Turbine Engine Hot Section Technology (HOST)

li IIIC) IIAltered Stress Geometry Many times the stress analyst wants to alter the part geometry to reduce his stresses, but the devia- tions will not, In the judgement of the heat transfer analyst,. affect the temperatures. We have In the ll past used lIad hoc procedures to transfer temperatures to the new stress geometry. This approach Is not optimum, but we do not plan to Include any capabilIty for this case In our transfer module.

This module, once it Is developed, will transfer thermal data from heat transfer meshes to stress analysis meshes. But it wIll have the capabIlIty to do much more •. The basic features of 3-D search and Interpolation will make It an outstandIng foundation for automatIc con~truction of embedded meshes, local element refinement, and transfer of other mechanical loadings.

OBJECTIVES

• TRANSFER TEMPERATURES FROM-A HEAT TRANSFER STUDY TO A STRESS ANALYSIS .

- I NDEPENDENT MESHES

.....

co \.0

- ACCURATE/EFFICIENT TRANSFER

- FLEXIBLE

•• 1 ,.

-l L1.J Q ~ c::: L1.J u..

en z: ~ I-- ~ L1.J :::t: Q r-t'\ ----~--~----~--------I • I~ I I I I I t f-- I I I

, I) !

t-----t-- I '-- I -~_____ I _ ~ : I : ... 3:i.

I I 1:_-: I .& It I I ....

I , '1.. • "1 i

t----+--- : f- -+--- i I • I - !

~ •• r', ;ii1.

I ,. 00 tI ~ I I i

1-----1---1 - I l I

i ..

. ~--I-- t -~ ~--4---l I

---~~~

1-----/0---+- ~-~ ~ ----' i

- ""'.~ll

I

~----~--~----~---l,.()-----'i

J I~--~I----~!--~I----~I--~ol

.. -.-~

HEAT TRANSFER

PROCESS

GEOMETRY ~ CORNER NODE

H T

~COEFFICIENTS

PEOMETRY

STRESS 110DEL ,'.-

LOCATE

POINTS I-- ~_SELECTED .

POINT

HT ELEI1ENT

IN

MODEL

.....

1.0 I'\) '---

COMPUTE

WEIGHTED

.-fINAL ~JOHTINO

-

COEFFS

..

COEFFICIENTS

CALC

AND

TEI1PERATURES FROI1

HEAT TRANSFER FORMAT

TEnPERATURE At.

RESTAiu STRESS POINT

TEMPS

THERMAL TRANSFER MODULE

OVERALL SYSTEM

09-~O-r.~ REVISION- 1.1 . ,. " LU e::: => ~ e::: LU Z l- e... a z: ~- LU I-

- a

I- <t: c..

U LU a en u -.J en <t: LU u..

c::: e::: l- => en U) <l

X

X

\ ~ \ a ::I:: \ I- LU -,

-

f- z:

-

c...

e::: -.J LU ~ - e::: LU ~ ~ e::: LU w u..

en z ~ f- I- <t: LU ::t: U) U) UJ e:::::

.-

~ U) U) z: > U) UJ - :3: e:::::

.-

Q U) z: N

-

:3: ........

U) z:

-

--'

i

e::::: ~ e::::: u..

UJ U) ::I: z: l- <C e:::::

.-

..

z:

-

!::i

--'

- =>

~ !:Xl ..

e::::: UJ ....J ::I:

.-

~ >- e::::: UJ ::I:

.-

N

FLEXIBILITY

_.

- TRANSFER MODULE NOT KEYED TO ANY SPECIFIC CODES

- INITIAL EFFORT DIRECTED AT TRANSFER OF THERMAL DATA

.......

lO U1

HOWEVER

BASIC TECHNOLOGY CAN BE APPLIED TO MANY AREAS.

- TRANSFER ~'ECHAN I CAL LOADS

- TRANSFER OF BOUNDARY CONDITIONS FOR MESH REFINEMENT

- COMPUTATION OF CONSTRAINT EQUATION COEFFICIENTS

" .

t-- Z lLJ :E: Z (,!J

- en

en c:r: en z z

- - t--

t-- c:r:

- u

~ Z --I

- 0

c...

u c...

c:r: >- c::: lLJ c:r: e::: ~ :=l Z t-- :=l :=l Ll- a:l U

-

~

t-- =:) c:r: COMPONENT-SPECIFIC MODELING Murray S. Hirschbein National Aeronautics and Spa:ce;Ad~in{strat:t.on Lewis Research Center Cleveland, Ohio 44135 As a result of the recent drastic increases in fuel costs, the aircraft gas turbine engine industry has placed much higher technical priorities on reducing engine weight and increasing engine efficiency. As part of this effort, engine temperatures, internal gas pressures, and rotational speeds are being increased, while the size and weight of the engine components are generally being reduced. The result is that ccmponents, in many cases, are operating closer to their structural limits. This places much greater importance on the ability to accurately structurally analyze engine components to assure that they can survive for their designed lifetime in an increasingly harsh environment.

The burner 1iner ,turbine blades and vanes are among the most structurally burdened and analytically complex components in the engine. High mean temperatures with severe transients, local hot spots, and steep gradients characterize the thermal environment of all three components. Additionally, the vanes and turbine blades are subjected to the highest gas pressures in the engine, and the turbine blades are loaded even further by strong centrifugal forces. The geometry of these parts is equally complex. The burner liner can be designed as overlapping stepped louvers with many cooling holes. The turbine blades and vanes may also have regions densely packed with tiny slanted cooling holes as well as complex internal gas paths along the span.

In addition, there are small radius fillets near the base of the blades and vanes. As a result of the geometry and thermo-mechanical loading, these components have locally steep stress and strain gradients, regions which undergo varying degrees of cyclic plasticity and creep deformation, and material properties which can vary significantly in time and space.

Irt order to increase the durability and life of hot section components, new high-temperature materials and fabrication techniques are being applied to these components. These new materials have significantly anisotropic material behavior, such as with single crystal blades and directional solidification.

This makes accurate and efficient structural analysis of hot section components even more difficult.

Currently, there are baSically two general approaches to structurally analyze complex engine parts. One is to use general purpose analysis codes, such as NASTRAN and MARC. These are extremely powerful tools which can be applied to a wide variety of problems. However, they are not tailored to the needs of anyone problem and depend very heavily on the user to adapt them to specific problems. Furthermore, these programs are designed as "one-shot" problem solvers based on the finite element method of analysis. That is, a problem is first modeled by whatever method the program is based on, and then the whole problem is solved at one time. It is up to the user to decide how the problem is modeled and whether the solution obtained is accurate to the desired degree. If the problem is to be re-solved, say with a finer finite element mesh, or if a non-linear approach is required, it is up to the user to remodel and to develop his analysis strategy. All sub-problems, approximate solutions, or sensitivity studies must be controlled by the user. To some extent this effort can be reduced by writing geometric and discretization pre-processors, and in the case of NASTRAN, new Rigid Formats can be added.

Essentially, with general purpose codes, the analyst tends to overpower very complex problems, but at great expense in computational effort and man-hours.

The second approach is similar to the first, except that the analysis codes are streamlined to meet the needs of more focused problems. These codes have built-in pre- and post-processors to reduce the modeling effort, facilitate sequencing of programs, and to make the display of data easier.

However, they still rely on a single analysis method and are still designed to model and solve the structural analysis problem in a single pass through the program. These p~ograms may have features to deal with component specific problems such as special axisymmetric shell elements or crack elements. Often the analysis method is greatly simplified -in order to reduce the cost of repeated analyses during a design process. In both cases analysis decisions su~h as local model refinement, or how and when to use linear and non-linear analyses are left to the analyst.

Either of these approaches, or a combination of them, may be adequate when a single, conservative, feasible design is acceptable. With the increasingly harsh thermo-mechanical environments expected for hot section components as engines are made lighter and more efficient, the structural limits of these components will be more severely challenged. Better, more efficient analysis methods must b-e developed in order to be able·to assure in advance that hot section components will survive for their design lifetimes.

Under the HOST (HOt Section Technology) program, advanced component-specific mOcrelTng methods, with built-in analysis capability, will be developed separately for burner liners, turbine blades and vanes. These modeling methods will mak-e maximum use of, but will not rely solely on, existing analysis methods and techniques, to analyze the three identified components. Nor will the complete structural analysis of a component necessarily be performed as a single analysis. The approach to be taken will develop complete software analysis packages with internal, component-specific, self-adaptive solution strategies. Each package will contain a set of modeling and analysis tools. - The selection and order of specific methods and techniques within the set to be applied will depend on the specific-component, the current thermo-mechanical loading, and the current state of the component. All modeling and analysis decisions will be made internally based on developed decision criteria within the solution strategies; minimal user intervention will be required. In this way, the structural analyses of burner liners, turbine blades and vanes may actually be comprised of a series of global approximate analyses and local detailed analyses which lead to computationally-efficient, total structural analyses with assured accuracy, and without extensive, time-consuming user intervention.

The software packages will be modular and open-ended to allow the addition or substitution of new modeling and analysis methods and to allow modification to or change of the solution strategies. The primary structural analysis method will be the finite element method. However, additional methods such as approximate closed form analyses, semi-analytical solutions, and boundary integral methods will be considered to develop the complete solution strategies. Linear and non-linear solution capability for static and dynamic responses will be included in each component-specific model. Automatic options within the solution strategies will include remeshing with optimization, substructuring of the mass and stiffness matrices, and automatic load step and time step control. During development the accuracy and computational efficiency of the component-specific models will be verified by comparison with established solutions and data sets.

To support the advanced structural analysis capability developed herein, advanced thermo-mechanical load mission models will be developed to predict detailed time-related pressure and temperature distributions in the burner liner, turbine blade and vane. These distributions will accurately represent conditions experienced during an arbitrary commercial aircraft mission cycle.

Also, as part of this load modeling effort, an advanced thermodynamic engine cycle model will be developed to predict the gross temperatures and pressures throughout the hot section of the engine as a function of the power lever setting. The advanced thermo-mechanical load modeling capability will be designed to interact with the component-specific structural models developed herein, as well as to supply detailed loading data for independent analysis programs. Acting together the structural and loads models will provide highly advanced capability to accurately predict the. loading and structural response of the burner liner, turbine blade, and vane over an entire arbitrary mission cycle.

This capability will be further enhanced by developing methodology to synthesize the loading and structural response histories of the hot section components over an arbitrary mission cycle. The synthesis process will involve developing methods which can use sparse, pre-computed, component-specific structural and mission cycle data to construct these histories with minimal interaction with detailed analysis codes. This technology will' significantly reduce the cost of predicting the loading and structural response histories of the burner liner, turbine blade and vane over complete cycles.

The thermo-mechanical load models and mission model synthesis methodology will be developed essentially independent of the component-specific structural models. As such, the progress of either part of the total program will not depend on the other. In this way, both parts can be designed more effectively to interact with independent loading or structural analysis programs.

COMPONENT-SPECIFIC MODELLING PURPOSE DEVELOP ADVANCED HIGH-LEVEL STRUCTURAL ANALYSIS CAPABILITY FOR HOT SECTION N a COMPONENTS a 1) BURNER LINER 2) TURBINE BLADE 3) VANE TWO MAIN THRUSTS 1) COMPONENT-SPECI~IC STRUCTURAL MODELLING A) GEOMETRIC MODELLING AND DISPLAY B) SELF-ADAPTIVE SOLUTION STR~TEGIES 2) THERMO-MECHANICAL LOAD/MISSION MODELLING N o ......

A) ADVANCED THERMODYNAMIC ENGINE MODEL B) DETAILED T~MPERATURE AND PRESSURE LOADING MODELS C) MISSION HODEL -DECOMPOSITION/ SYS~HESIS- METHODOLOGY • if ,.

TASK STRUCTURE BASE PROGRAM SURVEY DESIGN OF S-TRUCTURAL ANALYSIS ARCHITECTURE THERMODYNAMIC ENGINE MODEL SbFTWARE DEVELOPMENT MISSrON MODEL DEC'OMPOSITION STRUCTURAL ANALYSIS METHQDS EVALUATION THERMO-MECHANICAL LOAD/MISSION MODEL DEVELOPMENT ~COMPONENT-SPECXFXC HODEL D~VELOPMENT ~ OPTIONAL PROGRAM SURVEY STRUCTURAL ANALYSIS METHODS EVALUATION COMPONENT-SPECIFIC MODEL DEVELOPMENT MISSION MODEL DEVELOPMENT VERIFICATION TESTING 3-0 INELASTIC ANALYSIS METHODS FOR HOT SECTION COMPONENTS BRIEF DESCRIPTION Christos C. Chamis National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 BACKGROUND The most severe structural requirements imposed upon aircraft gas turbine engine components result from the extreme environmental conditions in the engine hot section. These conditions include very high temperatures with steep thermal gradients,high and fluctuating pressures, rapid transients, vibration, oxidation, corrosive and erosive atmospheres, and an assortment of structural loadings both from within the engine and as a result of engine/aircraft system interactons. Accurate prediction of structural response and life assessment of the components under these conditions require sound 3-D inelastic analytical methods. Present 3-D inelastic analysis methods usually rely on large volumes of input data to,define the problem, frequent user intervention during the analysis process, and considerable care in assessing the accuracy and interpreting the results. Most of these methods are parts of general purpose structural analysis programs which were not intended for the complex 3-D inelastic analysis problems associated with gas turbine 'engine components. Thus highly-skilled technical manpower is required to set up the problems and frequently, to interpret the results.

OBJECTIVE AND APPROACH The objective of this program is to develop advanced 3-D inelastic structural/stress analysis methods and solution strategies for more accurate yet more cost-effective analysis of components subjected to severe thermal gradients and loads in the presence of mechanical loads, with steep stress and strain gradients, and which include anisotropy and time and temperature dependent plasticity and creep effects. The approach is to develop four different theories, one linear and three higher order theories (polynomial function, special function, general function). The theories are progressively more complex from linear to general function in order to provide streamlined analysis capability with increasing accuracy for each hot section component and for different parts of the same component according to the severity of the local stress, strain and temperature gradients associated with hot spots, cooling holes and surface coating cracks. To further enhance the computational effectiveness, the higher order theories will have embedded singularities (cooling passages, for example) in the generic modeling region.

Each of the four theories consists of three formulation models derivable from independent theoretical formulations. These formulation models are based on (1) mechanics of materials, (2) special finite elements, and (3) an advanced formulation to be recommended by the contractor.

The mechanics of materials models shall be formulated for easily amenable solution (approximate calculations). The special finite elements will be formulated to be used as "stand-alone" modules and as modules integrated (using interfacing links) into general purpose structural analysis computer programs. The advanced formulation model shall be formulated to provide an alternate and complementary analysis capability to the special finite elements so that each theory can be used to check the other, and thereby minimize costly experiments that otherwise may be needed. In addition, each model shall be formulated to accommodate three different levels of constitutive theory with progressive levels of complexity. The three different levels of constitutive theory are needed in order to: (1) provide formulation flexibility, (2) provide for modeling different material behavior in different parts of the component, and (3) increased accuracy to assess the validity of the approximations made. Appropriate solution strategies with self-adaptive features shall be developed along with numerical solution algorithms with self starting and dynamic incrementation to further enhance the computational effectiveness of these theories.

All theories, including models and constitutive relationships, will be validated with respect to accuracy and computational effectiveness using available analysis results from simulated and actual hot section components.

In addition, the theories shall be verified using available experimental data and data generated under this program. The end product of these theories will be computer programs (modules) for stand alone use and for integration into other structural analysis programs. It is expected that the 3-D inelastic analysis capability being developed under this program will provide considerable flexibility for the solution of 3-D nonlinear structural problems; eventually it should lead to longer lifetimes and improved overall durability of the hot secti·on components made from present and future materials. Also, this capability will provide enhanced capability to experimentally evaluate constitutive relationships.

GENERAL SCOPE OF WORK This program is a four year, 45,000 man-hour effort.

HOT SECTION CmlPONENTS REQUIRING 3-D INELASTIC ANALYSIS -I

~-.----~~

MJ

N o (J'1 TURBINE TURBINE '-~ • .."'u1A~~'l;,:-..J!"""",,"'" SCHEr.IATIC ILLUSTRATING REGIONS FOR THE FOUR . DIFFERENT ORDER THEORIES DI SCONTINU IT I ES

GENERIC ~ '-I AXIAL

GENERIC t-l0DELING ./ I LATERAL t-IOUELING REGION REGION N o LINEAR POLYNOmAL 0'1 DISCONTINUITIES DISCONTINUITIES GENERIC ~,~' GENERIC AXIAL MODELING

MODELING [<?/~L--f- AXIAL

REGION I ~ LATERAL ~ LATERAL REGION o ," " " ..

J GENERAL FUNCTIONS SPECIAL FUNCTIONS ! II ,.

3-D I NELASTI C MI\LYSI S OF HOT SECTION COMPONENTS FLOW CHART '3iJ.se Program OPTION TASK I - Linear Theories TASK IV Special Function Theories o 3-Formulation Models o 3-Formulation Models .~ o 3-Constitutive Relationships o 3-Constitutive Relationships o No Embedded Discontinuities o a-Identical Embedoed Discontinuities TASK II - Polynomial Theories TASK V - General Function Theories o 3 -Formulation Models o 3-Formulation Models N o o 3-Constituitive Relationships o 3-Constitutive Relationships "-I o 2-Embedded Discontinuities o a-Different Embedded Discontinuities TASK III - Reporting I TASK VI - Reporting

I -

I

- NOTE: Each technical Task (I, II, IV and V) consists of three (3)subtasks describing the formulation models and constitutive relationships.

3-D INELASTIC ANALYSIS: TASK TIME SCHEDULE TASK DESCRIPTION - PERCENr--l TIME FRa-1 DATE OF CONTRACT YEARS EFFORT I 1 2 3 4 NO.

BASE PROGRAM I I Linear Theories 15

~---.j

[ II Polynomial Theories 22 A Common Requirements for Tasks I and II II Reporting (Base Program) 1 Option 1 IN '0 i \CXl 1-- IV Special Functions Theories 26 ----- V 34 General Functions Theories B Common Requirements for Tasks IV and V VI Reporting (Option 1) 2 - - ------------ -------- '-- ~ ;/ y y Base Program Option 1 (17,000 Man-hours) (28,000 Man-hours) LIFE PREDICTION AND CONSTITUTIVE BEHAVIOR -- OVERVIEW Gary R. Halford National AeronautfcS-an<rspace Aciniinistra tion Lewis Research Center Cleveland, Ohio 44135 One of the primary drivers that prompted the initiation of the HOST Program was the recognized need for improved cyclic durability of costly hot section components. All too frequently, fatigue in one form or another was directly responsible for the less than desired durability and prospects for the future weren't going to improve unless a significant effort was mounted to increase our knowledge and understanding of the elements governing cyclic crack initiation and propagation lifetime. Certainly one of the important ingredients was the ability to perform accurate structural stress-strain analyses to determine the magnitudes of the localized stresses and strains since it is these localized conditions that gov~rn the initiation and crack growth processes.

Consequently, the programs that evolved included high-temperature cyclic constitutive behavior work as well as cyclic life prediction methods development. Figure 1 lists the areas for which funding was sought. Our initial intent was to fund four programs, Life Prediction and Constitutive Modelling for Isotropic Materials and Life Prediction and Constitutive Modelling for Anisotropic Materials. The latter two have been combined into·a single program for a number of technical and managerial reasons. Another change to our plans is the possibility of funding two contracts for the Constitutive Modelling of Isotropic Materials. This is largely because of the relative newness of this research area and the proliferation of competing theories.

Furthermore, additional funding has permitted us to pursue programs in the area of high temperature cyclic crack growth~ Three such programs are in the planning stages: a contractual effort aimed at the problem of high temperature crack growth; and two University Grant activities, one being directed at the micromechanisms of high temperature crack growth~ and the other involving an interdisciplinary approach to the overall problem of crack initiation, crack growth~ and final fracture. Only the proposed contractual program will be discussed today. The University Grant programs will be reviewed next year. A milestone chart is shown in Figure 2 for the six contract and grant programs.

While the details of test programs will be given by the individuals that are intimately involved, I would like to emphasize the underlying objective of these programs:. The development and verification of workable engineering methods for the ca1cu1ation~ in advance of service~ of a) the local cyclic stress-strain response at the critical life governing location in typical hot section components~ and b) the resultant cyclic crack initiation and crack growth lifetimes.

A contract has been in existance with Pratt & Whitney Aircraft since the first of June 1982, and the P&WA Project Manager~ Vito Moreno~ will be making the presentation covering that work~ The other efforts will be described by the individuals responsible for creating the Request for Proposal Packages~ A Grant has been awarded to Professor H~ W~ Liu of Syracuse University tor studies of the mechanisms of high temperature crack growth •.

In addition, to the contract and grant programs, we are up-grading our in-house High Temperature Fatigue Laboratory capabilities as indicated in Figure 3.

FIG. 1 OVERVIEW LIFE PREDICTION & CONSTITUTIVE BEHAVIOR G. R. Halford, LeRC • CYCLIC CRACK INITIATION • CYCLIC CRACK GROWTH • CYCLIC CONSTITUTIVE BEHAVIOR • LeRC FATIGUE FACILITY UP-GRADING FIG. 2 LIFE PREDICTION & CO"NSTITUTIVE MODELLING- CONTRACTS &GRANTS PROGRAM ILEMENT FY83 FY8S FY86 FY87 FY88

FY8" I

wi ISOTROPIC CREEP-FAnGUE lIAS 3-23288-PUoIA (MlREIIJ) I LIFE PREDIcnON (C)

I W

ISotROPIC CONSnTtJrIVE ~,,~ CI!1P£nnVE-I'OOSIBLY 2 AWARDS I HOOELLING (C)

I~ ...IY

ANISotROPIC LIFE PREDIcnoN/coNSnrurIVE I

~ " ""l CI!1P£nnVE

I«lDELLING (CJ

I

FAnGIJ£ CRACK GRIM'fH SYRACUSE UNIVERSITY(LIIl) MECHANISMS (GRANT)

I

W HIGH t'EHPERA TIJR£ CI!1P£nnVE I~,,'-'I CRACK GRIM'fIf (CJ

I

...I.

INTERnISCI PLINARY TO 8E DETERMINED ~ CRACK LIFE (GRANT)

"-'

I FIG. 3 FATIGUE FACILITY ENHANCEMENT • Servo-Controlled Testing Machine/HI-Temp Crack Growth • Servo-Controlled Tension/Torsion Machlnes(3)-Blaxlal Studies • Host/Satellite Computer Installation - Data Acqulsltlon/Processlng/Storage/Retrleval • HCF/LCF Machines-Cumulative Damage Studies

CREEP-FATIGUE LIFE PREDICTION FOR ENGINE HOT SECTION MATERIALS (ISOTROPIC) Vito Moreno Program Manager Pratt & Whitney Aircraft Commercial Engineering OBJECTIVES: The objectives of this pr.ogram are the investigation of fundamental approaches to high temperature crack initiation life prediction, identification of specific modeling strategies and the development of specific models for component relevant loading conditions.

PROGRAM DESCRIPTION AND APPROACH: The sixty month technical program is divided into two sub-programs which contain a total of thirteen tasks. The basic program (Tasks I-IV) represents a 24-month effort. Task I includes a survey of the hot section material/coating systems used throughout the gas turbine industry. Two material/coating systems will be identified for the program. The material/coating system designated as the base system shall be used throughout Tasks I-XII. The alternate material/coating system will be used only in Task XII for furthe~ evaluation of the models developed on the base material. In Task II, candidate life prediction approaches will be screened based on a set of criteria that includes experience of the approaches within the literature, correlation with isothermal data generated on the base material, and judgements relative to the applicability of the approach for the complex cycles to be considered in the option program. The two most promising approaches will be identified. Task III further evaluates the best approach using additional base material fatigue testing including verification tests. Task IV consists of technical, schedular, financial and all other reporting requirements in accordance with the Reports of Wor~ clause.

This activity concludes the basic program.

The optional program (Tasks V-XIII) represents a 36-month effort. Specific crack initiation prediction models will be developed within the various tasks to address various aspects associated with hot section life prediction. Task V considers the development of thermal-mechanical fatigue models for uncoated and coated structures. Task VI addresses multiaxial stress state effects. Task VII considers a cumulative loading model to address sub-cycle and block loading effects. A screening of available environmental and protective coating models is conducted in Task VIII. Also, the extent of the problem for thermal-mechanical cycling will be quantified in this task. Tasks IX and X consider the development of environmental attack and protective coating models. Task XI addresses the effects of mean stress in the creep-fatigue initiation process. In Task XII final verification of the model(s) developed in the previous tasks will be conducted. In addition, baseline isothermal, TMF and biaxial testing will be conducted in the alternate material/coating system to ·assess the applicability of the approaches and models developed on the base material. Task XIII consists of technical, schedular, financed and all other reporting requirements in accordance with the Reports of Work clause.

-_.

PROGRAM STATUS: • Basic program started 5-27-82.

• Cast B1900 + Hf designated as base material.

• Wrought IN718 designated as alternate material.

• Diffusion Aluminide (NiAl) and Overlay (MCrALY) coatings sele~ted.

• Single heat (2500 lb.) of B1900 + Hf acquired.

• 24 bars cast for initial specimen fabrication.

• Review of life prediction approaches initiated.

CONSTITUTIVE MODEL DEVELOPMENT FOR ISOTROPIC MATERIALS Albert Kaufman National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 T!"'l\:! tr~nd toward incr~ased parfo:.'lnance for aircrc:.f·t gas -zt!rbine el:~inez has re~:;ul "ted in :'ligh:r" "tu('bin~ !JJ,zd(;: tip sp~~d$' and high~I" inlet gas telil~e::"a';~ur-:=:s ar..j l~t'~s::;!..~:--es. Tl:.ese l;~o!'e se",ere o~~r,r~~.nz cc~,,:!l:i tions Ij,~~\1,~ ~-2duccd ·t!~~ c'!.u:"''::~J:15 ... t.J 01 to·~ seC-C:'Oi1 co;.~::o:~en·:;~ ~r:.:~ h~"Je dei,1or:$·tr,~:~~~d "t2'lc nae~ -to i~":riJrOve upon t!~:: C1.l~rent. ~!'!'41y'~::!.~C11 ~~thof:"~$ used in tne uesit;n 0: tl-:ase co::t::or.c:nts.

U:1dct" :t ::~c~nJ.;ly in.s·~1:~u·t~u corr~(',-,::':; effor-i; I.Jn=er·i:~!~~n ~s par"t of th~ :·~os·r p~o;~:""al."., cr-c:;c!, in:i.·citi~ion lift;! pre::ic-:ion ~(!thocls uill b:~ c1evelo!led fer ho~ sec~ic~ cc:~~ol".~nts fC!:;r::'cc:r~.ed f::"o:n i:::otropic ma"teri<:l;;;. To a~~ly "these nlet~Qds L~ is firs';; r..ec2SS2.::"j' ·to deJ..;er~in~ the C01~pC~'2~t s~::-L1c':;un!l respor..:::c, s::;eci:::'c,'3.11j' t;le stI"':!:::s-strain histoI"!f a't ·t;h~ cri'~it::al crac!::ing location. The $'l;~uc';;ur:~::' a:1alysis i.1e·cnod r..uzt ~e ca!lable of accoun'i;in~ for' cyclic t!1~ri.:o:.~:=ch<::~:!.c:;l 10z.ciZ'!6, pl:!s'Cic flo,", curing thermal t::,,~nsien·ts, c::"eep and s~:-ezs rcl.:::.<z~ticn cll":r"ir.~ s·l;c:c:.dy-state operc:rl;ic;'l and inelas·tic strain I"atc~:~t'ti4:.~ a::-ld I" ~"/e:-sal due to r.ep':!~ted fligh"t cycles.

In .... ec::m·;; ~/2nI"$, r:cnlir.e~r f~ni·te cl'1~en"l; COli~~u'l;eI" cedes such as ri:',;$YS ar'-'l r~ll':C have beco::!e a"Jz::la!:lle for cyclic anal:,'si~ of co:::~onen'i;s imlOl"Jin.; in{:lastic strains. !he:::e c':)dc::: are based on classical plasticity th::on' ar..d use ul'!coupled cree!l ccns'l;i':;uti "Je. rr.od81s. T:'c classical I:.ethods ut::'::"ize sh.plifying assu;;.;;ticn'3 'i:OI" CO::l;::.\t,:,·;;~onal con";cniel"!ce. I'..:o:on~ these a$$Ui.1!ycions are (1) the c1~:::i.::ition of ,a speci::i.c yield sUI"f"ce with assaci:;rtec flow rules and harde~ing mcc''21s, (2) 't02 pc:rti tionin~ of incl:::::'t:i.c s·trains in;;;o -cir.:e-indapenden':; (pl;:',S'1;ic) <:nd t:i.r.'!e-d':"~pcr.d:m'c (creep) s'::rain::: and (3) the tm~Ouplin6 of tij:le-ind~~cl"~d~t"!t c:n:1 ti::!~-dep~nden-l:; ir.~laztic strain eff:octs. Th::t these classical :.127.{'Jods and ';:;!'1eir assu:.1~tioi1s do not realisticc:lly represent sU~~I"c:lloy l:laterial heh~vior unda:- cyclic loadin~ have h~cl"l de:::onstratetl in t~o pn:-HOS"!" pcog:::-o~s (-ehe! turbine bl2.de durability s·tuc1y ~-eported in NASi'!.

CR-165268 ar.:1 ';;;he combustor liner durabili'ty study reported in NASA CR-165250).

To.::! o!::ljecti'le ef this p';O~rai<1 is 'to deve~,op a unified ccns'I;:i"tU'l;iv~ li:odel for fi~i·1;e-cl<=~~:~n·~ s"truc~t!:",z.l a~lyzi3 of "turbin~ er!gil"'l~ h~r:; sec'i;ic~l com,'Onent:.s.

This efio::';; cc~,:;'t:!.tlrtes a di;!'::erer.'t approach for nor..li4"!eZlI" fin:i:tc-eleh:ent co~,!Ju";cr" cod~s. Wi1ico ha"!e hereta:':c:-e be'2n based on clas::ical inelz.stic me'thods.

A un::':;:ie~ cO;1::t:I:tu'c:!.·J€ ':;heor"Y H~ll 2.'!o:!.d the zir.::;>l:!.fy:i,r:.:; aszur,1,tiom: of classic.:.: ~~heor:r c:r.:! si- • .:.uld ~:~-::>re c:cCI.ll""<.'!;(~ly rc~;:-es:;::.~t t!lo. l::~~vior of zu,~rcl:'cy l::<r;:eri<:l:!.;:; under c~!clic lcadii"1; contli't;ions a::1:l high ter:tpeI"Q.t~I"e e!:wiI"ar::n'-':n·:;s. l'icG21 de'Je!o!"::',ent ~:1.11 b2 Gi::"ected tO~Tc:nl isotropic, ca~t n:!.c!~,~l-D:lS~ al:!.oy$ us~d for .;;.:lLcool~d "tu::"!:::inc bl.;;.des <:.r-..::1 vanes. The Ccn"!;ractor u:~ll S~!02C':; ~ Ease l'!c:ter-i21 fer r.:odal develop~~nJc ~r:d c:!l IU'c'2rnate l1ate::"i",l fo!"

v2I":i.:Z'ica:;ior. pU:"iJos~o:; :-ro~ a list at "thr~a alleys sp2cified by NASA. The c::!:.:~~cla-t.2 a:lo!t"s r(~~JLe::cnt a cros::-sec~ion of turbi~a hlade c:nd v.un~ In:ltcrials o:f 5,0':;2r~st ':;0 1:o'.;h laI"~,~ ar.j ::;r.l<:.ll si=e er..6in~ mZlr~factur"ers. tlate::"ial stoc!:: fa;:" ':;i:e 2:.sc C!l"!d :'.1 ~(;:!"'n;;:te fr::!tcric:ls H::ll be supplied to the Can':;::"actor by the G:..,,)~/ern:~'2rr:; • Th-:! co:~::;r2c-;;u~1 ei':o:-'t; ~.;ill be ccr::~ucted in tw-o phases, a !3c!.sic Progra~.l of tw-a ye2.I"s dL!l·z:~:1.on <:nd an optiena: fcll01-l-on r-I"ograll'l also of tuo yea::"s ~uration. In _-the Basic Program, a unified consti-tu-tive model will be eeveloped foe the pcediction of -the structural response of isoo;;ropic materials for the temperatures and strain ranges characteris-tic of cool~d turbine vanes in advanced gas turbine engines. A da-~a base of uniaxial and mul -tia:dal mate.iill proper-ti~s requir-ed for- -the cons-titutive !Rodel will be ob-tained for t!-:.e :&ls~ l1aterial. The constitutive model will be incorporated into a finite-elem.ent com~uter code. An evaluation will be made of the capability of the analytical method to ~redict structural respotlse for rnultiaxial stresz states and nonisothermal conditions by conducting the~momechanica1 loading and benchmarlc notch verification experim.ents and analyses. As a final evaluation of the analytical methods, a structural analysis llill be per~ormed for a hot section component fabricated of the Base Haterial for simulated engine operating conditions. In the optional program entitled Option 1, further development will be undert;;tken to consider thermal history effects and to corr-ec-!; any deficiencies ir,-dica~.;ed in the constitutive model or in the computational algorithms in the code. The material property test procedure lli11 be developed to miniiolize the amount of testing required, estimate the r.10dcl material constants from conventional prop'2rty data, and accoun-I; for coating e:ffects. In addition, the constitutive model development Hill be verified for an al-i;ernate material.

CONSTITUTIVE HODEL DEVELOPMENT

FOR ISOTROPIC MATERIALS

OBJECTIVE: ......

'"

'-J

TO DEVELOP A UNIFIED CONSTITUTIVE MODEL FOR

REPRESENTING CYCLIC INELASTIC BEHAVIOR OF

ISOTROPIC CAST HICKEL-BASE ALLOYS USED FOR

AIRCOOLED GAS TURBINE BLADES AND VANES.

II •.

BASIC PROGRAM SCREENING OF CANDIDATE CONSTITUTIVE MODELS N ~

SPECIMEN FABRICATION AND TESTING

UNIAXIAL EVALUATION OF CONSTITUTIVE MODELS

BASIC PROSRAt1

IMPLEMENTATION OF NODELS IN F. E. CODE

N .....

~

NUL TIAXIAL EVALUATION OF CONSTITUTIVE NODELS

BEOCHHARK NOTCH EXPERIMENTS

BASIC PROGRAH COMPONENT DEMONSTRATION PROBLEM N N o DELIVERY OF COMPUTER CODE TO NASA

OPTION 1

DEVELOPMENT OF HAT. PROP. TEST PROCEDURE

FINAL DEVELOPMENT OF CONSTITUTIVE MODEL

N N ,.....

HODEL VERIFICATION FOR ALTERNATE MATERIAL

LIFE PREDICTION AND CONSTITUTIVE HODELS FOR ANISOTROPIC MATERIALS Robert C. Bill Propulsion Laboratory AVRADCOM Research and Technology Laboratories Lewis Research Center Cleveland, Ohio The trend toward improved engine efficiency and durability is placing increased demands on gas turbine materials, especially in the hot·section. New materials and coatings are being developed to meet these demands. A particular area of challenge is in the turbine airfoil components. Here single (SC) and directionally solidified or recrystallized (DSR) polycrystalline materials are finding application. A difficulty impeding the full implementation of SC or DSR materials is the limited knowledge and understanding of failure (crack initiation) mechanisms and constitutive behavior.

The intent of this program is to develop a basic understanding of cyclic creep-fatigue deformation mechanisms and damage accumulation, a capability for reliable life prediction, and the ability to model the constitutive behavior of anisotropic SC and DSR materials employed in turbine airfoils. Four options comprise the program, and the work breakdown for each option reflects a distinct concern for two classes of anisotropic mate~ials, SC and DSR materials, at temperatures encountered in the primary gas path (airfoil temperatures), and at temperatures typical of the blade root attachment and shank area. Work directed toward the higher temperature area of concern in the primary gas path includes effects of coatings on the behavior and properties of the materials of interest. The blade root attachment work areas will address the effects· of stress concentrations associated with attachment features.

· CONTRACT: CYCLIC CONSTITUTIVE MODELING AND LIFE PREDICTION METHODS

FOR ANISOTROPIC MATERIALS.

OBJECTIVE: DEV~LOP AND VERIFY CYCLIC MATERIALS CONSTITUTIVE MODELS AND

N N ~

LIFE PREDICTION METHODS FOR COMPONENT SPECIFIC ANISOTROPIC

MATERIALS FOR USE IN STRUCTURAL ANALYSIS rOMPUTER PROGRAMS.

DURATION: 5 YEAR, 35 MAN-YEAR EFFORT.

FOUR PHASE PROGRAMS, EACH PHASE AN OPTION

PROGRAM STRUCTURE:

o PHASE 1 - COATED SC AIRFOILS

o PHASE 2 - COATED DS AIRFOILS

o PHASE 3 - UNCOATED SC BLADE ROOT

o PHASE 4 - UNCOATED DS BLADE ROOT

N ~ APPROACH:

o SELECT MATERIALS AND COATINGS

o SCREEN ADVANCED CONSTITUTIVE AND LIFE PREDICTION

MODELS; SELECT BEST

o INTEGRATE MODELS WITH STRUCTURAL ANALYSIS PROGRAMS

o VERIFY IN SIMULATED COMPONENT TEST

PROGRAM WILL ADDRESS:

o CREEP-FATIGUE

o THERMOMECHANICAL FATIGUE

o ORIENTATION EFFECTS

o COATING/SUBSTRATE INTERACTIONS

N N

o BIAXIAL LOADING

0\

o ATTACHMENT STRESS CONCENTRATIONS

HOST High Temperature Crack Propagation Thomas W. Orange National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 This proSram will attempt to build on the results o~ the Pre-HOST proSram I described ~esterda~, usins the latest anal~tical and experimental ~racture mechanics techniGues. The" anticipated proSram and i ts lui~ ~eatures are slJmmarized on the ~i rst sl ide. A ~lowchart is sho~n on the next slide, and I will describe the work to be done.

First we will attempt to extract additional in~ormation ~rom the literature and ~rom the Pre-HOST proSram. Specifically, nonlinear ~inite element (NLFE) anal~ses o~ the tubular specimens will be made usins GAP ELEMENTS to calculate crack openinS loads. Then, usins these loads, the data will be re-anal~zed to see i~ improved correlations result.

Various specimen confiSurations will be evaluated ~or elevated-temperature isothermal and TMF testins with tension-compression loadins, With a new constraint beins that now crack mouth openins displacement (CMOD) is to be measured. The most suitable specimen will be selected. Available methods ~or measurinS CYCLIC CMOD at elevated temperatures on the selected specimen will be evaluated. Possible methods ~or measurins near-tip displacements are also to be identified. An -analos material- will be select~d. This will be a material suitable ~or simulatins hish-temperature material behavior but at temperatures onl~ a ~ew hundred desrees F above ambient. The object o~ the analo~ material is to permit well-instrumented tests to be run at reasonable temperatures to ease the instrumentation and cost problems.

An experimental proSram similar in scope and approach to Task IV o~ NAS3-22550 will be run with additional instrumentation (at least CMOD) and more detailed anal~sis (a NLFE proSram with Sap elements).

The specimen, measurement methods, and material previousl~ identi~ied will be used, with both isothermal and simple (linear) TMF tests to be run. A limited series o~ tests usins a di~ferent specimen con~iSuration will be run to see i~ the resultins Srowth rate correlation is indeed specimen-independent.

We intend ~o evaluate several ~ormulations which have been proposed ~or nonlinear ~racture anal~sis in the presence o~ thermal sradients which result in material inhomoseneitw. These will include the anal~ses o~ Blackburn et aI, Ainsworth et aI, Wilson & Yu, Kishimoto et aI, and Atluri. Methods and stratesies for per~orminS the necessar~ calculations using a NLFE program will be considered. The five most promisins ~ormulations will be evaluated usins the simple analytical model o~ an edSe crack in a larse plate, with a linear temperature Sradient (and/or correspondinS modulus variation) in the direction of crack propaSation and a uniform distribution normal to the crack plane. The five ~ormulations will be compared and two selected ~or ~urther evaluation. One will be that Jud~ed to have the most technical merit, the other-the best compromise between technical merit and computational ease. Then these two formulations will be evaluated for further use b~ modeling the actual specim~n geometrY and temperature gradient to be used in later tests. Next we wish to determine whether the analytical formulations identified previously actuallY enable one to correlate nonlinear crack growth in" the presence of thermal gradients. Specimens of the analog material will be tested under monotonic and cyclic load in the presence of a simple (nearly linear) thermal gradient.

At this point it will be NASA's option to proceed with the optional optional program as planned, to technicall~ redirect the program by re-negotiation, or to t~rminate.

The optional program will consist of three main elements. We intend to extend the analytical effort to include a comparison of path-independent RATE integrals (or time derivatives of path- independent integrals), and this will b. done in much the same manner as before. Then we will attempt to verif~ these anal~ses using the analog material at a somewhat higher temperature than before (into the creep range). Finally we will attempt to appl~ the knowledge gained using the analog material to predict and correlate crack propagation in a nickel-base alloy at temperatures t~pical of combustor liners.

ELASTOPLASTIC CRACK PROPAGATION AT ELEVATED TEMPERATURES OBJECT: DEVELOP METHODS FOR CHARACTERIZING & PREDICTING CRACK GROWTH AT ELEVATED TEMPERATURES CONSIDERING NONLINEAR MATERIAL BEHAVIOR, THERMAL GRADIENTS & THERMOMECHANICAL CYCLING.

SCOPE: FOUR-YEAR TWO-PHASE CONTRACT.

FEATURES: • SURVEY & COMPARISON OF CURRENT PATH-INDEPENDENT INTEGRALS.

• COMPUTER TEST OF ~5 PATH-INDEPENDENT INTEGRALS USING SIMPLE PROBLEM.

• EXTENSIVE USE OF CRACK DISPLACEMENT MEASUREMENTS FOR BETTER UNDERSTANDING.

• ANALOG MATERIAL AT MODERATE TEMPERATURES (~500F) FOR PHASE I: NI-BASE ALLOY AT <2000F IN PHASE II.

• CYCLIC CRACK PROPAGATION·TEST WITH TEMPERATURE GRADIENTS, THERMOMECHANICAL CYCLING & CREEP.

BASIC PROGRAM RE-ANALYZE o EVALUATE SPECIMENS SURVEY , EVALUATE PRE-HOST DATA o CMOD. CTOD METHODS P-I INTEGRALS o ANALOGUE MATERIAL

+

"', EXPERIMENT: EXPERIMENT: I SO , TMF DA/DN TEMPERATURE GRADIEN OPTIONAL PROGRAM l ~ EVALUATE P-I EXPERIMENT: RATE INTEGRALS REPEAT WITH NI-BASE ALLOY

~

EXPERIMENT: DA/DT (SIMPLE) ~,

+

EXPERIf'ENT: EXPERIMENT: ..

...

VAIOT (COMPLEX) COMPLEX WITH , NI-BASE ALLOY 229 - SURFACE PROTECTION OVERVIEW Stanley R. Levine National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 Turbine engine hot section materials are subjected to aggressive chemical and thermomechanical environments. High temperature environmental attack of dollar intensive turbine components reduces turbine efficiency and can limit life. The bottom line, of course, is that high temperature oxidation and hot corrosion attack costs you money. The objective of materials durability research at Lewis is to understand the mechanisms of alloy and coating attack, and the effects. of interaction with the environment on mechanical behavior.

This base of understanding provides the foundation for developing life pre- diction methods and identifying strategies for controlling attack via advanced metallic and ceramic coatings. The Turbine Engine Hot Section Technology Project (HOST) augments the life prediction area of our program.

Our objective under HOST is to develop a first-cut integrated environmental attack life prediction methodology for hot section components. Under HOST we are concerned with oxidation and hot corrosion attack of metallic coatings as well as their degradation by interdiffusion with the substrate. The effects of the environment and coatings on creep/fatigue behavior are being addressed through a joint effort with the Fatigue sub-project. Finally, an initial effort will attempt to scope the problem of thermal barrier coating life prediction. Verificiation of models will be carried out through benchmark rig tests including a 4 atm. replaceable blade turbine and a 50 atm.

pressurized burner rig.

~'~~~;~;~:~,~~r~;~;O-C TURBINE ENGINE HOT SECTION TECHNOLOGY

NJ\SI\

~.R~"Cenl" SUR FA CE PRO TECT ION 0 V ER V I EW S. R. LEVINE MATERIALS DIVISION MATERIALS DURABILITY BRANCH

HIGH TEMPERATURE

ENVIRONMENTAL ATTACK

REDUCES EFFI CI ENCY

LIMITS UFE

rosrs$

SCHEMATIC OF MODES HI"GH·TEMPERATURE ArrACK

OXIDATION HOT CORROSION OXIDE SPALLING OXIDE FLUZING BY

#/e'

LIQUID SALT DEPOSITS , '- AI203 Na, CI, S CULPRITS: CYCLIC THERMAL STRESS SURFACE PROTECTION OBJECTIVE: DEVELOP AN INTEGRATED ENVIRONMENTAL AlTACK LIFE PREDICTION MElHODOLOGY FOR HOT SECTION COMPONENT LIFE APPROACH: • COMPILE FIELD FAILURE MODES DATA BASE • MODEL EFFECTS OF ENVIRONMENTAL ArrACK AND COATINGS ON CRACK INITIATION • CORROSIONr~ MODEL FOR AIRFOILS • COATING OXIDATION/DIFFUSION MODEL • COATING HOT CORROSION LIFE PREDICTION • THERMAL BARRIER COATINGS • MODEL VERIFICATIONIINTEGRATION COMMENTS: • THIS WILL BE A FIRST CUT • TIME, MANPOWER AND ARE INS UFFICIENT UNDER HOST TO DEVELOP A FULLY SATISFACTORY METHODOLOGY SURFACE PROTECTION I

I EXPECTED RES ULTS

84 I 85 PROGRAM ELEMENT FYBI 82 : 83 86 87

!

I !

ENVI RONMENTI MECHANICAL MODEL FOR EFFECTS OF ENVIR.

I PROPERTY INTERACTIONS ATTACK& COATINGS ON CRACK

I (TASKS OF CIF CONTRACTS) INITIATION I

I

I I MODEL TO PREDICT THE LOCATION & AIRFOIL DEPOSITION MODEL POTENTIAL SEVERITY OF CORROSIONI ATTACK OF TURBINE AIRFOILS COA T1NG LIFE PREDICTION CAPABILITY TO PREDICT COATING DEGRADATION ON BLADES, VANES, COMBUSTORS SURFACE PROTEC"TlON PRQG RAM aEMENT 81 82 83 84

I FY 85 86 87

ENVIRONMENT/MECH. PROP.

INTERACTIONS AIRFOIL DEPOSITION MODEL AIRFOIL MODEL

l

---

MODEL VERIFICATION (I) I

---

, COATING LIFE PREDICTION \

OXIDATION/DIFFUSION m

I I

.. __ 1

HOT CORROSION SURFACE CHEMISTRY I L. __ I : 1----- DUAL CYCLE A TIACK (I) --_.

,---

LIFE PREDICTION VERIFICATION L __ I

I

po-- THERMAL BARRIER LIFE PRED.

~ __ l II ]

___ J ____

I

RIG/ENGINE CORRELATION m

~

~---r---

I AIRFOIL DEPOSITION MODEL GOAL: DEVELOP THEORY TO PREDICT CORRODANT DEPOSITION ON TURBINE AIRFOILS DURATION: 36 MONTHS APPROACH: • GRANT - DAN ROSNER, YALE • EXTEND CHEMICALLY FROZEN BOUNDARY LAYER THEORY TO AI RFOILS - LAMINAR & TURBULENT FLOW - PRESSURE& TEMPERATURE OVER AIRFOIL • MULTI-COMPONENT CORRODANTS AIRFOIL DEPOSITION MODEL VERIFICATION GOAL: VERIFY DEPOSITION MODEL AND INTEGRATE WITH ALLOY CORROSION RATE MODEL .

DURATION: 48 MONTHS, IN-HOUSE APPROACH: USING 4 ATM REPLACEABLE BLADE TURBINE RIG - VERIFY AIRFOIL DEPOSITION MODEL - VERIFY ALLOY RATE MODEL (FROM R& T BASE) - INTEGRATE TO LOCATION/RATE MODEL COATING OXIDATION/DIFFUSION LIFE PREDICTION GOAL: TO DEVELOP AN IMPROVED METHODOLOGY FOR PREDICTING "OXIDATION LIFE OF METALLIC COATINGS DURATION: 36 MONTHS, IN-HOUSE APPROACH: SELECT, "PROCURE COATED SPECIMENS DETERMINE COATING LIFE VS TEMPERATURE FURNACE BURNER RIG MEASURE SPALLING PARAMETERS FROM ISOTHERMAL OXIDATION DETERMINE KINETICS OF DIFFUS IONAl DEGRADATION DEVELOP CYCLIC OXIDATION MODEL WITH MODIFICATION FOR COATING DIFFUSIONAL DEG"RADATION TEST MODEL BY VARIATION OF CYCLE FREQUENCY HOT CORROSION SURFACE CHEMISTRY GOAL: DETERMINE EFFECT OF SURFACE CHEMISTRY ON HOT CORROSION LIFE DURATION: 36 MONTHS APPROACH:- ANALYZE REPRESENTATIVE FIELD COMPONENTS - DETERMINE COATING LIVES FOR VARIATIONS IN ALLOY COATING COATING AGE AGING METHOD - MODEL RESULTS/SUGGEST METHODOLOGY FOR LIFE PREDICTION - VERIFY METHODOLOGY DUAL C Y C LEA TT A C.K GOAL: CHARACTERIZE THE EFFECT OF COMBINED OXIDATION/HOT CORROSION CYCLIC EXPOSURE ON LIVES OF METALLIC COATINGS DURATION: 36 MONTHS UN-HOUSE> APPROACH: DETERMINE MACH 0.3 BURNER RIG LIVES FOR COMBINED OXIDATIONI HOT CORROSION CYCLIC EXPOSURE - VARIATIONS IN DEPOSITION RATE - VARIATIONS IN TEMPERATURE LEVELS - VA RIATIONS IN COATING AGE FIT RESULTS TO EMPIRICAL MODEL LIFE PREDICTION VERIFICATION OBJECTIVE: VERIFY COATING LIFE PREDICTION METHODOLOGY WITH BURNER RIG BENCHMARK TESTS DURATION: 30 MONTHS APPROACH: PREDICT C()\ TING LIVES FOR SIMULATED LONG HAUL/SHORT HAUL COASTAL& INLAND MISSIONS SIMULATE MISSIONS&MEASURE COATING LIFE DIAGNOSE PREDICTIVE METHODOLOGY FOR DEFICIENCIES ", ;

THERMAL BARRIER COATINGS

\ ' .

" " " , " ' METAL SUBSTRATE • EXTEND LIFE • INCREASE TEMPERATURE AND lOR • REDUCE COOLING ENVIRONMENTAL COMPOSITION DEGRADATION • CERAMIC .PHASE CHANGES .eOND COAT • OXIDATION • DEPOSITION

THERMAL

BARRIER

COATING

PERFORMANCE

STRUCTURE • SPRAY PARAMETERS • POST SPRAY TREATMENTS • ALTERNATE PROCESSES TBC LIFE PREDICTION GOAL: TO DEVELOP AN IMPROVED DESIGN/LIFE PREDICTION METHODOLOGY FOR THERMAL BARRIER COATINGS DURATION: 36 MONTHS APPROACH: COMPILE COATING PROPERTY DATA PREDICT C()\ TING LIFE IN RIG& ENGINE VARIATIONS IN: COATING PROPERTIES COATING THICKNESS THERMAL CYCLE TEST PREDICTIVE CAPABILITY BY RIG (& ENGINE) TEST RIG/ENGINE CORRELATION GOAL: VERIFY CORROSION, EROSION AND COATING LIFE MODELS AT NEAR ENGINE CONDITIONS DURATION: 36 MONTHS (IN-HOUSE) APPROACH: VERIFY MODELS USING HIGH PRESSURE BURNER RIG DEPOS ITION/CORROSION METALLIC COATING LIFE THERMAL BARRIER COATING LIFE DETERMINE MODEL DEFICIENCIES ATIRIBUTABLE TO HIGH PRESSURE HIGH HEAT FLUX AIRFOIL DEPOSITION MODEL Fred J. Kohl National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 Aircraft gas turbine failures associated with sea-salt ingestion and sulfur~containing fuel impurities focus attention on salt deposition and the attendant hot corrosion and fouling of gas turbine blades. However, in the past, quantative understanding of deposition from gas turbine combustion gases has been impeded by the lack of a comprehensive yet tractable theoretical framework for organizing new deposition rate information. The present research program deals with the further development and exploitation of such a theory, and builds upon the foundation provided by previous NASA LeRC-sponsored research (Refs. 1-3 and references contained therein). The goal of this program is to develop the methodology to predict deposit evolution (deposition rate and subsequent flow of liquid deposits) as a function of fuel and air impurity content and relevant aerodynamic parameters for turbine airfoils.

The program is carri ed out under a HOST-supported grant, "Theory of Mass Transfer from Combustion Gases" (NAG 3-201), with Professor Da~iel E. Rosner and associates of the Chemical Engineering Department of Yale University.

The spectrum of deposition conditions encountered in gas turbine operations includes the mechanisms of vapor deposition, small particle deposition with thermophoresis. and larger particle deposition with inertial effects. In the present program the focus is on using a simplified version of the comprehensive multicomponent vapor diffusion formalism to make deposition predictions for (1) simple geometry collectors and (2) gas turbine .

bl.ade shapes, including both developing laminar and turbulent boundary layers. For the gas turbine blade the insights developed in previous programs are being combined with heat and mass transfer coefficient calculations using the· IISTAN 5 boundary layer code to predict vapor deposition rates and corresponding liquid layer thicknesses on turbine blades. A computer program is being written which utilizes the local values of the calculated deposition rate and skin friction 'to calculate the increment in liquid condensate layer growth along a collector surface. Preliminary results are now available for deposition and aerodynamic shear-driven flow of Na S0 on stationary 2 4 cylinders and turbine blades.

Detailed results of progress to date appear in several papers and preprints (Refs. 4-9), copies of which can be obtained from Professor D. E. Rosner at Yale University, Department of Engineering and Applied Science, New Haven, CT 06520.

REFERENCES 1. D. E. Rosner, B.-K. Chen, G. C. Fryburg, and F. J. Kohl: "Chemically Frozen Multicomponent Boundary Layer Theory of Salt and/9r Ash Deposition Rates from Combustion Gases," Comb. Sci. and Tech. 20, 87 (1979).

2. D. E. Rosner: "Thermal (Soret) Diffusion Effects on Interfacial Mass Transport Rates," Physicochemica"l Hydrodynamics 1, 159 (1980).

3. D. E. Rosner and K. Seshadri: "Experimental and Theoretical Studies of the Laws Governing Condensate Deposition from Combustion Gases," 18th Sympos. (Intl.) on Combustion, The Combustion Institute, 1981, p. 1385.

4. D. E. Rosner and J. Fernandez de la Mora: "Small Particle Transport Across Turbulent Nonisothermal Boundary Layers," ASME Trans., J. Engin.

for Power, in press, 1982.

5. S. G8koglu and D. E. Rosner: "Correlation of Thermophoretically-Modified Small Particle Deposition Rates in Forced Convection Systems with Variable Properties, Transpiration Cooling and/or Viscous Dissipation," submitted to Int. J. Heat and Mass Transfer, 1982.

6. R. Israel and D. E. Rosner: "Use of a Generalized Stokes Number to Determine the Aerodynamic Capture Efficiency of Non-Stokesian Particles from a Compressible Gas Flow," Aerosol Sci. and Tech., in press, 1982.

7. D. E. Rosner and J. Fernandez de la Mora: "Correlation and Prediction of Thermophoretic and Inertial Effects on Particle Deposition from Non-Isothermal Turbulent Boundary Layers," ASME Conf. on Particulate Laden Flows in Turbomachinery, St. Louis, MO, June 7-11, 1982.

. 1\ D. E. Rosner, D. Gunes, and N. Anous: "Aerodynamically-Driven Condensate 8.

Layer Thickness Distributions on Isothermal Cylindrical Surfaces," submitted to Chern. Engin. Commun., 1982.

9. D. E. Rosner, S. G~kqglu, and R. Israel: "Rational Engineering Correlations of Diffusional and Inertial Particle Deposition Behavior in Non-Isothermal Forced Convection Environments," Engineering Foundation Int. Conference on Fouling of Heat Exchange Surfaces, White Haven, PA, Oct. 31-Nov. 5, 1982.

AIRFOIL DEPOSITI"ON" MODEL

'GRANT NAG 3-201: II THEORY OF MASS TRANSFER FROM COMBUSTION GASES,"

WITH PROFESSOR D. E. ROSNER, ChE DEPARTMENT, YALE UNIVERSITY

EMPHASIS: TRACE SALT VAPOR DEPOSITION AND CORRESPONDING SHEAR-

DRIVEN CONDENSATE LAYER FLOW .

OBJECTIVE: N

• OVERAll - DEVELOP MODEl. TO PREDICT CORRODANT DEPOS ITIONON

.I:'- U1

TURBINE AIRFOILS

." 1ST YEAR - MODEL DEPOSITION RATE FOR SEVERAL SIMPLE GEOMETRIES

• 2ND YEAR - PREDICT AND DISPLAY LIQUID LAYER EVOLUTION ON TURBINE

VANES AS A RESULT OF VAPOR DEPOSITION AND LIQUID LAYER FLOW

CS-82-2576

[HOT CORROSION PROCESS'

COMBUSTION GASES 2 NaCI +. S03'+ H 0 ~ N SO 2 a2 4 + 2HCI

,________ J °Z' Na, S

DEPOSITION BOUNDARYlAYER---- --- COMPOUNDS Na2S04 , CORROSION N +=- 0\ CS-82-2624

CHARACTERISTICS OF DEPOSITION FOR SPECTRUM OF PARTICLE SIZES

MA.SS SIZE DEPOSITION DEPOSITION TRANSPORT RANGE* TRANSPORT CHARACTERISTICS SPECIES MECHANISM MODE Atoms and Molecules Fick Diffusion Vapor I-lOA

1. T < Te

dp Diffusion (Vapors) Soret Diffusion 2. low 11 and deposition on Eddy Diffusion side away from line-of- sight 3. low sensitivety to T e- Tw

4. Rate levels off for Tw « T

dp o -1 Heavy Molecules Brownian Diffusion IOA-lO ~ m Vapor 1. Tdp= Te (Condensate Aerosols, Diffusion Eddy Diffusion 2. lowestll Transition Clusters, Submicron Thermophoresis 3. High sensitivity to Te- Tw N Particles) ~ -....J 4. Rate nearly linear with Te- Tw -1 Macroscopic 10 - l(X)~m Inertial Inertial Impaction 1. No appa rent Tdp Particles Eddy Impaction 2. Highest 11 3. Independent of Te- Tw 4. Preferential d~position on side facing flow * Mode of deposition is not fixed by particle size alone 11 = deposition or collection efficiency, Tdp = dew point temperature, Te = gas mainstream temperature, Tw = wall temperature CS-1l1-1l82

PREDICTED DEPENDENCE OF SODIUM SULFATE

DEPOSITION RATE ON PARTICLE SIZE

P = 12 atm, Na2S04 DEPOSITION

Te = 1423 K

10-

* FRACTION

CAPTURED.

10-

CFBL

BROWNIAN 0.8 =r:==

DEPOSITION

I

THEORY

10-

, , , , , , [~ ' III

10- ' II

3 2

1 0

10- . 10-

10 10

PARTICLE DIAMETER, Jlm

CS-81-U78

C HEM I CALL Y .E R 0 ZEN J! 0 UN DAR Y .LA YER THE 0 RY- C FBL

GOAL: PREDICT THE DEPOSITION RATE FOR TRACE INORGANIC SPECIES AS A

FUNCTION OF SEED LEVEL, COLLECTOR GEOMETRY, THERMAL AND FLUID

DYNAMIC PARAMETERS, ETC.

BASIC ASSUMPTIONS:

1. NO CONDENSATION OR CHEMICAL REACTION WITHIN THE

MAS S TRANSFER GASEOUS BOUNDARY LAYER

I\.)

~ I,Q

2. CHEMICAL EQUILIBRIUM EXISTS AT THE VAPOR-CONDENSATE

INTERFACE

3. CHEMICAL SPECIES FOR TRANSPORT ACROSS THE BOUNDARY

LAYER ARE VERY LOW IN CONCENTRATION

4. TRANSPORT BY BODY FORCES AND PRESSURE DIFFUSION IS

NEGLIGIBLE

CS-82-2575 ,I '"

VAPOR DEPOSITION THEORY -CFBL

y. J;:\

I.e...-<~-""-·~

y .· ..• GAS ·BOUNDARY LAYER

o

EXTERNAL STREAM DEPOSITION SURFACE f-.)

Ln o

INTEGRATED DEPOSITION FLUX IS GIVEN BY

(So ret)

(Fick)

t-

le

e,

int 0 (Dt - F (tu rbl-FilSoretl- Nu , i*R ScI'"- [AWj + -FjlS~r~tI - wi, w]

m

\ J \. ~ '-- ----..y . v

DRIVING FORCE

TRANSFER COEFFICIENT

CS-82-2579

MACH 0.3 BURNER RIG DEPOSITION OF Na S0 ON ROTATING CYLINDRICAL COLLECTORS

2 4

CFBL THEORY

o EXPERIMENTAL POINTS

NaCI SEED 40 40 Na2S04 SEED 30 30

o 0 0 CD

N VI ..- DEPOSITION

o 0 ~ 8 0 0 0 IMP. Na2S04

8 .• .

RATE, 20 20 + M. p. Na2S04

mgh 10 10 o I 'f - -" I 0 ' I I I I ......... , I 500 600 700 800 900 1000 1100 600 700 800 COLLECTOR TEMPERATURE, 0 C GS-82-2570 ~ II ••

LIQUID DEPOSIT LAYER FLOW

GOAL: PREDICT THE DISTRIBUTION OF LAYER THICKNESS WHEN AERODYNAMIC

SHEAR IS THE DOMINANT MECHANISM OF CONDENSATE FLOW ALONG

mE SURFACE

BASIC ASSUMPTIONS:

1. FILM IS THIN AND FLOW IS LAMINAR

2. LIQUID IS NEWTONIAN AND SHEAR STRESS ACROSS LAYER IS CONSTANT

3. OTHER BODY FORCES, SURFACE TENSION, ETC. ARE NEGLIGIBLE

N lJ1 N

4. SURFACE IS ISOTHERMAL

CS-82-2568

AERODYNAMICALLY-DRIVEN THIN CONDENSATE LAYER FLOW

I) l (x+ L\x)

GAS

Bl

I) l (x)

LIQUID

LAYER ___

y = 0

"

"

COllECTOR

N VI W

LIQUID lAYER THICKNESS, I)l(x), IS GOVERNED BY

al) l + l.. ('T w(x) .1) 2) = _ mil (x)

at ax 2~l l P l

CS-tJ2-2S73

DEPOSIT THICKNESS AS A FUNCTION OF POSITION

VAPOR DEPOSITION AND LIQUID LAYER FLOW (NO SHEDDING) , , , PRESSURE NI Lni SURFACE ~I

~Q

Re = IcY

SUCTION

Tw = CONST

SURFACE

-1.0

-.5 0 .5 1

1.5 0 2 4 6 8 12

DISTANCE DISTANCE CYLINDER STATOR BLADE .,: .. :.1"'.

CS-82-2571 ! II '0

TRANSIENT EVOLUTION OF DEPOSIT LAYER

STATIONARY CYLINDER IN CROSSFLOW

VAPOR DEPOSITION AND LIQUID LAYER FLOW (NO SHEDDING) T WALL = CONSTANT N U1 U1 TO.56

T 0.00 T 0.16

T 0.86 T 1.06 T 1. 51

RELA TIVE TIME (1) CS-82-2572 ........ -..-.:.~"'-

TRANSIENT EVOLUTION OF DEPOSIT LAYER

STATOR BLADE

VAPOR DEPOS ITION AND LlQUI D LAYER FLOW (NO SHEDDING)

T WALL = CONSTANT

N VI (J\

TO.oo TO.31

T 1. 01

T 2.01 T 4.81

T 3. 21

RELA TIVE TIME (1) CS-82-2578 ~ II ••

FUTURE EMPHASIS OF DEPOSITION THEORY AND LIQUID LAYER FLOW

• BLADE ROTATION

• NONISOlHERMAL SURFACE TEMPERATURE DISTRIBUTION

• SALT SHEDDING; STEADY STATE

• MUL TICOMPONENT VAPOR TRANSPORT

• SEED LEVEL TRANSIENTS

N U1 -..J

• ALTERNATE DEPOSITION MECHANISMS

A. CONVECTIV~ DIFFUSION

B. lHERMOPHORETIC ENHANCEMENT

C. PARTICLE IMPACTION

CS-82-2S69 "11 •• COATING LIFE PREDICTION Michael A. Gedwill National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 Metallic coatings are widely used on hot section components of advanced gas turbine engines in order to take full advantage of the strength capabilities of turbine materials. Proper design to coating life limits can allow components to operate either for longer times or at higher temperatures, both of which are cost effective. However, costly engine inspections and component refurbishment or replacement are made many times on a conservative basis because component life and/or reliability are generally unknown. An analytical method for predicting life of metallic coatings on turbine airfoils should, therefore, result in substantial savings in maintenance and materials costs as well as providing an improved basis for initial design. The work to be discussed herein addresses itself to developing an improved methodology for predicting cyclic oxidation life of metallic coating on gas turbine airfoils.

A cyclic oxidation/spalling model was developed at LeRC that predicts long time cyclic furnace oxidation behavior of alloys. The computer inputs for the model are obtained from simple, short-time isothermal oxidation tests.

In the present study, the model is being applied to an aluminide coating on U-700, a low pressure plasma sprayed (LPPS) NiCoCrA1Y coating on U-700, and a monolithic LPPS NiCoCrA1Y. An empirical diffusion model to account for coating degradation will be integrated with the oxidation/spalling model to predict coating life in cyclic furnace oxidation. The integrated model will then be verified/adjusted to predict cyclic burner oxidation.

Further verification/adjustment will lead to a life prediction model for coated turbine airfoils. Preliminary results of isothermal and cyclic furnace oxidation of aluminide coated U-700 are presented.

CYCliC OXIOAIION VISUAlIZAIiON II~ ·0 nO' ° ' 0 f~. h~ • *XI ~~., A:1~::~? ... ~ _A#:;:{:¥~;~r:@~ COATING LIFE PREDICTION ,[ifJQJ~r.~i~~~~~~~W Jlr~:-;;;~::'::-l'\ If1JlZ-.-:,'I\ ' OBJECTiVE- )[f!J} DEVELOP IMPROVED METHODOLOGY FOR PREDICTING CYCLIC OXIDATION Ttf (NO (f FIRST lUTING (yeLl LIFE OF METALLIC COATINGS ON GAS TURBINE AIRFOilS

-&r:~~~jffr~~~~

JJjrrnIffi~ili~~~~~§?ili*~ NUA 100 ' .... lING

1.If(rtdJ~P~

AND COOtING ''flUS

Ttt: (,.0 (J n. FIRST

COOliNG (YCl( tv (j\ o SEM MICROGRAPHS OF SPALLED AI203 SCALES PREDICTIONS FROM ISOTHERMAL DATA AGREE WITH CYCLIC DATA NICrAIZr. 12000 C 2,0 .".~~ :"':"':"'l~ ~u~ '"...'. ,; '::r ~"~t, """:::',~ o 08SfRVlD ':::-- ... ~¥~ ,~, ~~>.

--PR£DICI(D ;,.a.;: _l· __ ,'- .. ~ [XPOSED1J"" :~~;e:r:.. .. , 0( 'l~T~~, !::;.,"';;':~ -~-; I' ':~-O~J'I:D' E'~ ," ~'~ ".,','1. 1\ ..

, -., ti~.1,..e..~ . . '~z~·,·~

~

SPECIFIC -LD

"t;, •• y',./., ~~":'t"!J':':':';~"

:-.1(.'1 ... ~ . 1'1. .......... "t, '. .. .. ", ..

MIGHT """,-~ ... "...'III .t:... ' • CI\AHC( • ~~~:.c.tJ " ,,} ':;: ..

2, ""gb -zg INM.

........ ..:. ' ....... ~;,-~ ~ AI o..fQlMER -~::':'-'~ 'Jf""""'~'~''''''' Kp ~RtlM ISOTH£RMAl nST ",,~,::,",,~ -~'" , -10 - -"i:;:5- :..~ .... ", ,",S , ';I'~, ... .,... ' QU',COII

~

h;-1,q:p;;;, ~ __ .:. ~.,~,~x -4.0 0, Ni -40 AI, lzooO C, 217 CYCLES ~O~, ____ ~ ____ ~ ____ ~~ __ -=~ ____ ~ ____ ~ o XXXIl·2 A PP ROACH NICoCrAIYfU-100 ENVIRONMENTAL AND SUBSTRATE REACTIONS SHORT liME ISOIII[RMAL FURNAC[ OXIDAIION DEGRADE COATINGS AT 1001' TO 1100" C PARAMmRS FOR CYCLIC OXIDA IIONI SPAlll'lG MODEL CYCLIC FURNACE OXIDATION FOR MODEL VERlfiCAliON PRHlhUNARY COATING i LJ[ PREDICTION MODEL • .I X250.

VIRJICA II0N/ADJUSWfNIISI AS-DEPOSITED NiCrAIY COATING Of MODEL FOR CYCLIC BURN£R OXIDAliOO N Q'\ .....

ISOTHERMAL FURNACE OXIDATION OF ALUMINIDE COATED U·700 COATING LIFE PREDICTION 1100 C LIFE • F (f0' '0) L5

WHERE '0 • OXIDATION COMPONENT FOR GROWTH/SPAlliNG ~ OXIDE SCALE

SPECIFIC WEIGHT '0 • DIFFUSION COMPONENT FOR CRITICAL ElEMENT(S) OF THE COATING CHANGE Img/cm2) XXXlI-3 CYCLIC FURNACE OXIDATION OF ALUMINIDE COATED U-700 ALUMINIOE COATED U·700 AFTER FURNACE OXIDATION AT llOO~ C IlOO" C 2.0 ~!I'I"'o;;o,.~-~ 'r ~ ... "'-, "', .-'" rt.!..! ' :;.' , .' ;~':kt\·, rqi5tr-, ~ .• '?-....q(W.'l~<~- .. -::~/~':i J.~...,.J"'. 'l :--".:..~i'l)I.r,:.(" : l! JI ~~ JJ~Q':l:~'~';, .~\ ~~d~'~ .~ .' ,- I ""'

\I ~, ,'. ' '. JI

SPECifiC WEICHT -to ,~ ~ •. ' f,,;Q c¢,' .:,,',.:. .• O;~, -v. , .. , ,- (,'( ,&-;: .;"" L..J c'" .. "', . v CHANG~ 'i'./- ,~,~ ..... ~ ,"- \,} (rngltm J ~4i,; [) 0': ~¢,o'~ Pelf.,. ~ (3- .. 1....). \..,. \..0 1 ! t(.'t"v <; 0 ( ~y;. Q' • ~ -2.0 "0 ,,- "'Coo<J g 00. "0.'. {r2 ,0:; ~1 \.. J." ,9 . u, ' ,0__ .. ~D ~_~'.\J'_""'~A!.~l.d.OI mm -lO 300 hr ISOTHERMAL 300-1 h r CYCLES I \I -4.0~ ~ N 0- N ALUMINIOE COATED U·700 AFTER FURNACE OXIDATION AT 1100 C ALUMINIDE COATED U·700 AFTER FURNACE OXIDATION AT llOO° C . i,'4i:o';;T'~.ll~\j··th,·~·,\·~·tj··'", ~c~~am ';(~'~;:"'iJ::'{."' ~'1': ~~:Zl

[

,- ~ ~1:.' ~ ~'~ .-" ':'. ',': /. I, ~ . " .". ," ,~;:<". .;' • .;~~~~~#i~~~~': j-. -.,;/l.. !'.' ,~, r , ... ""-:- ..... , ........ -)\...~;....j ,>.., - '~~~ " ) f) .":.,.

. )

) " ) ,,,:\,' I) ~ \.

~~' ... ti' , c ..

., ,... , , ti vi, ~ ..

( - •• ;'.~ • J . ) ~I 0

.. \ I. ~

~ I ,." "J <" - ( :) """(; .;, .. ' • oJ' : .. ," , ~: Jr":;~~'~ .01 mm .~ -?' ", ,~~, .. ~j".r'.

c .Olmm' II" ; ~. r. .

" 100-1 hr CYCLES 100 hr ISOTHERMAL 500-1 h r CYCLES 1000-1 h r CYCLES XXXII-4 EFFECTS OF SURFACE CHEMISTRY ON HOT CORROSION LIFE OVERVIEW John Merutka National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 This program con~entrates on analyzing a limited number of hot corroded components from the field and the carrying out of a series of controlled laboratory experiments to establish the effects of oxide scale and coating chemistry on hot corrosion life. This is to be determined principally from the length of'the incubation period, the investigation of the mechanisms of hot corrosion attack, and the fitting of the data generated from the test exposure experiments to an empirical life prediction model. It is a six task program.

GENERAL SCOPE OF WORK Task I involves the analysis of six field com~onents which were removed from service. The hot corrosion condition of these six will vary from slight to massive attack. Concurrent with the metallurgical analysis of field components in Task I, specimens of bare and coated alloys will be subjected in Task II to exposures in a high velocity burner rig (under conditions specified by NASA- LeRC) for not more than 1000 hours or until hot corrosion occurs. In Task III, the Contractor shall age specimens (bare and coated) in an inert atmosphere, in furnace oxidation (cyclic and isothermal), and in cyclic high velocity burner rig oxidation at 1100C (20l2F) for 100, 300, 600, and 1000 hours. In Task IV, the Contractor shall determine the effect of the various aging treatments on the hot corrosion mechanisms involved under the burner rig conditions specified in Task II.

Throughout Tasks I through IV, the results should be viewed not only in terms of identifying a model for the actual materials and test conditions run, but from the point of view of identifying a methodology whereby a life prediction model for other materials can be developed based on the results of one or more simple laboratory tests. After all the test exposures, the Contractor will review all the data and provide: (1) a preliminary hot corrosion life prediction model and (2) a recommendation of other test parameters to be evaluated so that simple laboratory tests can be used to predict hot corrosion life. The methodology to develop a hot corrosion life prediction technique shall be submitted to the NASA Project Manager for review and approval.

In Task V, based on NASA Project Manager's approval, the Contractor shall complete an experiment to determine the capability of the suggested methodology to predict the hot corrosion life of selected alloys and coatings.

Th~ last Task, VI, covers the reporting requirements.

2M .

RFP 3· 412m

OBJECTMa DETERMINE EfFECTS OF SURFACE OifMISTRY (If HOT CORROSION LIE BACXGROUN), PRIMARY 1Il0000ISMS CJ= HOT CORROSION. R.lJ)(1NG OFaxmE I ...

SCALES BY UQUID SALlS • RIG lESlS GIVE INCONSISTANT RESULlS • NEW lECHNl1lE DEVELOPS) TO • DE1'ERMItE mE INCU8A11Off1HRESHOLD PERIOD

• CARRY our REPRODUCIBlE HOT CORROSION 1ES1S

TASK I TASK II EVALUATION OF LABORAlORYHOT FIBD COMPONENTS CORROSION TESTS TASK III AG/te EXPERIMENTS TASK IV HOT CORROSION TESlS OF AGED SPECIMENS TASK V HOT CORROSION UFE PREDICTION TASK VI REPORTING • ANALYSIS OF HOT CORROSION COMPONENTS FROM THE RELD • C(WfROLLED LABORATORY EXPERIMENTS TO ESTABUSH EFF£ClS OF SURFACE CHEMISTRY ON HOT CORROSION UFE • DEVElOPMENT OF EMPIRICAL UFE PREDICTION MODB. BASED ON DATA GENERATED

EFfECTS OF SURFACE CHEMISTRY ON HOT CORROSION UFE

TASK I EYAWA110N fE FIELD COMPOfOTS SIX FELD COMPMNTS (1.1111£ 10 MASSI~ CORROSIONJ EYAWAl10N (METAllURGICAL AND CHEMlCAU 'TASK IllABORAlORY HOT CORROSION 1ES1S BUINR RIG CCJ4DmoNS: 0.3 MACH, pRE-CONDmONED AIR MAlERlALS, U700 AND CONTRAC'roR'S CHOICE - BARE AND COAlED (DUPUCATES' COAllNGS,. RT21 AWMlNlDE, LOW PRESSURE PlASMA NICoCrAly, CONTRAC'roR'S ./ CHOICE .

CYClE: 60 mfn HOT, 6 min AIR BlAST CPOL ..

SPECWIN SURFACE TEMPERATURE: 900C (l15Cf) TIIt£a DJJ HOURS OR UNTIL HOT CORROSION OCCURS RUN ADDmOOL SPECIMENS Im, 300, 500 'HOURS, nr.£ NOT 10 EXam 213rds

a: lHE TIME IN WHICH HOT CORROSION OCCURS

MlNI1OR: VISUAL AND INDUCTANCE MRY 20 CYCLES EVALUATION, IlETAiJ.uRGICAL AND CHfMlCAL (OXIDE, ALLOY AND COATING COMPOSmON AND STRUCTURE) TASK III AGlte EXPERIN£NTS TEMPERATURE: IlOOC (2DJ2fl MAlERlALSs AS IN TASK II (TRIPUCATES) AGING CONDmONS: TIME: lDO, ]X), 600, AND um h rs ENVIRONMENT: INERT; ISOnfERMAL FURNACE OXIDATION; CYClIC FURNACE OXIDATION; CYClIC BURNER RIG OXIDATION . MONITOR: INDUCTANCE CHANGES AND WEIGHT CHANGES AS APPROPRIATE CHARACTERIZATION: ONE SPECIMEN PER CONDmON AS IN TASK II

EFFECTS OF SURFACE CHEMISTRY ON HOT CORROSION LIFE

TASK IV HOT CORROSION 1ES1S OF AGm SPECIMENS , lEST CONDmON: AS IN TASK II, UNTIL HOT CORROSION OCCURS (DUPlICATES) AIlNlTOR: VISUAL AND INDUCTANCE EVERY 3J HOUR PERIOD HOT CORROSION OCCURS: VISUAL SIGNS FOR THREE 3J hr PERIODS.

EVAlUATION: (METALLURGICAL AND CHEMICAU PROPOSE: PRB.IMlNARY HOT CORROSION LIFE PREDICnON t1DDEL SUGGEST: ft£TH(lOOlOGY TO PREDICT HOT CORROSION LIFE BASED ON LAB ~PER~ .

TASK V HOT CORROSION UFE PRmlCTION • VERIFY UFE PRmlCTION MODEL • TEST METHODOLOGY TASK VI REPORTING REQUIRENiNTS FINANCIAL MONTHLY ANNUAL FINAL ORAL PRESENTATIONS COMBUSTION HOT SECTION TECHNOLOGY David B. Ercegovic Propulsion Laboratory AVRADCOM Research and Technology Laboratories Lewis Research Center . Cleveland, Ohio The overall objective of the Turbine Engine Hot Section " Technology Combustion Project is to develop and verify improved and more accurate analysis methods for increasing the ability to design with confidence the combustion system for advanced aircraft turbine engines. The analysis methods developed will" be generically applicable to combustion systems and not restricted to one specific engine or manufacturer.

This projects app~oach is to first assess and evaluate existing combustor aerothermal analysis models by means of a contracted effort initiated during FY '82. This evaluation effort will quantify known models strengths and deficiencies. A balanced contract and in-house program will then be conducted to support, focus, and accelerate the development of new methods to more accurately predict the physical phenomena occurring within the combustor. This balanced program will include both analytical and experimental research efforts in the areas of aerothermal modeling and liner cyclic life.

It is expected that the combustor model development effort will generate improved understanding in the areas of: high pressure flame radiation characteristics, model numerical methods and solution schemes, complex geometrical "boundary conditions, fuel spray - flow field interactions, combustion kinetics, flow and mixing of dilution jets, turbulence ~nd heat transfer, and soot and carbon formation. The primary in-house effort in this area will be the determination of high pressure flame radiation characteristics in a full annular combustor. This experiment will be conducted in the NASA LeRC High Pressure Facility with the results compiled into a comprehensive flame radiation and liner heat flux model.

In the area of liner cyclic life, HOST will develop a test apparatus to economically determine combustor thermal strains and cyclic life. This test apparatus will be run in-house at NASA LeRC and will be the test vehicle for many of the advanced high temperature instruments developed under HOST sponsorship.

The fundamental data generated in this project will be used to assess and develop current analytical liner life programs.

TURBINE ENGINE HOT SECTION TECfD'IOLOGY

NJ\SI\

OBJECTIVE TO DEVELOP HIPROVED ANALYTICAL ~10DELS OF THE INTERilAL CmlBUSTOR FLOH FIELD Arm LHIER HEAT TRA:ISFER AS A r1EAtlS TO SHORTEN COflBUSTOR DEVELopr1ENT TI ME AND INCREASE TURBINE ENGINE HOT SECTIml LIFE.

TURBINE ENGINE HOT SECTION TECfD'IOLOGY NJ\SI\

APPROACH • UTILIZE EXISTING f,10DELS - DETERMINE THEIR DEFICIE~lCIES • CONDUCT SUPPORTING RESEARCH TO IrlPROVE PHYSICAL f'10DELS • REFINE tl0DELS TO IMPROVE flUf1ERICS A~m tlUr1ERICAL DIFFUSION • HnEGRATE ilEI/ MID H1PROVED ROUTHlES uno EXISTING f10DELS AND VERIFY THEIR IMPROVED PREDICTIVE CAPABILITY I:-H2 0-11 '11.1 I','!

I'd' TURBINE ENGINE HOT SECTION TECHNOLOGY

NI\S/\

l t:WI: ."".1:.1/(.11 Ccntcr

COMBUSTION

I

FISCAL YEAR

PROGRAM EXPECTED

ELEMENT

81 82 83 84 85 86 87 RESULT

AEROTHERMAL MODELING KEY MODEL AND DATA

I I

ASSESSMENT DEFICIENCIES IDENTIFIED ~ 9 COMBUSTION MODELING NEW PHYSICAL MODELS

•

DEVELOPMENT AND COMPUTING METHODS N ......

'" Y MUL TIPLE JET DlLUTIDN EXIT TEMPERATURE PROFILE

•

•

I I

MIXING PREDICTION TECHNOLOGY (IH) FLAME RADIATION/HEAT HIGH PRESSURE FLAME I .

FLUX RADIATION AND HEAT FLUX (IH) DILUTION JET ANALYSIS JET MIXING MODEL

I

i (IH) i LINER CYCLIC RIG CYCLIC TEST FACILITY

I

I CO 82-Il18b

AEHOTHERMAL MODELIUG PROGHAM

E~~

PRINCIPAL INVESTIGATOR: S. K. SRIVATSAJ

GARRETT TURBINE ENGINE COMPANYJPHOENIXJ

ARIZONA

OBJECTIVE: ~ III »

ASSESS THE CURRENT STATE-OF-THE-ART AND IDENTIFY THE ""5 IT1

""5 ;0 CD 0 rT -I DEFICIENCIES IN CURRENT AEROTHERMAL MODELS FOR GAS- rT :I: IT1 ::r -I en :::0

"

os:::.

3: CD ""5 »

TURBINE COMBUSTORS

:::30'"7' r --' ........

N X:::3 3: '-I ..

CD en W ""5 0 » 1"T'l ....

IT1 ""5:::3< r .... ~ III .....

N .... rT :z O:::3Vl ~ :::3 CD III III n :::0

"

3 ~ "'C III ~ :::3 3: '< GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA

lffir PRUGRAM APPROACH OUTLINE - _ .. _J

TASK 1

1.1 MODEL UEFINITION

1.2 DATA BASE GENERATION

1.3 BENCHMARK TEST CASE DEFINITION

~ ~ ~

TASK 2

2.1 MODEL EXECUTION

2.2 MODEL ASSESSMENT

2.3 PROGRAM PLAN FOR MODEL IMPROVEMENT

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA " SUBTASK 1.1 - MODEL DEFINITION .

E;~

DESCRIBE COMPUTER PROGRAMS TO BE USED FOR ~10DEL ASSESSr1ENT

COMBUSTOR PERFORMANCE MODEL (2-D AND 3-D)

NEAR-WALL MODEL

DESCRIBE PHYSICAL SUBMODELS TO BE ASSESSED

t-.>

rURBULENCE MODELS

'-J I.J1

GASEOUS COMBUSTION MODELS

SPRAY COMBUSTION MODELS

SOOT FORMATION/OXIDATION MODEL

RADIATION MODELS

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORA liON PHOENIX, ARIZONA ......

ARRETI'

SUBTASK 1. 2 - DATA BASE GENERATION

.....

~

CLASS IFY AND REVI HI AVAILABLE DATA BASE

o COMPLEXITY OF FLOW FIELD

o PARABOLIC FLOW

o STREAMLINE CURVATURE

o RECIRCULATING FLOWS WID SWIRL

N -...J 0\

o SWIRLING FLOWS WITH AND WIO RECIRCULATION

o PHYS I CAL/CHHlI CAL PROCESSES .

o NON REACTING FLOWS

o REACTING FLOWS

o GASEOUS COMBUSTIUN

o SPRAY COMBUSTION

o . IDI:.ALIZED FLUW ELEt1ENTS AND PRACTICAL COMBUSTORS GARRETT TURBINE ENGINE COMPANY

A DIVISION OF THE GARRETT CORPORA liON PHOENIX. ARIZONA " .

~

SUBTASK 1.3 - BENCHMARK TEST CASE DEFINITION

~

o TURBULENCE MODEL

o FLOW OVER A FLAT PLATE:UJ U'2 uV'J K

J

-

'2 '2 '2 "

o HID STREAM I1IXING LAYER WITHOUT RECIRCULATION: U

J U J V J W J U V

o MIXING OF UNCONFINED COAXIAL JETS: UJ U'2 V'2 U'V'

J J

o DEVELOPING PIPE FLOW: U VJ U'2 V'2 U'V'

J

J J

N -...J -...J

o FLOW IN A CURVED DUCT: UJ VJ U'2 V'2 W'2 U'V'

J J J

o fLOW OVER A PLANE STEP: UJ U'2~. V'2 U'V'

J

'2 '2 '2 "

o r1IXING OF CONFINED COAXIAL JETS W·ITH RECIRCULATION:

U I VI U I. V I W I U V

o MIXING OF UNCONFINED COAXIAL SWIRLING JETS wIn RECIRCULATION:

'2. ,,'2 W'2

U V U

I I I Y I I U 'V' I U-'V-'- I V'W'

o SWIRLING FLOW IN A PIPE EXPANSION WITH RECIRCULATION:

- - '2 ~ -'2

U'V'~I U W'1 V~W~

U J VI U I V I W I

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORA liON PHOENIX. ARIZONA

BENCHMARK TEST CASE DEFINITION (CONT'D)

E~~

o GASEOUS COf1BUSTION MODEL

o I-D LAMINAR FLAT FLAME: FUEL 02) CO) CO ) H 0) H ) T

2 2 2

o 2-D LAMI NAR D I FFUS I ON FLAf1E WITH RECI RCULATI ON: CH4) C02) CO) H~) H20) 02) N~)

VELOCITY T

o 2-D TURBULENT PREMIXED FLAME WITH KECIRCULATION: VELOCITY) CO) CO~) C Hg) H~O)

O N2 T

o 2-D TURBULENT DIFFUSION FLAME WIO RECIRCULATION: VELOCITY) T) CH4/H2/C02) CO)

N

H 0) 02) N2

co " o 2-D TURBULENT DIFFUSION FLANE WITH RECIRCULATION: flIXTURE fRACTION 02' N , CH ,

2 4

CO, C02' H~O, T

o 2-D TURBULENT) S\HRLING DIFFUSION FLAME HIlH RECIRCULATION: CO ) CO) H ) VELOCIT'(,T

2 2

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA "

BENCHMAHK TEST CASE DEFINITION (CONT/D)

E;~

o SPRAY COMBUSTION MODEL

o 2-D TURBULENT EVAPORATING SPRAY WIO RECIRCUL~TION: DROP SIZE DISTRIBUTIOfL TJ

NIXTUHE FHACTION GAS VELOCITY

J

o 2-D TURBULENT REACTING SPHAY \'1/0 RECIRCULATION: TJ COJ C02J O~U N J H J H 0J

2 2 2

GAS VELOCI rY

o 2-D TlJRRULENT REACTING SPRAY HITH RECIRCULATION: DROP NUr1BER DENSITY TJ 02 COJ

J J

N -...J \.0

C0 J CXHYJ GAS VELOCITY

o 2-D TURBULENT SWIRLING REACTING.SPRAY WITH RECIRCULATION: DROPLET VELOCITY

AND SIZE DISTRIBUTION TJ COJ

J

CO 02J CXHYJ GAS VELOCITY

LJ GARRETT TURBINE ENGINE A DIVISION OF THE GARRETT COR I PHOENIX. ARIZONA

BENCHMARK TEST CASE DEFINITION (CONT'O)

e~

o SOOT FORMATION/OXIDATION MODEL:

o 2-D LAr1INAR DIFFUSION CH -FLAME W/O RECIRCULATION: SMOKE CONCENTRATION

o 2-D TURBULENT DIFFUSION CH ) C H FLAME W/O RECIRCULATION: SMOKE CONCENTRATION

g"

4 3

AND RADIATION

o GAS-TUKBINE Cor1BUSTORS - GARRETT IN-HOUSE DATA: N

o COMBUSTOR DESIGN CRITERIA VALIDATION PROGRAM: VELOCITY) CO) CO ) CXHy) RADIArION

o

o UPRATE T-76 COMBUSTOR: PRIMARY ZONE: T) CO) CXHy

o TPE331-15 COMBUSTOR: LINER WALL TEMPERATURES

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA . il •.

191 SUMMARY MID FUTURE WORK -I

o SUMMARY

o DATA BASE FOR TURBULENCE AND GASEOUS COMBUSTION MODELS FAIRLY ADEQUATE

o DATA BASE FOR SPRAY MODELS LESS SATISFACTORY STATE

o DATA BASE FOR SOOT AND RADIATION MODELS RATHER INADEQUATE

N

~ 0 FUTURE WORK

o Cor1PUTATIONS OF THE BENCHMARK CASES WITH AEROTHERMAL MODEL

o IDENTIFY MODEL DEFICIENCIES

o PREPARE PROGRAM PLAN FOR r10DEL IMPROVEMENT.

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA AEROTHERMAL MODELING - PHASE I A Progress Report G. J. Sturgess Commerical Engineering, Pratt & Whitney Aircraft Group The objective of the program is to develop the computational fluid dynamics tools needed to improve combustor design, analysis and development. In the first phase, current models will be evaluated, shortcomings identified, and improvements recommended. These recommendations will be implemented in the second phase. The approach adopted is to evaluate state-of-the-art numerical code and physical models. The evaluation consists of a step-by-step procedure using benchmark experiments.

The program is divided into three major tasks: Task 1 is concerned with defining the models, establishing a data base, and identifying test cases from the data base; Task 2 involves running ·the model, evaluating its performance, and formulating a program plan to achieve the necessary improvements; Task 3 is concerned with management and reporting activities.

The contract with Pratt & Whitney Aircraft went into effect on 13 July 1982.

Task 1 is nearly completed and Task 2 has been started.

Figure 1 outlines the calculation procedure. The modeling which has been selected represents the state of the art. The approach consists of a finite difference solution of the time-averaged, steady state, primitive variable, elliptic form of the Reynolds equations. Standard TEACH-type numerics are used to solve the resulting equations. These include hybrid differencing, SI~WLE algorithm for the pressure field, line-by-line iterative solution using the AD! method and the tri-diagonal matrix algorithm (TDMA). Convergence is facilitated by using under-relaxation. The physical processes are modeled by a two-equation eddy viscosity model for turbulence; combustion is represented by a simple, irreversible, one-step chemical reaction whose rate is influenced only by the time scale of turbulence; the radiating medium is assumed to be gray and a flux method is used for radiation together with a gas emissivity obtained from a four gray gas model. The liquid fuel spray is treated by particle tracking using the PSIC technique, and turbulent diffusion of droplets is accounted for by a stochastic approach. Provision is made for the fuel to be either a pure substance or mUlti-component.

SPECIES TRANSPORT INSTANTANEOUS NAVIER- STOKES EQUATIONS ENERGY EQUATION EQUATION STATISTICAL DESCRIPTION 1---_---1 OF TURBULENCE CHEMISTRY OF REACTION TIME AVERAGED: RADIATION MODEL ...J Cl TURBULENCE MODEL otz EQUATIONS OF MOTION ~:::l tl)w ENERGY EQUATION >-0 SPECIES TRANSPORT :%:0 COMBUSTION MODEl- (1.::E EQUATIONS REACTION RATES GEOMETRY BREAKUP 1--_--, GRID L_---r70:;!;;;;;--,

.-.-----4

BOUNDARY CONDITIONS ..........................................................................................•................................................................

...

::J (1.

5 GRAPHICS o POST-PROCESSOR Figure 1 Flow Diagram of Calculation Procedure The models will be evaluated against experimental data using a'data base currently being prepared. In order to avoid difficulties in separating effects to assess the performance of individual models, wherever possible only benchmark quality experiments Hhich deal with one physical process at a time are being considered. If real combustor flows are calculated, they will only be considered demonstrations of potential. Ideally, the comparisons will proceed from simple flows to'complicated flows. Complicated flows will only be used to study the effects .of interaction between different physical mridels.

Experiments are required to test each of the physical models. An "ideal experiment" has been defined, and experiments in the literature are being compared against this ideal to assess their qualification as benchmark test cases in the data bas~. An initial selection of test cases has been made. This selection covers co-axial jets with and without swirl in a confined sudden expansion, co- and counter-swirling co-axial jets, a widely-spaced co-axial jet bluff bodY'diffusion flame, and a single jet in a crossflow. These cases represent component flows typical of those in the gas turbine combustor.

Additional test cases are being selected to broaden the study.

Calculation of the initial selection of test cas~s has commenced, although it is too early at the present to comment on the results.

<0 o N UJ N cf.", N N ,... ~ N N -,00

en

.....,

en

-

:::J

en

I-

en

..t: 0)

zo, 0) (.)

>

~

.-

ca

.....,

W

en

~ (.)

I- ~ :::J

....., 0) .~ c.

.~

z

ca ca

....., .c c.

~

<C en

o·

U

•

• •

•

OBJECTIVES

Develop computational fluid dynamics tools needed to

improve combustor design, analysis and development, by the following means: N

• Define aerothermal models in the combustor design process

"

• Establish a suitable data-base against which to test models

• Identify shortcomings in the data-base

• Evaluate performance of models and identify their limitations

• Recommend future work to complete the data-base

• Recommend a course to result in improved models

J27289·3 821309 E206 ! II ,.

APPROACH

• Use state-of-the-art numerical code and

physical models

N

• Evaluate physical models using a step-by-step

approach utilizing benc.hmark-quality experiments·

• Be cognizant of the influence of numerical diffusion

J272894 821309 ! il ,.

STATUS

• Contract went into effect 13 July, 1982 (NAS3-23524)

N ~

• Work is on schedule and budget

• There are no current problems of a technical nature

27289-5 821309 E206 ..

! il ,.

PROGRAM SCHEDULE

. I I Task I Wntten summary 1.1 Model definition Written assessment procedure

1.2 Data definition I

: I I

Recommendation of experimental program 1.3 Test case

y J I

definition J ~ Task 2 o 2.1 Run model J 2.2 Assess model J Complete technical effort 2.3 Program plan

l.~ I I

I Final report

Work plan Oral report Oral report Exec. summary .Y 'VI sz.1 I \l Task 3 L J 1 1.

Annual workshop July Aug Sept Oct Nov Dec Jan Feb Mar April May June J27289-6 821309 E206 ! il ,.

CJ)

I-

..J

:J

CJ)

w

a: a:

o

..,

<C

~

(TASK 1.1)

FLOW DIAGRAM OF CALCULATION-

PROCEDURE

SPECIES TRANSPORT INSTANTANEOUS NAVIER- STOKES EOUATIONS ENERGY EQUATION EQUATION u..

OVl STA TlSTICAL DESCRIPTION >z ...JQ OF TURBULENCE al ...

~ct CHEMISTRY W:::l VlO OF Vl w ct REACTION N N

'"

TIME AVERAGED: RADIATION MODEL ...JeI TURBULENCE MODEL ctz EQUATIONS OF MOTION u- _...J ENERGY EQUATION Vlw >0 'SPECIES TRANSPORT xO COMBUSTION MODEL - a..~ EQUATIONS REAC110N RATES GEOMETRY BREAKUP , [5z GRID ... 0 .. r.nMPUTER .. SOLUTION :::l- a.. ...

ALGORITHMS ~3 BOUNDARY Vl U CONDITIONS ..........................................................................

...

:::l a..

...

GRAPHICS :::l o POST-PROCESSOR J27289-8 821309 E206 • j/ po

(TASK 1.1)

MODELING SELECTED AS FOLLOWS: • TEACH numerics - hydrid differencing,

SIMPLE, ADI, TDMA, under-relaxation

• K-£ turbulence model

• Fuel spray treatment - particle tracking

N \0

using PSIC for a thick spray,~ with a

W

stochastic approach for turbulent diffusion

of droplets

• Combustion model - simple, irreversible,

one-step chemical reaction, with eddy

breakup burning rate

• Radiation '- grey gas transport equation

containing particle terms, with Barteld's

4 grey gas emissivity modified for collision

broadening at high pressure

J27289· 821309 , u ..

TASK 1.2

EVALUATION PROCEDURE:

• To avoid difficulties in successfully

separating effects to assess individual model

performance, work with benchmark-quality

experiments dealing with only one physical

process at a time

• Experiments will be required in the areas of: N • Thermal radiation \0 ~ • Fuel spray development and flow field interactions • Combustion reaction and chemical kinetics • Fluid mechanics .

• Soot formation

• Proceed from simple flows to complicated

flows. Complicated flows will only be used

to study interaction effects between different

physical models

• Calculations of three-dimensional flows andl

or real combustor flows should only be

considered as demonstrations of potential

J27289·10 821409 E206 . I' , .

...

. (TASK 1.2)

CRITERIA FOR THE IDEAL EXPERIMENT

• Minimum flow dimensionality

• Well-behaved flows

N VI

'"

• Continuous variation of test parameters

• Progression of flow complexity

• Extensive instrumentation

J27289-11 821409 E206 . II ,.

(TASK 1.2)

BASIS OF COMPARISONS BETWEEN

MODELS AND EXPERIMENTS:

In addition to field variables, each flow will

be categorized, and characteristic quantities

identified:

Category : Recirculation zone

E.g., N

Characteristic quantities : Position of stagnation points

1.0 0'\

Maximum reverse velocity

Mass flow rate recirculated

Quantitative ability to predict the characteristic

quantities with variation of test parameters

can then be assessed

E.g.,

Category : Recirculation Zone

: Swirl number

Test parameters

Step height

Bluff body blockage

Heat release rate

J27289-12 821509 E206 ~ II ,.

(TASK 1.3)

RECOMMENDATIONS WITHIN THE SCOPE

OF THE PRESENT PROGRAM TO FilL VOIDS

IN THE EXPERIMENTAL DATA-BASE:

• Although there are many voids in the current data base

N

the nature of the ideal benchmark experiments

I.D

"

required to fill. them with respect to time,

instrumentation, facilities, and cost far exceed

the scope of the present contract

• It is recommended that no experimental work

be carried out as part of this contract

J27289-13 821609 E207 ! jl p,

(TASK·1.3)

INITIAL SELECTION OF TEST CASES

• Johnson (UTRC)

• Co-axial jets in confined sudden expansion

(NAS3-22771 )

• Co-axial jets with swirl in confined sudden expansior

N \0

• Gouldin (Cornell)

• Co and counter-swirl co-axial jets

(NSF-R ANN-GI-36538/NSG-3019)

• Roquemore (APL)

• Widely~spac.ed co-axial jet diffusion flame

• Greber (Case Western Reserve)

• Jet in a cross-flow (NGR-36-027-008)

J28289·14 821609 E207 • ,I "

(TASK 2.1.)

MODEL TESTING

Calculation of the initial test cases has been started

N 1.0 1.0 . Working is currently proceeding on Johnson,

Gouldin and Roquemo"re's experiments

J27289-15 821609 E207 . il .'

AEROTHERMAL MODELING PROGRAM Milton Kenworthy General Electric Co.

Evendale, Ohio General Electric began work on this effort in August, 1982. Rather than describing progress, therefore, this presentation will indicate some ~ignificant features .. of the planned approach.

Figure 1 schematically shows the individual computerized models utilized in General Electric's combustor aero design approach. The preliminary design module provides the overall envelope definition of the burner. The diffuser module provides the detailed contours of the diffuser and combustor cowl region, as well as the p,ressure loss characteristics into. each of the individual flow passages into the dome and around the combustor. The flow distribution module provides the air entry quantities through each of the aperatures and the overall pressure drop. The heat transfer module provides detailed metal temperature distribution throughout the metal structure as input to stress and life anlaysis that are not part of the aerothermo design effort.

The internal flow module, which entails 3-D elliptic flow field calculations, is not at present, used in General Electric's design method. It is planned, however, that it will be incorporated in the very near future to the extent that its usefulness is demonstrated. It is expected to be initially useful in providing improved hot gas side inputs to the heat transfer module and to help guide the development of combustor exit pattern factor. The capabilities include analysis of the mixing of the dilution region without chemical reaction, the treatment of fuel insertion, and the chemical reaction zone itself. While the phenomena in each of these regions is developed to a different extent of rigor, they all utilize the same basic 3-D elliptic framework.

General Electric's internal flow module, INTFLOW, has a basic core structure that can interchangeaply use either the 3-D combustor performance model from the Combustor Design Criteria Validation Program prepared by Garrett Corporation I (HC Mongia, RS Reynolds, and TW Bruce, 1979) or the newer Northern Research and Engineering Corporation's version of this same type of 3-D elliptic code.

These core packages are supplemented by special input routines.and output plotting routines that have been prepared at General Electric •. Figure 2 is the type of calculated output plots that are available from INTFLOW. The length of the arrows represent the relative velocity of the flows within this annular combustor, which includes swirl cups and a complex dilution pattern.

In the Aerothermal Modeling Program, comparisons are to be made with benchmark quality test data to permit evaluation of the accuracy of .the modules and to identify the sources of error or inaccuracy within the modules. Data from the FlOl series of General Electric combustors have been selected as a major basic source of this benchmark data. Extensive combustor liner metal temperature data and combustor exit gas temperature pattern data are available. These combustion aero thermodynamic data are particularly significant as these measured results are the important inputs in the stress and life analysis of the combustor structure and the turbine nozzles and vanes.

In addition, data fromlaboratoryexperiments, with more detailed flow data, will be utilized to help evaluate detailed features within the 3-D elliptic module. At General Electric, alternate computerized treatments are available as 2-D axisymmetric elliptic codes for: turbulence model modifications, numerics changes. kinetics treatment. and time fluctuation treatment. Hence, axisymmetric data will be selected for some of the computerized studies.

Also, a set of experiments will be conducted at General Electric with a test combustion sector having dilution hole characteristics like the General Electric FIOI combustor, but with wall boundaries compatible with the current 3-D elliptic code model capabilities. At present, the 3-D elliptic codes available at General Electric require that the combustor wall boundaries must be modeled along the cylindrical and axial grid lines; an axially curved wall such as the contraction at the aft end of most combustors cannot be accurately modeled.

Thus, these experiments will be done with a combustor having a flat dome and cylindrical walls to correspond to the current model capabilities and permit the separation of toe boundary shape effects. In addition, the experiments will be done with a series of tests of increasing complexity to help evaluate the error or inaccuracy due to each step in complexity.

Figure 3 indicates the steps in the exploration. A test will be done without burning and without swirl cups utilizing uniform dome flow. By utilizing a different temperature for the dilution flow, the dilution mixing can be documented with thermocouple measurements. Swirl cup flow is then introduced to examine the adequacy of modeling this complexity. The dilution holes complexity will also be varied. Burning tests will then be conducted first with gaseous fuel to avoid the question of spray drop size add vaporization and then with a liquid atomizing nozzle. A total of 15 different test setups are planned including either different configurations or different temperature traced regions.

Through comparison studies of model calculations with the type of data indicated, a program plan to improve the overall aero thermo model will be defined that will address the model deficiencies.

r .mternalJ:low_Module_(LNIF..LCLWlI

! ~ I Dilution I I

Region : • Exit Temperature Pattern Module Fuel

I

Insertion

I

Module

I

I • Combustion Efficiency

Combustion

Zone I • Emission Levels

Module I • Lightoff F I A

• Blowout FI A

L ..J

--------~------ -- w o .t-- • Metal Temperatures • Pressure Drop Distribution • Shaded boxe. are modules In current use In combustor design and development wo~k • • Open boxes are modules In General Electric's aerothermal model planned for u.e In the future In de.lgn and development work after adequate accuracy I. demon.trated or developed.

Figure 1. Internal flow module.

, it t.

..

CALCULATED VELOCITIES

-

-

~ ::::::::

--

~ . ::=::::::

--

- - ~

- ---.;

-

-

LENGTH IN • OR Y orRfCTrON. IN- •• D n/~£C 'UNEXY'- 1(0 16 • 17.1 19 2D • I I 4 '6 7 • 9 aD •• 'i! U .• 4 w o VI CALCULATED RADIAL TEMPERATURE PROFILES

u r 'r I ~ru mm ~mll1mj~j~ruru

,. ;".1 .. .2 ,~:. ,,, !". '. ::~ ,.I,,~ ,.,,~ ,: ". ,~.,,~ ':"':' •• ". .~.: ,. ,:. . ,. ';;" ,. ,,; ,}~¥

'0' t.. .... .... tI.' II.' .... ••.• .t.. If.. .... .t,' H.I '1., 'i'., 'I,' , .. '

Figure 2. Illustration of computer plotted output for General Electric's INTFLOW module.

, il ,.

GENERAL ELECTRIC TEST PROGRAM WITH COMBUSTOR TEST CONFIGURATIONS COMPATIBLE WITH ELLIPTIC MODEL NONBURNING TESTS: DOME FLOW TRACED WITH GAS AT DIFFERENT TEMPERATRUE THAN DILUTION AIR o PERFORATED PLATE DOME & FlOl DILUTION HOLE PATTERN o FlOl TYPE SWIRL CUP DOME - STANDARD FlOl LINER DILUTION HOLE PATTERN - ONLY ONE ROW OF DILUTION HOLES - DILUTION HOLE PLACEMENT TO MODIFY HOT STREAKS - DILUTION HOLE TOLERANCE EFFECTS BURNING TESTS o GASEOUS FUEL o LIQUID FUEL WITH PRESSURE ATOMIZING NOZZLE A TOTAL OF 15 SET UPS PLANNED INCLUDING EITHER CONFIGURATION CHANGES OR TRACE REGION CHANGES Figure 3. General Electric test program.

DILUTION ZONE MIXING STUDIES by J.D. Holdeman National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 Motivated by considerations of dilution zone mixing in gas turbine combustion chambers, NASA sponsored, in 1972 - 1975 contract and grant studies of the mixing characteristics of a row of jets injected normally into a duct flow of a different temperature (references 1, 2, & 3). Based on the favorable response to these studies, and the areas for further work which they identified, NASA Lewis is currently conducting a balanced program of contract, grant, and in-house research on various aspects of the ~jet in a confined crossflow~ problem. Included in these are: 'i) development of interactive computer codes for analysis of dilution jet mixing, and 2) extension of the experiments on jets in a confined crossflow.

From the data of References 1 & 2, an empirical model was developed (refs. 4 & 5) to describe the observed temperature distributions. The current interactive code provides a 3-D pictorial representation of the tempera~ure, as given by these correlations, for any user-specified downstream location, flow, and orifice parameters. Although calculations can be performed for (almost) any flow and geometric conditions of interest, they are, of course, most reliable for conditions within the range of the experiments. These codes will be improved and extended, and options added, as new data become available.

The experiments in References 1 to 3 dealt primarily with a single row of jets mixing into an isothermal flow in a constant cross-section duct. Variations in the mixing were observed as a function of jet-to-mainstream momentum ratio, orifice size, and spacing. The current experiments examine perturbations of this problem characteristic to gas turbine combustion chambers, namely: flow area convergence, non-isothermal mainstream flow, and opposed in-line and staggered injection.

Papers discussing grant studies on free-stream turbulence effects and reverse flow geometries, and'in-house analytical calculations of jets in crossflow will be presented at the Combustion Fundamentals Research Conference.

REFERENCES 1. Walker, R.E. and Kors~ D.L.: Multiple Jet Study Final Report. NASA CR-121217, 1973.

2. Holdeman, J.D., Walker, R.E., and Kors, D.L.: Mixing of Multiple Dilution Jets with a Hot Primary Airstream for Gas Turbine Combustors. AIAA Paper 73-1249, 1973.

3. Kamotani, V. and Greber, Isaac: Experiments on Confined Turbulent Jets in Cross Flow. NASA CR-2392, 1974.

4. Walker, R.E. and Eberhard, R.G.: Multiple Jet Study Data Correlations. NASA CR-134796, 1975.

R.E.: Mixing of a Row of Jets S. Holdeman, J.D. and Walker, AIAA Journal, vol.1S, no.2, with a Confined Crossflow.

Feb. 1977, pp243-249.

DXLUTXON ZONE MXXXNG STUDXES OBJ'ECTXVE TO CHARACTERXZE DXLUTXON ZONE MXXXNG XN SUFFXCXENT DETAXL TO=

* XDENTXFV AND UNDERSTAND THE

DOMXNANT PHVSXCAL MECHANXSMS GOVERNXNG THE MXXXNG PROCESS

* REFXNE AND EXTEND EMPXRXCAL

MODELS TO PROVXDE A NEAR-TERM COMBUSTOR DESXGN TOOL

* PROVXDE A DATA BASE FOR

VERXFXCATXON OF ANALVTXCAL MODELS DILUTION ZONE MIXING STUDIES

----""'"1-------------------------- _______ _

* EXPERIMENTS ON EFFECTS OF

FREE-STREAM TURBULENCE ON A JET IN CROSSFLOW (GRANT)

* EXPERIMENTS ON DILUTION JETS IN

REVERSE FLOW COMBUSTOR GEOMETRIES (GRANT)

* DEVELOPMENT OF INTERACTIVE

CODES FOR EVALUATION OF DESIGN ALTERNATIVES (HOST: IN-HOUSE)

* ~XPERIMENTS ON JETS IN A CONFINED

CROSSFLOW (HOST: CONTRACT) .

* ANALVTICAL CALCULATIONS OF

JETS IN CROSSFLOW (IN-HOUSE) DILUTION ZONE DESIGN COMPUTER PROGRAMS OBJECTIVE: DEVELOPMENT OF INTERACTIVE COMPUTER CODE FOR ANALVSIS OF MIXING OF JETS WITH A CONFINED CROSSFLOW FEATURES: PROVIDES A 3-D PICTORIAL REPRESENTATION OF THE TEMPERATURE FIELD PURPOSE: EVALUATE EFFECTS OF VARVING FLOW AND GEOMETRV GUIDE DESIGN TO REDUCE DEVELOPMENT TIME AND COST STATl!JS: CODES WILL BE IMPROVED AND OPTIONS ADDED AS NEW DATA BECOME AVAILABLE

FLOW SCHEMATIC

·····11

.•••. I· ••••• I' III

·····1·

••••• I' ••••• I' S

·····1·

zh

I

UM ·····1· TOP

••••• I' 1M VIEW ••.•• I' X ••••• I' ••••. I' .•••. t·

t

••••• I· , I) III •.•.. II

·····1·

t

••••• I.

I VJ

" 1 J ••••. I· •••.. I.

UM

SIDE .•••. I· H 1M VIEW

·····1·

••••• I·

I

Jet in a Confined Crossflow ...

· .. ··-7 .; ., ,,----,.- "'·-7 JJ •• y .5 H B .5 1 8= [l'M-l'J I [l'M-l'JJ 5/D=2 H/D=~ J=32DR=2 CD=.~~ X/H=.5 MR=.5 l'B=.33 PF=.3~ Typical Temperature Profile Distribution in Y - Z Plane MJ/M 00 =.25 (J = 32; (s/o) (HID) = 16) X/H ~ .25 .S 1 2 S/H HID .25 8 .5 5:6 1 4 l~ o 1 0 1

THETA = (Too - T)/(T 00 - TJ)

Variation 'in Temperature Distributions with Downstream Distance

XIII = .5; . (SID> (HID) = 16

J = 8 16 32

S/H HID

.25 8 .5 5.66 '1 ~---, 1 4 1 .

Varations in Temperature Distributions with Momentum Ratio

HID = 8 5.66 4

MJ/Ma> = .125 .25 .5

X/H

.5 y

H '1})I~ t'~ " 2

1 .

010 1 ·01 Variation in Temperature Distributions with Orifice Size

at Constant Spacing to Height Ratio (S/H = .5; J = 32)

NASA DILUTION JET MIXING - PHASE I

ffi

OBJECTIVE: ......

o COLLECT A DATA BASE ON MIXING OF A ROW OF JETS WITH A CONFINED CROSS FLOW

.-

G') c::: PI -i ......

"'1 "'1 a ro :z

o DEVELOP EMPIRICAL JET MIXING CORRELATIONS

c-t c-t c...

-0 I"T'J ~ --! --!

(/) o s::: ro "'1 3: ......

::30- ...J • ...... (/) x ......

X ::3 "'1 PARAMETERS INVESTIGATE~: .....

ro :z w ::3 G') ,.....

.....

:J::01'TJ L11 "'1 ::3 < ..... (0 PI N .....

VI ""0

o MOMENTUM RATIO (J)~ H/D~ S/D~ gj/goo

o ::3 PI :x: ::3 ro ::3 :J::o (/) PI ("") I"T'J (/)

o NON-UNIFORM CROSS-STREAM TEMPERATURE AND VELOCITY PROFILES

......

"0 PI ::3 PI CM ::3 '< 0.

o COLD/HOT JET INJECTION

......

......

o CROSS-STREAM FLOW AREA CONVERGENCE

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX. ARllONA

PHI~SE 1 TEST SfC I ION CONFHilJRATIO!!S

E::~

,'. ' ..

HO = 10.16 CM

TEST SECTIONS: I.J.)

......

0\ I.II!O 11l1li

l

-1.0 IIHII

__ "j,~D"

" __ ' I

, "

~

~t'f

.::::::" i 1 ~ . "f i

__ -'l1li H ~ , ~ ,: --- ' ['-----.

-:::.- :. ~

._~:~2 II

--- " .

-1.0 o 1.0

. ---1 ! I

IIHII o ' 1.0 V OppO"" ... 11 COftft'1ellCe V I IIIJICtlOll ... 11 cowa,...nce GARRETT TURBINE ENG IN It: COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX. ARIZONA " " .... ~ .... ----.

···-----1

SCHr·~F-M~T I C OF THE '11 LUTI ()~I JET r'I XI ~IG TEST ~l G

mI-········ ... __ ....

--- .. ~

~

PROFI LE DI UJT I ON GENERATOR JET RAKE Vol ......

MAIN STREA~M~ ____________ __ -...J RADIAL TEMPERATURE' PROFILE PROBE TRAVERSING MECHANISM AIRESEARCH MANUF"ACTURING COMPANY OF" ARIZONA A DIVISION OF" THI: GARRETT CDRPORATION

N.~S,~ DILUTION JF.T MIXING - PHl\SE II

E~~

OBJECTIVE:

o EXTEND THE DATA BASE ON MIXING OF SIN6LE-SIDED ROW OF JETS WITH A CONFINED

CROSS FLOW.

o COLLECT DATA BASE ON MIXIN6 OF TWO-SIDED ROW OF JETS WITH A CON~INED CROSS

FLOW

o DEVELOP EMPIRICAL JET MIXING CORRELATIONS

w ......

PARAMETERS INVESTIGATED: OJ

o CIRCULAR VS SQUARE ORIFICES) TWO-DIMENSIONAL SLOT

o MOMENTUM RATIO (J») HID) SID

o IN-LINE AND STAGGERED ORIFICE CONFIGURATIONS

o NON-UNIFORM CROSS-STREAM TEMPERATURE AND VELOCITY PROFILES

o CORSS-STREAM FLOW AREA CONVER6ENCE

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA

PHASE II TEST SECTIWIS A~lD ORIFICE CONFIGUPATIO~!S

E;:y

RHOj

=i~jl __ D-._

It-

T' I _

1.5Ho

I!o {' 10. 'B C" I I

....Ll -I '1= ~

~i~

C!

t.O 1.0

0- -1.0 ..-

X/Ho XlHo J X/Ho SYMMETRIC CONVERGENCE ASYMMETRIC CONVERGENCE Test Sections w ......

\0 -- ~ --0 0 0

~

g.1

°T

OS ~ OS

°T

oT

~ 0-' I

oT

O~

0 0 ---C:I

--

C D A B 2 4 4 SID- 2 8 8 8 HolD- 8 Orifice Configurations.

PHASt II SERIES 1 TFSfIMG

E~~

SCOPE

o

o COMPARE EFFECTS OF J) SiD) HID

RHOin 0

Um

o

W - 3Ho

o TWO-SIDED AND ONE-SIDED JET INJECTION

-

o

o IN-LINE AND STAGGERED ORIFICE

0,-

s CONFIGURATIONS

0-1-

W N a I RHOj T' =l1',--[-_D_--,

o DEVELOP CORRELATIONS FOR TWO-SIDED

JET INJECTION Ho -

~ ~HOm I

1 =J tl= .---------1

~ N XIHo GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA II ,.

PHASE 11 SERIES 1 TEST CONDITIONS

E~~

l

UM = 15 ~YS

TM = 6 J5 K Ho = J.O .16 CM

HID SID C()~IFIGlJf{ATInN ,hop

,JBOTT0t1

IN-LINE 6.81 6.8~

'2

INJECTION 25.0 2LL~

101.8

101.9

STAGGERED 6.53

6.54

w N

INJECTION 25.2

24.7 .-

99.3

99.6

. IN-LINE

7.85

1.81

4 INJECTION 27.9

27.3

108.3

107.0

STArjGERED

5.98 6.14

INJECTION 25.7

25.7

103.1

104.3

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX, ARIZONA

MEASUKED THETA DISTRIBUTIONS FOR SID = 2) HID = 8) X/Ho = 0.5

E~~

IMLINE ORIFICE CONFIGURATION

J = 6.81

J = 101.83

J = 24.95

, I hoc:::. .::::::>" PO.O Ii i 0.0

~ j

I . i [.

~I ~ ~ ..

0- 0- ...: I/) ~ i5 Q rJ.

a!

;R:~ = 1

i

i

F-§: 1.0 , L ~ .dl.O I 1.0 U-l

f f

N 1.6 -0.6 1.1i -0.6 N -0.6 1.6 TRfNiVERSE 015T. lIB TJIfNVERS£ 016T. liS TRANSVERSE OIBT. liS

STAGGERED ORIFICE CONFIGURATION

J = 6.53

J = 99.29 I . J = 25.16 j 0.0 0.0 f O~ ~ ~ ~

. . -

0- 0- ~~ Q i5 a! rJ.

In

;§ = ~

i

i

r -,., fi 5 t.O

r==-- ........-- =-' 1.0 1.0 i i T -0.6 1.0 -0.6 1.0 -0.5 1.0 ~SE 018T. 1I8 TRfINSVERSE 018T. liS TRfHMRSE 018T. liS .......

8 = (TM - T)/(TM - TJ) ir ••

MEASURED TIlETA DISTRIBUTIONS FOR S/U =·4-, -H/U ·~~,x;~:~-~~;- 1

E~

IrL-UHE ORIFICE CorW.IGURATION

J = 108.27

J = 7.85-

J = 27.92 I .,... J ...

I., I 0.0

J Y 0.0 > r, > ? < , 0.0 % .....

~ ~ ~ ... ..: .: ., I/) ~

- D -

D~

5 !

!

D D i ~ i 1.0 1.0 " 1.0 i a (.oJ N (.oJ -0.& 1.$ -0.& 1.1 -0.6 loS l'RFWSVERSE OIST. 118 TRANSVERSE 018T. ZlS TRANsvERs£ 015T. lIS

STAGGERED ORIFICE CONFIGURATION

J = 103.07

J = 5.97

. J = 25.68

I J J ~ L 7-=::::::;:::> ::::J 0.0

" ;# ........ )" =c::::::::< < 1'7 .:;10.0 I :::::==:=:=" """ 0.0 ~ ~ ~ .

0-

t3

~ =

o o

-

D -

rJ.

!

I

i

~ .L

r ___ 8 -- .......,. 1.0

f 2 ...e::: ~ ::...., J I I.D 1 i ==---, ,.0

I I 1.0 1.0 -0.6 -0.& 1.0 -0.5 TRANSVERSE OIST. !IS TRANSVERSE OISf. ZlS TRANSVERSE 016T. liS

8 = (TM - T)/(TM - TJ)

MEASURED THETA DISTRIBUTIO~ FOR PROFILED MAINSTREAM

E:=~

SID = 2J HID = 8

. J =24.63 J.= 23.77 .

J = 6.02 , I , 0.0 .----- ....... '----- , 0.0 I '~o l:.

~ '~ )- )-

S

.

":1 ..: ...

IIJ ~ C a ! ~

; l

~ i ~

1.0 1.0 I 1.0 w N .p.

-o.S I.S· -0.& lei -0.1 lei TRANSVERSE DieT. !II T~ DIST. liS ~SE DIST. !II J =23.62 J = 23.60

~l

'" I~o.o

0.0 ~ 7" /' '4' 0.0 ~ )- ..:

~~~

~

.c~~ ~

C ..J - s !

c ~ ~ a ' 1.0 c==::.' '>&>.Ji "I ~ <' .< I 1.0 • a 1.0 1& ...< -0.& 1.0 1.0 -0.& 1.0 -0.6 TRANSVERSE DI6T. ZI8 TRANSVERSE DIST. liS TRANSVERSE DI6T. liS

SIll = 4J HID = 8

8 = (TMAX - T)/(TMAX - TJ)

! il ,.

U

Iffil - PHASE I I SERIES 2 rESTS u

J

JET

SCOPE: ~ ., I I!I

0 TWO-SIDED JET INJECTION WITH PROFILED I!I

I I I!I I I!I

CROSS-STREAM

I ~ ~I!I.".

I- I!I is I!I i '- ~ I!I I CIC'VSS - STl<'E".A'i TEST CONDITIONS: I-~ I I!I e 'PI?OFIL': u I ~ I!I 0 I!I ~

UM = 15 f'lIS H = 10.16 CM., TJ = 300 K

o-liIl!I

e /

z 4V W U 0::: W N ~

HID SID CONFIGURATION JTOP JBOTTOM I I

Ln i-f 0.00 0.20 I!I I!I

8 I I!I

IN-LItlE 24.6 24.7 ~ III

. 6.21

8 2 STAGGERED 6.02

JET

STAGGERED 23.8 23.4

8 = (TMAx - T)!(TMAX - TJ)

23.6 2lL2

IN-LINE

4 STAGGERED 23.6 24.1

STAGGERED 99.5 99.3

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORATION PHOENIX. ARIZONA

FUTURE TEST PLAN ON PHASE II

E~~

o ONE-SIDED JET INJECTION

o TWO - DIMENSIONAL SLOT

o SQUARE HOLES

o TWO-SIDED JET INJECTION

UJ N '"

o NON-UNIFORM CROSS-STREAMrEMPERATURE PROFILES

o UNEQUAL JET INJECTION RATES

o CONVERGENT TEST SECTIONS (SYMMETRIC AND ASYMMETRIC)

SCHEDULED COMPLETION DATE ON PHASE II TESTS: DECEMBER 1982

GARRETT TURBINE ENGINE COMPANY A DIVISION OF THE GARRETT CORPORA liON PHOENIX, A~IZONA • iI ."

MEASURED THETA DISTRIBUTIONS WITH FLOW AREA CONVERGENCE FOR SID = 4, HID = 8 AT X/8 = 1

E~~

cp = 90 0, A1 I A'} = cp 104 0, A, I A? = 2 cf> = 97°, AlIA'} = 1.33

J = 21 .07

J = 26.34

7' 7 Y I I 0.0 • I o~ r r i 0.0 ~ ~ ~ >- ...

...

II) ~ ~ ° o C !

!

~ 5!

i

i

a.:: ~~ ______________________________ ~~j 1.0 W , , 1.0 ........ y :>V • ... ,1.0 N ~~ I~ .......

TRANSVERSE oIST. Z/S T~VERSE DIST. ZlS TRANSVERSE DIS T. liS cp =90°, AlIA? = 2 cp = 117°, AlIA? = 2

J = 26.59 °

J = 27.18

• • O.D ~-------KI----------------~'~-------'i 0.0 ~ ~ . .

....

...

II) ~ o c ri !

i ~

r, r ,t.O _0 '1.0

I.S '''' -0.5 TRANSVERSE DIST. ZlS TftANS"4:RSf DI6T. z/s , il

COMPARISON BETWEEN ONE-SIDED AND TWO-SIDED JET INJECTION

E;~

SID = 2~ HID = 8

I---------~

---I

TWO-SIDED INJECTION I ONE-SIDED- INJECTION --

I r·1JET/MMAI N

IN-nNE i STAG GE RE!lI __ H {Q=_8 __ +- _~/1l_:: -.!!.--I +-- PARAMETER

I ~ f t

. 1 ' !

I

I

5.74

MOMENTUM RATIO~ J i 6.81 0.53 25.3

0.176

0.139 0.169

0.198

8EB = CTM - TEB)/(TM - T J) !

0.36

0.2 0.19 0.40

r1AX JE.T PENETRATIotL Yc/H I

0.20

0.25 0.28

0.23 0.33

JET HALF WIDTH~W~/H~

AT x/H = 1

W N co

0.12 0.13

0.0 0.0

JET HALF WIDTH~ W~/H~

AT X/H = 1

2.06

8/8EB AT X/H = 1 1.82 1.94 1.89

i -T-

22.0

MOMENTUM RATIO~ J 25.0 25.2 107.8

8EB = (TM - TEB)/(TM - TJ 0.318 0.319 0.302 0.271

0.66 0.60

MAX JEf PENETRATION~ Yc/H 0.35 0.35

0.19 0.24

JET HALF WIDTH W~/H 0.47 O. O.

AT X/H = 1

0.27

0.2 0.28 0.26

JET HALF WIDIH~ W~/H

AT X/H = 1 i

8c/8EB AT X/H = 1

.--'----------J 1. 08 1 .10 1.17 1.. ~.:t

.,

COMPARISON BEfWEEN ONE-SIDED AND TWO-SIDED JET INJECTION'

,E;~

SID = 4) HID = 8

M JET/M J i:~~ ~ ~~ ED I ~~~~r~~~T~-~j~~~ I :ED i I N~~~T ~ 0:

PARAMETER MA1

MOMENTUM RATIO) J

7.85

5.97

26.3

6.1£1

SEB = (1M - TEB)/CTM - TJ)

0.112

0.102

0.105

0.107

MAX JE.T PENETRATlotL Yc/H 0.12)

0.37

0.24 O.Sll

(). 5/ 1

JET HALF WIDTH)W~/H)

0.05

0.13

0.27

0.20

w

AT x/H = 1

N 1.0

JET HALF WIDTH) W~/H)

0.29

0.11

0.23

0.24

AT X/H = 1

i ,

S/SEB AT X/H = 1

2.64

1. 77

1. 63

2.54

!

I !. - - _. . ....... -. .• . ............. ---- -, --- .... -----.-.

MGt-1ENTUf" RAT 10) J

27.9

25.7

109.0

26.7

SEB = (TM - TEB)/(TM - TJ)

0.190

0.ltl9

f). lSI

0.192

f1AX JET PENETRATIOfL Yc/H

0.50

0.50 1. f)

1.0

f).23

JET HALF WIDTH W~/H .

0.0

0.0

0.0

0.0

AT X/H = 1

JET HALF WIDIH) W~/H

,

f). 211

0.27

0.33

Q.45

I

AT X/H = 1

Sc/8EB AT X/H = 1 i J..43

1.37

1.41

1.75

.- ... - -.---- ... - ...... -. -". I - ... __ .. ___ . _____ . ___ ... __ .. _.

" .

COMBUSTION SYSTEM FOR RADIATION INVESTIGATIONS J. D. Wear National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 Description of Combustion System Hardware The combustion system·consists of an inlet interface flange, inlet diffuser, fuel struts and nozzles, combustor liner, liner housing and exhaust flange. The system will be installed in an existing test facility that can furnish combustion air at the conditions listed below. The system was designed for operation at 40 atmospheres inlet pressure, 900 K inlet temperature, and air flow to 80 kg/sec.

Six penetrations are provided in the outer pressure housing.

Adapters at the penetrations, permit use of various types of radiation instrumentation, such as total radiometers, spectral radiometers, porous plug and heat flux gages.

The primary zone of the combustor will have hardware modifica- .tions that will permit operation at different primary zone fuel-air ratios with constant overall fuel-air ratio.

Rotating exhaust instrumentation will be used to determine combustor performance in addition to the radiation data.

Hardware of an existing high temperature combustion system was modified to accept radiaton instruments.· Five total radiation radiometers and two heat flux gases were installed.

Data are presented showing total radiation at three axial positions of the combustor, and comparison of total radiation with data from a heat flux gage.

RADIATION INSTRUMENTATION &0, 210· FLOW -- DESIGN CYCLE CONDITIONS INLET TOTAL DIFFUSER COMBUSTOR Press. Temp. INLET F/A Exhaust MACH Temp.

MPa K NO. K TAKEOFF 4.05 889 0.328 0.0275 ClmB 3.47 849 .331 .0248 1669 CRUISE 1.72 815 .333 .0257 1674 TAXI IDLE .50 517 .337 .0114 972 RAOIA TlON INSTRUMENTATION FLOW -~ r- i!i·'::-r ., . I TOT AL RADIATION Combustor inlet Pressure--0.69 MPa 40 - Temp.--560 K Fuel-air ratio 20 - Mach number--0.:32 ,- ~ .04

Total : __ --L, ~~::~:::~g~~~~~==:::::;~"~~

radiation, OL. -

~O r

watts/sq. cm .05 20 - .04 , ....

.03 .02 o o 4 8 12 16 Centimeters downstream from fuel injection RADIATION AND HEAT FLUX Press.--O.o 9 MPa Temp.--S;;:;' K o Total radiation 60 r C Heal flux, Gardon , Liner temp.

I

--

Mach number 40- i 0.32

I

---- 20 - Total Liner metal radiation __ --<r.'--'---'"' temp., ~ n d he a I flu x, 0 .;.... ___ ===::..::: .. ===--~ __ --:.~_---.: K O OO watts/sq.cm

r •

60r number 0.25

l Mach

40 I

] I ./ i 20~ ....-0--- 0----.

L...

~---

I

:; -y- , ~ 1 .04 .05 .01 .02 .03 .06 Fuel-air ra tio - USING THE IN-HOUSE LINER CYCLIC RIGS Robert L. Thompson National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 The objectives of the HOST Burner Liner Cyclic Rig test program are basically twofold: (1) to assist in developing predictive tools needed to improve .

design analyses and procedures for the efficient and accurate prediction of burner liner structural response; and (2) to validate these predictive tools by comparing the predicted results with the experimental data generated in the tests. The data generated will include measurements of the thermal environment (metal temperatures) as well as the structrual (strain) and life' (fatigue) responses of simulated burner liners and specimens under controlled bou~dary and operating conditions. These data will be used to validate, calibrate and compare existing analytical theories, methodologies and design: procedures, as well as improvements in them, for predicting liner temperatures, stress-strain-responses and cycles to failure. Comparison of analytical results with experimental data will be used to show where the predictive theories, etc. need improvements. In addition, as the predictive tools; as well as the tests and test methods, are developed and validated, a proven, integrated analytical/experimental method will be developed to determine the cyclic life of a burner liner.

Figure 1 includes a list of the test rigs under consideration and the basic· liner segments or components to be tested in each rig. Each succeeding test rig and the tests to be conducted in that rig are increasingly more complex than the preceeding one, beginning with a flat plate, then a tube, then a subelement, and finally a full-scale liner test. Correspondingly, the structural analysis becomes more complex in thi's progression of test configurations. The Quartz Lamp Box Rig, from which experimental temperature and strain data will be obtained, will also serve as the test rig configuration for the evaluation of special instrumentation under development in the HOST program; for example, infrared camera for temperature mapping, thin-film thermocouples, thin-film strain gauges, laser speckle t~chniques, etc. The instrumentation with the greatest potential will be incorporated and used in the other rigs.

Test conditions and variab1~s to be considered in each of the test rigs and test configurations, and also used in the validation of the structural predictive theories and tools, will include: thermal and mechanical 'load histories (simulating an engine mission cycle, different boundary conditions, specimens and components of different dimensions and geometries, different materials, various cooling schemes and cooling hole configurations, several advanced burner linet structural design concepts, and the simulation of hot streaks. Based on these test conditions and test variables, the test matrices for each rig and configurations will be established with the intent ~o veri~y the predictive tools over as wide a range of test conditions as posslble uSlng the simplest possible tests. An illustrative flow chart for the thermal/structural analysis of a burner liner and how the analysis relates t?

the tests is shown schematic~llv ~n Ficure 2. The chart shows that several nonlinear constitutive theories-are toWbe evaluated.

Preliminary structural analyses in which several viscoplastic (unified) theories are being evaluated are underway for a flat plate, an axisymmetric combustor liner (tube) segment and a three-dimensional simulated combustor liner segment, each of which will be tested in its appropriate test rig. The basic elements required for a structural analysis of a flat plate are outlined in Figure 3. Analysis of the axisymmetric liner is just beginning. A representative finite element ring model and an imposed transient temperature distribution are shown in Figure 4. Analysis of a 3-D combustor liner constructed from stacked sheet metal louvers has been initiated. The construction of this combustor liner is shown in Figure 5. A representative symmetric finite element model of one of the segments with cooling holes is shown in Figure 6. In this example, less than 1 or 1/360th of the inner liner is being modeled. Typical temperature inputs to the structural analysis code, both steady-state and transient distributions obtained from measured data and a thermal analysis, are shown in Figure 7. The predicted strains will be compared with the experimental strains in order to validate the predictive theories for each of these test configurations. In addition, these types of preliminary analyses will also be useful in determining where both thermocouples and strain gauges should be located on the specimens in order to ensure that regions of steep thermal gradients and high stress concentrations are captured in the test measurements. A tentative schedule for completing the structural analyses of these test specimens is shown in Figure 9. As other test configurations are identified in this study, they will be added to this list.

STRUCTURAL ANALVSIS

OF COMPONENTS TO BE TESTED IN HOST LINER CYCLIC RIGS

r

1. QUARTZ LAMP BOX RIG

FLAT PLATE

2. QUARTZ LAMP ANNULAR RIG

SUBELEMENT OF COMBUSTOR LINER

w w 'J

TUBE

3. LO~ PRESSURE CYCLIC CAN RIG

FULL-SCALE COMBUSTOR LINER

FIGURE 1

NONLI NEAR THERMAL/STRUCTURAL ANALYSIS

- OF ADVANCED COMBUSTOR LINERS'

CONSTITUTIVE

RELATIONS

• CONVENTIONAL

• UNIFIED

MEASURED

• RHEOLOGICAL

TEMPERATURE

DATA

w w co

HEAT

LIFE

STRUCTURAL

TRANSFER

PREDICTION

• •• ANALYSIS I ~I

ANALYSIS ANALYSIS

MATERIAL

GEOMURY LOADS

DATA

FIGURE 2

COMBUSTOR LINER ANALYSES AND TESTS(HOST)

QUARTZ LAMP BOX .RIG

STRUCTURAL AtfALYSIS OF FLAT PLATE

18 ~n I" 2 "_ Ii 3 -.- -- "- I .. - -"l- Hi - I 6 I t

1I0T SPOT

I t

q In iLl 12 ] I ----:--...~ I t I I . " _. -2

~ - .-

I- - - - b ~ I w 2 II ..

w h '" ~ I .35E3 InpOSED TEMPERATURE HISTORY .

"

I = -5.3'1£" STRESS/STRA I tf COtfTOUR PLOTS

.](1"] 2 " - 'I.'i(E" .'fl{"] 1 . = ·3.67£"

.. • .. :£ ]

.. " -2.0:£" 5 ... in 5 = ·2 ..... £ ..

.'fBU -1.16£" .51E3 -3.251:3 B .5 :£3 5 .II£) .5tn .

.

q I .351:" 10= .59£3 • 10= 2 .1(£"

--

--

---------------~----------.----~ I="TGURE ~

FINITE ELEMENT MODEL OF AXISYMMETRIC COMBUSTOR 1"

SEGMENT

LINER METAL TEMPERATURES

w o ""'"

TEMPERATUf~E # F

", Zvc' &(II"~ <fG J\ lic) ~o c 20 bt"

TIME#SEC

FIGURE 4

TYPICAL .LOUVER COMBUSTOR CONSTRUCTION

COMBUSTOR

LINER

HOT COMBUSTION GAS

•

..

COOLING AIR

FIGURE 5

LOUVER METAL TEMPERATURES

MID-LIP, rLiP .L _1 I --- I ----- . : I

~K~L~KLE I

STEADY STATE DISTRIBUTION TRANSIENT RESPONSE

o THERMOCOUPLE DATA --- THERMOCOUPLE DATA

1800 r- HEAT TRANSFER PREDICTIO~OO HEAT TRANSFER PREDICTION LIP MID-LIP w ~ N METAL TEMPERATURE, 1200

of

800 ,--, ------------ DISTANCE ALONG LOUVER

o

20 40 60 80 100 TIME, sec

FIGURE 6

FlN\TE ELEMENT MODEL OF COMBUSTOR SEGMENT

/

/

/

--

/'

"

" /'

'"

0.577°

''-WELD

w .;:..

w

R

R

ENGINE

~

FIGURE 7

,... '.. ...

SCHEDULE FOR STRUCTURAL ANALYSIS OF COMPONENTS

..

COMPONENT FYJ83 FY 84

FLAT PLATE -------1-1 ----------

AXISYMMETRIC ·COMBUSTOR LINER --+-------------1

SIMULATED COMBUSTOR LINER(3D )-----+--------------

TUBE --------------------1

~ JT8D CAN

----------

""" ~

----- DEFINITION, MODELING, INSTALLATION, AND DEBUGGING INPUT DATASETS

----- PARAMETRIC STUDIES TO EVALUATE AND VERIFY ANALVSES

FIGURE 8

HOST Liner Cyclic Facilities Facility Description Donald Schultz National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 Three liner cyclic test apparatus are presently planned.

The first is a simple 5 x 8 jnch rectanglar box incorporating four 6KVA quartz lamps for cyclic heating of the test speci- men. This apparatus will include a stJicon win~ow viewport for IR camera temperature measurement. The quartz lamp box will be used to verify simple liner configurations ann to evaluate strain and temperature measuring techniques. This evaluation is scheduled to begin in June 1983.

The secon~ facility (shown schematically in Figure 1), is a twenty inch ~iameter annular O.D. liner simulator which is a joint NASA Lewis - United Technology Research Center program.

One hundred and twelve 6 KVA quartz 1a~ps will be cycled.

Power levels will be adjusted to simulate typical liner heat loadings. Air will be supplied to provide typical liner film and backside cooling. Two or three liner designs will be evaluated to compare with the modeling results. The in~icated modeling effort will be connucted independently by UTRC and NASA Lewis.

This apparatus in addition to 672 KVA of 480 Volt power, requires 7.5 1b/sec of 1000 F air at 30 psia, 3.5 1h/sec of ambient temperature air at 5 psig, 1.5 1b/sec of ambient temperature air at 1 psig ann 64 GPM of cooling water. This apparatus is scherlule~ to go under testing in April 1984.

The third facility; the low pressure liner cyclic can rig (shown schematically in figure 2); is being designed to fit the same test leg "as the annular quartz lamp rig. It will utilize a JT8D size can combustor operating on Jet-A or .possibly propylene fuel . (Propylene being a gaseous fuel of similar .

percent hydrogen as Jet-A, would significattly reduce fuel injector problems associated with cyclic operation while retaining simular flame radiation ch~racter5stics.) The test section of this rig would operate at about 2 1/2 atmospheres absolute pressure. A vitiated preheater will be use~ to supply 2.4 - 4.8 lbslsec of 800 - 10000F air to the test section. A torch ignitor will be used to minimize ignition problems associated with the cyclic operation. Testing of this apparatus is scheduled for November 1984.

Special Test Instrumentation Liner cold si0e temperatures will be measured using an IR-TV c~mera system. This will permit several hundred temper- ature measurements to be made in a relatively small area.

Liner hot side temperatures will be measured with thin film thermocouples. New technoJogy high temperature strain gauges will be used to obtain local strain mp~Sllrements.

Preliminary Small Scale Tests December 1981 saw the initial testing of a three lamp qllartz lamp apparatus pictured in figure 3. Limited success was obtained with this rig. A test plate temperatl1re of 2000 F was achieved. Lamp life~ however, appearerl to be limite~ for the standar~ commercial qllartz lamps then avail- able. Redesignerl lamps will be used in the two new quartz lamp facilities which should overcome the problems of the earlier lamps.

A preheater test is scheduled for December 1982 to design vitiated anrl non-vitiated preheaters required for the quartz lamp annular rig and the cyclic can rigs.

w .c- oo

H 0 S T II NER C YC L I C FA C I LIT I E S

I. QUARTZ LAMP BOX

II. QUARTZ LAMP ANNULAR RIG

III. LOW PRESSURE LINER CYCLIC CAN RIG

, il ,0

HOST CYCLIC LINER PROGRAM

SCHEDULE w ..,.. FY83 FY84 FY85 1.0

I. QUARlZ LAMP BOX

----

II. QUARTZ LAMP ANNULAR RIG

---------------- -------1---------------

III. LOW PRESSURE LINER CYCLIC .----·--------~---I ... ---------------ot

CAN RIG ------ DES IGN, PROCUREMENT, INSTILLATION -- TEST ~ iI •.

-

('t'\ ~, N I ..

~ I ( ...

V') .: I 4

J

U , t C.!:' 0:::

><

o

CO c...

:E \"oJ t- o:::: c:x: ~ 0' t- en 0:::: u..

UJ c:::: =» (.!)

~

«

c:::: c:::: UJ 0..

X 0 ::E UJ I:Q ~ UJ 0..

~

«

::E ....J

«

0..

....J ~ V) N

I=! ~

UJ c:::: LL..

!:::: ....J

«

O§ =»

0..

0- ::E N

«

~ ....J V") L.I.J :2 c::::

t=!

LL..

~

:E I:Q

-

0 ....J

•

•

HOST QUARTZ LAMP ANNULAR RIG SCHEMATIC·ECRL·l

672 KVA lIJO V POWER 64GPM COOLING WATER NATURAL GAS W

-

IN OUT I.J1 N 12.5 Isec

< 850 F EXHAUST

30 psia , LAMP t-----r--.,......---~~ MUFFLER I PREHEA TER I 7.5' I sec COMBUSTION

I • . 10000 F

RIG AIR 5 psig 15 II Isec 3.;511/sec 0 .... 100 F ..... 100 F ~ j I , •

HOST lOW PRESSURE LINER CYCLIC. CAN RIG SCHEMATIC-ECRl-l

NATURAL GAS JET A W \J1 W 9-12 1/1 sec 40-45 psia 1. ... 35-40 psia 2. 4-4.8 II/sec 800-1rooo F TEST <850 F EXHAUST COMBUSTION

I ""1000 F PREHEATER SECTION

t----r------.-... MUFFLER

I AIR 7.0 II/sec ""100 F . il ,.

SPECIAL INSTRUMENTATION

I. IR - TV MONITORING OF LINER TEMPERATURES

w U1 +:-

II. THIN-FILM THERMOCOUPLES

III. LASER STRAIN GAUGE

IV. HIGH TEMPERATURE STRAIN GAUGES

. II •.

FACILITY PREHEATER TEST

OBJECTIVE: W I.J1

1. DETERMINE IF A NATURAL GAS FIRED VITIATED PREHEATER

I.J1 IS CLEAN ENOUGH FOR USE WITH THE ANNULAR LAMP RIG.

2. EVALUATE PERFORMANCE OF LOW PRESSURE LINER CYCLIC

CAN RIG VITIATED PREHEATER.

• ii" THE FUTURE OF HOST Daniel J. Gauntner National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio 44135 a.

o ::c (J) ~ 0: o ~

--

0: (.!:) o I- z CJ) U <t: l-

-

0: e:::: ~ o I- o a.

0- z: UJ o W eo:: U 0::: ...J ...J <t: <t: => => z z: z: z: c::I: <C o o o National Aeronautics and

Space Administration TURBINE ENGINE HOT SECTION TECHNOLOGY

NI\SI\

Lewis Research Center

RESOURCES

$2.4f·'

FY81:

$4.m·l

FY82: w Ln $5. 6~i FY83: co

$7.5-$12M/YEAR (?????)

FY84-89: National Aeronautics and

Space Administration TURBINE ENGINE HOT SECTION TECHNOLOGY

NI\S/\

Lewis Research Center

FUTURE DIRECTIONS

o CONTINUING HOT SECTION DURABILITY RESEARCH o SELECTED INTERDISCIPLINARY GRANTS o STRONG NASA LEWIS IN-HoUSE RESEARCH EFFORTS w V1 \0 o COMPATIBILITY WITH DOD o ANAL YS I S r~lETHODS I S THE CORNERSTONE Nalional Aeronaulics and

Space Adminislralion TURBINE ENGINE HOT SECTION TECHNOLOGY

NI\S/\

Lewis Research Cenler

NEXT YEAR