Skip to main content

Full text of "NASA Technical Reports Server (NTRS) 19770013161: Preliminary design study of advanced multistage axial flow core compressors"

See other formats



r 


NASA CR 135133 
R77AEG222 




PRELIMINARY DESIGN 
STUDY OF ADVANCED 
MULTISTAGE AXIAL FLOW CORE 
COMPRESSORS 


Final Report 


D.C. msler, C.C. Koch, L.H. Smith, Jr. 


GENERAL ELECTRIC COMPANY 
AIRCRAirr ENGINE GROUP 
CINCINNATI, OmO 45215 

(MASA-CB-135133) PRELIMINARY DESIGN STUDY 
OP ADVANCED MULTISTAGE AXIAL PLOW CORE 
COMPRESSORS Pinal Report, May 1975 - Jan. 

1976 (General Electric Co.) 95 p 

HC A05/HP A01 CSCL 21E G3/07 

Pr^ared For 


N77-20105 


Unclas 

21777 


Natioaal Airoiittlcs aid Spact Adaiiaistratioa 


February 1977 

NASA Lewis Research Center 
Contract NA83-19444 




A 

P A« S77 % 

U NAMSIIFMIUK ^ 
\a MPUIMMIRN ^ 



1. Report No. 


NASA CR-135133 


2. Government Accanion No. 


3. Recipient's Catalog No. 


4. Title and Subtitle 

Preliminary Design Study of Advanced Multistage Axial Flow 
Core Compressors 


5. Report Date 
February 1977 


6. Performing Organization Lode 


7. Author(s) 

D. C. Wisler, C. C. Koch, L. H. Smith, Jr. 


8. Performing Organization Report No. 
R77AEG222 


9. Performing Organization Name and Address 
General Electric Company 
Aircraft Engine Group 
Cincinnati, Ohio 4 5215 


10. Work Unit No. 


11. Contract or Grant No. 
NAS 3-19444 


12. Sponsoring Agency Name and Address 

National Aeronautics and Space Administration 
Washington, D. C. 20546 


13, Type of Report and Period Covered 
Final May 1975 - January 1976 


14. Sponsoring Agency Code 


15. Supplementary Notes 
Project Manager, Robert S. Ruggeri 
NASA - Lewis Research Center 
Cleveland, Ohio 44135 


16. Abstract 


Identify an advanced core compressor for use in new 
high-bypass-ratio turbofan engines to be introduced into commercial service in the 1980’s An 

o better evaluate their aerodynamic and mechanical feasibility. Finally, a comprLsor was 


17. Key Words (Suggested by Author(s)| 
Compressor 
Core Compressor 
Preliminary Design 
Miltistage Axial Flow Compressor 
High-Bypass-Ratio Ttirbofan Engines 


18. Distribution Statement 

Unclassified - Unlimited 


19. Security Oasslf. (of this raport) 

Unci ass if cd 


20. Sacurlty Ctauif. (of this page) 
Unclassified 


21. No. of Pages 

93 


22. Price* 


For sale by the National Technical Information Service, Springfield, Virginia 22151 









FOREWORD 


This report was prepared by the Aircraft Engine Group of the General Electric Company, 
Cincinnati, Ohio, to document the results of a preliminary design study conducted to 
identify appropriate design parameters for an advanced core compressor to be used in 
high -bypass -ratio turbofan engines of the 1985 time period. Mr. Robert S. Ruggeri, 
NASA-Lewis Research Center, Fluid System Components Division, was Project Manager. 

The authors wish to acknowledge the valuable contributions made to this program by 
various supporting organizations within the General Electric Company. In particular, 
appreciation is extended to Messrs. R.E. Neitzel and P.W. Vinson for their assistance 
in conducting the economic analysis utilized in this study, and to Messrs. F. W. Tegarden 
and J.D. Hennessey for their efforts in accomplishing the mechanical design and analysis 
studies. 


TABLE OF CONTENTS 


Page 


SUMMARY 1 

INTRODUCTION 2 

AER(»YNAMIC STUDIES 3 

METHODS AND GROUNDRULES FOR SCREENING STUDIES 3 

Efficiency Prediction and Stall Correlation Model 3 

Technology Groundrules 4 

PARAMETRIC SCREENING STUDIES 7 

Parameters and Range of Parameters Studied 7 

Results of Parametric Screening Studies 8 

Discussion of Loading Level Screening Studies 13 

Discussion of One-Parameter Variations 13 

Discussion of Other Screening Studies 27 

FURTHER STUDIES OF HIGH EFFICIENCY COMPRESSORS 29 

p/p = 23 Compressor Configurations 29 

p/p = 14 Compressor Configurations 34 

Effect of State -of -the -Art Assumptions 34 

DETAILED STUDY OF THREE SELECTED COMPRESSORS 37 

Axisymmetric Flow Calculations 39 

De-Staged 23:1 Compressors 40 

Off-Design Analysis 40 

RECOMMENDED CONFIGURATION 44 

MECHANICAL DESIGN STUDIES 49 

COMPRESSOR MECHANICAL DESIGN TECHNOLOGY - 19P5 STATE 
OF THE ART 49 


Clearance Control 

Airfoil Surface Finish 30 

Blade Erosion 30 

Rear Rim Speed 31 

Inlet Hub Radius 31 

System Vibration and Engine Bearing Layout 51 

Additional Mechanical Design Features 31 

PARAMETRIC SCREENING STUDIES 33 




i;oT 


1 

( 


V 



1 


Page 

DETAILED DESIGN STUDY "5^? 

Blades and Vanf>s 59 

Dovetails 59 

Disks and Structure 59 

System Vibration 59 

Blade Erosion 62 

Revised Rotor Weight Estimates 62 

RECOMMENDED CONFIGURATION 62 

ENGINE SYSTEM STUDIES 64 

METHODS AND GROUND RULES 64 

Aircraft Mission Analysis 64 

Installation Effects 65 

Engine Performance 67 

Turbine Performance Effects 67 

PARAMETRIC SCREENING STUDY 69 

REFINED SCREENING STUDY 71 

DETAILED DESIGN STUDY 80 

ENGINE SYSTEM MERIT FACTORS SENSITIVITY STUDY 82 

ENGINES USING THE RECOMMENDED COMPRESSOR CONFIGURATION . . 82 

CONCLUSIONS AND RECOMMENDATIONS 84 

PARAMETRIC SCREENING STUDY FINDINGS 84 

DETAILED DESIGN STUDIES 85 

RECOMMENDED CONFIGURATION 85 

REFERENCES 86 




LIST OF FIGURES 


Figure Page 

1 Conservative Loading Compressor Design, 14:1 Total-Pressure 

Ratio, Configuration No. 1 14 

2 Nominal Loading Compressor Design, 14:1 Total-Pressure 

Ratio, Configuration No. 3 15 

3 Maximum Loading Compressor Design, 14:1 Total-Pressure 

Ratio, Configuration No. 2 16 

4 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 

Compressor Efficiency with Aspect Ratio 18 

5 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 

Compressor Efficiency with Average Solidity 19 

6 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 

Compressor Efficiency with Average Stator Exit Swirl Angle 20 

7 Variation in Mrect Operating Cost, Fuel Usage, Tip Speed, and 

Compressor Efficiency with Exit Mach Number 21 

8 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 
Compressor Efficiency with Inlet Corrected Flow per Unit 

Annulus Area 23 

9 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 

Compressor Efficiency with Inlet Radius Ratio 24 

10 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 

Compressor Efficiency with Exit Radius Ratio 25 

11 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 

Compressor Efficiency with Number of Stages 26 

12 Adiabatic Efficiency and Corrected Tip Speed Versus Number of 

Stages for 23:1 Total -Pressure Compressors, Configuration 26 31 

vll 


Figure Page 

13 Compressor Physical and Mechanical Characteristics Versus 
Number of Stages for 23:1 Total -Pressure Ratio Designs, 

Configuration 26 32 

14 Engine Economic and Physical Characteristics Versus Number 
of Stages for 23:1 Total-Pressure Ratio Compressors, 

Configuration 26 33 

15 Effect of Technology Level Assumptions on Efficiency of 23:1 

Total-Pressure Ratio Compressors 36 

16 Compressor Efficiency as a Function of Exit Radius Ratio and 
Assumed Level of Technology for 9-Stage 23:1 Total-Pressure 

Ratio Designs 38 

17 Stagewise Distribution of Rotor and Stator Mach Number for 

the Three Selected Compressors 41 

18 Stagewise Distribution of Rotor and Stator Pitchline Diffusion 

Factor for the Three Selected Compressors 42 

19 Estimated Performance Map for Recommended AMAC 23:1 

Total-Pressure Ratio Compressor 45 

20 Flowpath of Recommended AMAC Compressor, Ten Stages, 

23: 1 Pressure Ratio, Configuration 26e2 47 

21 Comparison of Actual and Predicted Blade Erosion Life 52 

22 Mechanical Layout of Configuration 26b2, Eleven Stages, 

23:1 Pressure Ratio 56 

23 Mechanical Layout of Configuration 26d5, Nine Stages, 23:1 

Pressure Ratio 57 

24 Mechanical Layout of Configuration 18c, Nine Stages, 14:1 

Pressure Ratio 58 

25 Campbell Diagrams for Rotors One and Nine, Configuration 

26d5 60 

26 Typical Mixed Flow Installation Used for the Study 66 

27 Nominal Turbine Flowpath for 14:1 Pressure Ratio Advanced 

Compressor 70 

viil 



Figure Pag e 

28 Nominal Turbine Flowpath for 23: 1 Pressure Ratio Advanced 

Compressor 75 

29 Effect of Number of Compressor Stages on Bare Engine 

Specific Fuel Consumption and Installed Drag 77 

30 Effect of Number of Compressor Stages on Bare Engine Weight 

and Price (Design Size) 78 

31 Effect of Number of Compressor Stages on Installation Weight 

and Price (Design Size) 79 

32 Layout of Engine Incorporating Reconr-mended AMAC Compressor, 

Configuration 26e2 83 


ix 


LIST OF TABLES 


Table Page 

I Design Parameters for Maximum Loading, Nominal Loading, 

and Conservative Loading Compressors 9 

n Summary of Aerodynamic Design Data for Parametric 

Variations 11 

in Aerodynamic Summary of Configuration 26 Designs, Maximum 

Efficiency, 23: 1 Pressure Ratio 30 

IV Aerodynamic Summary of Configuration 18 Designs, Maximum 

Efficiency, 14:1 Pressure Ratio 35 

V Comparison of De -staged 23:1 Compressors with 

Configuration 18c 43 

VI Detailed Design Study Results for 23:1 Total Pressure 

Ratio Compressors 46 

Vn Comparison of Recommended 10-stage AMAC Compressor 

with Configurations 18c, 26b2, and 26d5 48 

Vni Parametric Screening Study Mechanical Design Results 55 

IX Rotor Blade and Disk Mechanical Design Summary 61 

X Configuration 26e2 Compressor Erosion Summary 63 

XI Boosted vs. Unboosted Cycle Comparison 68 

XII Summary of Component Weight, Price, Installation, Main- 
tenance, and Specific Fuel Consumption Data for the 14:1 

and 23:1 Pressure Ratio Compressors (Design Size Engines) .... 73 

Xin Summary of Engine Characteristics and Engine Evaluation 

Results (with Erosion Effects) 81 


X 



SUMMARY 


/ 


A preliminary design study was conducted to identify an advanced core compressor for use 
in new high-bypass- ratio turbofan engines to be introduced into commercial service in the 
1980’s. The initial phase of the study involved x forecast of projected 1985 state-of-the-art 
technology in compressor and engine system aerodynamic and mechanical design areas. 

The turbine inlet temperature levels projected for use in 1985 vintage engines lead to 
optimum thermodynamic cycles that require an overall pressure ratio of the order of 40:1, 
To achieve this overall pressure ratio, two types of core compressor configurations were 
studied: boosted 14:1 pressure ratio compressors driven by single-stage turbines and un- 
boosted 23:1 pressure ratio compressors driven by two-stage turbines. Based upon the 
technology projections, a parametric screening study covering a large number of compres- 
sor designs was conducted in which the influence of major compressor design features on 
efficiency, weight, cost, blade life, aircraft direct operating cost and fuel usage was 
determinecl. Three high- efficiency, high-economic payoff compressors were developed 
using the trends observed in the parametric screening studies; these were then studied in 
detail to better evaluate their aerodynamic and mechanical feasibility. 

Finally, a compressor configuration was selected which demonstrated the best performance 
potential and good overall system economic payofl. The design selected for development 
was a 10-stage 23:1 pressure ratio compressor offering the best combination of the follow- 
ing advantages: high efficiency, low operating cost, low fuel usage, and acceptable develop- 
ment risk. It was found that this compressor with its first stage removed would also be an 
attractive 14:1 pressure ratio candidate for a boosted engine. 


I 


i 


INTRODUCTION 


A preliminary design study was conducted under NASA Contract NASS- 19444 to identify 
appropriate design parameters for an advanced core compressor for use in new high- 
bypass-ratio turbofan engines to be introduced into commercial service in the 1980’s. 
Although the core compressor in a modern turbofan engine is dwarfed in size by the high 
bypass ratio fan component, it remains a key element in the heart of the engine and has a 
large impact on system performance and operating cost. The high turbine inlet temperature 
levels projected for use in 1985 vintage engines lead to optimum thermodynamic cycles 
that require high overall pressure ratios, of the order of 40:1. The high pressure com- 
pressor component must produce the majority of this pressure ratio, and must do it 
efficiently and reliably. The compressor must also be designed so that engine surges will 
not occur in the operating envelope of the engine, even after thousands of hours of operation 
have led to some performance deterioration due to erosion, wear, etc. Because of the 
performance demands placed on the core compressor and the propulsion system’s overall 
dependence on the compressor meeting its design requirements, it was considered essential 
that the compressor design selection be based on an extensive preliminary design study 
which incorporated an assessment of the projected state-of-the-art advancements in the 
appropriate time period. The preliminary design study, together with its findings and re- 
sulting recommendations, are described in this report. 


2 


AERODYNAMIC STUDIES 


The aerodynamic studies phase of the program was structured to provide a systematic 
approach to the identification and selection of an optimum configuration for an Advanced 
Multistage Axial Flow Compressor (AM AC compressor). The effort ' ^as divided into five 
phases. First, the technology levels and ground rules used in the analytical design methods 
were selected to represent anticipated 1985 time period state of the art. Second, paramet- 
ric screening studies were conducted to determine tradeoffs between compressor efficiency, 
i :ze, weight, and cost. The range of design parameters studied varied from values typical 
of existing compressors, termed 'conservative loading” design3, to those that would be 
used in very highly loaded compressors, termed ”maximum loading” designs. The center 
point of the range was termed '’nominal loading” designs. Third, based on the results of 
the parametric study, several high efficiency compressor configurations were specified. 
Fourth, a more detailed study of the three most promising compressors was conducted 
using axisymmetric calculations and off-design performance estimating procedures. 

Finally, an optimum configuration was recommended for the AMAC compressor. This 
configuration was judged to have a high economic payoff for use in an advanced commercial 
engine, represented a substantial advance in the state of the art, and had acceptable de- 
velopment risk. 

METHODS AND GROUNDRULES FOR SCREENING STUDIES 

The parametric screening studies were conducted consistent with certain groundrules 
that relate to engine system constraints and 1985 state-of-the-art technology projections. 

In order to express tno aerodynamic technology projections quantitatively, it is first 
necessary to outline the aorodyuamic analysis methods that were employed. 

Efficiency Prediction and Stall Correlation Mode] 

Preliminary design studies of advanced multistage compressors at the General Electric 
Company rely on a computerized procedure, identified as the Compressor Unification 
Study, to estimate both efficiency potential and stall pressure ratio potential. The 
efficiency prediction model is intcnued inHiv!dte the potential peak efficiency of a well- 
designed compressor. It attempts to account for all known sources of loss except for 
those due to off -design operation, blading unsuited for the aerodynamic environment, or 
poor hardware quality. The losses are grouped into four sources: (1) end -wall boundary 
layers and end-wall region secondary flows anH leakage flows; (2) blade surface profile 
drag; (3) shocks on the blading; and, (4) part -.span shrouds. End -wall losses have been 
determined from hub aiid casing boundary layer measurements made on a number of 
multistage, low speed, research compressor configurations These losses are related to 
aspect ratio, solidi':y, scagger, tip clearance, blade row axial spacing, and aerodynamic 
loading level. Blace surface profile losses are related to suction surface diffusion. 





blade maximum thickness and trailing edge thickness, Reynolds number, surlace rough- 
ness, Mach number, and streamtube contraction. The shock loss model relates passage 
shock losses to inlet and exit Mach numbers and relates leading edge bow shock losses to 
inlet ^ ach number and leading eage thickness. The model for part-span shroud losses is 
based on measured shroua drag coefficients. 

A detailed description of this efficiency model and some compa isons showing the capabil- 
ity of the model to predict the efficiency of multistage compressors is given in Reference 
1. In most cases, the efficiency predicted by the model agrees with the efficiency deter- 
mined from test data within one point. 

The stall pressure ratio prediction method is based upon two groups of background data. 
The first group consists of the measured stall pressure rise capabilities of a large 
number of lew speed repeating stages covering a wide range of stage geometries. These 
experimental results are expj essed as a stall pressure rise coefficient, wliich is related 
to starve geometry parameters by a correlation. The second group of background data is 
from high speed multistage compressors. These data are presented in the form of a 
ratio, called effectivity, of an individual stagj pressure rise coefficient measured at 
design speed stall to that predicted by the low speed data correlation. The average values 
of effect! vity for multistage compressors lie generally in the range of 0.88 to 0.96. The 
best average effectivity ever obtained for a high pressure ratio multistage compressor is 
0.99. To date, the General Electric Company is not aware of any multistage compressor 
data from any source to indicate that an average effectiviby greater than this value has 
been achieved. 

In applying the stall prediction method to a new design, certain stage geometry parameters 
and appropriate effectivities are specified, and the correlation is used to deduce the 
pressure rise coefficjerit of each stage. Other input quantities are the distributions of 
axial velocity and stator exit flow angle at the design speed stall point, the airflow, pitch- 
line radii, and 3 n estimated speed. The computer program calculates the stage pressure 
ratios, stacks the stages to give an overall pressure ratio, and adjusts the speed until 
the desired overall pressure ratio is obtained. Hub and tin radii consistent with these 
results and other quantities of interest are then calculated. 

Technology Groundrules 


The levels of technology assumed in predicting the stall margin capability and efficiency 
potential for advanced multistage compressors of the 1985 time period are presented i.i 
the following discussion. 

Efficiency Prediction - Technology advancements leading to efficiency improvements can 
be classified as either aerodynamic or mechanical. Aerodynamic advancements result 
from the discovery of improved airfoil shapes and flowpath contours, and mechanical 
advancements result from such features as reduced clearances and improved surface 
finish. Both types of advancements were assumed for this study; the aerodynamic 
advancement element is described below and the mechanical advancements are described 
in a later section, Mecha nic al Design Studies. 


i 



Of the four aerodynamic loss sources previously identified, it was judged that only the end- 
wall loss source is likely to be reduced by any meaningful amount in the next few years . 
Profile losses are already quite low in well-designed and well- matched compressors that 
are represented by the current efficiency potential model. The shock loss model used in 
the study also represents the loss levels of the most efficient transonic/supersonic stages. 
The part-span shroud loss model is relatively unimportant because most of the configura- 
tions studied did not employ such shrouds. 

For this study, it was assumed that the end- wail loss was 15 percent lower than that yielded 
by the model at any given tip clearance level. Research efforts currently underway and 
anticipated in the near future should provide the knowledge needed to achieve this loss re- 
duction. 

The efficiency model discussed so far has been concerned with losses in the compressor 
blading only. Since the exit Mach number is a parameter that was varied, it was neces- 
sary to pr^ict the diffuser losses and the diffuser length needed to reduce the Mach 
number to the level required by the combustor. Size and perform .mce estimates for an 
advanced split-flow diffuser were made as functions of the Mach number (based on an 
effective -area coefficient of 0.9) at the compressor outlet guide vane exit. These 
estimates were done as part of the compressor aerodynamic aesign studies in order to 
determine the length and performance of ail compressors on the basis of the same (0.06) 
combustor inlet Mach number . 

Inlet guide vane losses were incorporated by employing a lose coefficieiit (based on exit 
dynamic pressure) that increased with deflection angle from a value of 0,035 at zero 
deflection to 0.050 at 20 degree deflection. This increase was in accordance with the trend 
of two-dimensional cascade data. In addition to the inlet guide vane loss model, core com- 
pressor inlet duct pressure loss and length correlations were also utilised. These duct 
correlations were consistent with General Electric experience. 

Stall Prediction - In order to carry out the study, it was necessary to establish design- 
point stall margin values. For many military applications, considerable stall margin is 
required to handle the large inlet distortions that can result from high angle -of-attack 
operation, off-design operation of supersonic inlets, armament firing, etc. These dis- 
tortions can occur while the compressor is operat ig at high corrected speed, and high 
stall margin must therefore be provided at this operating speed. Subsonic commercial 
transport engines do not experience such severe distortions and, therefore, lower values 
of high speed stall margin may be employed. It is believed that 15 percent stall margin 
will be adequate for most commercial transport applications at high corrected speeds. 
However, part-speed stall margin requirements for commercial transport engines are 
similar to those of military engines for starting and rapid acceleration of the engine. In 
fact, commercial transport requirements may be more stringent than military require- 
ments because the longer service life of a commercial compressor ultimately leads to its 
operation with deteriorated performance due to blade and casing erosion. Experience has 
shown that a part speed stall margir. above the steady-state operating line of 25 to 30 
percent is needed to assure satisfacr.ory engine operation in the 70 percent airflow region 
of the compressor map, A cursory stability analysis, conducted concurrently with this 
study, confirmed that this stall margin level is reasonable for 1985 commercial engines. 


5 


1 


J 




Since the paran'etric screening studies were carried out at the high speed design point, it 
was necessary to specify sufficient stall margin at that point to assure that adequate stall 
margin is available for part-speed operation. A review of General Electric experience 
indicated that the amount of design speed stall margin specified should not be the same for 
all compressors. This was concluded because plots of stall pressure ratio versus percent 
design airflow differ for different compressors. Although many design variables affect 
the stall line shape on such a plot, there is a general tendency for the stall line to be 
higher at intermediate airflows for compressors that have low solidities, low radius 

ratios, increasing hub radii through the compressor, and somewhat reduced effectivities i 

in the front stages as well as in the front stages of the fixed rear block. * Since these are 

similar to the characteristics of the conservative loading compressors of this study, it 

follows that the design point stall margin of the conservative loading compressors can be 

less than those of the nominal and maximum loading compressor types so that all three 

compressor types have the same part-speed stall margin. 

In view of the foregoing discussion, and relying on past experience to aid in selecting 
numerical magnitudes, the following values were selected for use in the screening studies: 


Parameter 

Design Point 
Stall Margin, % 

Average Stall 
Effect! vity 


Maximum 

Loading 


Nominal 

Loading 


Conservative 

Loading 


Compressors Compressors Compressors 


22 


19 


16 


0,975 


0.95G 


0.925 


The stall effectivity distributions through the compressors used in this study had a 
minimum value for the first stage and a maximum value of 0.99 for the rear stages. As 
the loading level was increased, the effectivity of the front stages was increased, result- 
ing in the higher values of average stall effectivity for the maximum loading designs. 

Casing treatment was considered as a possible means for increasing effectivity levels. 
However, at present, there is no hard evidence that casing treatment is capable of in- 
creasing the average stall effectivity of a multistage compressor beyond the levels 
employed in this study. Also, it seems likely that if it were found that the use of casing 
treatment could improve stall effectivity, it would also be found that such treatment would 
cause an efficiency penalty. This would probably be an unfavorable trade for the com- 
pressors of this study. 


♦The fixed rear block is that group of stages whose rotors are not preceded by variable 
stators. 


6 


Design Constraints - The compressor parametric design studies were subject to several 
mechanical design constraints. A minimum core compressor inlet hub radius of 16.51 cm 
(6.50 inches) in the 147, 000 n (33, 000 lb) thrust engine study size was established in 
order to permit the low pressure compressor (fan) drive shaft to pass through the center 
of the core compressor with sufficient clearance foi bearings, core compressor structure, 
etc. A maximum physical rear stage hub speed of 381 m/sec (1250 ft/sec) was estab- 
lished as an upper limit for the parametric screening study. Indications are that above 
this speed, elaborate rotor cooling schemes of very massive structures would be required 
for structural integrity. This constraint was later relaxed in the detailed study phase, 
and the weight and cost penalties of high rim speeds were factored into the analysis. A 
maximum physical speed of 17, 000 rpm was established as an upper limit based upon 
high pressure turbine stress considerations. 

The ground rule used to estimate axial spacing between blade and vane rows was based on 
General Electric experience. The axial spacing required to avoid blade interference due 
to blade deflections was assumed to be a function of the rotor blade height of the stage and 
the stage number, as long as the axial spacing was greater than some minimum spacing. 
This minimum value was set to avoid blade interference caused by differential thermal 
growth of rotors and stators, and was made a function of compressor overall length. For 
compressors in the 32-45 kg/sec (70-100 Ibm/sec) corrected flow size used in this study, 
the absolute minimum allowable spacing was assumed to be 0.635 cm (0.25 inch) in order 
to provide room for necessary structural details such as rotor blade retainers. 

PARAMETRIC SCREENING STUDIES 


A series of preliminary compressor aerodynamic designs was carried out in which key 
parameters were varied systematically in order to determine the trade-offs between 
compressor efficiency, size, weight, cost, life, etc. The parametric studies were 
conducted in three parts. The first part defined three compressor designs for each of two 
levels of total pressure ratio, 14:1 and 2.3:1. These three designs were a nominal loading 
compressor, a maximum loading compressor, and a conservative loading compressor. 

In the second part of the parametric study, the nominal loading 14:1 pressure ratio 
compressor was used as a center point. Each significant design parameter which defined 
the compressor was varied in two steps, one in the direction of a maximum loading con- 
figuration and the other in the direction of a conservative loading configuration. In the 
third part of the parametric evaluation, the trends determined from the earlier cases 
were used to guide the parameter selections for designs that focused on ore particular 
characteristic, such as long life. Since compressor efficiency was found to be an 
important parameter affecting economics, additional studies were carried out in an 
attempt to maximize efficiency. These studies are presented in a later section. Further 
Studies of High Efficiency Compressors. 

Parameters and Range of Parameters Studied 

The most important design variables that define the compressor and affect its performance 
are aspect ratio, solidity, swirl angle (reaction), exit Mach number, inlet flow/annulus 
area, inlet radius ratio, flowpath shape, and number of stages. Values of these design 
variables selected to define maximum loading, nominal loading, and conservative loading 


7 





compressors ^or each of the tvo levels of total pressure ratio are listed in Table I. Tip 
speed is not listed because it is determined by the stall margin requirement rather than 
being a direct design specification. 

The nominal values of aspect ratio, solidity, and stator exit flow angle (middle column of 
Table I) are generally consistent with those used previously in General Electric high- 
stage-loading compressors. Values in die maximum loading column were chosen as a set 
of extremes aimed at achieving a very high average stage pressure ratio. The conserva- 
tive loading values were selected to represent compressors currently in service. 

Stagewise distributions of aspect ratio, solidity, stator exit flow angle, and stalling axial 
velocity were established based on past experience and judgment. Generally, the trends 
of aspect ratio, solidity, and flow angle for the conservative loading compressors are 
consistent with those of the CF6 compressors, while the trends of these variables for the 
nominal loading compressors are consistent with those of General Electric high-stage- 
loading compressor designs . For each design type, the average of the rotor and stator 
stagewise distributions of each parameter is equ^ to the average value of that parameter 
listed in Table I for the configuration. From general considerations of structural 
adequacy, length and weight, the aspect ratios tend to decrease from the inlet to the outlet 
of a compressor, and this trend was modeled in the parametric study. Since aeromechan- 
ical considerations indicate that the first rotor aspect ratio cannot be much above 1.5 
without shrouding, a part-span shroud was specified for this blade row in the conservative 
design and a substantially higher aspect ratio was employed. The distributions of rotor 
solidity generally had the highest solidities iu the front stages where the relative Mach 
numbers were highest. The stator solidities were kept somewhat smaller than average in 
those front stages that were expected to be variable, and were highest in the rear stages 
where aerodynamic loadings were high due to axial velocity diffusion. 

A stagewise distribution of stator exit flow angle that maintained a moderate level of 
rotor inlet Mach number and fairly low reaction ratios was chosen for the conservative 
loading design. The level of swirl increased to a maximum through the front half uf the 
compressor, remained constant, and then decreased rapidly through the last few stages. 
The distribution for the nominal loading design had less variation and lower levels, lead- 
ing to higher Mach numbers and reactions. A constant zero-level of swirl was used for 
the maximum loading design. 

The stagewise distributions of stalling axial velocity for the maximum loading designs were 
characterized by an acceleration in stalling axial velocity through the first half of the 
compressor, estimated to provide a nearly constant axiad Mach number at the design point, 
followed by a rapid diffusion in the last three stages. The nominal and conservative load- 
ing designs employed more moderate distributions with less average axial velocity diffu- 
sion per stage. 


Results of Parametric Screening Studies 


A detailed listing of key aerodynamic and mechanical design parameters and economic 
analysis results for all three parts of the study Is given in Table II. Configurations 1-3 
in Table II are the basic 14:1 pressure ratio designs, and Configurations 20-22 are the 





Table I. Design Parameters for Maximum Loading, 

Nominal Loading and Conservative Loading Compressors. 



Maximum 

Nominal 

Conservative 


Loading 

Loading 

Loading 

Parameter 

Compressor 

Compressor 

Compressor 

Number of stages 

p/p * 14 

6 

9 

12 

P/P = 23 

7 

10 

14 

Average aspect ratio 

1.0 

1.5 

2.25 

Average solidity 

1.8 

1.35 

0.9 

2 

'‘low per annulus area, kg/ sec m 

200 

186 

171 

(Ibm/sec ft^) 

(41) 

(38) 

(35) 

Exit Mach number 

0.40 

0.34 

0.28 

Average stator exit flow 

angle, degrees 

0 

10 

20 

Inlet radius ratio 

0.75 

0.65 

Min. hub radiu: 
16.5 cm (6.5 ii 

Flowpath shape 

Max. hub speed 

Constant 

Exit radius 


in rear stages 

pitchline 

ratio = 0.91 


381 m/sec 
(1250 ft/sec) 

radius 

(or constant 
hub radius) 


9 





Table II. Summary of Aerodynamic Desigp 

Lengt 


Adiabatic 

Configuration Efficiency Incl. 

No. Parameter Variation Diffuser Losses 

Physical 
Speed, rpm 

Corrected 
Tip Speed, 
m/sec (fps) 

Rear Hub 
Speed (Phys.) 
m/sec (fps) 

Rotor 1 
to Diff 
Exit, m 

I, Pressure Ratio 14:1 Configurations 







1. 

Conservative Loading 12-Stage 

0.859 

15, 150 

415 (1360) 

336 (1101) 

0.642 

(Z 

2. 

Maximum Loading 6-Stage 

0.785 

14,410 

454 (1490) 

381 (1250) 

0.413 

r 

3.* 

Nominal Loading 9-Stage 

0.850 

n,510 

385 (1263) 

344 (1127) 

0.554 

(2 

4. 

Lower Aspect Ratio (1.0) 

0.832 

12,650 

360 (1182) 

321 (1053) 

0.806 

(. 

5. 

Higher Aspect Ratio (2.25), 1st Stage 
With Part-Span Shroud 

0.842 

14,280 

407 (1335) 

363 (1191) 

0.415 

(li 

6. 

Higher Solidity (1.8) 

0.845 

12,810 

365 (1197) 

326 (1068) 

0.563 

(2 

7. 

Lower Solidity (0.9) 

0.840 

14,430 

411 (1349) 

367 (1203) 

0.545 

(2 

8. 

Lower Swirl (0®) 

0.842 

13,490 

384 (1261) 

342 (1124) 

0.343 

(2 

9. 

Higher Swirl (20“) 

0.849 

13,610 

388 (1272) 

346 (1134) 

0.560 

(23 

10. 

Higher Exit Mach Number (0.4) 

0.846 

13,250 

377 (1238) 

338 (1109) 

0.566 

( 

11. 

Lower Exit Mach Number (0.28) 

0.856 

14,000 

399 (1309) 

T54 (1160) 

0.555 

(2 

12. 

Higher Flow per Annulus Area (200 (41)) 

0.842 

13,350 

376 (1232) 

339 (1113) 

0.529 

(21 

13. 

Lower Flow per Annulus Area (171 (35)) 

0.855 

13,750 

398 (1305) 

350 (1147) 

0.579 

( 

14, 

Higher Inlet Radius Ratio (0.75) 

0.844 

11.240 

369 (1210) 

352 (1156) 

0.461 

(1 

15. 

Lower Inlet Radius Ratio (min, Radius 
Hub) 

0.853 

14,940 

397 (1304) 

337 (1106) 

0.620 

(2^ 

16. 

High Rear Radius Ratio (max. hub speed) 

0.845 

11,820 

337 (1105) 

359 (1179) 

0.509 

(2 

17. 

Low Rear Radius Ratio (0.91) 

0.847 

15,590 

444 (1457) 

322 (1057) 

0.594 

(2 

18c. 

Maximum Efficiency 9-Stage 

0.868 

15,980 

437 (1435) 

374 (1227) 

0.536 

(2 

19b. 

Maximum Life 9-Stage 

0.849 

15,030 

428 (1405) 

311 (1019) 

0.715 

(2) 

24. 

Nominal Loading 8-Stage 

0.849 

14,680 

419 (1375) 

374 (1226) 

0,484 

(li 

25. 

Nominal Loading 10-Stage 

0.847 

12,520 

357 (1170) 

318 (1043) 

).633 

(» 

11. Pressure Ratio 23:1 Configurations 







20. 

Conservative Loading 14-Stage 

0.844 

14,360 

492 (1615) 

315 (1034) 

0.937 


21. 

Maximum Loading 7-Stage 

0.759 

11,730 

491 (1610) 

381 (1250) 

J.473 

(1 

22. 

Nominal Loading 10-Stage 

0.831 

11,320 

428 (1404) 

357 (1171) 

0.626 

(2. 

23. 

Lightly Loaded Front Stage 9-Stage 

0.822 

14,540 

474 (1555) 

370 (1213) 

0.612 

(2 

26b2. 

Maximum Efficiency 11-Stage 

0.856 

13,340 

457 (1500) 

355 (1163) 

0,767 

(. 

26d5. 

Maximum Efficiency 9-Stage 

0.849 

14,440 

480 (1575) 

393 (1290) 

0.630 

r 


*( One- Parameter Variation Center Point) 

tTranacontlnental trijet aircraft mission (relative to STEDLEC Baseline Engine) 






I 


Aerodynamic Design Data for Parametric Variations, 


^ar Hub 
ed (Phys.) 
tec (lp.s) 

Length 

Rotor 1 Inlet 
to Diffuser 
Exit, m (in.) 

(1101) 

0.642 (25.3) 

(1250) 

0.413 (16.2) 

(1127) 

0.554 (21.8) 

V (1053) 

0.806 (31.8) 

(1191) 

0.415 (16.3) 

(1068) 

0.563 (22.2) 

(1203) 

0.545 (21.5) 

(1124) 

0.543 (21.4) 

(1134) 

0.560 (22.1) 

(1109) 

0.566 (22.3) 

(1160) 

0.555 (21.9) 

(1113) 

0.529 (20.8) 

(1147) 

0.579 (22.8) 

(1156) 

0.461 (18.1) 

(1106) 

0.620 (24.4) 

(1179) 

0.509 (20.0) 

(1057) 

0.594 (23.4) 

(1227) 

0.536 (21.1) 

(1019) 

0.715 (28.2) 

(1226) 

0.484 (19.1) 

(1043) 

0.633 (24.9) 

(1034) 

0.937 (36.9) 

(1250) 

0.473 (18.6) 

, (1171) 

0.626 (24.6) 

(1213) 

0.612 (24.1) 

(1163) 

0.767 (30.2) 

(1290) 

0.630 (24.8) 


A 

Direct 


Inlet 
Tip Dia. 

No. of 
Blades 

Average 

AWeightt 

A Priced 

Operating 

Costt 

AFuel^ 
Usage % 

m (in. ) 

& Vanes 

Reaction 

kg (lb.) 

% 

% 

0.583 (22.9) 

1523 

0.69 

25 (56) 

0.3 

-0.44 

-0.41 

0.672 (26.4) 

1511 

0.83 

16 (35) 

0.4 

1.08 

3.83 

0.607 (23.9) 

1827 

0.69 

-1 (-2) 

0 

-0.37 

0.17 

0.607 (23.9) 

1265 

0.64 

44 (98) 

2.1 

0.58 

2.15 

0.607 (23.9) 

2537 

0.72 

-13 (-29) 

-0.4 

0.01 

0.34 

0.607 (23.9) 

2648 

0.65 

0 

0.7 

0 

0.79 

0.607 (23.9) 

1198 

0.73 

5 (12) 

-0.4 

-0.24 

0.70 

0.607 (23.9) 

1827 

0.82 

2 (5) 

0.2 

-0.14 

0.69 

0.607 (23.9) 

1827 

0.55 

2 (4) 

0.1 

-0.33 

0.21 

0.607 (23.9) 

1837 

0.67 

2 (5) 

0.4 

-0.16 

0.59 

0.607 (23.9) 

1620 

0.71 

1 (3) 

-0.2 

-0.56 

-0.26 

0.607 (23.9) 

1911 

0.67 

-5 (-11) 

0 

-0.20 

0.56 

0.616 (24.3) 

1698 

0.71 

8 (17) 

0.1 

-0.45 

-0.11 

0.697 (27.5) 

2707 

0.69 

3 (6) 

0.8 

-0.19 

0.08 

0.567 (22.3) 

1462 

0.68 

15 (33) 

-0.1 

-0.51 

-0.11 

0.607 (23.9) 

2419 

0.69 

10 (23) 

0.6 

-0.27 

-0.11 

0.607 (23.9) 

1379 

0.69 

12 (27) 

-0.3 

-0.35 

-0.30 

0.583 (23.0) 

1626 

0.75 

2i (44) 

-0.5 

-0.78 

-0.93 

0.607 (23.9) 

1212 

0.68 

32^ (71) 

0.6 

-0.38 

-0,40 

0.607 (23,9) 

1571 

0.71 

-2 (-5) 

-0.7 

-0.52 

-0.02 

0.607 (23.9) 

2010 

0.64 

12 (27) 

1.0 

0.06 

0.96 

0.676 (26. 6i 

1597 

0.69 

89 (197) 

6.0 

0.92 

-1.30 

0.825 (32.5) 

2329 

0.84 

68 (151) 

7.7 

3.22 

3,74 

0.753 (29.7) 

2592 

0.72 

28 (63) 

6.7 

1.47 

-0.51 

0.643 (25.3) 

1747 

0.73 

45 (99) 

4.8 

0.88 

-0.54 

0.676 (26.6) 

2087 

0.73 

45 (100) 

5.5 

0.33 

-2.41 

0.656 (25.8) 

1839 

0.74 

38 (84) 

3.9 

0.14 

-2.27 


MBEDING PAOi 


• ^ 'vr 

I ’ A V ' A 




three basic 23:1 pressure ratio cases that constituted the first part of the study. Of the 
other designs tabulated, Configurations 4-17, 24, and 25 are cases in which the primary 
compressor design parameters were varied individually to values above and below those of 
Configuration 3, which is the nominal loading 14:1 pressure ratio configuration. These con- 
stituted the second part of the study. The maximum efficiency, maximum life, and lightly 
loaded front stage Configurations 18, 19, 23, and 26 represent the third part of the study. 
The efficiencies presented in the table include the diffuser losses and the tabulated lengths 
include the diffuser length. The efficiencies have also been adjusted for stage matching 
effects to recognize that some compromise in design-point efficiency will usually result 
when a compressor is designed and developed to best match the needs of the overall engine/ 
aircraft system for a particular mission. The adjustment to those efficiencies calculated 
by the efficiency potential computer model involved an adiabatic efficiency reduction of 0.7 
point for the 14:1 pressure ratio compressors and a reduction of 1.0 point for the 23:1 
pressure ratio compressors. The mechanical design and economic analysis study methods 
are described in later sections of this report. Compressor weight and cost plus overall 
direct operating cost and fuel usage are give i as delta values relative to the engine system 
and compressor described in a NASA/GE study program entitled, ”Study of Turbofan 
Engines Designed for Low Energy Consumption'^ (STEDLEC), Reference 2. 

Discussion of Loading Level Screening Studies 

Flowpaths for the three basic 14: 1 pressure ratio designs defined in the first part of the 
parametric screening study are shown in Figures 1 through 3. The three basic 23:1 
pressure ratio compressor flowpaths were similar in appearance. As indicated in Table 
II, a primary result of this part of the study was that compressor efficiency, direct 
operating cost, and fuel usage can become very unfavorable if loading is increased to an 
extreme. Designs having the highest loadings (Configurations 2 and 21) were found to have 
no weight advantage, although they were much shorter, and their slightly lower cost did 
not offset the large efficiency penalties incurred.* 

An unforeseen result from these loading level studies was that the conservative loading 
designs required higher tip speeds than the nominal loading designs in order to achieve 
the required stall margin. This resulted from specifying parameters that gave low values 
of allowable stage static-pressure-rise coefficient for these designs, combined with 
numbers of stages that were not particularly large. 

Discussion of One -Parameter Variations 


For the second part of the parametric study, the 14:1 pressure ratio, nine -stage, nominal 
design (Configuration 3) was used as a center point, ai^ each significant design parameter 
defining the compressor was varied about this center in the direction of .maximum loading 
and in the direction of conservative loading. A total of 16 one -parameter variations was 
made from this nominal design in which two other levels of aspect ratio, solidity, swirl 
angle, exit Mach number, inlet flow/annulus area, radius ratio, flowpath shape, and 
number of stages were s^died. The results of the one-parameter variations are pre- 
sented in Table n as Configurations 4 through 17 and Configurations 24 and 25. 


♦Subsequent studies carried out under part three of the parametric screening study suggest 
that a more favorable flowpath than that shown in Figure 3 could have been found for the 
maximum loading cases, but the efficiency penalty would still be substantial. 


pm6EDlNG TA 


13 






14 


Figure 1 Conservative Loading Compressor Design, 14:1 Total-Pressure Ratio, 




15 


Figure 2 Nominal Loading Compressor Design, 14:1 Total-Pi essure Ratio, 
Configuration No. 3 





I 


Figure 3 Maximum Loading Compressor Design, 14:1 Total -Pressure Ratio, 
Configuration No. 2 



The results of var 5 ang the average aspect ratio (Figure 4) show that the best efficiency 
is obtained with average aspect ratios in the range from 1.3 to 2.0. Aspect ratios below 
and above this range lead to decreased efficiencies. As aspect ratio is increased, a 
corresponding increase in corrected tip speed is required to maintain stall margin. This 
raises Mach numbers, which causes the cascade losses to increase. At the same time, 
however, the higher blade speed leads to high-stagger blading which increases the passage 
aspect ratios* and decreases the end -wall losses. However, the improvement in end-wall 
losses at higher average aspect ratios is more than offset by declining cascade efficiencies. 
At low average aspect ratios, the increase in end-wall losses is sufficiently rapid as to 
more than offset a reduction in Mach -number -related losses at the lower tip speed. In 
addition, the low aspect ratio compressor is longer with increased weight and cost com- 
pared to the nominal design. All of these factors, when combined, indicate that the low 
aspect ratio compressor has a one percent higher direct operating cost and a two percent 
higher fuel usage than the nominal aspect ratio compressor. 

The solidity variation results shown in Figure 5 gest that near -nominal average 
solidities, ranging from 1.2 to 1.5, provide the best efficiency. As solidity is increased, 
the passage aspect ratio increases and end -wall losses are reduced. However, the 
resulting greater number of blades, with their associated wakes, decreases cascade 
efficiency. This occurs despite a reduction in tip speed. It appears that when the first 
stage rotor tip inlet relative Mach number is less than about 1.4, shock losses in the first 
stage have only a small effect on overall efficiency. For the remaining stages, cascade 
losses do not appear to be a strong function of tip speed. For solidities increasing from 
0.9 to 1.2, the end-wall losses decrease faster than the cascade losses increase, which 
resuHs in an overall efficiency improvement. For solidities larger than 1.5, the end- 
wall losses do not decrease so rapidly. Thus, the cascade losses increase faster than the 
end-wall losses decrease, which results in a decrease in overall efficiency. The reduced 
efficiency for both the low and high average solidity compressor designs is reflected in 
higher fuel usage and higher direct operating costs. 

Swirl angle variation results are shown in Figure 6. The low swirl (high reaction) case 
suffers in efficiency with no other clear beneht. This reduced efficiency results from the 
higher rotor Mach numbers at the low swirl angles. Efficiency and the economic param- 
eters appear to be best when average swirl angles from 9 to 20 degrees (reactions from 
0.7 to 0.5) are used. The corrected tip speed, rear rim speed, length of the compressor, 
and total number of airfoils remained essentially unchanged as the average swirl angle 
was varied. 

The results shown in Figure 7 indicate that a low value of compressor exit Mach number 
is desirable. As exit Mach number is increased from 0.28 to 0.4, the efficiency based on 
diffuser exit conditions continuously decreases. This results partly from higher diffuser 
losses. Also, as exit Mach number is increased, the blade height decreases, which 
increases end -wall losses by increasing the tip clearance /blade height ratio. Furthermore, 
the relatively high axial velocity /blade speed ratio associated with higher exit Mach 
numbers leads to somewhat reduced cascade efficiencies for the reai stages. Again, the 
consequences of lower efficiency are higher fuel usage and higher direct operating cost. 


♦ Passage aspe cl: ratio is defined as the blade height divided by the pitchline staggered 
spacing between blades. 






Figure 4 Variation in Direct Operating Cost, Fuel Usage, Tip Speed, and 
Compressor Efficiency with Aspect Ratio 




CORRECTED TIP SPEED 








Figure 5 


Variation in Direc t Operating Cost, Fuel Usage, Tip Speed and 
Comprei^scr Efficiency with Average Solidivy 

19 




CORRECTED TIP SPEED 






EXIT MACH NUMBER 


Operating Cost, Fuel Usage, Tip Sp6ed, and 
icy with Exit Mach Number 









The effects of variation in inlet corrected flow per luiit annulus area are presented in 
Figure 8 . These results clearly indicate that low inlet specific flow is desirable for the 
general class of designs considered. 

The results of varying the inlet radius ratio are shown in Figure 9 . Each design had a 
constant pitch radius type flowpath, but at a different radius level. The data show that 
high inlet radius ratio penalizes efficiency. The larger diameter of the high radius ratio 
design results in larger clearances and smaller blade heights, both of which increase end- 
wall losses. Cascade losses are not significantly reduced despite the reduction in tip 
speed that the higher radius ratio allows. The low inlet radius ratio design is longer, but 
has fewer airfoils, weighs less, and costs le^s than the nominal or high radius ratio 
designs . These factors, coupl^ with the higher efficiency, result in the lowest fuel usage 
and direct operating cost for the low inlet radius ratio design. 

The direct operating cost and fuel usage data in Figure 9 present an interesting result. 

For the high inlet radius ratio compressor, the rpm decreased to the extent that an 
excessive high pressure turbine loading was encountered. As a result, the turbine flow- 
path diameter had to be increased to reduce the loading to a reasonable level. This flow- 
path change increased the low pressure turbine diameter as well, and increased the low 
pressure turbine efficiency by 0. 7 point. This more than compensated for the reduced 
compressor efficiency, therefore decreasing fuel usage. The discontinuity in the curve in 
Figure 9 reflects this change in turbine flowpath diameter. However, the high radius ratio 
design had the largest weight and cost which caused an increase in the direct operating cost 
even though fuel usage decreased . 

The use of different exit radius ratios and flowpath shapes is explored in Figure 10. Con- 
figuration 16 has a constant tip diameter and an exit radius ratio of 0.957, while Config- 
uration 17 has a constant hub diameter and an exit radius ratio of 0.913. As the exit 
radius ratio is increased, the rpm and corrected tip speed are rapidly reduced because the 
work can be input at the larger average radius with a lower rpm. The cascade losses 
decrease as the wheel speed is reduced due to a reduction in shocK and Mach-number 
associated losses. At the same time, as the exit radius ratio is increased, the blade 
heights become smaller, making the tip -clearance /blade -height ratio larger which, in 
turn, increases end-wall losses. As shown in Figure 10, there is an efficiency improve- 
ment as the exit radius ratio is increased from 0.913 'o 0.934. This is due to the fact 
that the cascade losses decrease faster than the end -wall losses increase. For exit 
radius ratios above 0.934, the tip speed has become so low that no further improvement 
in cascade efficiency is realized, and the increasing end -wall losses dominate, thus 
reducing adiabatic efficiency. The low speed of the high exit radius ratio design required 
a turbine flowpath change similar to that needed for the high inlet radius ratio case de- 
scribed earlier. Resulting turbine efficiency changes are again responsible for the dis- 
continuities in fuel usage and direct operating cost. 

The effect of varying the number of stages in the compressor was also studied, and the 
results are shown in Figure 11. The interesting result is that the efficiency was not 
strongly affected by stage number, at least for the particular combination of stage param- 
eters investigated. With fewer stages, a higher speed is required, of course. The 
cascade losses are nearly constant, but the end -wall losses decrease due to increased 


22 



CORRECTED TIP SPEED 







CORRECTED TIP SPEED 


USAGE 


fj 




H 

W 

s 


2 


% 


w 


Pm 

<1 


1 


0 






Figure 9 Variation in Direct Operating Cost, Fuel Usage, Tip Speed and 
Compressor Efficiency with Inlet Radius Ratio 


24 


} 


! 









CORRECTED TIP SPEED 








blade stagger angles (higher passage aspect ratios) as the speed increases. With fewer 
number of stages, rotative speed increases, thus increasing high pressure turbine 
efficiency and counteracting the slight decrease in compressor efficiency. The result is 
a slight improvement in fuel usage for the eight -stage compressor compared to the nine- 
stage. However, for tlie 10 -stage compressor, the rotative speed has been significantly 
decreased, causing the high pressure turbine loading to increase and its efficiency to 
decrease. The reduced turbine and compressor efficiencies result in higher fuel usage 
for the 10-stage compressor compared with the nine-stage. The trend of the direct 
operating cost follows that of the fuel usage. It should also be noted that the rear rim 
speed of the eight -stage compressor is approaching the assumed mechanical limit cf 381 
m/sec (1250 ft/sec). 

In summary, the findings of the one -parameter variation studies show the following 
trends, although it is recognized that for a different class of compressors, such as turbo- 
jets, different values of the optimum parameters might result: 

1. Best efficiency is obtained when core compressors are designed with: 

(a) Medium average aspect ratios (1.3 -2,0) 

(b) Medium average solidities ( 1 . 2-1 . 5) 

(c) Medium -to-high reactions (0,5-0. 7) 

(d) Low exit Mach number (^0.28) 2 o 

(e) Low inlet flow /annulus area (171kg/sec-m (35 Ibm/sec fO ) 

(f) Low inlet radius ratio (i.e., minimum, practical value within physical and 
structural constraints) 

2. High blade speed does not penalize performance until front stago bp Mach number 
is greater than about 1.4. 

3. High rpm can increase turbine efficiency, often without reducing compressor 
efficiency, 

4. Fewer stages are less expensive, but not necessarily lighter, and need not 
involve an efficiency penity provided that tip speed does not become excessive. 

5. Medium -to-high rear radius ratio can be beneficial, provided that it helps main- 
tain the front stage relative tip Mach number below the level at which high shock 
losses are encountered. 

These findings were used to guide the selections of design parameters for two series of 
high efficiency compressors, one group having a pressure ratio of 14 and the other group 
having a pressure ratio of 23. These designs will be discussed in a later section. Further 
Studies of High Efficiency Compressors. 

Discussion of Other Screening Studies 

Maximum Life Configurations - Configurations 19a and 19b, the Maximum Life designs, 
were based on Configuration 17, the nine -stage, constant hub radius, nominal loading, 

14:1 pressure ratio design. Modifications to increase the rotor tip trailing edge thickness 


27 



and reduf'e blade speed were made in order to achieve longer blade erosion life. Config- 
uration 17 was chosen as a starting point because its rear rim speed was low. Lower 
aspect ratios and high solidities reduced the tip speed required for 19 percent stall margin. 
Use of lower aspect ratios also increased the trailing edge thickness for a given thickness - 
to-chord ratio. A reduced exit Mach number was also used to increase rotor blade 
heights and, for a given aspect ratio and trailing edge thickness-to-chord ratio, to increase 
the blade trailing edge thickness. High solidity enabled the blade passage aspect ratios 
to be kept relatively high as the chord was increased, thus avoiding high end -wall losses. 

In Configuration 19a only the rotors were given longer chords and high solidities, and in 
Configuration 19b the stator chords and solidities were increased as well to give a further 
reduction in speed. As a result, the relative blade life of the limiting stage was increased 
from 2.87 in Configuration 17 to 6.22 for Configuration 19b, an improvement of over 200 
percent. This increase was accomplished with a slight improvement in efficiency of 0.2 
point and only moderate increases in cost and weight, as seen in Table II. The study 
showed that compressor life can be significantly increased without adversely affecting 
performance by proper design of the limiting stages. 

Lightly Loaded First Stage Configuration - A nine-stage, 23:1 pressure ratio compressor 
was designed utilizing a front stage that was quite lightly loaded relative to its tip speed. 
This design. Configuration 23, was investigated in order to allow use of the double- 
oblique-shock type of rotor aerodynamic design reported in Reference 3 under NASA 
Contract NAS3-13498. This type of design could possibly result in a higher first stage 
efficiency than would be obtained if the first stage were loaded to its full capacity. Con- 
figuration 23 was obtained by zero-staging Configuration 24, a nominal loading eight -stage, 
14:1 pressure ratio compressor. This procedure consisted of selecting an operating con- 
dition at 97 percent corrected speed as ^e match point for the lightly loaded fan stage, 
scaling the fan so that its tip diameter was compatible with the flowpath of the existing 
eight-stage design, and cutting off the hub portion of the zero stage so that an inlet 
corrected flow of 46.86 kg/sec (103.3 Ibm/sec) was obtained. At this match condition, 
the corrected tip speed of the fan stage was 474 m/sec (1555 ft/sec), total pressure ratio 
was 1.64, and efficiency was estimated to be 0,85. The corrected flow and tip speed at 
the inlet to the rear block were approximately two percent lower than the original design 
values for this eight-stage compressor, so a new blade geometry definition, efficiency 
estimate, and stall margin check were made for these stages. The efficiency of the front 
stage reported in Reference 3 at the match point was 1 . 5 point better than predicted by the 
efficiency model, so the overall compressor efficiency was adjusted upward by 0.3 point 
to reflect this. The resulting overall adiabatic efficiency of 0.822, however, was not 
significantly better than expected from the trend of efficiency versus number of stages 
defined by the other 23:1 pressure ratio compressors studied. A summary of key design 
and performance parameters for Configuration 23 is contained in Table II. This table 
also allows a comparison with the other 23:1 pressure ratio configurations. This com- 
pressor is seen to be compact, light, and inexpensive; features which give the engine 
using it a low direct operating cost. The fuel usage, however, is not as good as the 
engine using the higher efficiency, conservative loa^ng, 14 -stage compressor (Config- 
uration 20). 


28 



FURTHER STUDIES OF HIGH EFFICIENCY COMPRESSORS 


The results of the parametric screening studies reported in the previous section were 
used to define two families of compressors with high efficiency potential: the Configura- 
tion 26 group with a total pressure ratio of 23; and, the Configuration 18 group with a 
total pressure ratio of 14. The approach used was to follow the trends seen in the one- 
parameter variations that led to higher efficiencies. These compressors, therefore, have 
low inlet and exit axial Mach numbers, medium blade aspect ratios and solidities, medium - 
to-high reaction ratios, low inlet radius ratios, and medium exit radius ratios. Aerody- 
namic design details of these configurations are presented in the following sections. 

P/P = 23 Compressor Configurations 


Aerodynamic design and performance parameters for the 23:1 pressure ratio high- 
efficiency compressors are presented in Table III. A number of different compressor 
designs were evaluated in an attempt to fine-tune the selection of parameters to yield the 
best combination of high efficiency, low fuel usage, and low direct operating cost. The 
number of stages ranged from nine to 14, and the corrected tip speeds required ranged 
from 451 m/sec (1480 ft/sec) to 541 m/sec (1775 ft/sec). The diffuser exit adiabatic 
efficiency for Configuration 26a of 0.858 shown in Table III was 1.4 points better than the 
efficiency of Configuration 20, the best of the initial 23:1 pressure ratio configurations. 
The efficiency values listed in Table III and shown in Figure 12 decreased appreciably as 
the number of stages was reduced, mainly because tip speeds became very high and high 
rotor shock losses were encountered, ""nd-wall losses remained about constant, and the 
overall adiabatic efficiency at the diffuser exit followed the trend of the freestream air- 
foil cascade efficiency. 

Configurations 26b2, 26e, and 26d2 were evaluated in an effort to regain the efficiency that 
was lost as the number of stages was reduced. These configurations had the same number 
of stages used previously, but utilized flowpaths having larger exit radii which allowed the 
use of considerably lower tip speeds. The results presented in Table III and in Figure 12 
show that an efficiency improvement of about 0.4 -0.5 point was obtained in each case. 

This efficiency improvement came about because the significantly lower tip speeds reduced 
the Mach -number -associated losses. This effect overshadowed the end -wall loss in- 
creases caused by the larger tip-clearance-to-blade-height ratios that resulted from the 
larger exit radius ratios. 

In order to further improve efficiency, the previous best nine-stage design, Configuration 
26d2, was modified by decreasing the inlet radius ratio from 0.546 to 0.502, thus creat- 
ing Configuration 26d5. This modification reduced the inlet tip diameter, and although the 
required rpm increased, the corrected inlet tip speed was reduced. The resulting reduc- 
tion in Mach number losses brought about a 0.3 point improvement in adiabatic efficiency 
compared to Configuration 26d2. 

Compressor mechanical design and aircraft/engine system economic analysis data are 
presented in Figures 13 and 14, respectively. As the number of stages is reduced, the 
rear rim speed increases, with the higher exit radius ratio designs having higher rear 


29 



f 

\ 

i 

I 


Table III* Aerodynamic Summary of Configuration 26 Designs, 
Maximum Efficiency, 23:1 Pressure Ratio 


Configuration Number 
Ntimher of Stages 

26a 

14 

26b 

11 

26c 

10 

26d 

9 

26b2 

11 

26e 

10 

26d2 

9 

26d5 

9 

Corrected Inlet Tip Speed, m/sec 
(fps) 

451 

(1480) 

500 

(1642) 

522 

(1712) 

541 

(1775) 

. 457 
(1500) 

472 

(1547) 

491 

(1610) 

480 

(1575) 

Physical Rear Hub Speed, m/sec 
(fps) 

301 

(987) 

340 

(1116) 

358 

(U74) 

382 

(1252) 

354 

(1163) 

366 

(1200) 

388 

(1273) 

393 

(1290) 

Physical Speed, rpm 

13,155 

14,596 

15.218 

15,767 

13,333 

13,751 

14,301 

14,436 

Inlet Radius Ratio 

0.488 

0 488 

0.488 

0.546 

0.488 

0.488 

0.546 

0.502 

Inlet Flow/Annulub Area, 
Rgm/sec-m^ (Ib/sec-ft^) 

171 

(35) 

171 

(35) 

171 

(3Si 

186 

(38) 

171 

(35) 

171 

(35) 

186 

(38) 

186 

(38) 

Inlet Tip Diameter, m 
(in. ) 

0.676 

(26.6) 

0.676 

(26,6) 

0.676 

(26.6) 

0,677 

(26.6) 

0.676 

(26.6) 

0.676 

(26.6) 

0.677 

(26.6) 

0.656 

(25.8) 

Kxit Radius Ratio 

0.910 

0.913 

0.915 

0.915 

0.930 

0.931 

0.934 

0.934 

Exit Tip Diameter, m 
(in.) 

0.480 

(18.9) 

0,488 

(19.2) 

0.491 

09.3) 

0.502 

(19.8) 

0.546 

(21.5) 

0.545 

(21.5) 

0.555 

(21.8) 

0.556 

(21.9) 

Length to Diffuser Exit, m 
(in.) 

1 03 
(40.7) 

0.773 

(30.4) 

0.743 

(29.2) 

0,987 

(23.1) 

0.767 

(30.2) 

0.717 

(29.2) 

0.603 

(23.7) 

0.630 

(24.8) 

Number of Airfoils 

2237 

178] 

1677 

1651 

2087 

1925 

1915 

1839 

Average Aspect Ratio 

2.0 

1.85 

1.78 

1.79 

1.85 

1.70 

1.70 

1.70 

Average Solidity 

1.30 

1 . 14 

1.34 

) .40 

‘1.3^ 

1.46 

1.49 

1.49 

Average Swirl, Degrees 

17.1 

15.3 

14.7 

13.7 

15.3 

14.7 

13.7 

13.7 

Average Reaction 

0.692 

0.734 

0.748 

0.761 

0.732 

0.736 

0,749 

0,744 

St.Ul Margin, % 

17 

17 

18 

20 

17 

18 

20 

18 

Adiabatic Efficiency at 
Difluser Exit 

0.858 

0.852 

0.849 

0.841 

0.856 

0.853 

0.846 

0.849 

Compressor Weight, kg 
(lb) 

294 

(648) 

257 

(566) 

253 

(558) 

242 

(534) 

271 

(598) 

269 
( 594) 

258 

(568) 

252 

(556) 

\Compressor Price, of STEDLEC 

Compressor Price 

17.0 

21.2 

16.6 

K.2 

2 / . 3 

22.0 

1 1.2 

12.1 

Minimum UlmJc Life, hr 

6000 

1 700 

1300 

<1000 

3500 

1900 

•1600 

•1600 

\Engine Price*, of STEDLEC 

Total Engine Price 

7.1 

l.H 

•1.5 

3.7 

5.5 

5.0 

1.0 

3.9 

\Fucl Usage*, V 

-2.06 

-2.11 

- 1 , 90 

-1.5;: 

-2.11 

-2 . 26 

-2.07 

-2.27 

ADirect Operating Cost*, 

0.92 

0.36 

0 . 1 K 

0.31 

0.33 

0. U 

0.23 

O. 1 1 


*TranHcont liu'iita l trijet aircraft mlsslun, 

t 


( 

i 


L 

\ 

i 


30 




Figure 12 Adiabatic Eificiency and Corrected Tip Speed Versus Number o 
Stages for 23:1 Total-Pressure Compressors, Configuration 26 











^ INSTALLED 
ENGINE WTIIGHT 








rim speeds despite their lower rotative speeds. It should be noted that the originally 
assumed mechanical limit of 381 m/sec (1250 ft/sec) for the rear rim speed was relaxed 
for Configuration 26d5, for which the value of this parameter is 393 m/sec (1290 ft /sec). 
The compressor weight, length, and number of airfoils all decrease as the number of 
stages is reduced, which has the effect ot reducing the engine cost. The overall engine 
installed weight remains relatively constant even though compressor weights are reduced 
by the use of fewer stages, because with fewer stages the rotati'^e speeds have increased, 
requiring heavier turbines. The trend of fuel usage at a giver, exit radius ratio followed 
the trend of compressor efficiency, as expected, and the lighter engine weight and im- 
proved efficiency made possible by the use of a high rear radius ratio flowpath also 
reduced fuel consumption. The 11 -stage design, Configuration 26b2, had the lowest fuel 
usage. Direct operating costs reflect both fuel usage and engine cost, and these combine 
to slightly favor the compressors with few'er stages. 

P/P = 14 Compressor Configurations 

Aerodynamic design parameters for the 14:1 pressure ratio high-efficiency compressors 
are presented in Table IV. As for the 23:1 pressure ratio designs, several designs dif- 
fering in number of stages were evaluated in an attempt to select parameters that yielded 
high efficiency, low fuel usage, and low direct operating cost. The results of this effort 
are summarized in Table IV. The efficiencies are as high as 0.869, one point better than 
any of the 14:1 pressure ratio configurations that were identified in the initial parametric 
studies (Table II), and are within 0.5 point of the same level with as few as eight stages. 

As the number of stages is reduced, the corrected tip speed increases and the shock 
losses and Mach-number-associated losses increase. For the 12-stage> 10-stage, and 
nine-stage compressors, increases in cascade loss are offset by decreases in end-wall 
loss, resulting in nearly constant adiabatic efficiency. As the number of stages is re- 
duced below nine, the increasing cascade losses caused by the increased tip speed 
dominate and the adiabatic efficiency decreases. 

Mechanical design data and economic analysis data are also presented in Table IV. As the 
number of stages was reduced, the rear rim speed increased, and rim speeds above 381 
m/sec (1250 ft/sec) were required for some configurations. No engine system data were 
given for the seven-stage compressor, Configuration 18f, because its extremely high 
rotative speed was judged to result in unacceptably high turbine stresses. The compressor 
length, number of airfoils, weight (except for Configuration 18f), and cost decreased as 
the number of stages was decreased, resulting in a decrease in engine cost. Overall 
engine weight varied less than compressor weight, partly because turbine weight increased 
with the higher speeds which were required as the number of compressor stages was re- 
duced. Fuel usage was found to increase for fewer than 10 stages, due to declining com- 
pressor efficiency and increasing turbine cooling flows. The trend of direct operating 
cost shown in Table iV is due to both the trends of fuel usage and engine cost, resulting in 
minimum direct operating cost for an engine having a nine-stage core compressor. 

Effect of State-of-the-Art Assumptions 

The sensitivity of the predicted efficiency of the high-efficiency 23:1 pressure ratio de- 
signs to variations in the assumed level of aerodynamic and mechanical design technology 
was evaluated. In Figure 15, the upper curve of compressor efficiency versus number of 



Table IV. Aerodynami''. Summary of Configuration 18 Designs 
Maximum Efficiency, 14:1 Pressure Ratio 


Configuration Number 
Number of Stages 

18a 

12 

18b 

10 

18c 

9 

18d 

8 

CO 

Corrected Inlet Tip Speed, m/sec 
(fps) 

387 

(1270) 

415 

(1360) 

437 

(1435) 

460 

(1510) 

489 

(1604) 

Physical Rear Hub Speed, ra/sec 
(fps)- 

314 

(1029) 

348 

(1143) 

374 

(1227) 

398 

(1306) 

416 

(1364) 

Physical Speed, rpm 

14,145 

15,148 

15,983 

16,819 

17,866* 

Inlet Radius Ratio 

0.566 

0.566 

0.566 

0.566 

0.566 

Inlet Annulus Area, 

kg sec-m** (Ibm sec-ft'^) 

171 

(35) 

171 

(35) 

171 

(35) 

171 

(35) 

171 

(35) 

Inlet Tip Diameter, m 
(in. ) 

0.583 

(23.0) 

0.583 

(23.0) 

0.583 

(23.0) 

0.583 

(23.0) 

0.583 

(23.0) 

Hxit Radius Ratio 

0.908 

0.907 

0.911 

0.913 

0.916 

Exit Tip Diameter, m 
(in.) 

0.465 

(18.3) 

0.485 

(19.1) 

0.491 

(19.3) 

0.495 

(19.5) 

0.485 

(19.1) 

Length to Diffuser Exit, m 
(in.) 

0. 737 
(29.0) 

0.604 

(23.8) 

0.:36 

(21.1) 

0.470 

(18.5) 

0.443 

(17.4) 

Number of Airfoils 

1899 

1780 

1626 

1486 

1211 

Average Aspect Ratio 

2.09 

2.00 

1.93 

1.88 

1.71 

Average Solidity 

1.20 

1.34 

1.37 

1.40 

1.44 

Average Swirl, Degrets 

16. n 

14.4 

14.0 

14.1 

13.3 

.Average Reaction 

0. 705 

0.733 

0.747 

0.756 

0,762 

Stall .Margin, ^ 

17 

17 

20 

21 

17 

Adiabatic Efficiency 
at Diffuser Exit 

0.868 

0.869 

0.868 

0.863 

0.858 

Compressor Weiglit, kg 
(lb) 

224 

(493) 

208 

( 458 ) 

198 
(4 36) 

191 

(421) 

197 

( 434 ) 

\Conipi*cssor Price, of STEDLEC 

Compressor Price 

13.1 

1.6 

- 3,2 

- 10.9 

- 12.9 

Minimum Dladc Life, lir 

6300 

5300 

■1900 

3900 

1200 

A Engine Pricet, ^ of STEDLEC 
Total Engine Price 

1.3 

0 

- 0.5 

-1. 1 

— 

\Euel Csaget, i 

- 0,73 

-1.01 

- 0.95 

- 0 . 60 

— 

\nire('t Opera tiiUi Cosi^, 

-0. 17 

- 0.71 

- 0 . 7 « 

-0.6h 

.... 


♦Tnaccopt able because of turbine stress 
t’l !\\nscont i nenta 1 trijet aircraft mission 

35 


V 








sfiijia IV ADNaiDiiija dixvhviqv 


Figure 15 Effect of Technology Level Assumptions on Efficiency of 23: 
Total-Pressure Ratio Compressors 




stages, labeled ’’advanced clearance, advanced aero”, gives the efficiency for the advanced 
level of technology assumed throughout this study. The consequences of not reducing end- 
wall losses by the assumed 15 percent and not achieving the assumed improvement in rear 
stage rotor blade surface finishes, combined with a 15 percent increase in shock losses 
above the assumed ’’best current experiences” level, are shown in the middle curve 
labeled '’advanced clearance, current aero” to result in about a one point reduction in 
predicted efficiency. A further reduction in efficiency of about one point is predicted if 
clearances cannot be reduced below current levels* ,It should be noted that all four com- 
pressors shown in the figure were affected nearly equally by changes in the level of 
technology assumed. It was thus concluded that, for this type of engine/aircraft system, 
the selection of the optimum number of core compressor stages should not be particularly 
dependent upon achieving specific advances in either aerodynamic or mechanical design 
technology. Although not shown in the figure, use of current technology clearances would 
result in approximately four to five points less stall margin than if advancement in clear- 
ance control were achieved. 

Additional studies were conducted to determine if variations in assumed technology level 
would affect the choice of compressor rear radius ratio. The nine-stage 23:1 pressure 
ratio compressor was employed for this investigation. As shown by the upper curve in 
Figure 16, predicted efficiency is highest at an exit radius ratio of about 0.95 if advanced 
technology clearances and aerodynamics are assumed as was done throughout this study. 
This trend is caused by the large reduction in shock losses that accompanies the lower 
rotative speeds which result from the use of high exit radii. If current technology clear- 
ances are assumed, as in the middle curve of Figure 16, the larger rear stage clearance- 
to-blade height ratios that result from use of high exit radius ratios reduce the most 
favorable exit radius ratio to approximately 0.935, This value of exit radius ratio still 
appears to be about optimum if shock losses aie assumed to be 15 percent higher than 
used throughout the study, as shown by the lower curve in Figure 16. It was concluded, 
therefore, that the exit radius ratio should be chosen to be in the 0.930 -0.935 range in 
order to keep front stage tip speeds from becoming so high that overall efficiency would 
be compromised if low shock losses could not be achieved. Yet this radius ratio is not 
so high that failure to achieve significant clearance reductions below current levels might 
severely increase rear stage losses and reduce overall efficiency potential. 

DETAILED STUDY OF THREE SELECTED COMPRESSORS 


Three of the most promising compressors identified in the final screening studies were 
selected for further, more detailed, aerodynamic and mechanical design studies. These 
three configurations were the 11-stage and nine-stage 23:1 pressure ratio designs and the 
nine -stage 14:1 pressure ratio design, Configurations 26b2, 26d5, and 18c, respectively. 

The two 23:1 pressure ratio designs were selected because engines employing these com- 
pressors had the lowest fuel usage, and because it was realized that a good 14:1 pressure 
ratio compressor could be derived from each of these designs by removal of the first 
stage. The 11-stage and nine-stage designs covered the range of stage numbers where 
overall unboosted engine system performance was best. Since the optimum number of 
stages was not clearly evident from the parametric study results, it was desired to study 


37 





o 


o 

ITJ 

o 



CO 

CO 

W 


X 

X 

X 

X 

X 

X 


iixa aasnjdia xv AONaioiaaa oixvaviov aossaadwoo aaoo 


16 C inpressor Efficiency as a Function of Exit Radius Ratio 
Assumed Level of Technology for 9-Stage 23:1 Total-Pressi 
Ratio Designs 




two cases that differed in this parameter to see if some subtle difference, perhaps not 
adequately identified in the parametric study, might be uncovered. For example, there 
was concern that the higher rotative speed of the nine -stage design might lead to bearing 
problems or to blade and disk stress problems. Part-speed stall margin and acceleration 
time could also be of concern for this configuration. On the other hand, the greater length 
of the 11-stpge design might lead to bearing span, rotor deflection, or system vibration 
problems . 

The nine-stage 14:1 pressure ratio compressor was selected for further study because the 
engine based upon this design had the lowest predicted direct operating cost of any con- 
figuration studied. It also had a fairly high rear rim speed and a very high rotative speed, 
thus providing mechanical design challenges in the areas of compressor disk stresses and 
turbine stresses and cooling. 

The major aerodynamic design parameters for the three selected compressors, as well 
as a comparison of these parameters with the other configurations identified in the para- 
metric studies, are presented in Table II. Cross-sections of the three designs, indica- 
ting the aerodynamic flowpaths and major mechanical design features of these configura- 
tions, are presented in a later section, Detailed Design Study. 

The detailed aerodynamic design studies consisted primarily of an analysis of vector 
diagram parameters and aerodynamic loading levels in order to determine if any severe 
problems existed that might m^e it unlikely that the predicted stall margin or efficiency 
could be achieved. In addition to studies of design point operation, a combination of 
analytical and semi-empirical methods was employed to generate compressor perform- 
ance maps that would represent estimated part -speed flow, efficiency, and stall line 
magnitudes. 

Axisymmetric Flow Calculations 


Calculation of circumferential average values of vector diagram parameters and fluid 
properties along nine design stream surfaces was performed using the General Electric 
Circumferential Average Flow Determination computer program. Flowpath annulus 
geoiaeUy, blade geometry, inlet temperature, and pressure at the engine maximum 
cruise thrust sizing point, flows, and rotative speed were input for the three selected 
configurations. These inputs were based upon data generated during the parametric 
screening studies using the pitchline analysis procedure. Stagewise distributions of 
average pressure ratio, stator exit flow angle, efficiency, and wall boundary layer block- 
age were also based on results from the screening study. 

Nonconstant radial distributions of loss coefficient, stator exit flow angle, and total 
pressure were input so as to simulate the general characteristics of profiles actually 
measured in high speed compressors. This design approach gives an indication of the 
severity of blade end aerodynamic loading and incidence angles that will exist in the com- 
pressor, and allows the designer to select appropriate blade sections for these conditions. 

After the axisymmetric flow analysis was completed, preliminary values of incidence 
angles, deviation angles, and camber and stagger angles were determined for each blade 
and vane row using General Electric correlations. In general, the analysis of each 



configuration was carried sufficiently far to determine if major aerodynamic problem 
areas existed, but not so far as to represent a finished design. 

Selected results of this analysis are presented in Figures 17 and 18. Mach numbers are 
seen to be moderate after the first one or two stages. The pitchline diffusion factor plots 
show that all three designs have highly loaded rear stators. Although these loadings are 
somewhat larger than usual and may require further design refinement in axial ve- 
locity ratios, swirl angles, and work input distributions, they are not believed to be 
so excessive as to present an unsolvable problem in achieving stall margin and effi- 
ciency goals. 

De-Staged 23:1 Compressor 


By examining the stagewise pressure distributions of the two 23:1 pressure ratio com- 
pressors, it was found that both designs had the potential of being useful as a building- 
block type compressor. For example, if the first stage of Configuration 26d5 were re- 
moved, an eight -stage, 13.9:1 pressure ratio compressor would result. If the last stage 
were removed, an eight -stage 18.3:1 pressure ratio compressor would be obtained. 
Similarly, the 11-stage design. Configuration 26b2, could form the basis of a 10-stage 
compressor with pressure ratios of either 14.5:1 or 18.9:1 by removing the first or last 
stage, respectively. 

The results presented in Table V show that there would be little or no sacrifice in efficiency 
with the de-staged compressors when compared with the efficiency of Configuration 18c, 
the high efficiency 14:1 pressure ratio compressor. Another significant conclusion was 
that even though the de-staged compressors are heavier, cost more, and have more air- 
foils, they are still competitive with Configuration 18c on a direct operating cost and fuel 
usage basis. In fact, the direct operating cost and fuel usage of the eight -stage com- 
pressor and of Configuration 18c are nearly identical. This results because the lower 
speed of the de-staged compressor allows the turbine weight and cooling flow to be de- 
creased without a significant sacrifice in turbine efficiency. Based on these results, it 
was concluded that a de-staged 23:1 pressure ratio compressor was competitive with the 
14:1 pressure ratio compressor previously selected as an optimum design. 

Off -Design Analysis 

In addition to determining the estimated design point performance, a combination of semi- 
empirical and analytical methods was employed to generate performance maps for the 
23:1 and 14:1 pressure ratio compressors. Stage characteristics were constructed con- 
sistent with stage design points and the estimated stall points. These stage characteristics 
were then employed in a stage stacking computer program capable of accounting for the 
effects of bleed and variable stators, and the performance was determined at several 
part-speed conditions for various bleed flows and stator schedules. The results of the 
study showed that sufficient part speed stall margin could be obtained with reasonable 
stator schedules and without the use of additional bleed flow, other than that normally 
needed for turbine cooling and aircraft cabin pressurization or air conditioning. It is 
expected that additional bleed flow will be required for engine starting. 


40 









Table V 


Comparison of De-staged 23:1 Compressors with Configuration 18c 


Conf inuralion 

Stages 2 - 9 of 
Configuration 
26d5 

18c 

Stages 2 - 11 of 
Configuration 
26b2 

Pressure Ratio 

13.9 

14.00 

14.5 

Corrected Flo*, k^/sec 

30.5 

31.0 

31.7 

( Ibm 'sec) 

(67.3) 

(68.4) 

(69.9) 

Number of Stages 

8 

9 

10 

Inlet Corrected Tip Speed, m/sec 

425 

137 

406 

( f t /see) 

(1395) 

( 1435) 

(1331) 

Inlet Physical Tip Speed, m/ sec 

•176 

488 

453 

(It sec) 

(1562) 

( 1601) 

(1488) 

Kxit Physical Hub Speed, m sec 

393 

371 

354 

(ft sec) 

(1290) 

(1227) 

(U63) 

Plus leal Speed, rpm 

M, 136 

15. 983 

13,333 

Inlet Radius Ratio 

0.656 

0.566 

0 . 624 

Inlet Corrected Flo* per Annulus Area. 
kK'sec tii^ 

171.3 

170.9 

158.2 

(Ibni'sec tt^) 

(3:1.7) 

(35.0) 

(32.4) 

Inlet Tip lU.^meter, m 

0 . 625b 

« ) . 5832 

0.6547 

(in) 

(21.63) 

( 22.96) 

(23. 12) 

Kxii Radius Ratit* 

0.936 

0.911 

0 . 930 

FxiL Mach Number 

0.26 

0 . 26 

0.26 

Kxit Tip Oi.ur.eter, m 

0.55MH 

0. 1905 

0.5 158 

( 1 n) 

(22.00) 

(19.31) 

(21 . 19) 

ben^th. Rotor 1 Inlet to ()utl<'t Gunie Vane, 

III 

0. n »2 

0 .51.18 

0. 5860 

( in) 

(17. 19) 

(20.23) 

(23.07) 

Number ol blades and Vanes 

1770 

1 b26 

2016 

Vdiabatic Kt t icieticv at Outlet Guide Vane 

0. h71 

0,875 

0.876 

Adiabatic t.lficieic\ at liitfuser Kxit 

0 . 861 

0.868 

0.869 

Compressor Weight, kg 

217 

i98 

235 

Ob) 

(179) 

( 136) 

(517) 

\Coinpressor Price, ‘r c»f STKDbKC 

-1.2 

-5.2 

12,7 

Compressor Pr\« e 
\l)l rei t tiperating Cost*, 'f 

-0 . 75 

-0.78 

-0 . 51 

\Kuel Csage*, 

-0.96 

-0. 9s5 

-0 . 92 


*1 raimvont inontul Irijet airciaft tai-tors rulatlvu to STFIM Kr baviMitu*. 


43 


A 23:1 pressure ratio compressor performance map was constructed using the stage 
stacking results, combined with other General Electric experience, and is presented in 
Figure 19. It is seen that relatively low flows are delivered at part speeds. This is a 
desirable feature because it leads to high part-speed stall margin and reduced accelera- 
tion time from idle airflow to full airflow. The 14:1 compressor performance map is 
similar in character. In general, these performance maps indicate that sufficient part- 
speed stall margin and good part-speed efficiency are obtainable. 

RECOMMENDED CONFIGURATION 


The configuration for the Advanced Multistage Axial- Flow Compressor recommended as 
a result of this study program is a 10-stage compressor with a pressure ratio of 23:1. 

This compressor is a i efinement of Configuration 2Pe that was developed during the 
screening studies. The two key issues that arose in selecting the recommended config- 
uration were to determine which pressure ratio should be selected and how many stages 
should be selected . 

The 14:1 pressure ratio compressors were driven by single-stage high pressure turbines 
and required booster stages to achieve the overall cycle pressure ratio. The 23:1 pres- 
sure ratio compressors had two-stage high pressure turbines and were unboosted. 

Results of the detailed design study showed smaller engine system economic differences 
between engine types than predicted in the parametric screening study. Engines using 
the 23:1 pressure ratio compressors had moderate fuel savings over those using the 14:1 
compressors, although they had a slight direct operating cost penalty. In summary, 
there was no clear-cut advantage for either the boosted or the unboosted configuration 
based on the engine studies. The 23:1 pressure ratio compr^^ssor was selected primarily 
because it will provide technology directly applicable to either type of engine. If de -staged, 
as previously discussed, this compressor would form a good 14:1 pressure ratio com- 
pressor for a boosted engine and could be achieved without the extensive additional devel- 
opment that would be anticipated if the high pressure system pressure ratio were increased 
by adding a front stage. 

The second issue to be decided was the number of stages to be recommended for the 23:1 
pressure ratio compressor. In selecting the number of stages, certain key results of the 
detailed design studies shewing the variation of performance and mechanical design 
factors with stage number were examined. Table VI lists comparative data for the nine- 
stage and 11 -stage compressors studied in detail, and also presents estimated data for a 
lower speed, more refined version of the 14 -stage Configuration 26a design. It can be 
nc^ed in Table VI that the 14 -stage compressor has better efficiency and acceleration 
time, but higher direct operating cost and fuel usage. Consequently, there is little benefit 
in selecting more than 11 stages. It is believed that a 10 -stage compressor, when com- 
pared to the nine-stage, has the potential of achieving an overall core engine weight saving 
with better efficiency, faster acceleration time, and more acceptable development risk, 
all with little or no economic disadvantages. Selection of a 10-stage compressor, there- 
fore, represents a reasonable compromise between the nine-stage and the 11 -stage 23:1 
pressure ratio compressors. 

The flowpath of the recommended 10-stage compressor is shown in Figure 20. Compar- 
isons of aerodynamic, mechanical, and economic parameters for the 9-stage, 10-stage, 
and 11 -stage 23:1 pressure ratio designs are given in Table VII and reflect final weight 
and engine performance data at the completion of the detailed design studies. Final data 
for the nine-stage 14:1 pressure ratio design Configuration 18c is also given in the table. 







I 


Table VX . Detailed Design Study Results for 23:1 Total-Pressure 
Ratio Compressors 


Advanced Technology Efficiency, % 

Current Technology Efficiency, ?o 

Physical Speed, rpm 

Turbine Efficiency, A points 

Rear Rim Speed, m/sec 
(ft/sec) 

Physical Rotor 1 Tip Speed, m/sec 
(ft/sec) 

Relative Accel Time 

Compressor Weight, kg 
(lb) 

Turbine Weight, kg 
(lb) 

AWeight (Compressor & Turbine) 

for 5% rpm Growth, kg 

(lb) 

Total Weight After Growth, kg 
(lb) 

ACompressor Price, % of STEDLEC 
Relative to Base Compressor Price 

ADirect Operating Cost, *7o 

AFuel Usage, *% 


9-Stg 

H-Stg 

14-Stg 

84.9 

85.6 

85.8 

82.7 

83.4 

83.9 

14,425 

13,336 

12,178 

+ 3.1 

+ 3.0 

+ 3.0 

393 

355 

308 

(1290) 

(1164) 

(10101 

496 

472 

431 

(1626) 

(1549) 

(1415) 

1.19 

1.00 

~0.9 

275 

294 

~338 

(606) 

(648) 

~(745) 

384 

343 

~316 

(847) 

(757) 

--(ege) 

52 

35 

-32 

(114) 

(78) 

-(70) 

711 

673 

-685 

(1567) 

(1483) 

-(151): 

12.2 

27.4 

-47 

-G.40 

-0.18 

-0.45 

-1.93 

-1.92 

— 1.3 


♦Transcontinental triJet aircraft mission factors relative to STEDLEC baseline 


46 




Table VII, Comparison of Recommended 10-Stage AMAC Compressor With 
Final Data for Configurations 18c, 26b2 and 26d5. 


Configuration Number 

18c 

26b2 

AMAC 

26d5 

Total Pressure Ratio 

14 

23 

23 

23 

Number of Ntages 

9 

11 

10 

9 

Corrected Inlet Tip Speed, m/sec 
(f ps) 

437 

(1435) 

457 

(15C0) 

469 

(1540) 

480 

(1575) 

Physical Inlet Tip Speed, n/sec 
(fps) 

488 

(1601) 

472 

(1349) 

485 

(1590) 

496 

(1626) 

Physical Rear Hub Speed, m/sec 
(fps) 

374 

(1227) 

354 

(1163) 

370 

(1214) 

394 

(1292) 

Physical Speed, rpn 

15,983 

13,336 

13,900 

14,425 

Inlet Radius Ratio 

0.566 

0.486 

0.496 

0.503 

Inlet Specific Flow, kg/sec-m^ 
(Ib/sec-f t^) 

171 

(35) 

171 

(35) 

178 

(36.5) 

186 

(38) 

Inlet Tip Diameter, m. 
(in.) 

0.5832 

(22.96) 

0.6761 

(26.62) 

0.6660 

(26.22) 

0.6563 

(25.84) 

Exit Radius Ratio 

0.908 

0.925 

0.930 

0.932 

Exit Macfi Number 

0.26 

0.26 

0.26 

0.26 

Exit Tip Diameter, m 
(in, ) 

0.4912 

(19.34) 

0.5466 

(21.52) 

0.545O 

(21.48) 

0.5588 

(22.00) 

Len'»rh to OCV Exit, m 
(in. ) 

0.5169 

(20.35) 

0.7592 

(29.89) 

0.6614 

(26.04) 

0.5982 

(23.55) 

Length to Diffu.ser Exit, m 
(in.) 

0. 5367 
(21.21) 

0.7699 

(30.81) 

0.6900 

(27.16) 

0.6317 

(24.87) 

Number tf Airfoils 

1637 

2035 

1959 

1810 

Average Aspeci Ratio 

1.87 

1.80 

1.72 

1,6d 

Average S«»lidiiy 

1.39 

1.39 

1.40 

1.48 

Average Swirl, Degree 

19.5 

19.4 

20.4 

19.2 

Average Reaction 

0.714 

0.715 

0,695 

0.701 

Stall Margin, X 

20 

17 

18 

18 

Adiabatic Efficiency at OGV Exit 

0.875 

0.86) 

0.860 

0,856 

Adiabatic Efficiency at Diffuser 
Exit 

0.868 

0.856 

0.853 

0,849 

Compressor Weight, kg 
(lb) 

225 

(496) 

294 

(648) 

269 

(592) 

275 

(606) 

ACompressor Price, 't of S’’ EDLEC 
Compre.ssor I’rice 

-5. 1 

27.2 

’H.O 

12,2 

A Engine Price*, '< of STEDLEC 
Total Engine Price 

-0.5 

5 . 5 

2.7 

3,9 

\ Fuel Usage, T* 

-1.02 

-1.02 

-1.99 

-1.93 

A Direct Operating Cost, '"r* 

-1.02 

-0. 18 

-0.35 

-0.40 


*Tran.scuntlncntal trljot aircraft missluns factors r^*latlvo to .STEDLEC 
l)as<$l Ino 


48 



MECHANICAL DESIGN STUDIES 


Compressor mechanical design studies paralleled the aerodynamic design effort. An 
estimate was made of 1985 time period state of the art in the areas of clearance control, 
blade erosion resistance, achievable blade surface finish, and in the availability of 
materials and allowable stresses for blades, disks, and static structures. These assump- 
tions were incorporated into preliminary design techniques used during the parametric 
screening study to estimate compressor weight, manufacturing cost, blade erosion life, 
and resulting blade replacement costs. Based upon the data generated in the parametric 
study, three compressor configurations were selected for further detailed mechanical 
study. This phase of the program was intended to determine if there were any structural 
problems that might lead to significant chaises in the conflguration, thereby affecting the 
efficiency, weight, or cost of the compressor. It also served to identify any problems 
which would require a major change in the engine layout, such as an additional frame at 
the compressor discharge. 

COMPRESSOR MECHANICAL DESIGN TECHNOLOGY - 1985 STATE OF THE ART 


Compressor mechanical design and development efforts at the General Electric Company 
have resulted in mechanical design innovations and techniques that have improved overall 
aerothermodynamic performance characteristics and provided significant advancements in 
the state of the art. The innovations included introduction of bore entry cooled compressor 
rotors, high strength low thermal mass disks, thermally shielded casing structures, low 
hysteresis variable stator mechanisms, removable casing liners with abradable rub coat- 
ings, and low loss, low heat generating seals . These innovations have provided improved 
clearance control and compressor efficiency and permitted higher rim speeds, higher 
stage pressure ratios, and higher compressor exit temperatures. 

The 1985 state-of-the-art mechanical technology ground rules established for this study 
were based on application of these innovations, plus expected advancements in materials, 
manufacturing, and similar design disciplines. The following topics were specifically 
addressed: cooling and clearance control; blade surface finish; blade erosion; rear rim 
speed; inlet radius ratio; and system vibration as it relates to bearing placement. 

Clearance Control 


Realization of the full flow, pressure rise, and efficiency potential of any compressor is 
dependent upon maintaining small operating radial tip clearances. Recent experience at 
Genera' Electric has indicated that significant future progress in compressor development, 
especially for subsonic turbofan applications, can be achieved by using advanced clearance 


49 













control features. These features include use of thermally shielded casing structures, 
cooled outer casing supports, use of advanced materials that would permit better matching 
of rotor and stator thermal growths, and use of internally cooled rotors. Analysis 
indicated that use of such clearance control techniques in the 1985 time period assumed 
for this study would enable compressors to operate with clearances approximately 25% 
smaller than those in current designs of the same type . 

A tip clearance model was used to establish a base level of minimum running clearance -to- 
diameter for each stage in the compressor at the sea level takeoff operating condition 
where compressor physical speed, inlet temperature, and inlet pressure were highest. 
This minimum clearance was then adjusted to the altitude cruise operating condition where 
compressor performance was evaluated. This base level of clearance accounted for 
thermal and stress growths, rotor deflections from vibration or maneuvers, and manufac- 
turing tolerances. Stage -to -stage differences in local temperature, materials, and type 
of rotor or stator construction were accounted for in the base clearance -to-diameter 
ratio established for each stage. Additional tip clearance was sdso provided as needed 
for each stage to allow for blade vibratory or stall deflections and for rotor/stator differ- 
ential axial movement. This extra clearance was a function of rotor blade aspect ratio, 
blade height, and casing slope . 

Airfoil Surface Finish 


Airfoil surface finish has a significant impact on aerodynamics, especially in the latter 
stages of the compressor where Reynolds numbers are high. Recent advances in the field 
of electro-mechanical machining (ECM) have made it possible to forecast that a rotor 
blade airfoil surface finish of 0.25 micron (10 microinches) can be achieved and this level 
has been assumed for the rear stages in the analysis. Current capability is about 0.4 
micron (16 microinches), and this base has been used for the forward stage rotor airfoils. 
Stator vane airfoil surface finishes have been maintained at the current level of 0.8 
micron (32 microinches). 

Blade Erosion 


Blade erosion causes aerodynamic performance degradation, mechanical deterioration, 
potential engine loss due to foreign object damage, increased maintenance costs, and 
blade replacement costs. A new correlation of in-service data from commercial engines 
was developed to predict blade life and to identify key stages in the compressor where 
blade geometry could be adjusted to achieve increased life. 

The erosion life estimations were based on data from the CFG -6 engine. A correlation 
was developed based on the consideration that blade erosion is a function of: 

1 . Time in hours or cycles on a wing position 

2. Operating environment 

3. Axial location in the compressor 

4. Blade geometry 

5. Blade tip speed 

6. Blade material 


50 



Factors such as particle size, particle hardness and sharpness, and average impact 
angle were not considered explicitly in the analysis, but were assumed to be typical of 
those encountered by CFG engines. 

The erosion life estimation was based on rotor tip trailing edge thickness reduction. 

Blade life is assumed to be used up when a 50 percent reduction in trailing ec^e thickness 
occurs in the tip region. 

Figure 21 depicts actual and predicted erosion lives for the CF6-6. Agreement is 
generally satisfactory, except for the rear stages where the actual life does not appear to 
be affected by the material change as would be predicted . 

Rear Rim Speed 


For compressor discharge temperatures and materials considered in this study, rear rim 
speeds are primarily limited by the increasingly severe weight penalties associated with 
an increasing rim speed. Therefore, for the purpose of the study, it was recommended 
that a rear rim speed of 380 m/sec (1250 ft/sec) not be exceeded unless the associated 
aerodynamic improvement offset the additional weight penalty. This limit was adhered to 
in the first phase of the screening study. However, it was subsequently relaxed, and the 
additional weight required was employed in the economic analysis. 

Inlet Hub Radius 


The effect of inlet hub radius on structure capability was studied to determine the minimum 
value that could be used. The level of torque to be carried by a fan shaft was estimated, 
and the corresponding size of the shaft thus dictated the minimum diameter of the stage 1 
disk bore. As the inner flowpath diameter was decreased with the disk bore remaining 
fixed, a mechanically less efficient disk design resulted. Weight was added to overcome 
this inefficiency up to the point at which the design became untenable. For this study 
involving engines in the 147, 000 n (33, 000 pound) thrust size, the compressor flowpath 
inlet hub radius was restricted to a minimum value of 16.5 cm (6.5 in.). 

System Vibration and Engine Bearing Layout 

Current General Electric practice is to use two bearings to support the core rotor: one 
located at the front end of the compressor; and one located aft of the high pressure turbine. 
With this system, no core engine rotating parts are cantilevered. A more rigid rotor 
system is realized, and clearance changes due to maneuver loadings are minimized . A 
preliminary system vibration analysis was performed to determine if the two-bearing 
concept could be maintained for all the compressors to be studied. The analysis, which 
was later refined, indicated the two-bearing arrangement would be satisfactory. 

Additional Mechanical Design Features 

In addition to the above considerations, general ground rules were established for blade 
maximum thickness and edge thickness that were expected to lead to aeromechanically 
acceptable designs for the maximum, nominal, and conservative loading compressor 
configurations. 




LENGTH PARAMETER 


I 


H 


Additional mechanical features were also assumed during the study in order to provide 
the potential for achieving the desired clearance control for aerodynamic performance, 
while providing a rugged, reliable, long life design. They included: 

1 . Casing liners with rub coatings 

2. Shrouded stators with honeycomb interstage seals 

3. Abrasive coated rotor spool interstage seal teeth 

4. Horizontally split casings 

Materials selected were 1985 state-of-the-art nickel- and titanium-base superalloy 
materials to permit high stress levels, while maintaining defect tolerances through 
increased fracture toughness. All components of the rotor were assumed to be designed 
using a selected combination of corrosion-resistant, high strength superalloys. 

PARAMETRIC SCREENING STUDIES 


The parametric screening studies were conducted to establish relative merits of the 
various configurations using a consistent set of assumptions. For the compressor rotor, 
the weight estimation assumptions included: 

1. All spools would be fully inertia welded . 

2. Disk temperature was assumed equal to stage exit temperature . 

3. Disk size was set by equating average tangential stress to 0.02% yield strength. 

4. Disk bore radii were set by manufacturing criteria for inertia velded spools. 

5. Minimum disk web thickness was set at 0.165 cm (0.065 in.) . 

6. Spacers were 0. 178 cm (0.070 in. ) thick . 

A computer program was developed utilizing these assumptions to: size the flowpath from 
the aerodynamic input; scale the rotor blade and disk rim weights; calculate the disk web 
thicknesses from the rim loads and allow^le stress criteria; and calculate the disk, 
spacer, and shaft weights . The screening study initial costs for the rotor were based on 
the 250th unit cost for the FlOl core compressor, quoted in 1974 dollars. The FlOl 
compressor was chosen as a baseline because it is typical of an advanced, highly loaded, 
core compressor. Costs were scaled based on compressor weight, number of stages, 
number of blades, ai.d included the effects of incorporating bore entry cooling. 

Parametric studies of the compressor stator structure also utilized the FlOl core com- 
pressor as the base for both weight and cost estimates. The mechanical configuration 
was assumed to be similar to that of the FlOl. The point where steel casi' ; material had 
to be substituted for titanium was assumed to occur at approximately the s. \e pressure 
and temperature as exists in the FlOl. Subsequent conceptual layouts of th. selected 
candidate compressors validated the earlier cost and weight scaling assumptions. 

Static component items scaled in the weight assessment studies included casint,3, liners, 
the radial support for the rear compressor case, airfoils, shrouds, lever arms, and 
actuation rings. Small components with minor influence upon the total assembly weight, 


53 


( 




such as the variable stator actuator assembly, were assumed constant for all configura- 
tions. Scaling factors accounted for length, diameter, and quantity differences. Addi- 
tional assumptions employed in the weight analysis included: 

1 . The inlet guide vanes and Stage 1 through 3 stator vanes were assumed to be 
variable with all remaining stages stationary. (For the detailed design study 
the number of variable stators required was related to the other compressor 
aerodynamic design variables using General Electric correlations.) 

2. The required thickness of the forward compressor casing would be limited by 
msnuracturing capability rather than compressor size and internal pressure. 

3. The required thickness of aft casings is greatly influenced by stress limitations 
and, therefore, largely dependent upon pressure loading and size. 

The base for stator cost projections was an estimate established by the General Electric 
Development Manufacturing Engineering Operation for the 250th unit of the FlOl com- 
pressor in 1974 dollars, individual component costs were ratioed by quantity and size 
differences to arrive at the individual component and total costs of the candidate com- 
pressors. 

A tabulation of mechanical design data for all the compressors studied in the parametric 
screening study is presented in Table Vni. 

DETAILED DESIGN STUDY 


The mechanical layout of the three compressor configurations selected for detailed design 
study are shown in Figures 22, 23, and 24 for configurations 26b2, 26d5, and 18c, 
respectively. The major mechanical characteristics of each are also tabulated in these 
figures. These configurations were analyzed in detail to assure no problems existed that 
could not be eliminated by prudent application of current design practices. Specifically, 
these included blade stresses, blade natural frequencies, blade stability, dovetail stresses, 
disk stresses, stresses in related shafting and attaching shell structures, and system 
vibration characteristics. 

The designs were based on achieving the objectives of high reliability, durability, safety, 
ease of maintenance and ease of assembly. Design requirements were consistent with 
those of current engines: 

1. 15% margin over 2/rev excitation 

2. Life - 36, 000 cycles /36, 000 hours 

3 . Material properties were average minus 3 sigma 

4 . Blade and disk dovetails stronger than the blade airfoil 

5. Disk burst speed greater than 122% speed 

6. No engine system critical speeds in the operating range 

7. Blades designed to be "stall protected" from aeromechanical instabilities 


54 



Table VIII, Parametric Screening Study Mechanical Design Results 


Corrected Physical 

Hotor One Inlet Rear Rim A Rotor A Stator A Total Minimum 





Number 

Tip Speed 

Speed 

Design 

Weight 

Weight 

Weight 

Relat ive 

Configuration 

Pressure 

of 

m ^sec 

m/sec 

Physical 

kg* 

kg • 

kg* 

Blade 

Number 

Ratio 

Stages 

(ft sec) 

(ft/sec) 

rpm 

(lb) 

(lb) 

(lb) 

Lifet 

1 

14; 

a 

12 

415 (1360) 

336 (HOI) 

15148 

-12 (-28) 

14 (32) 

2 (A) 

2.04 

2 



( 

s 

454 (1490) 

381 (1250) 

14411 

26 (56) 

-17 (-37) 

9 (19) 

2.83 

3 




9 

385 (1263) 

344 (1127) 

13513 

0 (0) 

0 (0) 

0 (0) 

2.83 

4 





360 (1182) 

J21 (1053) 

12646 

31 (68) 

8 (19) 

39 (87) 

4.83 

5 





407 (1335) 

363 (1191) 

14283 

-16 (-36) 

-4 (-8) 

-20 (-AA) 

1.65 

6 





365 (1197) 

326 (1068) 

12807 

3 (6) 

5 02) 

8 (18) 

3.17 

7 





411 (1349) 

367 (1203) 

14433 

-4 (-9) 

-6 (-13) 

-10 (-22) 

2.43 

8 





384 (1261) 

343 (1124) 

13491 

1 (1) 

0 (0) 

1 (1) 

2.83 

9 





388 (1272) 

346 (1134) 

13609 

1 (2) 

0 (0) 

1 (2) 

2.78 

10 





377 (1238) 

338 (1109) 

13245 

1 (1) 

o (0) 

» (1) 

3.04 

11 





399 (1309) 

353 (1160) 

14005 

3 (5) 

1 (3) 

A (8) 

3.00 

12 





376 (1232) 

339 (1113) 

13351 

-3 (-7) 

-3 (-5) 

-5 (-12) 

2.74 

U 





398 (1305) 

350 (1147) 

13755 

5 (1!) 

1 (6) 

8 (17) 

3.09 

14 





369 (1210) 

352 (1156) 

11273 

21 (46) 

12 (?7) 

33 (73) 

3.17 

15 





397 (1304) 

337 (1106) 

14935 

-7 (-17) 

-6 (-13) 

-13 (-30) 

2,65 

16 





337 (1105) 

159 (1179) 

11822 

14 (30) 

7 (15) 

21 (A5) 

2.57 

17 





444 (l'i57) 

322 (1057) 

15588 

-8 (-17) 

-7 (-1) 

-lA (-31) 

2.87 

18a 



12 

387 (1270) 

313 (1028) 

14145 

-2 (-3) 

22 (48) 

20 (A3) 

2.83 

18b 



10 

415 (1360) 

348 (1143) 

15148 

-4 (-9) 

8 (17) 

4 (8) 

2.30 

18c 




9 . 

437 (1435) 

374 (1227) 

1598 3 

-6 (-14) 

0 (0) 

-6 (-14) 

2.13 

18d 




8 

460 (1510) 

398 (1306) 

16819 

-4 (-9) 

-9 (-20) 

-13 (-29) 

1.70 

18e 




8 

492 (ihl4) 

387 (1270) 

17534 

-5 (-11) 

-9 (-19) 

-14 (-30) 

1.70 

18f 




7 

489 (1604) 

416 (1364) 

17866 

8 (17) 

-15 (-11) 

-7 (-16) 

1.K3 

19a 




9 

442 (1451) 

321 (1053) 

13524 

1 (2) 

-5 (-10) 

-4 (-8) 

5,4 3 

19b 




9 

428 (1405) 

311 (1019) 

15032 

4 (9) 

(10) 

9 (19) 

6.22 

19c 




9 

435 (1428) 

316 (1036) 

15278 

4 (7) 

2 (5) 

6 (12) 

5,96 

20 

23 

:l 

14 

492 (1615) 

315 (1034) 

14356 

4 (9) 

70 (155) 

75 (164) 

2.00 

21 




7 

491 (1610) 

381 (1250) 

11731 

91 (199) 

n (30) 

104 (229) 

2.70 

22 



10 

428 (1404) 

357 (1171) 

11355 

50 (109) 

33 (74) 

Hi (183) 

2.57 

23 




9 

474 (1555) 

3?n (1213) 

1 4 544 

9 (20) 

6 (14) 

15 (14) 

1.83 

24 

14 

:1 


8 

419 (1375) 

374 (1226) 

14676 

2 (-6) 

-9 (-19) 

-11 (-25) 

2.48 

25 

14 

:1 

10 

357 (1170) 

318 (1043) 

12517 

6 (12) 

9 (21) 

15 (33) 

3.39 

26a 

23 

:1 

14 

451 (1480) 

301 (987) 

13155 

in (66) 

60 ( 1 52) 

90 (198) 

2.61 

26b 



1 

1 

501 (1642) 

340 0116) 

14596 

22 (47) 

11 (69) 

53 (116) 

2.0A 

26c 



10 

522 (1712) 

358 (U74) 

15218 

2) (50) 

26 (58) 

49 (108) 

1.87 

26d 




9 

541 (1775) 

382 (1252) 

15767 

25 (54) 

n 5.10) 

38 (84) 

1.74 

26e 



10 

472 (1547) 

366 (1200) 

1 1751 

16 (7h> 

10 (66) 

65 (144) 

2 , n 

26b2 



11 

457 (1500) 

154 (1153) 

1 3i;»> 

10 (66) 

37 (82) 

67 (148) 

1.96 

26d2 




9 

491 (1610) 

188 (1273) 

14 101 

14 (74) 

20 (44) 

54 ni*-') 

2.00 

26d5 




9 

480 (1575) 

394 (1292) 

14447 

10 (66) 

18 (40) 

48 (106) 

2. (JO 


*Kolative to Configuration Number 3 
t Based on GC CF6<»6 Experience 



Figure 22 Mechanical Layout of Configuration 26b2, Eleven Stages, 23: 
Pressure Ratio 






Figure 24 Mechanical Layout of Configuration 18C, Nine Stages, 14: 
Pressure Ratio 


Blades and Vanes 


The airfoils for the selected configurations were analyzed to determine their vibratory 
characteristics. Figure 25 shows Campbell diagrams for stages 1 and 9 of the 9-stage 
23:1 pressure ratio design (Configuration 26d5). The diagrams are typical of the remain- 
ing stages of this compressor, and also of the other two configurations. In some in- 
stances, particularly in the first rotor, blade geometry might have to be adjusted to avoid 
resonances in the operating range. However, it was believed that only small changes to 
the chord or thickness would be required to increase blade flexural frequency enough to 
assure avoidance of resonances. 

The aeromechanical stability of the rotor blades was also studied. The blading was 
checked to assure the blades were "stall protected"; i.e., the blades would stall prior to 
encountering blade aeromechanical instability. The analysis indicated that most rotors 
had adequate stability margin and that only small refinements to blade geometry would be 
required to achieve adequate stability margin on all rotors . 

Table IX presents a summary of rotor blade mechanical design parameters for the three 
configurations. 

Dovetails 


The dovetails were analyzed to see if the blade and disk dovetails were stronger in fatigue 
than the airfoils. The dovetails were of conventional design (axial and/or circumferential 
straight line) with anti -fretting coatings to permit high levels of crush stress. No 
problems were uncovered in this area. 

Disks and Structure 


The disks were analyzed for basic stress levels and burst speed margins. Design 
allowables used were typical of advanced materials now in use. A summary of stage 1 
and stage 9 disk mechanical design data for the three configurations is also given in Table 
IX. No untenable problems were found to exist in the stages analyzed. 

System Vibration 


A preliminary engine system vibration analysis was performed on the three compressor 
configurations recommended for further evaluation in order to evaluate potential vibration 
problems and to determine preference, if any, for one of the compressor configurations. 

Analytical system vibration models of the three core compressors were established using 
the mechanical design information determined in the study. The engine system modeled 
was the NASA/General Electric STEDLEC Baseline Engine, with proper modifications to 
fit each compressor design. STEDLEC component weight estimate.*!, and NASA/General 
Electric QCSEE or FlOl component stiffnesses were used where applicable. The vibration 
analyses were conducted using the General Electric System Vibration and Static Analysis 
Program. The program was used to determine engine system critical frequencies, modal 
deflection patterns, and system response due to unbalance. These parameters were 
determined for critical frequencies of vibration which were synchronous with the core or 
synchronous with the low pressure rotor speed. 


59 














Table IX. Rotor Blade and Disk Mechanical Design Summary 



18c 


26b2 


26d5 



P/P ■= 

14 

P/P ^ 

23 

P/P = 

23 

HOTOK BLADE 

Stage 1 

Stage 9 

Stage 1 

Stage 11 

Stage 1 

Stage 9 

Number uf blades 

28 

98 

24 

128 

28 

140 

Material 

Tl-8-1-1 

1718 

Ti-8-1-1 

1718 

Tl-8-1-1 

1718 

Average blade height, cm 

11,6 

2.26 

K 0 

2,03 

14. G 

2 . (X) 

(in.) 

(4,569) 

(0.891) 

(6.280) 

(0.799) 

(5.765) 

(0, 788) 

Chord (Tip), cm. 

7 25 

1.88 

9.97 

1.69 

9. is 

1,82 

(in. ) 

(2.855) 

(0.742) 

(3.923) 

(0,666) 

(3.603) 

(n,71b) 

Chord (Hoot), cm. 

7.25 

1.88 

9.97 

1.69 

9. 15 

1,82 

(in. ) 

(2,855) 

(0.742) 

(3.925) 

(0.66b) 

(:^603) 

(0.71fa) 

Max thickness to chord ratio (Tip) 

0.025 

0.042 

0 025 

0,044 

0,025 

0.042 

Max thickness to chord ratio (Rout) 

0 . 090 

0.090 

0.090 

0.090 

0.090 

0 . 090 

Solidity (Tip) 

1.119 

1. 4il i 

1.143 

1,252 

1.238 

1.446 

Solidity (Root) 

1.870 

1.313 

2. 1H3 

1,352 

2.260 

1 . r 58 

Stagger (Tip) 

62.6 

63. 1 

64.2 

61,0 

61.5 

61.1 

Stagger (Root) 

no . 3 

56.7 

20,9 

55.2 

20,9 

57.8 

Camber (Tip) 

M.5 

7.0 

7.1 

10.0 

8.7 

10.2 

Camber (Rout) 

39. 7 

19.2 

jl). 3 

2'.. 1 

37.3 

17,0 

Hoot Stress MPa 

.197 

18J 

411 

IjO 

HI 

151 

(KSI) 

(57.5) 

(26.5) 

(59, b) 

(18.9) 

(64.0) 

(22. t) 


ROTOR DISK 

Material 

Ti-17 

R95 

li-17 

P95 

Tl-17 

R95 

Bore Temperature C 

216 

499 

216 

510 

216 

199 

< K) 

( 120) 

(930) 

(420) 

(950) 

(420; 

(930) 

Mat*! Prop 

,2'i ^leld Stress at Bore Temperature MPa 

7(j5 

U38 

765 

M3H 

765 

1 13H 

(K^^I) 

(111) 

(165) 

(111) 

» 165) 

(111) 

(185) 

I'ltlmatc Tensile Strength at Uor»* 

Temperature, MPa. 

945 

1 1H9 

945 

1 JHO 

945 

1 189 

(KSI) 

(137) 

(216) 

(137) 

(216) 

( 137) 

(216) 

.2'; Plastic Ci'eep (36,000 hrs) , MPa 

724 

965 

724 

965 

724 

965 

(KSI) 

(in.5) 

(MO) 

( 105) 

( t 10) 

( 105) 

( 1 10) 

Stre.'i.^.es f'* I>esign Pt. 

Uoi e Stresses, MPa. 

631 

938 

631 

903 

634 

045 

(KSI) 

(62) 

(13b) 

(92) 

( 1 3 U 

(92) 

(137) 

Average Tangential Stress, MPa. 

— 

611 



593 



669 

(KSI ) 

— 

(93) 


(M6) 

... 

(97) 

keb Ktfective Stress, MPa, 

— 

738 



703 

- -- 

721 

(KSI) 


(107) 

--- 

< 102) 


( 105) 



! 


A possible core compressor excitation of the low pressure system (fan) shaft was found to 
be the most significant vibratory mode. This response sensitivity of the shaft to core 
rotor excitation is characteristic of two-spool engine systems using two main frames. 
Operating experience has shown that vibration problems have not resulted, because current 
state-of-the-art balancing techniques have been quite effective in holding vibration levels 
within safe limits. No severe core bending modes were found in the operating speed range 
of the core, and the system vibration spectrum for the low pressure rotor excited modes 
was found to be free of problems for all three compressor configurations. 

For the three compressor configurations evaluated, no preference could be established 
based solely on the system vibration analysis. The low pressure system shaft excursions 
resulting from core rotor excitation are not expected to present a vibration problem for 
any of the three core compressor configurations. 

Blade Erosion 


Blade erosion estimates were revised for the three compressor configurations based on 
the final blade geometry data. In some instances, trailing edge thicknesses were made 
slightly thicker to attain a goal of 20 percent improvement in blade life over the current 
CF6-6 compressor rotor. For the recommended compressor configuration, a 50 percent 
improvement was specified. 

Revised Compressor Weight Estimates 

At the completion of the detailed design study, when all parts in the compressors had been 
sized so as to be structurally adequate, a final revised estimate of compressor weight 
was made. This was done partly as a check on the accuracy of the preliminary estimating 
method used in the parametric screening study, and also to account for detailed refine- 
ments in blade chord and thickness that had been made for structural or erosion life 
reasons during the detailed study. The revised weights were approximately 10 percent 
greatc tha given by the preliminary method for all three configurations. Since the 
change in estimated weight was essentially the same for each configuration, it was con- 
cluded that the relative merits of the various designs studied woulo not be affected. 

RECOMMENDED CONFIGURATION 


Configuration 26e2, recommended for design, manufacture, and test, is a 10-stage com- 
pressor with a 23:1 pressure ratio. This configuration was not studied in detail, since its 
mechanical design parameters were roughly midway between those of the nine- and 11- 
stage designs that had been studied. However, preliminary weights and costs were 
generated along with estimates for erosion life. Table X lists the estimated erosion lives 
of the compressor blades. Other compressor characteristics for the recommended con- 
figuration are listed below. 


Pressure Ratio 23:1 

No. of Stages 10 

Design Speed (rpm), Uncorrected 13,900 

Inlet Radius Ratio 0,496 

Rotor One Inlet Tip Speed m/sec (ft/sec), Uncorrected 485 (1590) 

Rear Rim Speed m/sec (ft/sec), Uncorrected 370 (1214) 

Compressor Weight kg (lb) 269 (592) 


62 









Table X. Configuration 26e2 Compressor Erosion Summary. 


■K 

0> 

0) «4-l 


11 

iH 


CM 

CM 

CM 

CM 

CM 

-O' 

O 

vO 

CM 


\0 

ro 

MT 

O 

00 


m 

r- 

CJN 

O 

CO <U 

a 

a 

a 

a 

• 

• 

• 

• 

• 

• 

rH 

m 

m 

CM 

CM 

f-H 

i-H 

1-H 

CM 

CM 

cn 


Q) CO 
Ptf iH 
PQ 


<D 

T3 


H 




y-v 


/— s 

yrs 

yr^ 

y«-S 

y— s 


yTN 



CM 

CM 

00 

NO 


cn 

CM 

rH 

rH 



cn 

CM 

iH 

rH 

fH 

fH 

rH 

fH 

rH 

c 

O 

O 

O 

O 

o 

o 

O 

o 

O 

o 

(0 *H 

a 

• 

a 

• 

a 

• 

a 

a 

• 

• 

CO 

o 

o 

o 

o 

o 

o 

o 

o 

o 

o 

0) 



N-y* 

v-y” 

w 


V— ^ 




c 











X 











o 

•H 

cn 

o 

NO 

NO 

o 

nO 

CM 

o 

00 

ON 

J2 

CJ^ 

00 

m 


Mf 

cn 

m 

cn 

CM 

CM 

H B 

o 

o 

o 

o 

o 

o 

o 

o 

o 

O 

u 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 


o 

o 

o 

o 

o 

o 

o 

o 

o 

o 












o 











(U 



y-N 

<r«s 

yry 

y^ 


yrs 

yTs 

y~s 

(0 

«H 

C^ 

<r 


CM 

in 

Mf 

m 

o 

in 

TD 

CM 

vO 

cn 

ON 


nO* 

CM 

o 

ON 

00 

0) 4J 

lO 



m 

cn 

cn 

m 

cn 

CM 

CM 

(U U-4 

»“4 


rH 

rH 

rH 

rH 

rH 

rH 

rH 

rH 

a'-' 

V— «' 

'w' 


v_y 

'w' 

y^ 

V— o' 


N-y 


C/3 











a 











«H o 











H <u 


00 


NO 

00 

O 


00 

m 

CM 

(0 

vD 

-t 

m 

CM 

rH 

rH 

o 


(3N 

C3N 




<r 

-O' 



-3- 

cn 

cn 

cn 

B 















y^ 

/TN 

yr> 


y~\ 

y^ 



m 


00 

o 

m 

MT 

m 


m 

NO 

C 

vO 

1-H 


CM 

o 

<3N 

00 

r-. 



•H 

ro 

CM 

rH 

rH 

rH 

O 

o 

O 

O 

O 

'O 



s— / 

V— «• 


s_^ 

y-^ 

'w/ 

Na^ 

Naax' 

V4 











o 











6 

00 

NO 

nO 

in 


00 

rH 


o 

m 

E 

CM 

m 

r- 

o 

NO 

m 

rH 

ON 

ON 

CJN 

o 

<Js 

in 

m 

cn 

CM 

CM 

CM 

rH 

rH 

rH 

>% 











■u 











•H 

cn 

NO 

CM 

nO 

00 

rH 

CM 

cn 

St 

in 


CM 

CM 

cn 

cn 

m 


-3* 

<T 

Ht 

'd' 

•H 

• 

• 

a 

• 

a 

• 

a 

a 

a 

a 


rH 

rH 

iH 

rH 

rH 

rH 

rH 

rH 

rH 

rH 


O 

10 


0) 

OO 

nj 

4J 

CO 


fO 


vO 


00 


63 








w 


*Rela«:ive to CF6-6 average blade life experience. 


ENGINE SYSTEM STUDIES 


METHODS AND GROUNDRULES 


Since the relative merit of each of the advanced compressor configurations is influenced 
by the performance of other engine components and by installation effects, engine systems 
studies were performed in order to arrive at the net impact of each configuration on 
typical aircraft mission performance. Aircraft mission merit factors used to measure 
performance were direct operating cost (DOC), return on investment (ROI), and fuel 
consumed (WF). Since DOC and ROI include the effects of fuel consumption on aircraft 
economics, they were considered to be the primary indicators in determining the relative 
merit of each compressor configuration. 

Factors considered in the study (in addition to compressor performance) were: 

1. Effect of engine performance, weight, and price on typical aircraft mission 
performance 

2. Installation (including pylon) weight, price, and drag as influenced by engine 
length 

3. Engine performance, expressed as specific fuel consumption, as influenced by 
engine type (boosted or unboosted) 

4. Number of required high pressure turbine stages as influenced by core com- 
pressor pressure ratio 

5 . High pressure turbine weight, price, efficiency, and cooling flow requirements 
as influenced by core compressor rotational speed 

6. Low pressure turbine efficiency and cooling flow requirements as influenced by 
engine type (boosted or unboosted) 

The methods and ground rules used in evaluation of each of the above factors are described 
in detail in the following sections . 

Aircraft Mission Analysis 


The mission analysis procedure used in the study was identical to that used in prior NASA 
STEDLEC studies (Reference 2), Two baseline aircraft were defined: a domestic 3-engine 
trijet aircraft with a design range of 5550 km (3000 nm) and a total gross weight of 



101,200 kg (223,000 lb); and an international <. gine quadjet aircraft with a design range 
of 10, 180 km (5500 nm) and a total gross weight of 145.200 kg (320, 000 lb). Both aircraft 
were sized for 200 passenger capacity and appropriate fuel reserves. A parametric air- 
craft sizing procedure was used with varying wing loading and engine thrust to arrive at the 
minimum aircraft gross weight consistent with mission requirements. (Xher aircraft per- 
formance ground rules that were influential in aircraft sizing were a takeoff balanced field 
length, a sea level takeoff thrust per engine of approximately 88,900n (20,000 lb), and a 
minimum rate of climb of 1,52 m/sec (5 ft/sec) at the nominal cruise altitude of 10,600 m 
(35, 000 ft) and 0. 8 Mach number. Advanced technology aircraft features, such as a high 
aspect ratio wing (AK = 12), a high average cruise lift coefficient (Cl » 0. 52), low cruise 
drag, and a high cruise lift/drag ratio (L/D w 17), were assumed. 

Finally, each of the aircraft was exercised on an average mission of 1300 km (700 nm) for 
the trijet and 3700 km (2000 nm) for the quadjet using a load factor of 55 percent to determine 
the impact of engine performance quantities such as specific fuel consumption, weight, 
price, and maintenance cost on aircraft merit factors (DOC, ROI, and WF). Fuel costs 
were assumed to be 7.9 cents/^ (30 cents/gal) for the trijet and 11.9 cents// (45 cents/gal) 
for the quadjet. 

Engine maintenance costs for each component, other than costs for compressor blade erosion 
were estimated by using a component Parts Index* based on CF6-6 service experience. 

This index, together with the component initial price, then established total maintenance 
material costs accrued by each component. Summing the component maintenance costs and 
spreading them over the aircraft service life then determined the average hourly main- 
tenance material costs. In addition to material replacement costs, maintenance labor costs 
equal to 2/3 of the material costs were added, again based on CFG service experience. 

These costs were spread over a 15 year aircraft service life with average yearly utiliza- 
tions of 3020 flight hours and 3840 flight hours, respectively, for the domestic and inter- 
national range aircraft. 

Compressor blade erosion life was handled separately from other engine maintenance costs. 
Average compressor erosion blade life and blading prices as defined in previous sections 
were used to compute engine maintenance costs due to blade erosion. Blade replacement 
costs associated with erosion were based on reworking or replacing all blade rows whose 
erosion life would have been used up before the next compressor maintenance. The time 
between compressor maintenance was determined by the stage having the minimum erosion 
life. Based on experience, the assumption was made that the compressor would be available 
for maintenance work approximately at these Intervals 1. use of other required engine 
work not resulting from compressor blade erosion. Tb< re, the erosion costs did not 
include the labor involved in an engine removal and tearu. 

Installation Effects 


A typical mixed flow installation as shown in Figure 26 was used for this study. Aircraft 
accessories were mounted inside the pylon to permit utilization of a cylindrical cross 
secticii, minimum-drag nacelle. Since the variations in booster, core compressor, and 


♦Parts Index = number of times a part is replaced during the life of the engine. 


65 













high pressure turbine staging inherent in this study could affect engine and installation 
length, the impact of engine length changes on installed drag was determined. The 
computed drag change with engine length assumed the nacelle inlet and exhaust sections 
were unaffected by engine length; only the center cylindrical portion of the nacelle was 
varied. The net effect of nacelle length, diameter ratio, nacelle surface area, and pylon 
surface area was then represented in terms of installation drag. 

Engine Performance 


Nominal engine cycles were estiblished for both the boosted (14:1 pressure ratio core 
compressor) and unboosted (23:1 pressure ratio core compressor) engine types as shown 
in Table XI. The boosted engine cycle and nominal components were identical to those 
defined in the prior NASA-STEDLEC study (Reference 2), while a new unboosted cycle 
was defined to provide a basis for evaluation of the 23:1 pressure ratio core compressors. 
Note that the unboosted cycle retained the same overall cycle characteristics (overall 
pressure ratio ^ 38, T^ = 137T C (2500^ F) @ max climb) as those of the boosted engine and 
was sized for the same fan corrected airflow. A two-stage, high pressure turbine was 
used for the unboosted cycle, since the turbine pressure ratio requirement was higher 
than practical for a single stage. 

Since these cycles were established for nominal component performance levels, and 
component performance was an inherent variable of tnis study, cycle derivatives (influence 
coefficients) were determined to provide a basis for evaluating component performance 
changes. For ease of evaluation, the influence coefficients were established at constant 
engine thrust; i.e., fan and core compressor size were varied in order to maintain 
constant core engine energy extraction and constant use as component performance 
levels were changed. The net effects on required fan and core compressor size and on 
specific fuel consumption could then be determined. 

Turbine Performance Effects 


The basic factors that arise in considering the turbine performance effects are: 


1, High pressure turbine staging effects as influenced by core compressor 
pressure ratio 

2, High pressure turbine efficiency and cooling flow requirements as influenced by 
core compressor rotational speed selection 

3, Low pressure turbine efficiency and cooling flow requirements as influenced by 
core compressor pressure ratio 

The high pressure turbine efficiency variation with stage loading was based on a General 
Electric correlation of the observed performance of a number of specific designs. The 
two-stage turbine for the 23:1 pressure ratio core was assessed as having an efficiency 
potential of 2.6 points greater than the single-stage turbine for the 14:1 pressure ratio 


67 


Table XI. Boosted Versus Unboosted Cycle Comparison. 
Mixed Flow - Nominal Components 


Boosted (STEDLEC) Unboos ted 


Pouer Setting 

Maximum 

Climb 

Maximum 

Cruise 

Take-Off 

Maximum 

Climb 

Maximum 

Cruise 

Take-Off 

Altitude/Mach No., m 

(ft) 

10.7K/0.8 

(35K/0.B) 

I0.7k/0.8 

(35K/0.8) 

0/0 

(0/0) 

l0.7k/0,8 

(35K/0.8) 

l0.7k/0.8 

(35K/0.8) 

0/0 

(0/0) 

\Temperature relative to 
Standard Dav, "U 
('F) 

10 

(+18) 

10 

(+18) 

15 

(+27) 

10 

(+18) 

10 

(+18) 

1 5 

(+27) 

Bare Engine Thrust, n 

(lb) 

Kelatlvo Specific Fuel 
Consumption 

38, 900 
(8740) 

1 . 001 

35,600 

(8010) 

1 .00 

147.000 

(33160) 

38,600 

(8670) 

0.9879 

35, 500 
(7990) 

0.9833 

149.000 

(33600) 



Overall Prc“'-.ure Ratio 

3H 

36 

30 

38 

36 

30 

High Pressure Turbine Inlet 
Temperature. "C 

TF) 

1371 

(2300) 

1326 

(2420) 

1427 

(2600) 

1371 

(2500) 

1 326 
(2420) 

1427 

(2600) 

Bypass Uatlt> 

8.9 

7.1 

7.5 

7.0 

7.2 

7.6 

Corrected Fan Flow, kg^sec 
( 1 bm ' sec ) 

5bK 

(1253) 



56H 

(1253) 



Fan Pressure Ratio 

1.71 



1.71 



Boost Pr«»ssuro Ratio 

2.75 



1 67 (Fan 

Hub) 


Core f'ltnipressor Pressure 
Ratio 

1 1 



23 



Corrected Core Compressor 
Flow, kg 'sec 

(Ibm sec) 

31.0 
(68. 1) 



•16. H 
( 103.2) 



No. High Pressure Turbine 
Stages 

1 



2 



High J’ressurc Turbine 
Pres‘*ure Ratio 

3.8 



4.1 



No, Low Pressure Turbine 
Stages 

Ifl 2 



1 ♦ 1 /2 



L()w Pressjire Turbine 
Pressure Ratio 

3.(i8 



4 . 9 




68 














... -.y- W 1 1^.. 






I 




core compressor. This performance differential was based upon the observed perform- 
ance of typical cooled designs and is due to a number of factors, two of which are reheat 
effects and cooling effects, with the two-stage turbine showing a performance advantage 
from both effects. 

Turbine cooling variations with staging and stage loading were computed with the aid of 
other correlations of General Electric data. Ground rules were selected to be consistent 
with achievable cooling systems and the availability of advanced materials and to provide 
a consistent basis for computing cooling flow trends. Use of these correlations and 
ground rules, together with turbine velocity diagram calculations to define a gas path heat 
load parameter, provided the required estimates of cooling flow variation with turbine 
loading. 

The low pressure turbine efficiency and cooling flow requirements were established in a 
similar manner. The low pressure turbine for the 23:1 pressure ratio compressor 
required less cooling air than that used with the 14: 1 pressure ratio compresj or by virtue 
of the reduced inlet temperature. The major impact was in the first stage blading, which 
was uncooled for the turbine used with the 23:1 pressure ratio compressor. This reduc- 
tion in cooling air impacted turbine efficiency, since less expansion energy in the down- 
stream stages was available. Hence, with the 23: 1 pressure ratio compressor, the low 
pressure turbine had a slightly reduced efficiency potential. 

Detailed turbine weight and price estimates were made for the turbine configurations 
used with the nominal 14:1 and 23:1 pressure ratio compressors, and empirical scaling 
reladonships were used to evaluate the effect of rotational speed changes or these 
nominal values . 

A maximum turbine rotational speed of approximately 17, 000 rpm was established for both 
the 14: j and 23:1 pressure ratio compressors. This limit was established based on 
mechanical feasibility of the turbine rotor blades and blade attachments in the 147,000 n 
(33,000 lb) thrust engine design size, 

PARAMETRIC SCREENING STUDY 


As previously discussed, a series of preliminary compressor aerodynamic designs was 
carried out in vvhich key parameters were varied systematically. The results of this 
parametric screening study were used to evaluate the different compressor configurations 
in typical aircraft missions. 

A nominal single-stage high pressure turbine and a compatible low pressure turbine flow 
path were established for the core engine having a 14:1 pressure ratio compressor as 
shown in Figure 27. This turbine flowpath was found suitable for all of the 14:1 pressure 
ratio core compressor configurations with the exception of Configurations 14 and 16, The 
relatively low rpm of these compressors exceeded the allowable single-stage loading limit, 
and for these two cases, the turbine diameter was increased to bring the high pressure 
turbine stage loading within acceptable limits. This also resulted in a less heavily loaded, 
more efficient, low pressure turbine. 


Chargeable cooling air is assumed to completely bypass the turbine in the cycle calcu- 
lations, The adjustment to turbine efficiency accounts for the fact that some work is 
done by the chargeable cooling air. 


69 



I 


Similarly, a nominal two-stage high pressure turbine and matching low pressure turbine 
flow path was established for the core engines having a 23: 1 pressure ratio compressor 
(Figure 28). the rotational speed variation was less than that of the 14:1 pressure 

ratio compressor, this flowpath was found to be suitable for all the 23:1 pressure ratio 
core compressors. 

A summary of the component and engine weight and price variations, installation effects, 
and maintenance costs in the design size for each of the compressor configurations is 
presented in Table XII. It can be seen that the high pressure turbine weight variation far 
exceeds that of the compressor and is generally of an opposite sign. In terms of com- 
ponent and engine price variation, the compressor has the dominant effect, while turbine 
price is relatively insensitive to compressor configuration. Installation effects include 
the pylon, and maintenance cost data ‘»"*-^ude parts and labor, as previously discussed. 
Although not shown in Table XII, an irt »*esting result of the analysis was that the com- 
pressor efficiency variation, as lirtnd in Table II, was far greater than the variation in 
turbine efficiency. Turbine cjolirg flow and intercompressor duct pressure loss varia- 
tions were also fairly small. Thus, bare engine fuel consumption trends tended to be 
dominated by variations in core compressor performance. 

Component performance levels that established the bare engine fuel consumption and the 
data contained in Table XII represent the total input to the aircraft system evaluation. 

The data given in Table XII are for the design size engine. Two final scaling steps were 
required on engine and installation weight and price to obtain the data presented in the last 
four columns of Table II. The first scaling step was required to adjust the core engine 
and the installation weight and price for the noted component performance differences, 
while the second scaling step was required to adjust the weight and price data to the 
engine size appropriate for the mission. The missions were 88, 900 n (20, 000 lb) sea level 
static taKe-off thrust and 93,300 n (21,000 lb) thrust for the trijet and quadjet, respectively. 

The overall results of the engine system analysis have been summarized previously in 
Table II. Data in Table II are given only for the domestic transcontinental range trijet 
aircraft mission, since the trends seen for this case were typical of the international 
range mission as well. All aircraft system performance data given in Table II include 
the economic effects of compressor blade erosion. 

REFINED SCREENING STUDY 

The trends of compressor efficiency and engine system merit factors versus compressor 
design parameters established during the parametric screening study were used to define 
two families of high-efficiency compressors for further study. These families were the 
14:1 pressure ratio Configuration 18 series for use in boosted engines and the 23:1 pres- 
sure ratio Configuration 26 series for use in unboosted engines. Because of the high 
efficiencies of these compressors, engines were obtained having superior economic and 
fuel usage ratings compared to those studied in the earlier parametric screening study. 

In addition to using design parameters shown to give high compressor efficiency, consid- 
erable attention was also devoted in this phase of the study to determining the effects that 
the number of compressor stages had on overjill engine system performance and economics. 


71 


) 






Table XII. Summary of Component Weight, Price, Installation, Maintenance and Specific Fuel Cc 
Pressure Ratio Compressors (Design Size Engines). 


AHlgh A;.ow ATotal AHlgh 

Pressure Pressure Installed Pressure 


ACcrmpressor 

Compressor Humber Weight 


Coafigurstlon Stages kg (lb) 


Base 

9 


0 

1 

12 

-4 

(-8) 

2 

6 

3 

(7) 

3 

9 


0 

4 

9 

34 

(75) ■ 

5 

9 

-25 

(-56) 

6 

9 

3 

(6) 

7 

9 

-15 

(-34) 

8 

9 

-5 

(-11) 

9 

9 

-5 

(-10) 

10 

9 

-5 

(-11) 

11 

9 

-2 

(-4) 

12 

9 

-11 

(-24) 

13 

9 

2 

(5) 

14 

9 

28 

(61) 

15 

9 

-19 

(-42) 

16 

9 

15 

(33) 

17 

9 

-20 

(-43) 

18A 

12 

14 

(31) 

I8B 

10 

-2 

(-4) 

18C 

9 

-12 

(-26) 

18D 

8 

-19 

(-41) 

19A 

9 

-9 

(-20) 

19B 

9 

3 

(7) 

19C 

9 

0 

(0) 

24 

8 

-17 

(-37) 

25 

10 

10 

(21) 

26B2 

10 

25 

(55) 

"De-staged" 

26D5 

8 

8 

(17) 

''De-staged** 

20 

14 

69 

(152) 

21 

7 

98 

(217) 

22 

10 

78 

(171) 

23 

9 

10 

(22) 

26A 

14 

84 

(186) 

26B 

11 

47 

(104) 

26B2 

11 

62 

(136) 

26C 

10 

44 

(96) 

26D 

9 

33 

(72) 

26D2 

9 

48 

(106) 

2605 

9 

43 

(94) 

26B 

10 

60 

(132) 


Turbine 
Weight 
kg (lb) 

Turbine 
Weight 
kg (lb) 

A Installation 
Weight 
kg (lb) 


0 


0 

0 

(0) 

29 

(63) 


0 

21 

(47) 

14 

(31) 


0 

-30 

(-66) 

-5 

(-12) 

-5 

(-12) 

0 

(0) 

-11 

(-24) 


0 

31 

(68) 

12 

(26) 


0 

-17 

(-37) 

-9 

(-20) 


0 

-1 

(-3) 

15 

(32) 


0 

0 

(-1) 

0 

(0) 


0 

-1 

(-2) 

1 

(3) 


0 

1 

(3) 

-4 

(-8) 


0 

2 

(4) 

7 

(16) 


0 

0 

(0) 

-2 

(-4) 


0 

-4 

(-8) 

4 

(8) 


0 

6 

(13) 

-11 

(-24) 

10 

(22) 

-22 

(-48) 

24 

(53) 


0 

18 

(40) 

-4 

(-8) 

10 

(22) 

-5 

(-11) 

39 

(85) 


0 

S 

(11) 

10 

(21) 


0 

28 

(62) 

29 

(63) 


0 

15 

(34) 

49 

(108) 


0 

8 

(17) 

77 

(170) 


0 

0 

(0) 

37 

(82) 


0 

11 

(25) 

26 

(58) 


0 

20 

(45) 

32 

(69) 


0 

20 

(44) 

19 

(41) 


0 

-8 

(-17) 

-12 

(-27) 


0 

10 

(21) 

-2 

(-4) 


0 

0 

(0) 

15 

(32) 


0 

-18 

(-40) 

167 

(368) 

-18 

(-39) • 

29 

(65) 

95 

(210) 

-18 

(-39) 

-47 

(-103) 

88 

(195) 

-18 

(-39) 

-23 

(-50) 

176 

(388) 

-18 

(-39) 

-14 

(-31) 

127 

(280) 

-18 

(-39) 

41 

(91) 

179 

(395) 

-18 

(-39) 

10 

(21) 

132 

(290) 

-18 

(-39) 

9 

(19) 

212 

(468) 

-18 

(-39) 

6 

(13) 

257 

(567) 

-18 

(-39) 

-20 

(-43) 

166 

(365) 

-18 

(-39) 

-17 

(-38) 

172 

(380) 

-18 

(-39) 

-10 

(-21) 

144 

(318) 

-18 

(-39) 

3 

(6) 


Engine ACompressor Turbine 


Weight 

(1).(3) 

(lb) 

Price 

X 

Price 

X 

0 

(0) 

0 

0 

46 

(102) 

0.1 

0 

-13 

(-28) 

-0.3 

0 

-11 

(-24) 

-0.1 

0 

54 

(119) 

0.8 

0 

-30 

(-67) 

-0.3 

0 

-8 

(-17) 

0.4 

0 

-1 

(-3) 

-0.6 

0 

-6 

(-13) 

-0.1 

0 

-2 

(-4) 

0 

0 

-7 

(-15) 

0 

0 

5 

(12) 

-0.1 

0 

-16 

(-36) 

-0.1 

0 

12 

(26) 

0 

0 

5 

(11) 

1.0 

0 

23 

(51) 

-0.5 

0 

16 

(36) 

0.5 

0 

24 

(53) 

-0.5 

0 

52 

(114) 

0.9 

0 

42 

(93) 

0.1 

0 

45 

(99) 

-0.4 

0 

59 

(129) 

-0.8 

0.1 

40 

(87) 

-0.2 

0 

50 

UlO) 

0 

0 

51 

(113) 

-0.1 

0 

-6 

(-13) 

-0.6 

0 

7 

(15) 

0.5 

0 

23 

(51) 

0.9 

0 

4 

(9) 

-0.2 

0 

145 

(319) 

1.8 

5.8 

26 

(58) 

2.1 

6.5 

23 

(50) 

2.1 

6.7 

51 

(113) 

0.3 

5.8 

132 

(291) 

2.9 

6,2 

115 

(254) 

1.4 

5.8 

81 

(179) 

1.9 

6.0 

141 

(311) 

1.1 

5.8 

150 

(330) 

0.6 

5.9 

76 

(167) 

1.0 

5.8 

85 

(187) 

0.9 

5.9 

86 

(190) 

1.6 

6.0 


(1) A Awelght of -103 kg (-227 lb) was applied to the unboosted engine weights In addition to the 

above component Aweights to obtain the A total engine weight. This was done to account for the 

fact that these engines did not have booster stages whereas the base configuration did. 

(2) A Apric. of -3X appUtd to tht unboost.d engine price In addition to the above component 

prlcee to obtain the Atotal engine price. This was done for the same reasons as discussed 

in footnote (1), 

(3) These totals are for design uiae engines. 


V‘' ' \ 





Specific Fuel Consumption Data for the 14:1 and 23:1 


"pressor 
Price 
X 

0 

0.1 
-0.3 
- 0.1 
0.8 
-0.3 
0.4 
- 0.6 
- 0.1 
0 
0 

- 0.1 

- 0.1 

0 

1.0 

-0,5 

0.5 

-0.5 

0.9 

0.1 

-0.4 

- 0.8 

- 0.2 

0 

- 0,1 

- 0.6 

0.5 

0.9 

-0.2 

1.8 

2.1 

2.1 

0.3 

2.9 
1.4 

1.9 
1.1 
0.6 
1.0 
0.9 
1.6 


Ion to the 
unt for the 
did. 


component 

•cussed 


AHlgh 

Pressure 

Turbine 

Price 

X 


0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0.1 

0 

0 

0 

0 

0 

0 

0 

5.8 

6.5 

6.7 

5.8 
6.2 
5.8 
6.0 

5.8 

5.9 

5.8 

5.9 

6.0 


^Low 

Pressure 

Turbine 

Price 

X 


0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0.1 

0 

0,1 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 

0 


0.3 

0,3 

0.3 

0.3 

0.3 

0.3 

0.3 

0,3 

0.3 

0.3 

0.3 

0.3 





Alnstallation 

Price 

% 

ATotal 

Installed 

Engine 

Price (2), (3) 
X 

ACompressor 

Blading 

Maintenance 

Costs 

$ /Flight Hour 

A All Other 
Engine 
Maintenance 
Costs 

$/Flight Hour 

ABare 

Engine 

Specific 

Fuel 

Consumption 
at Cruise. Z 

Alnatallsd 

Drag, 

Drag/Thruat 

X 

0 

0 

0 

0 

0 

0 

0.4 

0.5 

-1.52 

+0.03 

-0.7 

40.12 

-0.6 

-0.9 

-1.98 

-0.11 

+3,7 

-0.16 

0 

-0.1 

-1.85 

-0.03 

+0.2 

0 

0.6 

1.4 

-2.30 

+0.27 

+1.6 

40.19 

-0.3 

-0.6 

-0.25 

-0.09 

+0,5 

-0.10 

0 

0.4 

-1.67 

40.12 

+0.8 

0 

0 

-0.6 

-2.02 

-0.18 

40.6 

0 

0 

-0.1 

-1.85 

-0.03 

+0.6 

0 

0 

0 

-1,83 

-0,02 

+0.2 

40.01 

0 

0 

-1.82 

-0.01 

+0.5 

+0.01 

0 

-0.1 

-1.92 

-0.04 

-0.3 

0 

-0.1 

-0.2 

-1.77 

-0.05 

+0.6 

-0.02 

0.1 

0,1 

-1.90 

+0.01 

-0 2 

40.02 

-0.4 

0.7 

-1.58 

+0.31 

+0.2 

-0.11 

0.4 

-0.1 

-1.97 

-0.14 

-0.3 

40.10 

-0.1 

0.5 

-1.50 

+0.18 

-0.1 

-0,04 

0.1 

-0.4 

-1.95 

-0.14 

+0.2 

40.02 

0.6 

1.5 

-1.60 

+0.29 

-1.0 

40.16 

0.3 

0.4 

-1.58 

+o.o: 

-1.2 

+0.09 

0,2 

-0.2 

-1.67 

-0.10 

-1.1 

+0.04 

0 

-0.7 

-1.57 

-0.23 

-0.7 

0 

0.2 

0 

-1.60 

-0.08 

0 

+0.06 

0.4 

0.4 

-2.30 

-0.02 

+0.1 

+0,12 

0.4 

0.3 

-2.32 

-0.01 

+0.2 

+0.11 

-0.1 

-0.7 

-1.87 

-0.18 

0 

-0.04 

0,2 

0,7 

-1.90 

+0.15 

+0.8 

40.06 

0 ' 

0.9 

-1.27 

+0.27 

-0.9 

0 

-0.3 

-0.5 

-1.36 

-0.04 

-0.8 

-0,10 

0.6 

5.5 

-1.38 

+5.05 

-1.8 

+0.18 

0.9 

5.0 

-1.50 

+6,00 

+3.5 

-0.24 

-0.5 

5.6 

-1.50 

+6.30 

-0.4 

-0.13 

-0.3 

3.1 

-1.28 

+4.20 

-0.6 

-0.09 

0.8 

7.1 

-1.43 

+6.02 

-2.6 

-0.25 

0.2 

4,7 

-1.37 

+4,88 

-2.3 

40.06 

0,2 

5,4 

-1.32 

+4.73 

-2.5 

40.05 

0.1 

4,3 

-1,40 

+4.65 

-2.2 

+0.01 

-0.4 

3.4 

-1.47 

+4.36 

-1.7 

-0.10 

-0.3 

3.8 

-1.40 

+4.21 

-2.0 

-0.10 

-0.2 

3.8 

-1,60 

+4.36 

-2.3 

-0.06 

0 

4,9 

-1.43 

+5.15 

-2.3 

+0.01 






73 







■ . s . 






101.6 CM (40 IN) 
DIAMETER 



figure 28 Nominal Turbine Flowpath for 23:1 Pressure Ratio Advanced 






I 


For both series of en^nes, the variation of core compressor efficiency with number of 
compressor stages was the major component performance trend affecting engine un- 
i.istalled performance. As shown previously in Figures 12 and 15, core compressor 
efficiency varied by about one point over the range of number of stages studied, with the 
lowest efficiencies being for the designs with the fewest stages and the highest speeds. 

High pressure turbine efficiency, however, was essentially constant for all configurations 
within each series, since rotative speeds were sufficiently high in all cases that aero- 
dynamic loading effects in the turbine were negligible. Although the unboosted 23:1 
pressure ratio coie compressors had polytropic efficiencies that were slightly lower than 
those of the 14:1 pressure ratio compressors, the higher efficiency of tl j two-stage 
turbine used to drive the unboosted compressors more than offset this condition. As seen 
in Figure 29, the resulting uninst;illed specific fuel consumption data show that the un- 
boosted engines had an advantage over the boosted engines and that, in general, the 
compressors with the fewest stages gave the poorest fuel consumption. 

As is also shown in Figure 29, both engine types have similar installed orag values. The 
trend toward reduced drag at low number of compressor stages is a consequence of re- 
ductions in overall engine length. For a given engine type the improvement in installed 
drag obtained as the number of compressor stages was reduced nearly offsets the i.. crease 
in bare engine specific fuel consumption. 

Uninstalled engine weight and cost trends versus number of compressor stages are showm 
in Figure 30 for the 147, 000 n (33, 000 lb) thrust engine size. The engine weight for the 
unboosted engines was greater tiian for the boosted engines primarily because of the use 
of a two-stage high pressure turbine. The fact tliat the turbine v>eight increased rapidly 
with rotative speed was responsible for the increase in bare engine weight as the nui iber 
of compressor stages was reduced. Bare engine cost was reduced as the number of 
compressor stages was reduced, however. 

Weight and cost trends for engine installation items (such as the nacelle, pylon, and 
thrust reverser) are «^own in Figure 31. Both weight and cost for engine installation 
•^ere reduced as the number of compressor stages, and thus engine length, wa^ 'cf.uced, 

. reduction in installation weight nearly counteractec the increase in bare engine 
V' , ‘ .t with fewer compressor stages shown in Figure 30. For unbocsted engines, for 
unple, the minimum bare engine weight is obtained by using the best 11 -stage com- 
pressor, Configuration 26b2. If the best 9-stage compressor (Configuration 26d5) were 
used, this minimum bare engine weight would be exceeded by 23 kg (50 lbs). However, 
the installation-related weight of the engine with the nine-stage compressor is 18 kg 
(40 lb) less than that of the engine with the 11 -stage compressv^r, thus nearly offsetting 
the bare engine weight penalty resulting from use of the nine -stage compressor. 

The final significant variable in the engine systems evaluation vvas maintenance cost. 
Although compressor blading maintenance cost due to erosion did decrease slightly as the 
number of compressor stages was reduced, the cost difference was not significant be- 
cause each configuration was designed to have a similar blade life. Remaining engine 
maintenance costs increased slightly with an increase in number of compressor stages 
due to a corresponding increase in the number of stator hub seals, abradable rotor tip 


76 



(i^BARE ENGINE SPh*” Ci^INSTALLED DRAG/THRUST 

FUEL CONSUMPTION. PHUCENT UATIO, PERCENT 


0.4 


0.2 4 


OH 


-0.2 H 


-0.4 J 



# UE-STAGEI) 23: 1 CONFIG 26b2 ™ONT 
^1 I I STAGE 

■ DE-STAGED 23: 1 CONFIG 26d5 J REMOVED 

DESIGNS WriH EXIT lUDIls' UATIO - 0.91 

DESIGNS Wrni ' TT RADIUS RATIO = 0.93 



NUMBER OF COMPRESSOR STAGES 


Figure 29 Effect ol Number ot Compressor Stages on Bare Engine 
Specific Fuel Consumption and Installed Drag 


77 



^^BAKE ENGINE V.i;iGHr 
LliS 



8 9 10 11 12 13 14 

NUMBEK OF COMPKESSOK STAGES 


Figure 30 Effect of Number of Compressor Stages on Bare Engine Weight 
and Price (Design Size) 



3 . 0-1 


H 

Z 

W 

a 

a: 

u 

cu 


M 

O 

t-H 

c: 

cu 

s:; 

o 

1-4 

2 

w 


2.0 


l.OJ 


-l.O 


- 2.0 


CONFIG 26a 



DESIGNS WITH EXIT RADIUS RATIO - 0,91 

DESIGNS WITH EXIT RADIUS RATIO = 0.93 

I I I I 



NUMBER OF COMPRESSOR STAGES 


Figure 31 Effect of Number of Compressor Stages on Installation 
Weight and Price (Design Size) 


79 




I 


I 


} 


i 


a 

5 ■ 


liners, etc., that would be required. This trend was accounted for by the direct relation- 
ship of maintenance costs with engine initial price. The unboosted engine maintenance 
costs were significantly higher than those for the boosted engine due to the additional high 
pressure turbine stage. 

Economic and fuel usage results for engines using the high-efficiency compressors de- 
fined in the refined screening study have been presented previously in Figure 14. As 
discussed in a previous section. Further Studies of High Efficiency Compressors, the 
principal results were that the unboosted engines had the lowest fuel consumption, while 
the boosted engines had the best economic merit factors. It was also found that no strong 
preference could be seen for a particular number of compressor stages for either engine 
type, as several compressors could be found in each case that gave near -optimum per- 
formance. Direct operating cost considerations slightly favored use of the fewest possible 
compressor stages, but fuel usage considerations favored selection of a less highly loaded 
compressor. In order to cover the spectrum of engine types and a range of near optimum 
number of stages for a given pressure ratio, the nine-stage 14:1 pressure ratio com- 
pressor and the nine-stage and 11 -stage 23:1 pressure ratio compressors (Configurations 
18c, 26d5, and ‘^6b2, respectively) were selected for fui'ther detailed design studies. 

DETAILED DESIGN STUDY 


Three of the most promising compressors, identified in the refined screening study dis- 
cussed above, were selected for more detailed design studies. Compressor aerodynamic 
and mechanical design refinements were made to each configuration, and the turbines used 
in each engine were evaluated as well. Finally, revised engine merit factors were com- 
puted which reflected the results of these detailed studies. All three of the turbines were 
examined in more detail, but only the turbine for the 9-stage 14:1 pressure ratio com- 
pressor, Configuration 18c, was changed. The high pressure turbine diameter was re- 
duced slightly from that used in the screening study in order to better match the high ro- 
tational speed of this configuration. Turbine velocity diagram data and performance 
estimates were then made for the turbines of the three engine configurations. The prin- 
cipal result obtained from the turbine detail design analysis was that an increase in cool- 
ing flows was required for the two-stage high pressure turbines compared to those used 
in the screening study. 

Component weight, relative price data, and relative mjuntcnance costs for the three 
selected configurations w^ere reevaluated using the same methods as discussed in the 
pr:vious section. Parametric Sc reening Studies. The unboosted engine component and 
maintenance costs were determined to be slightly lower relative to the boosted engine than 
was determined in the screening study, 

A summary of final engine characteristics for t!iO three selected compressors is presented 
in Table XIII. Data are also given for the 10-stage 23:1 pressure ratio compressor 
recommended for further development, plus a de-staged version of this compressor having 
nine stages and a 14:1 pressure ratio. Configuration 18c was selected as the reference 
configuration for this comparison, since it had the best direct operating cost and return 
on investment of all the configurations studied. While the unboosted engines enjoyed a 
specific fuel consumption and fuel usage advantage over the boosted engines, they were at 


80 

f 


V 



Table XIII. Summary of Engine Characteristics and Engine Evaluation Results (with Erosion Effects). 


I 


• ’ * 


T3 O 

W <M 0\ O ^ 

cd 0) o • 

so <t 00 *0 

WcMO\f-nr>oom + 

I 

(U 

a 


00 o ro o <r 

vD 00 m o • 

Ch m m o • o 

.H •-« ro rH I 


CM 

CO O CO 
O O • rH 

• • o • 

O O I o 


o 

O /-N 

CM ON O f-l 

0) •'-d' o • 

vO O CO 00 • i-H 

CM rH CM CO 00 lO I 


00 O MD O 

\0 vO o o o o 
o to IH • 

CM rH CO CO o 


m 

n£> iH ON O 

\0 • • • 

• O O iH 
O I 1 1 




O 














NO 



O 











NO 

iH 

ON 

o 

m 


ON 

O 


iH 

ON 

O 


o 



CM 

iH 

ON 

o 

iH 





o 

• 

o 

tn 

ON 

00 

o 

o 

vD 

• 

• 

• 


vO 

CO 

00 


• 

f— 1 

iH 

vO 

CO 

o 

• 

• 

• 

O 

o 

rH 

00 

CM ON 

CM 

CO 

00 

in 

1 

CM 

<r 

iH 

CO 

<r 

o 

O 

1 

1 

1 



CM ON O 

\0 iH CO 00 

CM iH CM CO 00 


o 

o ^ 

ON O 01 

cj <r o w 

00 -<j- oo • CO 

»— I ON rH CO 00 IT| CO 


CO o o^ o 

iH CO O O 

o m <r »H • • 

CM 'd' iH CO sO «H 


ON o NO o <y oi 
O CO o o w w 
o - 3 * i-H (d td 

CM HT »H CO CQ pQ 






o 

NO 

CM 




CM 

ON 

O 

ON 



00 

• 

• 

• 




• 

O 

o 

f—i 

CM 



o 

1 

1 

1 

<r 

T3 







0) 







0) 


<11 

0) 

Q) 

(U 

Q> 

Zi 


Oi 

(0 

(0 

tf) 

(0 



CO 

ta 

<0 

<d 

CO 

<u 


PQ 

PQ 

PQ 

PQ 

PQ 

c 







•H 







00 







c 







w 












c 


Q> 





o 


N 





•H 


-H 





4J 


CO 





a 



NH 




jS 


C 





3 


O 

u 




(0 



•H C 
»H O 
O O - 



cn 

<1 

25H 

S a Vi 

50 

44 


(0 •«-) 

(0 

VI 


a 

3 /-> 




a u 

u 



•« 

VI *H 

o 

G 



Vi *0 


M 


v-x 

4-» o 


A 

A 0) 

rH 

tJ 

03 


u 

CJ rH 


0 

<u 

X 

cd »H >; 



0) o 

3 H 


a 


•H 

v-^ 


(0 


4-i CO 

M-l cd 

4J 

JC 

U »H 

(0 

50 

VI 


iVi 

X 


(0 

3 

W S 



50 

•H V: 

4l rH 

c 

CO 


•H 

Cd 50 


a# 

lA 

3 

a cj 

50 

•H 

Vi a 

a Cd 

•H 

03 


O 

X a 


M 

(0 

0 

<U o 

•H 

<U 

a 

VI 

Vi 

> ^S 

03 



& 


^ 00 

50 aoo 


:2 

c 

G G 

Cd 

G 


ao Vi 


s 

u 

H iH 

Cd to iH 



<13 O 

O 03 

M 

M 

A 

c/3 

00 


o 

a 

+ 

u + 


c 

C -H 

•H C 

03 


03 


rH 50 


o 


<U Q Tl 

0) 

0 

*H VI 

VI «H 

a 

c 

00 *3 

+ 

V. 


w 

3 O 

(U o 

c 

♦H 

50 Cd 

Cd VI 

o 

0 

Cd 

c 

03 

o 

M-i 

o 

•H 00 

*0 rH 00 

•H 

4J 

C rH 

3 C 



CO 

rH 

o 3 

0) 

0 

cn 

50 • 

<y IH • 

50 

Cd 

a rH 

»-l o 

u 

G 

D 

rH 

00 

cn 


(f. 

c o 

I-H Cd o 

3 

iH 

Cd 

Cd u 

u 

Vi 


Cd 

• rH 

(U 

Ij 

0) 

W 

rH 4J 

U 

fH 

03 4J 

> CO 

03 

3 

rH 

VI 

O Cd 

u 

<y 

Vi 


Oi M 


Cd 

V4 CO 

a c 

Vi 

VI 

03 

CO 

Vi 


9* 

QJ vo 

4J C UO 

<13 

VI 

cd G 

Cd 

•H 

03 

3 

G 

a 0 

6 

a 

a 

Vi CO 

to v-i CO 

Vi 

CO 

PQ M 

03 Vi 

O 

a 

a 

H 

lA H 

0 

3 

0 

<d N^ 

C 

Cd 

c 


G H 





CO 

o 

z: 

o 

PQ 

M <3 

PQ 

M 

< < 

•H 

<3 

<3 

<3 

< 

< 


A Total Price, % Base 4.1 2.2 3.0 0.80 

A Compressor Blade Maint. Cost, $/FL Hr. Base -0.04 -0.31 -0.20 -0.10 

A Total Maint. Cost $/FL Hr. Base 3.45 2.90 3.18 0.24 


I 


an economic disadvantage. These results were consistent witn those of the screening study; 
however, the differences between the boosted and unboosted engines were reduced as a re- 
sult of the more detailed analysis. The direct operating cost advantage of the boosted 
engines was due to lower engine price and maintenance costs, while weight and compressor 
blading erosion effects were negligible. The fuel usage advantage of the unboosted engines 
was due primarily to the specific fuel consumption advantage attributed to the higher 
efficiency of the two-stage high pressure turbine. 

ENGINE SYSTEM MERIT FACTORS SENSITIVITY STUDY 

Since there is some unavoidable degree of uncertainty in the estimated performance levels 
of the various components, sensitivity studies were conducted to determine the effect of 
variations in component performance, weight, and cost on the engine evaluation results. 
Several conclusions were drawn from the sensitivity studies. The first conclusion related 
to the engine evaluation’s sensitivity to compressor efficiency: Changes in compressor 
efficiency of two to three points would be required in oi'der to change the relative ranking 
of the configurations. Since the compressor efficiency model is expected to predict the 
efficiency potential of each configuration to within ±1.0 point, it is not expected that the 
trend of the engine evaluation results would be influenced by uncertainties in compressor 
efficiency. The second conclusion was relative to the effect of liigh pressure turbine 
efficiency on the boosted versus unboosted engine comparison. It was found that a relative 
shift in the attainable efficiency of the single versus two-stage high pressure turbine of 
about three points was required to change the trend of the direct operating cost or return on 
investment results, while an approximate 1.5 point shift was required to change the trend 
of the fuel usage results. The result was that the single-stage high pressure turbine 
efficiency would have to be about six points lower than that of the two-stage configuration to 
eliminate the direct operating cost and return on investment advantage of the boosted 
engine. Conversely, the single-stage turbine efficiency would have to be within 1.5 points 
of that of the two-stage high pressure turbine for the boosted engine fuel usage to approach 
that of the unboosted engine. The third conclusion was relative to the impact of fuel cost, 
engine weight, and compressor price on the evaluation results. It was found that fuel 
costs of 18.5 and 21, 1 cents per liter (70 and 80 cents per gallon), with all other costs 
held at 1974 levels, were required for the single-stage and two-stage turbines, respectively, 
to eliminate the economic advantage of the boosted engine. Similar lack of sensitivity to 
uncertainties in engine weight and compressor price were obtained. Therefore, it was 
concluded that anticipated uncertainties in these parameters would not be expected to 
change the trend of tin^ results. 

EmmES USIN^’niE RETO .COm;*RESSOR CW 

An improved version of the 10-stage 23:1 pressure ratio c ompressor identified in the re- 
fined screening study, Configuration 26e2, was selected as the design recommended for 
further develoj)ment, A layout of an unboosted engine using this recommended compressor 
is shown in Figure 32. Weight, price, and performance factor.s used in the evaluation of 
this engine and of a boosted engine using a de -staged version of this compressor arc also 
presented in Table XIII, along with the engine evaluation results. These results are 
compared to the "esults for the boosted engine using the nine-stage, Configurat 18c, 
compressor. The data indicate that aJi unboosted engine using the recommended com- 
pressor would have a direct operating cost comparable to one using the nine-stage com- 
pressor, Configuration 26d5, and a slightly lower fuel usage than engines using ither the 
nine -stage or 11 -stage compressors that were examined in the detailed design study phase 
of the program. 


82 





CONCLUSIONS AND RECOMMENDATIONS 


PARAMETRIC SCREENING STUDY FINDINGS 


The Parametric Screening Study effort identified a number of factors leading to high core 
compressor efficiency for the general class of compressors considered: 

1. Low inlet radius ratio is beneficial. 

2. Medium levels of stator exit swirl, lO*" . lead to high efficiency. 

3. Low inlet and exit axial Mach numbers improve o/erall compressor efficiency. 

4. Shock losses do not significantly penalize overall efficiency for first rotor tip 
Mach numbers below about 1.4. 

5. Use of fewer stages does not significantly penalize overall efficiency until 
increases in speed raise first rotor tip Mach numbers above about 1.4. 

6. High exit radius ratio can be beneficial for efficiency provided it aids in 
reducing an excessively high tip speed. The optimum exit radius ratio is 
likely to increase as the number of compressor stages is reduced. 

7. Medium aspect ratios give best overall efficiency. Low values cause increased 
end-wail losses, while high values require either high tip speed or more stages 
to maintain stall margin. 

The effects of the advanced technology assumptions made for this study were: 

1. Relative ranking of compressors is not greatly affected even if ’'current’* rather 
than "advanced” technology is assumed. 

2. "Advanced" aerodynamic and mechanical technology assumed in this study is 
responsible for a 2. 0-2. 5 point increase in predicted efficiency compared to 
"current" technology compressors. 

Other findings of the parametric screening studies related to the overall engine system 
were: 


1. Fewer stages give less expensive and shorter compressors, but core engine 

weight does not necessarily decrease and engine acceleration time may increase. 


84 


\ 


[ 

I 


2, Blade erosion life can be improved significantly with virtually no performance 
penalty if low-life stages are identified and improved. Erosion itself has only 
a small effect on average engine economics, provided it does not become the 
reason that an engine must be removed prematurely from the airplane for over- 
haul. 

3. Boosted engines have better direct operating cost but poorer fuel usage than un- 
boosted engines, mainly because boosted engines use less expensive, but less 
efficient, single-stage high pressure turbines compared to the two-st; ' ir- 
bines in unboosted engines. While the magnitudes of these difference., wre 
relatively small, they persist despite any reasonable variations in aircrnit 
mission, turbine efficiency, turbine cost, fuel cost, and compressor :ost 
assumptions. 

DETAILED DESIGN STUDIES 


Aerodynamic design analysis results were: 

1. No severe aerodynamic design problems were identified in any of the three 
cases studied that might invalidate the estimates of their performance potential. 

2. The nine -stage, 23: ' i.ressure ratio compressor configuration had higher rotor 
and stator inlet Mach numbers and liigher diffusion factors than the 11-stage 
23:1 pressure ratio configuration or the nine -stage 14:1 pressure ratio con- 
figuration, and thus would be more difficult to develop. 

3. Off-design studies of the 23:1 pressure ratio compressor designs indicated that 
a part-speed stall margin of about 25 percent could be achieved without using 
bleed in the normal viigine operating range, although starting bleed probably 
would be needed. 

Mechanical design analysis results were: 

1. No severe mechanical problems were discovered in any configuration. The 
high rear rim speeds, low inlet radius ratios, two-bearing rotor layout, and 
medium front rotor blade aspect ratios all were specifically examined and found 
to be acceptable. Final weight estimates were somewhat higher than in the 
screening studies, but the estimated weights increased about equally for all 
three coi^igurations and did not change the relative ranking of the compressors. 

RECOMMENDED CONFIGURATION 


A 23:1 pressure ratio design was recommended for further development because: 

1. It provides an engine having the lowest fuel usage. 

2. Lower pressure ratio versions could be derived from this configuration by 
removing front or rear stages, and the resulting compressors would still be 
near-optimum for use in boosted engines having excellent economic ratings. 

A 10-stage 23:1 pressure ratio compressor was selected as the configuration with the 
best combination of advantages: high efficiency, low operating cost, low fuel usage, and 
acceptable development risk. 




85 



REFERENCES 


1. Koch, C.C. and Smith, L.H., Jr.; "Loss Sources and Magnitudes in Axial-Flow 
Compressors," Transactions of ASME Journal of Engineering for Power, Vol. 98, 
Series A, No. 3, July 1976, Page 411 

2. Neitzel, R.E., Hirschkron, R. , and Johnston, R. P. ; "Study of Turbofan Engines 
Designed for Low Energy Consumption," NASA CR-135053, 1976 

3. Ware, T.C., Kobayashi, R.J. , and Jackson, R. J. ; "High-Tip-Speed, Low-Loading 
Transonic Fan Stage, Part 3 - Final Report." NASA CR-121263, Februr.ry, 1974 


86