAN RADIUS) • 0.1295 C • 1.40 RAIl (80.37·)

8 - 0.77 RAIl (44.18·)

8MS - 1.122 RAIl (64.27') C • 0.995 RAIl (56.99") MS ).4 - 1.19745

FIGURE 9. I-Dimensional Vector Diagram Comparison (IOO-Percent

and 50-Percent Inlet Corrected Flows).

o The first assumes that the majority of the cooling flow is discharl:led in the rotor-shroud region. This would require pumping into these regions. Depending on the magnitude of cooling flow required, this would result in a performance decrement of 2 to 3 efficiency points.

0 The second approach assumes that the cooling flow is discharged at the same location. However, the pumping penalty is offset by a corresponding reduction in tip- ~ " clearance leakage effects. The feasibility of this approach was demon8trated using axial-turbine test data t with tip-discharge cooling.

i

~; The third possibility assumes that the majority of the ;; cooling flow could be ~ischarged at the rotor trailing- edge. Under these conditions, the cooling-flow penalty ~ would be in between the previous two effects due to t replacing the lower-momentum, trailing-edge wake region with the higher-momentum ccoling flow.

These assumptions not onli result in predicted performance uncertainty, but also indicate the importance of a comprehensive test program.

Even for uncooled rotors, a certain amount of cooling along the rotor backface (approximately 1. 0 percent) is required to prevent hot gas from recirculating in this region. This effect was evaluated during the GTP30S-2 test progra.ll. (7) The test results showed that the required pumping work (to the scallop rotor speed) was offset by expansion of the cooling flow through the rotor.

The utilization of variable-area stators will also result in a certain amount of endwall leakage flow. At high-temperature levels, the minimum feasible clearance will require careful mechanical analysis. Previous Garrett variable-area stator radial turbine designs have used full-vane rotation similar to conven- tional axial stator designs. Test results conducted on these i ..

designs have shown similar leakage effects. This implies that sidewall clearances of from 0.0051 to 0.0127-cm (0.002- to 0.005-in.) are required to achieve equivalent axial stator per- formance with 0.0127-cm (0.005-in.) clearance. On this basis, alternate variable-area concepts must be carefully evaluated during Task I.

I

f

, •

4.0 'l't\$K I - $'l'A'l'ON. ANl> ltO'l'Oa CONCEPTS EVALUATION 4.1 Variable-Area Radial Turbine Requirementa The basic requirements and goals established for the Cooled, Variable-Area Radial Turbine program were: Single-stage radial turbine defined as the gas- o generator (high-pressure) turbine in the variable-flow- capacity engine; o Turbine specific work is between 407 and 698 kJ/kg (115 and 300 Btu/lbm); o Turbine-inlet temperature goal: Cooled configuration • 1644K (2500 t) Uncooled configuration - 1478K (22000~) o Turbine mass flow is nominally 2.27 kg/s (5.0 Ib/sec); 'l'urbine mate'r ial properties: Based on 1988 metal o technology; o TurbL.d mission-cycle life goal: 4000 hours.

• o The turbine will operate at constant speed and pressure ratio over a range of flows from 50- to lOa-percent maximum engine power.

" The objective of Task I was to define and evaluate potential variable-area stator and internal rotor cooling concepts. An engine configuration was defined, and cycle performance was evaluated over a range of radial turbine inlet temperatures and cycle pressure ratios at lOO-percent power. A preliminary parametric study was then performed over a range of power settings from 50-to laO-percent power. The preliminary study allowed early evaluation of the variable-area radial turbine and indicated trends that were used as guidelines for selecting a representative stage conf iguration for the aerodynamic and mechanical concept evaluation.

The var iable-i:low-capacity turboshaft engine configuration selected cons isted of os 2-stage, var iable-di ffuser centr i fugal compressor, a reverse-flow annular combustor, a single-stage vari- able geometry radial turbine, and a 2-stage, variable-geometry free-power turbine. with this configur~tion, the variable- geometry radial turbine specific work would fall between 419 and 512 kJ/kg (ldO and 220 Btu/lbm) and compressor pressure ratio would be between 14:1 and 20:1.

4.2 100-Percent Power Cycle Study A cycle analysis with estimates of engine/component perform- ance was conducted at 100-percent power for cooled rotor configur- 0 0 ations at l533K (2300 F), l589K (2400 F), and l644K (2500 F) rotor inlet temperatures. In addition, an uncooled configuration at 1478K (2200 F) was also evaluated. The following basic assump- tions were used for these analyses: o Engine mass flow s 2.27 kg/s (5.0 lbm/sec) o Burner efficiency • 0.995 o Burner total pressure loss (including a 1. 57-radian

(90-degree) bend upstream of the turbine stator) = 0.041

o Interturbine duct total pressure loss: 0.015, o Two-stage, variatJ]·:: -:;C!ometry power-turbine total-to-

diffuser-exit static efficiency = 0.8751

o Gearbox efficiency • 0.98r o Gas-generator spool mechanical eff.iciency • 0.9951 o Two-stage, variable-geometry, centrifugal compressor efficiency levels based on compresaor design study pro- jections (Table I), and o A range of variable-geometry radial turbine stage total- to-total efficiencies from 0.84 to 0.90.

The 100-percent power parameters used for the cycle analysis are also listed in T.!ble I. This table in~ludes the two-stage compressor eff iciency, leakage flow, and turbine cooling flows used to define cycle performance as a function of compressor pres- sure ratio. The results of the cycle analysis are presented in Figures 10 through 12 for the cooled rotors, while Figure 13 shows the results for the uncooled configuration.

4.3 Preliminary Parametric Study (Stator/Rotor Concept Evaluation) 4.3.1 Design Point Study

I

• Using the design conditions establishpd in the cycle ana- lysis, a preliminary parametric study was performed over a wide range of cycle pressure ratios. The objective of this study was to examine the relationship between turbine performance, geom- etry, duct loss, and rotor relative temperat~re over a range of

TABLE I. ln~-PERCENT POWER PARAMETERS USED FOR CYCLE ANALYSIS

--

W ng 'W Cooli A T4 T4 'N T4 =

=

'" =

\'lLe~~-!<3.e l6HK 1589K l5]3K 1478K I' 1] • c W (250QoFl (2400°F) (2300 F) p:200~** ~\:

A ---

10 0.829 0.0064 0.062 0.052 0.042 0.010 12 0.823 0.0075 O.Ob9 0.058 0.045 0.010 14 0.817 0.0088 0.074 0.063 0.049 0.010

I

16 0.812 0.010 0.079 0.068 0.052 0.010 18 0.0112 0.084 0.806 0.072 0.054 0.010 0.0125 0.087 20 0.800 0.075 0.056 0.010 -") -) 0.795 0.0138 0.090 0.0775 0.057 0.010

.. -

24 0.790 0.0150 0.091 0.078 0.057 0.010 *Total-to-total efficiency levels based on compressor design study projections minus l.O-point for variable geometry.

**Uncooled rotor: 1.O-percent cooling flow for rotor back face

HP1 1l1-1/ _ 6.11-0.

...

l: <1 (100) (230)

~C(220)

~ (210) , ~ (200) 0.90 ~ ~ (190) 0.88 oM: 0.86 (180) 0.84 (0.46) 0.28 ~ l: ......

~ (0.44) l: ......

0.84 CD ..l

--

~' (0.42) ~ 0.86 ...

-

0.25 ~ oM: 0.88

-

u' (0.401

.. 0.24

,.

en 0.90 (0.38) 0.23 12 14 HPC PRESSURE RATIO Figure 11. Cooled, Variable-Area Radial Turbine Study,

T4 = lS89K (2400 F) At Maximum (lOO-Percent)

Power.

ii (300) ..I 1l1-1/HP1 ::t

-

-

tii

0.84 ~ 0.90

-

.. (200) ~ ~.

% (100) <l (240) -; (230) A.

(220) 0.90

=

.

(210) 0.88 i.

0.86 ~ (200)

!

0.84 '(190) (0.46) ....

~ 0.27

~0.44)

1l1_11HP1

i

0.26 0.84 ....

.::( 0.42) ~.

0.86 0.26 ~ ..

0.88 ~ ~(0.40)

- 0.24

0.90 u' &I.

fI) (0.38) 0.2~ 10 12 14 18 18 20 22 24 HPC PRESSURE RATIO .

Figure 12. Cooled, Variable-Area Radial Turbine Study,

T4 • 1644K (2500 F) At Maximum (100-Percent)

Powet.

ii (300 -' ::;, HP1

- eoo

1l1-1/ t- ID

-

. - 600

0.84 .... 0.90 (200 c:i ~ 400

:;I --..-

--

% <l (100 (210 :: (200

i4(

A.

::t

- 0.90

.

0.88

i

oM 0.86

~ (170

oM 0.84 (160 0.28

-

Q; 0.27 ~ (0.

«

1'I1_1/HP1 0.84

i

0.26

.=

(0.

~ 0.26 0.86 ~ ..

'i.

:! 0.88 0.24 cJ &L en 0.80 0.23 10 12 14 16 18 HPC PRESSURE RATIO Figure 13. Uncooled. Variable-Area Radial Turbine Study,

T4 • 1478K (2200 F) At Maximum (lOa-Percent)

Power.

I

, f

cycle pressure ratio~ and rotor-inducer tip speeds. This study i also provided the characteristic turbine stator and rotor geometry f, used in the detailed concepts evaluation.

i The l644K (2500 F) cooled configuration was based on a 1-

f

dimensional aerodynamic analysis and account~d for the following

t

l effects:

t

~ .

o Required turbine operating conditions for oach selected t cycle pressure rati01 f ~, o A rotational speed for each cycle pressure ratio result-

t

ing in a specific speed to maintain peak attainable ~ efficiency.

f

I o Effects of rotor inlet incidence for a range of rotor- r inducer tip speeds.

!

I f o Reynolds number effects.

o Rotor-clearance effects based on a shroud clearance of 0.038 cm (~.015 in.) and a rotor backface clearance of 0.078 cm (0.030 in.). The backface-clearance effects were updated in the model to account for backface scallop saddle-to-inducer tip radius ratio. This radius ratio was set equal to the specified exducer tip-t'l- inducer tip radius ratio.

o Blade number effects • 14 full blades.

o Rotor exit-hub radius • 3.9 cm (1.55 in.). This radius was representative of previous designs and accounted for a bore radius of 2.2 cm (0.85 in.) plus a 1.8-cm (0. 7-in.) disk between the bore and rotor exit hub contour.

o Rotor backface disk friction effects.

o Rotor reaction effects.

o System performance (stage plus illterturbine duct) was evaluated as a function of rotor e~it swirl with a mini- mum loss coefficient (w) of 0.2001 o Effects of rotor exit blockage will be accounted for in the detailed parametric study (S~ction 7). (However, with the large rotor exit-hub radii required with the bore, rotor-exit blockage was not a significant factor)~ --_._------. __ ...

o Current state-of-the-art performance levels; o Radial rotor blades.

Cycle pressure ratios of 12:1, 16:1, ~nd 20:1 were selected for the preliminary parametric analysis. This pressure-ratio range covered the minimum turbine stage work level of 407 kJ/kg (175 Btu/lb) and the minimum SFC. The results of this parametric study are presented in terms of both syst~m and stage results • System results were taken from stations 2.0 to 6.0, while the stage results were taken from stations 2.0 to 5.0. (Total condi- tions in both cases.)

Detailed results of the parametric study are presented in Figures 14 through 16 for a range of s~ecified rotor inlet work

coefficients (A4 = Vu 4/Ut4)*' As expected, peak performance was

achieved for a given cycle pressure ratio and specified tip speed.

Figure 17 shows the results obtained when the loci of peak system efficiencies were plott~d as a function of cycle pressure ratio.

Although the results shown are for a turbine inlet temperature of l644K (2500 F), lower temperatures reaolts would be similar.

Also, all results are for maximum power. Therefore, the char- acteristics over the duty cycle have yet to be determined.

Rotor geometries for each cycle pressure ratio are presented in Figure 18 and show the effects of increased cycle pressure ratio on reduced rotor inlet corrected flow. The rotor flow paths shown are for relatively low exducer tip-to-inducer tip radius ratios (minimum disk designs). Changes in rotor geometries as a function of radius ratio for a cycle pressure ratio of 16:1 are presented in Figure 19. These geometries were defined as charac- teristic flow paths for the cooled rotor conoept evaluation.

Figure 20 shows the effects of rotational speed on turbine stage efficiency, specific speed, and geometry for cycle pressure ratios of 16:1 and 20:1.

The basic parametric study was performed with a specified stator ex it angle of 1. 26 radians (72.0 degrees). Rotor clear- ance, reaction, and stator-geometry effects (as a function of sta- tor exit flow angh·) are presented in Figure 21. The repre- sentative rotor geometry shown in Figure 19 was used to evaluate the stator variable-geometry concepts.

*Dashed lines in the figures refer to parameters on the right side of the plots.

DASHED LINES REFER TO RIGHT SIDE VARIABLES

•

, SOLID LINES .. EFER TO LEFT SIDE VARIABLES ,.

•

NOTATION ON CURVES REFERS TO INDUCER ; ,

•

TIP SPEED (FTlStC) i

~

IS t ,.

0.12

S~

J

ig

0.10 o.a ~ : ~ W wa: ..

0.7 Ii

0.01

f

2::1-

-I"

0.8 ...

0.01 Iw<l ::I a:

1 0.1 ~i

o.CM

~ ..

~ a:~ 0.02 We

i. 0.4 e~

~~ , !O O~ 0.0 ~ 0.3 a:5

'"

w f:J' c

r

0.10 (-10) ..J ~ _N

-o.ao

"

(-iO) a:.

U) ~

t

~J; ~ -0.10

~ . (-30, ... !

~.

0> o.

:~' --

-GAO ~~

(..JO,: I

i Ow ".

~- -0.20 'U

~ (.,o'!1

~- ~ elL

~, ... "-

0.0 0.80

OW (0' i

t ~ " I"- , 0.10 > (2440) " u !.

w· w a: (2340)~ IL " IL C 0.86, w a: ~ "-

f

2 ~ ~.

(2240)Io~

., , ~ '''0......

~ eo

f

N ~~ " , >

e

~ 0.80 tit ... .

'i " ',. ""0 .......

w~ ) ~J; (2140'i; "

.' '~ '~ ........ '6oo

,', ~ - ...

..

a: " 0.75 llO4O'~

~

a: C ~ 0.70 ~ (1840' 0.8 0.1 1.0 1.1 1.2 1.3 1.4 ROTOR INLErWORKCOEFFICIENT, x..

Figure 14. Preliminary Parametric Study for Cooled Rotor Concepts, 12:1 crc1e Pressure Ratio,

T4 = 1644K (2500 F), N = 6178 rad/s (59,000 RPM)

_. ------------------- ._-

fQIU: • DASHED LINES REFER TO RIGHT ItDE VARIADLES • SOLID LINES REFER TO LEFT SIQE VARIABI.ES ~ • NOTATION ON CURVES REFERS TO INDUCER ~ til TIP SPEED (FT/SEe) 0.'

0.14

!; ~

e~

~i 0.1 0.12 0.7 0.10

ig 10 ~

'" ~; ~~j

0. • ...

Zw )(~> i5a: 0.& 0.01 ~;:)

"'3

a:- ~a ot- 0.04 0.4 t-- ~'"

",f 03

a: t- 0.02 0.3

!

0.0 ~ (-60) !

-0.'

(..eo) I _

es • -0 ••

(-30) ~ ",' !

)(~ -0.4

(-20) '" 1-

a: 1

-0.2 (-10)

e

(0) a: -0.0

•

(2440) 0.80 Uloo >' (J Z ~ (2340)

'"

"'0

~

>-

0 •• "- i=~ "-

~ ~

'"

(2240)

:. ~

.t- ..

~

.,. t- •

0.10 N

....... o.!Boo >

~~ (2140) .,. "=' !~ "0 ....... 0...

~t" C a: ~ O~

" ....... ~

0.75 t- (2040)

St

a:", t- t- ..;, 4( (1INO) t- 0.7 O t- 1.3 1.4 1.5 1 •• 1.0 1.1 1.2 0.8 0.'

ROTOR INLET WORK COEFFICIENT, X • .

Preliminary Parametric Study for Cooled Rotor

Figure 15.

Concepts at 16:1 Cycle PreS8ure Ratio, T, ;; 1644 (2500·F) N II 6807 rad/s (65,000 RPM).

NOTU: • DASHED UNES REFER TO RIGHT IlDE VARIABLES • IOLiD LINES REFER TO LEFT SIDE VARIABLES • NOTATlON ON CURVES REFER. TO INDUCER TIP SPEED (FT/lEC' ;

, ~

I

•

G.I § i

~

0.7 i

~~ '0'

0.1 ~ Q

'= "

'n ~~

~ t:iJ.

~

u ~g>

o.

r 0.4 ! E

).

1-0 c ~ 0.3 a: U ~.

~ ;.

(-60' ..l !

, -0.8 c_ ~ -G.1

( .. "e"

,

-0.1 Ie

..

(-3e"

-0.5

tf5

f -0.4

.

( .. ,ze" ~!

~ -0.3 -0.2 (-'0)

-0.1 el

(-0' It t.

, ..

~

I

(2340) w 11&5 at ~ C

(2240) i:L

wt,.

,.

(2140) t- '" t; .J ~ w z a: (2CMO) - t- at ,.

~ (UMO) C I I.

1.4 1.5 , ..

1.3 1.1 1.2 G.I 1.0 ROTOR INLET WORK CO!FFICIENT. >A

Figure 16. Preliminary Parametric Study for Cooled Rotor

Conc~pts at 20:1 Cycle Pressure Ratio, T4 • 1644K

(2500.F), N • 7435 rad/s (71,000 RPM).

NOTES: • DASHED LINES REFER TO RIGHT SIDE VARIABLES \ • SOLID LINES REFER TO LEFT SIDE VARIABLES • NOTATION ON CURVES REFERS TO INDUCER TIP 0.27 SPEED (FT/SEC) 0.44 J: ....

~A.

0.26 -:

--

.f. ~ 0.42

0.25 ~ -:::E yO CD 0.40 0.24 ~..1

(1)=

w 0.38 0.23

>

i= ~ ~ a: (2340)

w""

a:wjL ~ ~ct. (2240 i~~ -a: a: w (2140)

e~

i ~ (2040)

0.75-'--....--....--....--....--...--...---r---r--....- 16 17 18 19 20 12 13 14 15 CYCLE PRESSURE RATIO Figure 17. Parametric Study for Cooled Rotor Concepts, 1644K (2500 F) Case.

<~ __ .... - ... "~ .... ~ _...., .... A-. _ __ l"~-':-~ -. -.,_~~ ~_~ '" '" _~'. ___ . ..- __ ,... ___ .. _~_~~ .~_

PR = 12:1

PR = 16:1

PR = 20:1

N = 6178 rad/s (59,000 RPM)

N :: 6807 rad/s (65.000 RPM)

N = 7435 rad/s (71,000 RPM)

P/P)T - T 2-5 = 2.807

PlPl -1" 2-5 = 3.57

P/P)T-T -5 ,. 4.49 T

Wy'()16 = 0.4472 kg/s (0.986 LB/SEC)

Wy'016 = 0.331 kg/s (0.729 LBS/SEC)

Wy'OI6=0.26kg/s (0.575 LBISEC)

AH = 384 kJ/kg (165 BTU/LB)

AH = 460 kJ/kg (198.0 BTU/LB)

AH = 528 kJ/kg (227 BTU/LB)

~4 = 0.950

A4 = 1.040

A4 = 1.075

l1T-T 2-5 = 0.90'

l1T-T 2-5 = 0.884

l1T-T 2-5 = 0.873

T REL :: 1513K (2264 F) o TREL :: 1469K (2185 F)

T REL = 1437K (2127 F)

b:: 0.50 ern

-O_b. 1.06 em

(6) J 13

--0- b = 0.80 em

- (0.259 IN.)

(0.418 IN.)

(0.314 IN.) -D

11 ·U 610 _ -,--- u .640 T - -rUT" 670

- T T4 - 120001

T 4 121001 -: (4) T 4 (22001 ~ RIR = 0.665 RIR = 0.683

-

~r75

(3)

fi -L

U; ~--.L

::;) 6

-

(2) c ca: a:

.-

RBORE .. 2.16 cm(O.8S IN.)

1 2 ~ ~ J 1 ~ --1\r-T'"""'T'"""T...,.- ...._ ..

o 1 2 4 5 6 1 01234561 , --- • j .. -_._y- •

CO) e" (2) (0) e 1) (2) (0) C;) (2)

em (IN.) em (IN.)

em (IN.'

*U - m/.IFTISECI T

"'"

Figure 18.

Effect of C'lcl~ Pressure Ratio on Flow-Path Geometry.

"'"

~ _ ...,., - ~-. ... --- - -----

w ..

RELATIVELY LOW RADIUS RATIO RELATIVELY HIGH RADIUS RATIO FLOW-PATH DESIGN (SMALLER DISK, FLOW-PATH DESIGN (LARGER DISK, LARGER BLADES) SMALLER BLADES) (5)

12""U • e 71 mi. 0

T T 4 C2200 FTJlEC)

U • 640 mh 0

4 !21oo FTIlECI 11 -4 / (4) 10...J-'----

/ -

t

•

\ R/R • 0.670

RIR c 0.760 Z ::::. (3)

l

g

0~ :;)

E (2} 6

«

a:: (1) RSORE r.-;.1; ern (0.86 tN.)

(0)

o J

I I I .- I I I ~

1234 &87

G 1 2' ~ l .... L ....... '~t -- -1'

I I I i I

(0) (1.0) a.O) (0) <1.0) (2.0) LENGTH, em (IN.) LENGTH, em (IN.)

Figure 19. Characteristic Rotors for Evaluation of Concepts.

0.90 0.89 .... >' 4{ ZU 0.88 ;- ~w , O-In ,

!:l2 N 0.87

o~ ,

~ ~ tt 0.86

4{w~

b ~ 0.85

~~ en 0.84 , i , 5400 5800 6200 6800 7000 7400 7800 , , i (50,OOO) (55,0001 (80,0001 ,.,0001 (70,0001 (76,000) ROTA TlONAl SPEED' N, rl!1d/s (RPM)

Figure 20. Effect of Rotational Speed on Turbine

Performance and Geometry at 1644K (2500 F).

I

\

1.4 (1.2) (O.DI 3.0 Q 1.2 .D "" II:

~ 0

<:J (0.41 2.8

11.11 ~ i

1.0 u;-: ~-

- -

%~

~ B

II:

""

2.8

~ !

0.8

10.31 (1.01 g ~

>

II: ~ 2.4 ;

~

(D.l1 2.2 0.5 II: ~

0-

j:~ II: ~~ 0.4 1I:4( ~ ~

t)~

(jI) ~w ~ > OZ

~~

0.3 II::A 11:11:

""w

CDZ o~ .

~~ 1 ~- ;:,> z

~

0.2 w ~. 0.80 yZ ~ z"" ~~ w

t

II: ell: e we,) w t!> 0.&0 ~w II: yO ~ ~Ul ~> ee,) ~z

~

o!!!

0.40 ~e,) ~ 1.00 1.06 1.10 1.15 1.20 1.a 1.30 1.. 1.40 1.4& 'U50 "- w I (85) (10) (80) ABSOLUTE EXIT FLOW ANGLE AT ROTOR INLET, a. Adient (DEGREES) •

Figure 21. Effects of Stator Exit Angle on Stage and

Nozzle Design at 16441 (2S00 ¥), 7435 [ad/s

(71,000 RPM) and at 671 m/s (2200 Ft/Sec).

4.3.2 Turbine Off-Design Performance Characteristics The preliminary parametric study established the turbine per- formance, geometry, duct loss, and rotor-inlet relative- temperature characteristics for a range of cycle pressure ratios and rotor-inducer tip speeds. Since these results were for the maximum power point, the objective of the off-design analysis was to evaluate the feasi bili ty of maintaining constant per formance over a range of engine powers using constant speed and pressure ratio. In addition, the effects of tip speed and rotor exducer tip-to-inducer tip radius ratio were also '!valuated. Figure 22 shows the three cases selected for evaluation. The first case was based on the characteristic flow path with an inducer tip speed of 640 mID (2100 ft/sec), a radius ratio of 0.75, and peak system efficiency. This condition was equivalent to an optimized design point at maximum power.

The off-design study was based on aerodynamic effects alone; the effects of stator and rotor cooling and leakage were not included. However, these effects were taken into consideration in the detailed parametr ic analysis. Figure 23 presents the pre- dicted off-design characteristics for Case No. 1 from 50- tOl 100-percent engine power. The stage efficiency and rotor inlet relative temperature decreased uniformly from 100- to SO-percent power. However, when the effects of the interturbine duct loss were included in the stage efficiency, the system efficiency was fairly uniform down to 80-percent power. Below 60-percent power, however, duct loss significantly reduced system efficiency.

For the Case No.2, the rotor radius ratio was maintained at 0.75, but inducer-tip speed was increased to 67 m/se~ (2200 ft/sec), and rotor-exit swirl was increased from -0.31 to -0.53 radian (-17.8 to -30.4 degrees). Therefore, this configur- ation was equivalent to optimizing thE design point at a lower engine power point. The off-design characteristics for this case are presented in Figure 24. Although the turbine stage efficiency decreased uniformly from 100- to 60-percent power, system effic- iency was fairly flat from 100- to 70-percent power (±1.0 point).

For Case No.3, the effect of rotor inducer-to-exducer tip radius ratio was examined in relationship to Case No.2. Decreas- ing the radius ratio from 0.75 to 0.6431 produced the results shown in in Figure 25. For this case, a fairly uniform system efficiency was achieved (±1.2 pOints) between 60- and 90-percent power. A comparison of the I-dimensional vector diagrams at 60- t and 100-percent power is presented in Figure 26 for Case No.3.

The general trends resulting from the off-design analysis

i

are:

i

I

CASE NO. 2 - MAXIMUM POWER, NONOP1'IMUM EFFICIENCY, ROTOR EXIT SWIRL· -0.631 rad (-30.4°) RADIUS RATIO - 0.75

~

0.90 CASE NO. 1 - MAXIMUM POWER, PEAK z w EFFICIENCY, ROTOR EXIT SWIRL U • 0.30 red (-17 DEGREES), RADIUS RATIO

-

&L

:--.-8--- ..... 0.75 j &fa

&L 0.88 W

-- .!..~100

:IE i I w '" 'J i ~ «90.86 I > N

" "

(1)_ I

"

..J~

. I

~ I; 0.84

, CA3~ NO. 3 - MAXIMUM POWER, NONOPTIMUM EFfiCIENCY, ROTOR EXlT SWIRL - -0.70 red

~

(-39 9 DEGREES) :1ADIUS RATIO = 0.64

~

0.82 ~

e

0.80 ....... ..,..----r----r----r-- __ --_--_-

0.85 0.90 0.95 1.0 1.05 1.10 0.80 ROTOR INLET WORK COEFFICIENT, "'4 Selected Maximum Power Design Points for Off-Design Figure 22.

Analysis. 16:1 Cycle Pressure Ratio, T4 • 1644K (2500 F) and N • 6807 rad/s (65,000 RPM) •.

. '.'

• "" . .. ->Y ..... ,., •••• ,' _, ,. •• , ....... _. • ~~.' •• ' .... , ••• ~ ••••• ,.

'. .

. .

TOTAL·TO- TOTAL SYSTEM EFFICIENCY, '1T.T.2~ AND ROTOR EXIT ABSOLUTE CRITICAL INTERTURBINE DUCT tIQ STAGE EFFICIENC~'1T.T 2-6 MACH NO., VIAe,.

~. , TOTAL PRESSURE LOSS, fJ'1P .0 c;:

.., o

o o o o o CD o 0 CD CD o 0 0 CD CD CD CD

w w !='

o N ~ al

o 0

o o 0 :.. ...

•

0 N ..

N ~ • 0 N I I W I I I I f

-

•

, I: IIJOIO ::J~,""

/

0. '"" t-J I al

/

0\(J'I0 0 I,., ~~CD o~.

I

~~.

III .om ....

/

ZO '_::J / ~Q GIN to)

q

U'lto:1z ~

/I

-oc;:m

I ~

NO"'~!

.... 00"

I~

,.

O'l!l ~. /

~ O-::Jm

- ~

2In " CD 2r

•

•

~O'IO:,!

p ...

......... co::rm

*' rt ~

at

tnollJ~ ..

~ -..J'" Q ....

-t1~~8

. ~ ~

J I I I I I

o.fD

,.., ............

ON ....

en ~.

ROTOR EXIT HUB TANGENTIAL CD BLOCKAGE, B. PERCENT -~ (J'I ~.

U1 ,,) ~ - I 1 -r-- I , , I I I I I i iii r , Ii' iii " .

........ ~ -- ..

o cco~~~~:'" 0,""

i § & iii .. N N" ~ • c

~$:~i.1ie~ t

::0'" .!. ~ .1 1 r 1 I I

i~

J 1 I I

-. :;; :;; ~ :;; ;;J iii

I i ~ ~

" CD

~~g~~i~~ ON.. • • 0

........... _----

- - ~ ~

g: ROTOR INLET RELATIVE ROTOR EXIT ABSOLUTE STATOR EXIT VANE •

TEMP, T REL' KCoF) SWIRL ANGLE, (lEXIT' rad (DEGREES) ANGLE, Qv. ,. (DEGREES'

....

W \0

- .- .. -. .......... ---~ .......... ---.-,.-,."-

0.07 PRC·18:1 0.06 R/R·0.7&O 0.06 0.04 0.03

- t!)

~ 0.02 1.00 Q (50)

-

0.01 0.80 (40) 0.0 0.80 (30) 0.50 0.40 (2Ot (10) 0.20 0.40 (0) 0.0 (-10) 0.30 -0.20 (·20) -0.40 (-30) 0.20 (2240) 0.80 ui a:

~

a:

(2180) i

1480 ~-

~ ~~

(2140) S w:~

u.t u.t a: ... a: ...

u.t (2010) ...

!

a:

1400 e

o (2040) a: OJO~~----~----~--~----~----~~ 100 10 70 80

eo

ENGINE POWER, PERCENT

Off-Design Turbine Characteristics for Case No. 2

Figure 24.

at lS44K (2500 F), 6807 rad/s (65,000 RPM) and

671 roVs (2200 Ft/Sec).

0.86 PR 18:1 C • 0.80 ~ R/R • 0.84

0.65 ~

4(CJ

0.60 ~ In

~cr: 0.46 ~ 0.02 4(

0.40 14

0.01 (60) 0.7 (40) 0.8 (30) 0.6 0.60 0.4 (20) 0.3 0.2 (10) 0.40 0.1

o

(0) -0.1 (-10) -0.2 ViAe, 0.30 -0.3 (-20) -0.4 -0.5 (-30) -0.6 (-40) (2240) ,,; :I w t- w

>-

(2190) 5

t

w~ a: ~ w t- wa: (2140) it- a:

e

o

(2090) II: 60 70 80 ~ ENGINE POWER, PERCENT

Figure 25. Off-Design Turbine Characteristics for Case No. 3

at 1644K (2500 F) 6807 rad/s (65,000 RPM) and

671 m/s (2200 Ft/Sec\.

:; 1~ERCENT ~ER 6O-PERCENT POWER PIP T-T'2~ - 3.572 PIP T- T '2-6 .. 3.572 U ... 670 m/s (2200 FT/SEC) u • ... 670 m/s (2200 FT ISEC' T T c.N915 .. 0.203 kg/s (0.448 lB/SEC) wJQfS .. 0.331 kg/s (0.729 lB/SEC) ~/AND ~ 1.0 ~/AND = 0.60 W/~r • 0.220 Q - 1.3 red (72.0°, P • ·3.3 red (.18.SO, V/ACT .. 1.016 W/ACf .. (;,214 ViA CT = 0.207 f3 - ·1.11 rad (-63.So)

p. ".09 rMI (-62.50)

0' - .a.54 fad (.3O.9a,

I" lIu/ACT - 0.996 ~~

VIA CT '" 0.257 ViAe; 0.396

I ~WulACT = 0.438

l

I UIA .. 0.551-

CT

VJACf" .Q.u1 ~

~.- U/~CT - 0.567

r-- Wu/ACf - -0.795

I RSTG .. 0.43]

biG .. 0·&41

ACf'. = 730.0 m/s (2395.1 FT/SEC) Ac~5 = 644.3 rals (2113.8 FTISEC) R ... 9.8513 em (3.8785 IN .• R .. 6.2740 em (2.0764 IN.)

M5 Figure 26. Case No~ 3 Turbine I-Dimensional Vector Diagrams for 60- and 100-Percent Power at 1644K (2500 P), 6807 rad/s (65,000 RPM)

and R/R = 0.643 •

. 4 WI o Optimizing ~he turbine design pC'int at maximum power resulted in significant performance penalties at lower power levels for Case No.1.

o All cases showed that limiting the flow reduction to 60 percent with a constant speed, pressure ratio, and temperature concept is desirable.

o All cases snowed a reduction in rotor inlet celative temperature (higher inducer loadin~) in excess of 3llK (lOO°F) occur ted from 100- to 60-percent power at a con- stant turbine-inlet temperature.

o The change in rotor exit Bwirl is on the order of 1.05 radians (60 degrees) between maximum power and 60-percent power. [For example; Case No.3 was -0.54 to 0.51 radian (-31.0 to +29.0 degrees»).

o The change in rotor exit swirl dnd rotor inlet work coefficient is minimized between 100- and 60-percent power with lower radius ratio designs.

o Rotor reaction decreases as engine power decreases (see Figure 25).

o A relatively constant system efficiency (fl.l points) was maintained between 60- and 97.5-percent power for Case No.3.

4.4 Stator Concepts Aero/Mechanical Evaluation 4.4.1 Vane Profile Design for Stator Concepts A vane profile design for the stator concepts analysis was d~rived from the maximum-power l-dimensional vector diagram shown in Figure 26 and the vane meridional flow path shown in Figure 19.

The key stator design parameters used to optimize the vane loading were: o R2INLET· ll.50 cm (4.92 in.); o R3EXIT = 9.896 cm (3.896 in.);

o b • 0.798 cm (0.314 in.); I

o ~ • 18 vanes o Stator inlet flow angle • 0.0 radian (0.0 deg.); o Stator exit flow angle • 1.23 radisns (70.4 deg.)J

o Statl)r inlet critical velocity ratio • 0.1424 (with paral1til endwalls)J o Stator exit critical velocity ratio • 0.907.

The basic design procedure used for a r3dial stator is to define the profile in the axial plane (at a section corresponding to the nozzle trailing-edge radius) and thf!n to transform this section to the radial plane. This conformal transformation increased the geomE:tric throat dimension so that the radial- profile section burface had to be modified to reestablish the original throat dimension. The final a~ doil shape was then established by successive iterations using the transformed vane shape and blade-to-blade velocity distributions. The stator vane profile design derived for the aerodynam~c and mechanical stator concepts is shown in Figures 27 and 28. The design objective in both cases was to load the stator lea~~ng-Edge. This minimized the blade-to-blade pressure gradient in the trailing-edge region, sinc~ loading increases at reduced nozzle-area settings in this region. The resultant high loading at the st.ator leading-edge can then be reduced by en~wall contouring.

A screening study was then condu~ted to review the variable- area stator designs. The objectives ot this study were: To define variable-area nozzle methods that would allow o arbitrary stator endwalls; o To eliminate the need for upstream struts for structural support1 o To minimize stator leakage1 and o To allow realistic mechanical actuatiol'l.

For a constant section vane profile with parallel endwalls, the stator throat area is: At • Ny.a.b where~ Ny g Number of vanes; a • Throat dimension; b • Vane height.

Twelve possible stator concepts were defined using the three yariables that affected the total throat area. An aerodynamic and mechanical screening pzocess then reduced this number to four.

___ 't:' _______ ~ ..... ,...,... •• - ... - - ""'~""~ .. ' ._.., - _. ~~ .. - ~ 12.0 (C.7) 11.& (C.&) 11.0

-

~ (c.3)

I

10.1 ~ ~(c.ll \ , to.O J"t' , ~ THROAT , ~(3") ,

§

1.& > (3.7) 1.0 (3.6) 1.6 -

--I.h

rr-------~'~------~,~------~,~------·~f~------~r~------~i~------~I~------~.

(3.3} 0.0 1.0 2.0 3.0 C.O &.0 1.0 7.0 8.0 r- __ -r ____ r- __ -r ____ r- __ -r ____ r- __ -r __ ~~--~i~--~--~~--~ i i. • (0.0' (1.0' (2.0' (3.0) X COORDINATE. em (IN.'

Figure 27. Stator Profil~ Design for Variable Area Concepts Evaluation (19 Vanes) •

..

U'I ~L. EB -

Figure 28. Stator Ring at Maximum Power Setting.

' 4.4.2 Full-Vane Rotation Concept The full-vane rotation concept is considered the conventional approach, an~ has been successfully used in both axial turbines and compressors. For this particular application, a selected num- ber of vanes would remain fixed for structural support of the vane sidewalls as shown in Figure 29. The major advantage of this approach is that it provides a single-pie~e, easily cooled vane with solid, stationary sidewalls.

The vane suction- and pressure-side sucface velocity distri- butions for 60- and laO-percent power are also shown in Figure 29.

While these velocity distributions do not take into account the effect of maintaining a portion of the vanes fixed, the extremes in vane-to-vane velocity variation would be expected to occur between 60- and lOa-percent power. The aerodynamic and mechanical effects associated with the higher circumferential velocity gradients were diff icult to assess. Previous experience wi th inlet scrolls has indicated that if significant turbine circum- ferential maldistribution is generated by the scrolls, overall stage performance would be decreased more than would be expected from scroll pressure loss alone. Mechanically, circumferential maldistribution could result in inducer blade vibration.

4.4.3 Articulated Trailing-Edge Concept This concept was based on rotating the vane trailing-edge only. This allowed the vane leading-edge section to remain fixed for structural support as shown in Figure 30. The vane inlet sidewall could then be contoured to minimi~e vane loading. The sidewall leakage should be significantly less than that of the full-vane rotation concept.

Velo~ity distributions at 60- and lOa-percent power indicate that the major drawback to this configuration is the high trailing-edge loading that results from rotating the vane to 60- percent power. The high static-pressure gradient in this region increases vane leakage unless an effective sidewall seal is pro- vided. Mechanically, the smaller size of the articulated trailing-edge could complicate routing of internal cooling airflow to this region.

4.4.4 Insertable Minivane Concept In the insertable minivane concept stator flow area is reduced by inserting minivanes between fixed primary vanes. This results in stationary sidewalls and single-piece vanes that can be readily cooled. Thf! inscrtable minivane concept is compatible with contoured sidewalls, and provides support for the sidewalls, and should minimize leakage floWS. Two methods of insertion were

T-

12.54 (4.936) 0.797 (0.314) 9.87 (3.808) 1.20 G ~ 1.00

>

.

0 0.80

z

:I: 0.60 CJ

<

~ ..J 0.40

<

CJ

t=

0.20

-

a: CJ , , 0.0 60.0·70.0 80.0 90.0 0.0 10.0 20.0 30.0 40.0 50.0 RADIAL CHORD, PERCENT Full Vane Rotation, Nozzle Area Variation Concept.

Figure 29.

•

!

!

~ ..

I' ~ ~

r

f

f

If !

...

~ t (' t ~' }

"

" t

12.54

f

(4.936) t

_.-c

..

r

0.80 ~ (0.314) , 1.20 ~

~b 1.00

., ~

>

t: O· 0.80

.-_.- .'/

z MAXIMUM ~~I z 0.60 FLOW," (.)

<

;,- I

~

.... 0.40 6O-PERCENT'· I

FLOW i"

< (.)

i= 0.20 a: (.)

0.0 0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 'RADIAL CHORD, PERCENT Figure 30. Articulated Trailing-Edge Nozzle Area Variation Concept.

i

i

considered feasible. In the first method, the immersion depth of all minivanes would be varied in unison. In the second method, a selected number of minivanes would be inserted along the full vane width. Sequencinq the number of fully immersed mini vanes results in a step change in turbine flow. This concept is presented in Figure 31.

The minivane design depends on maintaining a convergent pas- sage on either side of the mini vane. This results in a rather large trailing-edge thickness that significantly increases vane- exit blockage. Mechanically, the periodic vane trailing-edge wakes would vary from 0.31 to 0.63 rad i an (18 to 36 degrees), which could create serious rotor blade vibration problems. In addition, mechanical linkage would be complicated if an incre- mental insertion were used.

4.4.5 Locally Movable Sidewall Concept This concept reduces the stator throat area via a local change in stator passage width as shown in Figure 32. A passage constriction is readily achieved by axial trenslation of a portion of the sidewall. The most favorable vane loading -- from 60- to lOO-percent power -- results with this method. This design is compatible with contoured sidewalls and has many of the advantages of the inserted-minivanes concept--without the disadvantage of forced periodicity at the stator-exit flow. However, previous cold-air test data show that stator"pas~age width reduction results in relatively high dump losses due to the sudden expansion in passage width at the rotor inlet.

Subsequent evaluation of this concept showed that alternate methods could be used to restrict the passage width and thus eliminate the high dump losses. Two alternate methods were defined: a purely rotating sidewall segment and a rotating- translating sidewall segment. In each method, the sudden expan- sion at the stator exit was replaced with a smooth ramp. However, both methods would require more complex actuation mechanisms.

4.5 Stator Material Evaluation 4.5.1 Airfoil The material selection for the cooled stator airfoil for the engine was based on the following parameters: o Stress-rupture capabilitY1 o Coated-oxidation resistance1 o Overtemperature capabilitY1 and o Thermal-fatigue resistance.

I

10.634 (4.187) 12.637 (4.938) I , 0.798 ~3~)-+--~ (0.314)

~~

1.20 t; ~ 1.00

>

.

ci 0.80 z % (.)

0.60 c( :I ..J 0.40 t= 0.20 a: (.)

0.0

o 10.0 20.0 30.0 40.0 60.0 60.0 70.0 80.0 90.0 100.0

RADIAL CHORD, PERCENT ,

'

, , .

Figure 31. lnsertable Minivane, Nozzle Area Variation concept.

t

~

i

•

"

9.896

".884

(3.896) (4.60) 0.478 (MIN.)

(0.188) -+---+-- 0.798 (MAX.l

\

9.896 (0.314) (3.896) 1.20 ~ u ~ 1.00

>

.

0.80

z

~ (.)

().60 ex: ~ .oJ 0.40 ex: (.)

...

0.20 a: (.)

100.0 80.0 20.0 40.0 60.0 0.0 RADIAL CHORD. PERCENT Locally Movable Sidewall, Nozzle Area Figure 32.

Variation concept.

· .............

~ /

i

t .- " The following material candidates were considered for the stator airfoils: Equiaxed Mar-M 247 0 DS Mar-M 247 0 MA6000E SC NASAIR 100 4.5.1.1 Stress R~~l1.!!.

Stress-rupture capabilities of the cand idate alloys (17 ,18) are shown in Figures 33 ~nd 34. The temperatures for 1000- and

t

4000-hour rupture lives at 68. 948-MPa (lO-ksi) stress leve Is are

t

shown in Table II. Table II also indicates that, with the pos- sible exception of equiaxed Mar-M 247, all materials considered

i

were viable candidates for the proposed vane application. It ~ ~ should also be noted that use of a thermal-barrier coating on the \.

stationary airfoil either would result in improved durability, or ~: ~. would allow for higher gas temperatures.

i<- t to;' ':t,., ; TABLE II. STRESS-RUPTURE CA~ABILITIES OF :~" CANDIDATE TURBINE SUPERAT,LOYS , i h· RUPTURE LIFE '.

10CO-Hr 4000-Hr 69.948 MPa (10 Ksi) TEMPERATURE K ~ 1394 1358 SC NASAIR 100 (1985 ) (2049) ....

~' l' 1324 1290 DS Mar-M 247 (1862) (1923 ) 1290 1257 Equiaxed Mar-M 247 (1803 ) (1862) MA6000E (2200) (2200) 4.5.1.2 Coated Oxidation Resistance The potential of oxidation resistance is a critical aspect for this application. Current coating systems have coating lives on the order of about 250 hours per 0.0254 mm (0.001 inch) of coating

J

,

J

~. ' • • If ',' •• ,.~, '.' • (100) (90) 821 (SO) (70) 483 (60) 414

en

~ (50)

-

III Q" (40) 276 :I r .

~

(30) 207 LLI a:

Iii

EQUIAXED (20) 138 MAR-M 247 LARSON-MILLER PARAMETER Figure 33. Larson-Miller Parameter Showing Stress Life Capabilities of Candidate Superalloys.

S4

f \

, o 1000-HOUR RUPTURE LIFE AT 138 MPa (20 KSI)

, .

.

~ 4000-HOUR RUPTUfiE LIFE AT 138 MP. (20 KSI)

(1800)

- ~

CL ~ 1225 >.

5 (1700)

ii

<

~ <J w ~ (1800) t-

<

a:: w Go ~ w t- (1500) ASTROLOV EQUIAXEO OS MAR·M 247 MAR·M 247

f

Figure 34. StresS-Rupture Capabilities of Candidate Turbine Superalloys.

ss

..

thickness at 1366K (2000 F) (19), the projected vane hot-spot temp- erature. However, by 1987, it is expected that the lives of pro- tective coatings will be increased by a factor of four. Conse- quently, the life of a 0.102-mm (0.004-in.) coating should be about 4000 hours.

4.5.1.3 Overtemperature Capabilit~ • The capability of the stator assembly to withstand br ief overtemperature conditions is dependent on the incipient melting temperature of the superalloy. Table 111 indicates that incipient melting points of the candidate superalloys are in the range of 1505 to l603K (2250 to 2425°F). Thus, all alloy candidates have overtemperature capabilities well above the proposed design requirements.

III. OVERTEMPERATURE CAPABILITIES OF

I TABLE

CANDIDATE STATOR MATtRIALS Incipient Melting Alloy Temperature~ K {OFi Mar-M 247 1505 (2250 ) NASAIR 100 1603 (2425 ) MA6000E 1569 123GSl 4.5.1.4 Thermal Fat~ Studies of thermal-fatigue cracking of DS superalloys have indicated that resistance to thermal-fatigue crack initiation and propagation is increased as the elastic modulus is decreased (20 ,21) • Elastic moduli of the ~andidate alloys are listed in Table IV. This table indicates that OS Mar-M 247 and SC NASAIR 100, with the ccyp.tallographic axis or iented parallel to the vane axis, are preferable from an elastic modulus standpoint.

- ..

TABLE IV. ELASTIC MODULI (R OOM TEMPERATURE) !

~ OF CANDIDATE SUPE RALLOYS ;; , ~~ Modulus " ~.~ Allo~ kPa Dsi " Equiaxed Mar-M 247 212.36 x 10 (30.8 x 10 )

.'

}, [ 6 6 OS Mar-M 247 (Longitudinal) 10 (21.0 x 10 ) 144. 7~ x ~ se NASAIR 100 6 6 (Longitudinal [001]) (21.0 x 10 ) 144.79 x 10

i'

t 6 6

MA6000E (Longitudinal) (29.4 x 10 ) 202.71 x 10 se Mar-M 200 [001] * 124.11 x 10 (18 x 10 )

I

I 6 6 (32 \ se Mar-M 200 [110]* 220.63 x 10 x 10 ) 6 6 ~ SC Mar-M 200 [111]* 296.47 x 10 (43 x 10 ) i~ '"

f

,r " *Kear and Piearcy - Pratt , whitney Aircraft, 1967 ~: i' ,,.

t~ t~ !~ Although HA6000E has a high modulus parallel to the ~ extrusion/rolling direction, it should be noted that initial ~.

thermal fatigue test results were comparable to OS and ( SC Mar-M 200 data (22). Based on tl,ese results, OS Mar-M 247, se NASAIR 100, and HAGOOOE are all considered ~1able candidates for vane application. Oxidation resistance of the coating is expected to be life-limiting.

4.5.2 Sidewall Since the sidewall will have biaxial stresses, the preferred material for this application is cast equiaxed Mar-M 247. Hot- spot conditions of l366K (2000 F) indicate that oxidation resis- tance may be life-limiting for this applic3tion. It should also be noted that the impingement-cooled sidewall design may be com- plemented wi th the use of a thermal-barrier coating to reduce metal temperature and increase component life.

4.6 Recommended Vaciable Stator Concepts After considecing all aspects of each design, the full-vane rotAtion and minivane concepts were eliminated. The articulated , IiIIT •

" ."

trailing-edge design was slightly more favQred than the movable- sidewall design and waR ultimately selected as the primary design candidate. However, in the March 27, 1980 Task I Review at NASA it was recommende" that the local movable sidewall concept be retained for further investigation. This recommendation led to an add-on contract to the detailed stator substantiation phase for examination as an alternate candidate.

The fabrication method selected for the stator concepts con- sists of laminated v~nes inserted and brazed into separate side- walls. Depending on future engine design analysis, the sidewalls could be either continuous rings or two vane segments. The lami- nated vanes could be constructed from either equiaxed Mar-M 247

material or from os Mar-M 247 (if the lamindtes can be oriented

perpendicular t~ the flow direction to increase vane strength).

If high-temperature integrity beyond Mar-M 247 capabilities is

necessary, oxide dispersion-strengthened alloys (such as os

nickel) are alternllte candidates. The sidewalls would be cast from equiaxed Mar-M 247. If the final ytator-design cooling configurations were simple, conventional vane casting could be used for the articulated trailing-edge vane pieces as an alternate to laminates. An additional option is the use of thermal-barrier coatings, particularly in the sidewalls (where allowances can be made for coating thickness).

The turbine stator detailed mechanical-design substantiation was accomplished by ensuring that acceptable temperature levels and gradients were achievec It should be noted that life pre- diction requiring 3-dimens onal stress anelysis, LCF calcula- tions, llnd creep-deformation predictions were considered beyond the scope of the existing program. Preliminary design cooling configurations were established for the articulated trailing-edge vane, the movable sidewall vane, and the sidewalls. This included predicting metal temperatures and cooliny flow-rate requirements at 60- and 100-percent engine power. All work was accomplish~d using a rotor-inlet temperature of l644K (2500·F), which was later reduced for the final design. The cooling configurations for the vane and band were similar to that selected for the final deSign, and Are discussed in Section 6. The final design demonstrat~~ the feasibility of achieving satisfactory temperature distributions in a variable-area stator. However, cooling-flow usage at 60-percent power was quite high in the final design.

4.7 Turbine Rotor Materials Evaluation The principal requirements for the radial-turbine materials system are:

o High tensile strength in the hubJ o Good biaxial-creep strength and thermal-fatigue strength in the rimJ and o High creep strength in the inducer.

To meet these requirements a dual-alloy rotor is recommended.

A cooled, laminated radial turbine developed by Garrett suc- cessfully demonstrated the use of a laminated bonding technique.

This technique was selected for development of the dual-alloy rotor blade ring. The rotor blade ring will be constructed from

~amtr;tedp!~~-: 2!~t:rr~jr w~~~ O~i~~~~~er ~i~a:_ete~~~~_~~~;~g:~

material) will be used for the rotor disk. Candidate materials considered for the cooled rotor concept relative to an Astroloy baseline are shown in Figure 34.

The viability of this dual-alloy radial turbine approach was demonstrated by both Garrett and another engine manufacturer(24).

A solid, dual-alloy ~Mar-M 247/Astroloy) radial turbine rotor with OS blade tips is also being developed at Garrett with NASI·lOO!:: sponsorship under Contract No. DEN 3-167. Figure 35 shows the tensile properties of the dual-alloy rotor hub materials.

4.8 Cooled Rotor Concept Aero/Mechanical Evaluation Selection of a rotor concept was based primarily on mechan- ical considerations, since the goal was to maximize allowable tip speed while satisfying the required duty-cycle life. In addition, the magnitude and location of rotor :nternal-cooling flows would be similar for either a cast or laminated cooled rotor. No unique aerodynamic advantage was identified from the rotor concepts eval- uate-d.

with uncooled blading previous Garrett experience in radial- t'Jrbine rotor design has shown that a practical limit of about l422K (2l00·F) exists for turbine rotor inlet temperatures. This limit reflected the use of m~terials with near-term applicability ; to tne design. ~otor tip spe~ds beyond the current state-of-the- !.

art, and life requirements consistent with program goals (1589 to l644K (2400· to 2500 f) total inlet temperature) require subst~~ tial cooling flows and high thermal performance internal COOli&lg geometry. Radial turbine configurations with no provision for ..

inserted blades require either an integrally cast or laminated design. Pr ior Garrett 4txper ience wi th integrally cast turbines with cooled blading has been discouraging, dlJe to the cooling com- plexity and casting yield. However, recent imp{ovements in

(240) AF 115 l C .. , Ul rIMA TE (220) (200) ASTROLOY, ULTIMATE

- 1300

en :=.=: (180) ca 0- ~ AF 115 LC, YIELD (0.2 PERCENT)

~

(160) 1100 w a:: ...

en

~

(140) ASTROLOY, YIELD (0.2 PERCENT)

-

~ (120) AF 115 LC, ELONGATION 'tit 20 TROLOY ELONGATION .

z

~

Z ...I W 300 400 500 600 700 800 900 1000 I I I I I 0 (200) (400) (800) (800) (1000) (1200) (1400) TEMPERATURE, K (qF) Figure 35. 'l'ensile Properties of Hub Materials, Dual-Alloy Turbine Rotor.

. "".- -- ..

------

casting technology have been promising in regards to the feasi- bility of such an approach. Laminated sheet technology has also been encouraging. Garrett, in cooperation with the U.S. Army, has designed a radial turbine that meets similar, but less ambitious goals.

With the increase in turbine inlet temperature experienced with conventionally cooled radial-turbine designs, the stress- rupture life in the inducer portion becomes a primary design restriction. The ability to reduce metal temperatures below cur- • rently attained values with added cooling flow is limited. There- fore, material property improvement is necessary. For a typical radial-turbine inducer stress, Figure 36 compares the potential metal-temperature increase (over an Astroloy baseline) for several alloys at a 4000-hour mission life. The temperature increase afforded by the selection of DS Mar-M 247 was deemed a realistic goal for the 1988 timeframe. However, both the cast or laminated approach would require a moderately aggressive manufacturing tech- nology program of approximately equivalent effort to permit pro- duction within that timeframe.

An effort was made to compare the cooling-system complexity achieved with laminates to a cored casting. A typical laminated design with a high level of complexity is presented in Figure 37.

This figure shows the USARTL turbine design previously completed at Garrett. Discussions were held with a casting vendor on the feasibili ty of achieving s1.: i lar geometL"y features in a radial turbine by 1988. It is believed that time and expense, in con- junction with advanced casting technology such as the use of quartz-rod cores, could produce the radial design shown in Figure 38. The dimensions indicated are believed feasible, and would be also technically acceptable for a high-thermal- performance design. It is believed that thp.re would be no dis- cernible difference between thp. cooling effectiveness achievable with the laminated approach using current techniques and the cast blade ring yet to be developed. However, both approaches will necessitate improvements in material technology to reach the pro- gram total inlet temperature goals.

I

One possible method to produce laminates with the high rup- j ture strength of a DS Mar-M 247 casting is to simply slice sheets

I

from a bar of material with directional grains in the outermost J region. This would provide tensile and fatigue properties in the disk bore region equivalent to those of an equiaxed casting. It is proposed that both these approaches (laminated and cored cast- ~ng) be used to produce a rotor blade rir.g of material containing cooling passages. Two potential candidates for the powder-metal hub design are powdered Astroloy and AF 115. The powder metal hub

..

(300) LEVEL 3

•

• FULLY LAMINATED AD\fANCf;D ROTOR ..J < ~- ..

WL&.

~«!..

a: ~ (250) o , ~W Oz a:- ..J WW ..J(I) alct

«aI

~O (200) o po.

..J > ..Jw

<> -

....

..J

-

z ....

LEVEL 2 «

-«

w..J c INDIVIDUAL BLADES AND w ~ ~ (150)

• (,)

POWDER-METAL DISK Ww z a: a:

«

(,):::l lEVEL 1 > c ~po.

BLADE RING AND

«

.... ea: ..

POWDER-METAL

«ffi

DISK ~ ~ (100 WW ...

~po. -

a: ~ ~ Ci N en

«

z ~ c" (50) ~ W~ ~

«

X ~ «:E -I ct en :::la:

,.....0-

Q p- Oct w~ N L&.

ea: (0) 0 CANDIDATE MATERIALS Figure 36. Comparison of 4000-Hour Mission ~ife and Density- Corrected Stresses at Inducer Tip at 640 mls (2100 Ft/Sec).

0. 5% 0.5% We BLADES ONLY = 4. 94% Wgg /INDUCER 1. 06% ~EXDUCER We TOTAL :: 5. 84% Wgg 0.88% 0. 12% 1 74 C1l 0.14% . III o T BULK", 992K (1325 F) METAL ALL FLOWS SHOWN IN PERCENT OF COMPRESSOR FLOW 4.94% Figure 37. USARTL Cooled. Laminated Rotor Design.

•

•

•

•

NOTES:

APPROXIMATE SCALE = 2X

PASSAGE THICKNESS ~ 0.064 - 0.076 CM (0.025 - 0.030 IN.)

HOLE SIZES: 1 DUST HOLES 0.038 - 0.051 CM (0.015 - 0.020 IN.) DIA.

2 DUST HOLES 0.076 (0.030 IN.) OIA.

3 LEADING-EDGE DISCHARGE 0.064 - 0.076 CM (0.025 - 0.030 IN.) DIA .

4 IMPINGEMENTS HOLES "" O. 38 - 0.051 CM (0.015 - 0.020 IN .) DIA .

5 INDUCER DISCHARGE SLOT 0.07 x 0.254 CM (0.030 x 0 .1 00 IN.)

6 TIP DIS HARG E ~.07 6 x 0.406 CM (0.030 x 0.100 IN .) SLOT 7 TIP DISCHAR GE "" 0.076 x 0.102 CM (0.030 x 0.040 IN.) SLO TS 8 COR E SUPPORT SLOT = V RIA BLE WIDTH PIN FINS: 0.063 eM (0.025 IN.) DIA ; MI NIMUM S PA CING ~ 0.152 CM (0.060 I .)

Figure 38. ~re limin a ry oole d Rad i al Rotor Design.

would then be bonded to the blade ring using a hot-isostatic- pressing (HIP) process. A powder-metal hub with ~niformly fine- grain structure would produce higher burst-margin capabilities and fatigue strength in high-stress regions.

4.9 Recommended Cooled Rotor Concept The recommended cooled rotor concept is a dual-alloy lami- nated rotor. The rotor blade ring will be constructed from lami- nated Mar-M 247 sheets wi th os inducer blade tips. An AF l1S powder metal (or other similar high-tensile-strength material) would be utilized for the rotor disk. Comparison between lami- nated and advanced cast os Mar-M 247 blade rings showed approxi- mately equal cooling-flow capabilities, development cost, and '" risk. Therefore, a clear choice between these two methods is not " ," possible at this time. The laminated method was selected on the basis of cost effectiveness for the follow-en roto[ test program.

The potential rotor-inlet temperature achievable using os Mar-M 247 os blades was evaluated using the internal cooling-flow geometry from the existing USARTL laminated rotor and the mission life defined for the cooled variable-area radial turbine. Table V shows that mission life decreases rapidly between 1533 to l644K (2300 to 2500 F) rotor-inlet temperatures. On the basis of pos- sible improvements to the USARTL inducer cooling-flow scheme (e.g., film cooling in the critical region), a maximum rotor-inlet temperature of l589K (2400 F) is recommended to achieve the 4000-hour mission life goal. The recommended cooled rotor concept I' is presented in Figure 39.

I

l ~ ~~

•

J

'.

!

~ :' ~ I , "}

f

t

i

.i t , ~ , ,

<" <" TABLE V. EVALUATION OP COOLBD RO'l'OR LIPE AS A PUNC'1'IOM OP ROTOR-INLE'l' '1'DlPERATUJUI:, DUTY CYCLE, AND TIP SPEED.

Uu '1'4 100-Pe~cent- Ratio of

_'II

It Powe~ Life Misllion Life Mill8ion LUe' .!!L.

Configllration tt.terial (-P) (hra) aec (hu) a.lleline LUe Ballellnel USOUtTL Cooled Aatroloy 1513 Inl 573 8455 1.0 ta.inaled Radial Tu~blne (~300) (18110) DS Mar-M 247 15H 29,240 USAl<'l'L 573 146,000 17.0 (2300) (1880) Variable-Area Realal 1513 4470 573 16 ,300 2.0 'l'u~bine (2300) (1880) Variable-Area Radial 1533 640 1360 4910 0.6 'l'urblM (2300) (2100) Va,lebl.-Area Raaial 1589· 640 315 1110 O.ll Tu~bi_ (2400) (2100) Va~l.b1e-Are3 Raaial 16U 640 80 290 0.03 1urbine (2500) (210(1) o POI' eonatant life, 'l'~tal DS ~ar-M 247 • 'l'ftOtal Aatroloy + 6l.81t (llS-P) o Pol' 1611 Cooled, Variable-Area Radlol Turbine, life at lOO-percent power. 0.28 • _llIlIlon life • ~~a.aended aa.i.ua ~otor-Inlet te_perature for cooted rotor

• COOLED LAMINATED MAR-M 247 BLADE RING ------l------~L~ WITH DIRECTIONALLY DIRECTIONAL SOLIDIFIED INDUCER BLADE TIPS.

SOLIDIFICATION REGION

__ ~ __ l ____ _

Figure 39. Recommended Cooled Rotor Concept.

5.0 TASK II - DETAILED PARAMETRIC STUDY 5.1 Selected Speed and Cycle Conditions The objective of the detailed parametric study was the selec- tion of the turbine stage configuration for the detailed design.

The study consisted of examination of the effects of stage cooling flows, stator leakage, and rotor/exducer blockage on stage eff iciency from 60- to 100-percent power for both cooled and uncooled turbine configurations.

• The results of the Garrett Small Axial/Centrifugal Compressor Design Study (conducted for the NASA-Lewis Research Center under Contract No. NAS3-21621) were used to evaluate turbine/compressor rotational speed compatibility. For a 2.27 kg/s (5 lb/sec), 2-stage centrifugal compressor, the selected pressure ratio in the compressor study was 17:1 with a rotational speed of 5601 rad/s (53,486 rpm). The rotational speed selected for the turbine pre- liminary parametric study was 6964 rad/s (66,500 rpm) with a pres- sure ratio of 17 :1. This speed was selected on the basis of a minimwn specific speed of 0.468 (60.0) to maintain peak attainable efficiency. However, examination of the 17:1, two-stage, centrifugal-compressor efficiency characteristics showed a rapid decrease in performance as speed increased. It was determined that turbine specific speeds below 0.468 (60.0) were required to achieve optimum turbine/compressor rotational speed. The compres- sor and turbine efficiency characteristics as a funct'on of rota- tional speed are preser.ted in Figure 40. As shown in Figure 41, the efficiency product is relatively flat between ~550 rad/s (53,000 rpm) and 5969 rad/s (57,000 rpm), as shown in Figure 41.

Therefore, a design rotational speed of 5969 rad/s (57,000 rpm) was selected for the detailed parametric study. This result, in conjunction with the preliminary parametric study and the stator and rotor concepts ~valuation, resulted in the following cycle and turbine conditions for th~ detailed parametric study: o Cycle pressure ratio: 17:1; o Rotational speed: 5969 rad/s (57,000 rpm). (Although peak compressor/turbine efficiel1cy occurred at 5760 rad/s (55,000 rpm), increasing the rotational speed to 5969 rad/s (57,000 rpm) had relatively little effect on efficiency; o Rotor inducer tip speed: 640 m/s (2100 ft/sec); o Maximum rotor-inlet temperature; Cooled rotor = l589K (2400 F); Uncooled rotor = l478K (2200 F);

....... .... 0: ...... ,.".,-;r ,.- ........... ' ...... ""',. :-o-t""....-"""~ "'...-_.v .. ~.:.~~ ........ ' .

. ' .....

-RADIAL TURBINE 0.89 0.88 0.87 > u 0.86 z ~'

- w • 2.27 kg/I (6.0 LBS/SEC)

~ 0.85 P • 17:1 RC ~ W ...J 0.84

g

2-STAGE CENTRIFUGAL 0.83 / COMPRESSOR CFIXED GEOMETRYI ~ ...J 0.82

-

~

-

.......

--- -

2-sT AGE COMPRESSOR

--

"""-'-

--

(WITH VARIABLE GEOMETRY) 0.81

--e

I

"'

0.80

"

"

J

n79 • --

~. ~ 4200 4600 5000 5400 6800 6200 - 6600 I I I I I .~-.

-r

(40) (46) (50) (66) (60)(66) (36) ROTATIONAL SPEED. N. radls (RPM X 10-3, Figure 40. Cooled, Varia~le-Area k~dial Turbine Compressor Speed Study.

0\ ~

W • 2.27 kg/I (5.0 LR/SEC)

•

P • 17:1 R C 0.71 ....

..... _- RPM M T ~EAK I ..

PERFORMANCE s:- )( u 71 s:- ~ELECTED TURBINE

t.: 0.89

....-- ROTATIONAL SPEED • (J :J 6989 radII (57,000 RPM) C o· a: A.

>

~ 0.87 w • ~ u.

u.

w

O.H ~~----~------~------,--------r------~

8000 6400 8800 4800 5200 I I (50) (55) (80) (65) (45) ROTATIONAL SPEED, radII (RPM X 10-3) Figure 41. Compressor Turbine Efficiency Product Charac~eri3tics.

, o Variable-area stator: Articulated trailing-edge and locally movable sidewall configurations; o Cooled rotor: Laminated dual-alloy configuIation; o Uncooled rotor: Dual-alloy configuration.

5.2 Effects of Stage Cooling Flows A detailed schematic of the stage cooling flow circuits for the cooled variable-area radial turbine is presented in Figure 42 • • Since the effects of internal rotor-cooling flow on stage perform- ance were not available, the cooling-flow model relied heavily on the resultG available from existing axial turbine concepts and from correlation with the tested effects of p.xternal radial rotor- cooling flow results. The situation with th~ tltator was similar, except test results for trailing-edge discharge flow were avail- able from the Model GT601 radial gas generator turbine.

At Garrett, the accounting procedure for cooling flows is to asslJme all stator pr imary and cooling flows are available to do work in the rotor, and the available rotor-inlet temperature is based on a primary and cooling-flow mixed value. Therefore, the l589K (2400 F) inlet temperature established for the cooled, variable-area radial turbine is defined as a rotor inlet absolute mixed temperature (T4, Mix) • Conversely, rotor -cooling flow is assumed to do no w~rk 1n the turbine and thus bypasses that stage.

However, the rotor-cooling flow is availablp. to do work in down- • stream stages. Thus, the cooled turbine efficiency defined in this manner is a hybrid between truly primary and thermodynamic efficiency. That is, the available energy of the stator cooling flow is accounted for in the isentropic expansion, but the rotor- cooling flow work is added or subtracted from the actual work of the r('tor-inlet flow. On this basis, the expression for the cooled turbine efficiency becomes: • W ~H4,Mix + ~Wi~Hi 4,Mix (1) ~T-T Cooled (W 6H ,Mix)isen 4,Mix 4 where:

= W 6Hp + W ,EW "v,EW + WVI~"vl

V p W ,Mix

~ N ARTICULATED TRAILlNG- Wp / EDGE VANE DETAILED COOLING FLOW CIRCUITS

-- '- z.::-~-7 WY)'S

WYEW--fL.JJ!l W ~~

, .. V,EW W ~ -,__ V,SS

WBFS -- -11 '.n ROTOR INLET MIXED STATION

WV,TE

fiI ,"""",a W4 - WP+WVI+WVEW

T ,MIX • 1589K • WR1,EXT MAXIMUM POWER ~R' (24O(JOF) \\\.\ I - , COOLING FLOW (%) ~~J Fa lEADING-EDGE\ ~ IMPINGMENT =- 2.5 WV,EW FOLLOWED BY FILM COOl- - 1.9 Wv,ss WRBF~-- _ >~ING WV,PS - 2.0 WRI--- (e: WRI - 1.5 WV,TE I WRB .. __ ..... --_-=.-_-_-_.....i!l ROTOR BLADE DETAILED

- n.5

WefS

L ~LING FLOW CIRCUITS i _ _

WRS - 1.0 WRBF • 0.8 GAGE COOLING FLOW =- 1.5 SCHEMATIC WRI,IND • 4.0 WRI,EXT W • 0.3 RS Figure 42. Turbine Cooling Flow Model for the Detailed Parametric Study •

•

•

A cooling-flow model was then established based on the following component evaluation: WV,EW - Vane si~ewall impingement and film-cooling flow o was estimated at 2.S percent for lOa-percent power, a~d 4.1 percE:nt for 60-percent power. Al though in-house test results f(om radial nozzle inlet veil cooling indi- cated no penalty, the location of the sidewall film- cooling discharge holes indicated the effect would be similar to vane leading-edge film-cooling flow. NASA test results indicated that fllm-cooling in the vane leading-edge region can be as high as 80-percent effec- tive relative to the ideal kinetic energy at the vane exi t. Therefore, vane sidewall-cooling flow was con- sidered 80-percent effective.

o Wv,SS' Wv,PS - Vane suction-side and pressure-side surface discharge cooling flows were considered to be 80-percent effective. (This was based on the discussion in the preceding paragraph.) The t~tal cooling flow was estimated at 3.9 percent at laO-percent power, and 5.1 percent at 60-percent power.

Wv TE - Vane trailing-edge cooling flow was considered o l06-percent effective (based on in-house test results).

WBFS, WRS - Rotor backface seal and shroud cooling flows o enter the mainstream flow in the reg ion of the rotor inducer, and are, on the average. 55-porcent effective.

The estimated cooling flow is 0.60 percent for the rotor backface seal, and 1.0 percent for the rotor shroud.

Since a preliminary heat-transfer analysis was not per- formed for these regions, the estimated rates were assumed valid for all power settings.

WRBF - Rotor backface cooling flow prevents hot gas from o recirculating on the back side of the rotor disk and enters the mainstream flow in the scallop saddle region.

The detailed analysis of the Army l~inated rotor showed that 0.6-percent cooling flow was required for this reg ion. Test results conducted on the Model GTP30S-2 radial rotor (7) indicated that, without preswirl, the pumping work required to achieve the scallop-saddle roLor speed was offset by the cooling flow expand ing through the remaining portion of the rotor flow path.

Since the cycle had bypassed this cooling flow and the

, .... ..,., .. - ...... -..---, .. -~. _ ..... _ ...

pumping p~nalty was offset, no additional penalties were incurred.

WRI IND - without ~reswirl, the rotor cooling flow in o the' inducer ragion is pumped to inducer tip speed.

Except for location, this is also true for both the rotor backface flow (which m~st be pumped to scallop wheel speed) and rotor shroud and backface seal cooling flow (which enters the rotor in the inducer reg ion) • However, in terms of main:atream flow disturbance, the rotor inducer cooling flow was considered similar to the cooling flow entering the scallop region. Therefore, the rotor inducer cooling-flow pumping penalty was con- sidered offset by expansion of the cooling flow through the rotor after discharging in the inducer region. F~r thermore, since the cycle already bypassed this flow, no additional penalty was incurred.

o WRI,EXT - Rotor internal cooling discharged into the exducer tip region - Based on axial turbine test results with tip discharge, the pumping penalty was again offset by a reductiun in exducer clearance penalty.

o WRB - Rotor bore cooling flow - Analysis of the Army cooled laminated rotor showed that a 0.30-percent cooling flo~' was required in this region. In addition to bypassing the rotor, a pumping penalty based on the rotor exit hub radius was applied to this cooling flow.

From the preceding considerations, the 100-percent power cooled turbine efficiency is expressed as: 4,Mix

= 0.9884 6H

1}T-T Cooled 6H ,Mix)isen W U H5 RB (2) using Garrett test data and the difference between the rotor- cooling flow assumed by the cycle and the estimated rotor-cooling flow, the penalty for rotor-shroud and bac~face-seal cooling flow was estimated at 0.6-percent efficiency. The final expression for the cooled turbine efficiency is:

1}T-T) Cooled = (0.9884) (1}parametr iC) - 0.006

\ Study G-L.

In general, the total penalty for all cooling flows was on the order of 2. a-points eff iciency at lOa-percent power. Cooling flOtIS were then assumed to vary linearly to 50-percent power, and the cooled efficiency was varied accordingly.

5.3 Effects of Stator Leakage Two ~t~~or sidewall leakage m01els were derived for the arti- culated trailing-edge configuration. Both m01els were based on the current Garrett model for axial nozzle leakage with full vane rotation. This wos modified to account for leakage in the trailing-edge regi.on alone. The axial turbine model related the performance decrement by a ratio of clearance area to nozzle throat area in the following manner: ~L _ ALeakane - = 1 4.::.

K • ~ AThroat The constant, K, was evaluated from test results. Applying the model to the full-vane rotation concept resulted in the characteristics shown in Figure 43, and illustrates why this concept was eliminated from consideration. For the articulated trailing-edge concept, no leakage occurred until the pivot point was reached (slightly downstream of the nozzle suction-surface throat point). The leakage area was no longer based on the annular leakage area at the vane trailing-edge, but was reduced by cos £r p i v 9t. If all the tangential momentum generated up to the nozzl~ P1VOt point is assumed lost, the expression for the leakage model becomes (Model No.1):

= 1 - Cos ~vot . 2C

(5) --Cos Q3-' b where:

apivot = Average stator angle at articulated pivot point

a = Stator-exit angle

NOTE: £rpivot and Q are shown in Figure 78, Section 6.3 • • Assuming £rpivot is equal to 1.13 radians (65 degrees) and Q3

j

is equal to 1.2] radians (70-degrees) leakage, Model No.1 char- acteristics are as shown in Figure 43. Thes~ results show signi- ficant performance decrements--even with only O.09-radian

I

I

SIDEWALL LEAKAGE PENALTY FOR ARTICULATED -.a 1.0 0\ T.E. CONCEPT BASED ON RECOVERY OF TANGENTIAL MOMENTUM GENERATED TO PIVOT POINT.

0.98 MAXIMUM POWER DETAILED PARAMETRIC STUDY

\

.£ ~ 0.96 ~

~ \

"

~

c

Z SIDEWALL LEAKAGE PENALTY FOR ARTICULATED au 0.94 ~ T.E. CONCEPT BASED ON ALL TANGENTIAL au MOMENTUM UP TO PIVOT POINT IS LOST.

\

~

¥ C au AXIAL STATOR LOSS ~ 0.92

r ImDEL APPLIED TO

\ FULL VANE ROTATIO!~

"

CONCEPT 0.90

o 0.05 0.10

0.15 0.20 0.25 0.30 0.35

"

.- , I I I I I~

(0.012) (0.014) o (0.002) (0.004) (0.006) (0.008) (0.010)

STATOR SIDEWALL CLEARANCE. C (ONE SIDE) mm (IN.'

L Figure 43. Articulated Trailing-Edge Nozzle, Sidewall Clearance Loss Models • ...

(S-d~grees) downstream t'Jrning. If the tangential momentum is • recovered, the performance decrement can then be approxlmated with the ~ollowing e~ression (Model No.2): Sin ~ivot • 2C (6)

+ sin u -·-b --

3 3

Stator sidewall leakage effects are presented in Figure 43, and are b~sed on the: pivot and exi t anqles assumed for Model No.1. If a O.Oll-cm (0.005-in.) sidewall clearance is selected for the m~ximum power point, the effect of stator leakage at part- power is as illustrated in Figure 44. With the sidewall clearance held constant, the reduction in performance at lower power levels is due to the increase in a with a fixed Upivot' This effect can be minimized with a stator designed for SO-percent power. The articulated trailing-edge would then be required to actuate in both an open and closed mode, but the total excursion ~n each direction would be reduced by 50 percent. This approach was examined in the detailed design phase.

5.4 Cooled Rotor Aerodynamic Analysis The performance model used for the detailed parametric study was updated to calculute the radial variation of flow properties inside the trailing-edge plane for a given minimum blade tip th ickness, maximum al10wable blade stress, and spec if i ed rotor exit mean work coefficient. This was achieved by solving the non- isentropic radial equilibrium equation for the s~)ecified rotor ex i t hub and shroud radi 1. The pr imary reaul ts of the solu tion included the rotor exit blockage, relative total pressure loss due to the blockage, and radial relative velocity distribution.

The maximum power results for the cooled 1589K (2400 F) rotor configuration are presented in Figures 45 through 47 foL' rotor inlet-blade angles of 0, 0.17, and 0.35 radian (0, 10, and I 20 degrees), and a radius ratio of 0.65. The following parameters

!

were held constant during the analysis:

o Stator exit angle: ('w 3 • 1. 31 rad ians (75. a degrees)

specified at the rotor-inlet station - A higher stator exit angle (compared to the preliminary study) was required to achieve reasonable rotor inlet blade heights

I

with the lower shaft speed of 5969 radls (57,000 rpm).

However, the stator vane loading and tt:aili~·.g-edge blockage were maintained at conAtant values by adjusting the vane number and radii chord;

I

o Rotor blade number: 14; t

I

, \

I

•

!

I

I fl !

\ I 1.0 a: w ,.,.

~

~ 0.98 ,- ~ .f: ~ ARTICULATED 0.96

--~

~

TRAILING-EDGE ..I I (DETAILED OFF- c(

z • i

DESIGN STUDY) w ~ w

O. l

CJ c( I lIiII: c( w

0')2 ~

..I a:

I

~ c( t;

o·~t

.- , 1.00 0.80 0.90 0.70 0.60 0.50 STATOR AREA RATIO. AN/~D Figure 44. Effect of Stator Leakage on Off-Design Performance.

.

•

!

. ~ t.

W

f

(!)

I

~ <t N 0.88 ~

- en

~ ...I ~ ~

J <t

~ 0.86 0>' ..

IjU OZ ... W IU ...1- 0.84 «u.

SELECTED WORK ... u.

OW COEFFICIENT (OFF- I- DESIGN ANALYSIS) 0.86 ~.

~

I

j

0.84 t

t

I

i .

0.82

t

t

•

• 0.80 I ~ I

j

0.78 L,----~----~.----~---~ 0.95 1.00 1.05 1.10 1.15 ROTOR INLET WORK COEFFICIENT. A4 ..

1589K (2400 F) T4

•

..

5969 ntd/l (67,000 RPM) N

•

II 640 mil (2100 FTISEC) U

•

T4 ..

N8

•

..

0.65 RIR

•