nasa transply report

8/2/2019 NASA Transply Report

1/21

N A S A Technical Paper 1806TP1806c . 1

Performance of Semi-Transpiration-Cooled Liner in High-Temperature-RiseCombustor

Jerrold D. Wear, Arthur M . Trout,and John M. Smith

MARCH 1981


2/21

TECH LIBRARY KAFB, NM. ." 0134771N A S A TechnicalPaper 1806

Performance of Semi-Transpiration-CooledLiner in High-Temperature-RiseCombustor

Jerrold D. Wear, Arthur M. Trout,andJohn M. SmithLewis ResearchCeuterCleveland,Ohio

National Aeronauticsand Space AdministrationScientific and TechnicalInformation Branch1981


3/21

SummaryA combustor liner fabricated from Lamilloy wastested at inlet air pressures to 8 atmospheres, an inletairemperaturef 894 K , andxhaust gastemperatures to 2430 K . Results are compared withresults obtainedwith a conventionally designed, film-cooled step-louver liner.Temperature-indicating paint on theLamilloy linershowed uniform color, except for panel welds, whenoperatedat he 894 K inlet air emperatureanda1700 K exhaustgas emperature.Thisuniformityindicated minimum thermal gradients throughout theliner. The color of the paint on the welds indicated atemperature of about 180 kelvins above the inlet air

temperature.heaintolornhe Lamilloyadjacent to theeldsndicatedoweretaltemperature.When the exhaust gas temperature was increasedto 2215 K , the aint olorstreaked,ndicatingsubstantialhanges i n metalemperature.Thermocouples installed in the Lamilloy panel weldsshowed emperaturesabout 380kelvins above heinlet air temperature. A step-louver liner operated atcomparableconditionsshowedmetal emperaturesabout 460 kelvins above the inlet air temperature.A change in fuel module design and blockage wasmade to increase the pressure loss differential acrossthe Lamilloy liner in order o ncrease hecoolingairflow through the liner. This change increased theLamilloy cooling airflow from 9.2 to 13 .7 percent ofthe total airflow.Exhaust gas temperatures to 2430 K were obtainedwith hese fuel moduleswith he 894 K inlet airtemperature. At these conditionsLamilloy linermetal temperatures were about 280 kelvins above theinlet air temperature.Comparable mokedensitydata were obtainedwith both the Lamilloy and step-louver liners. Smokenumbers were below 5 for a fuel-air ratio range of0.024 to 0.044; they then increased to about5 as thefuel-air ratio was increased to 0.058.

IntroductionExperimental investigations were conducted with asemi-transpiration-cooled combustor liner fabricatedfromLamilloy'material.Results were obtainedat

pressures to 8 atmospheres and high combustor inletandxhaustemperatures.xperimentalaluesobtained with the Lamilloy liner are compared withresults obtainedwith a conventionally designed, film-cooled step-louver liner.The rendforgas urbinecombustors is towardhigh exhaust temperatures at increasing compressorpressure ratios. The need for combustor liners to bedurableunder heseconditionshas alwaysbeen aproblem of paramount importance. As pressure andtemperature levels have increased, the availability ofexcess airorncreased liner filmoolingasdeclined. naddition, he recent proposal by theEnvironmental Protection Agency (EPA) to controlengineemissionshasseverelyaffected heairflowdistribution in combustors ref. 1 ). Moreeffectiveliner cooling schemes are needed as the availabilityofair for liner cooling decreases.Transpiration cooling of liners is a technique thathasheotentialomaintain lowinerwalltemperatures with educedcoolingair. In thepastporousmetalor wire structureshave beenused.Theseaveotlwaysxhibitedheurabilityrequiredbecause hevery mallpassagesbecomepluggedwith dirtand hesurfacemetalgraduallyoxidizes,educinghelowrea. An alternativeapproach as been taken by theDetroit DieselAllison (DDA) DivisionofGeneral Motors with amaterial they call Lamilloy. This material consists oftwoormore layers ofmetalbonded ogether.Airenters from one side through regularly spaced holes,flows through small etched passageways to holes inthe next layer of metal, and continues this processuntil the air exits through regularly spaced holes onthe flame side of the liner. The liner is cooled by acombination of effects: the air passing through theliner,and he exit airactingas a film.Since hematerial surface is only partially covered by film air,this approach maybe thought of as semitranspirationcooling.The resultsobtainedfrom estsof heLamilloyliner are compared with the results obtained with afilm-cooled step-louver liner. Comparisons are madeon the basis of measured iner temperatures and alsoaccount or hevariation in fuel-injectormoduletype.Testonditionswere ressures f 5 to 8atmospheres at an nlet air temperature of 894 K andexhaust asverageemperaturesrom 400 to2460 K . ASTM Jet-A fuel was used in all the tests.

' Lamilloy is a registered trademark of the General Motors Corp., 3044 West Grant Blvd., Detroi t , Mich. 48202.


4/21

ApparatusCombustor

The test combustor, hownn igure 1, is anannular design4.5entimetersong fromhediffuser inlet to he ombustor exit plane.Fuel-injector modules are arranged in two circumferentialrows, 24 in each row. An indicationfhisarrangement an be seen in figure 2. The inletdiffuser (fig. 1) is 5.2 centimeters long and hasan exitto inlet area ratio of 1.379. The ratio of the annularflow areaat heplane of the fuel modules o hediffuser exit area is 6.7. The reference area for thiscombustor is 0.2474 squaremeter. All theairflowpasses through the fuel modules except that requiredto cool the liner. The test combustor is shown with

the Lamilloy liner in figure 1 and with a step-louverliner in figure 3.

Combustor inlet plane

,-Fuel t u b e-"--i '?- " - .'\

CD-12209-07Figure 2. - Fue l modules with step-louver liners.

Fuel Module DesignTwodifferent uel-injectormodule designswere

used in these tests and are shown in figure 4. Eachfuel-injectormoduleconsistedof woconcentric-vanedairswirlers hatswirled heair in oppositedirections to create a zone of high shearing action.Fuelwassupplied toeachmodule by a fuel tubelocated in thecentralcavityof each module fig.4(c)). Fuel flowed from the fuel tube and impingedon a splash plate mounted on the downstream facefeach module. This splash plate broke up the fuel jetand directed it radially outward, where the fuel wasfurther atomized by air passing through the inner airswirler. Additional fuel atomization occurred in theshearingegionetweenhelows exiting thecounterrotating air swirlers.As ndicated in figure4 he fuel moduleswereadjusted in size so thatequalnumbersofmodulescould be installed in eachcircumferential ow. Asshown in figure 4(a) the model 1 assembly consistedof type A fuel-injector modules. The arrangement ofthese modules was such that a corotational flow ofair was generated by the addit ive flow exiting fromthe outer swirlers at he nterface of he nner andouter rows. The model 2 assembly used type A fuel-injector modules n the outer row and type modulesin the inner row, as shown in figure 4(b). The type Bmodules had the swirler exit flows opposite to thoseof the type A modules. Thus, when both type A andtype B modules were used (fig. 4(b)), no corotationalflowwas nduced by the swirlers. Instead localized

c t'-Fuel 34.5 cm -Film cwling air

tjyFigure 3. - Combustor est section with step-louver iner. (Not to scale.)

2


5/21

-.r Cooling air slots

1.5- 1.12.6__ 2 .13.6 ~ C 3. 04.9 -d 3.9 \ ' '\\LCwling ai r holes.E?]

(a) Model 1 assembly comprised of type A fuel modules. Ib) Model 2 assembly comprised of type A module s for outer rowand type B modules for inner row.

4.3 ----e- 3.43.5 " ~~ 2.91.3 -- g ~ 1. 3.8 ~-

Type A

h- . a

Added venturi-,Splash pae" e-"

Fuel modulet y p e A o r B----,--

(c) Mix ing vent uris added to module discharge plane of model 2 assembly.Figure 4. - Sketch of combustor fuel module assembly. (Dimens ions are in centimeters. )

3


6/21

high hear egions were created by theopposingswirler flows from adjacent modules.The modules used in the model 2 assembly with ashort mixing venturi installed on thedownsteam sideof each module are shown in figure(c). The purposeof the venturi was twofold: First, to reduce the flowareaof heouter swirler and hus increasehecombustorpressure loss; and econd, oenhancemixing of he fuel andair n hewake of eachmodule.Test esults rom heExperimentalCleanCombustorprogram refs. 2 an d 3) and rom n-house tests conducted in a simple tubular combustor(ref. 4) have hown hat uchventuriscancausereducedmissions and lowerombustor linertemperatures.Combustor Liners

Design.-The liners used for this investigation weredesigned to the pecifications of table I . Although thefacility used forhepresent tests had nupperpressure limit of 8 to 10 atmospheres,he 40-atmospherespecification was ncluded so that heliners coulde usedn a futureigh-pressurecombustor facility.The mechanical design of the Lamilloy iner isshownschematically in figure 1. This liner and hestep-louveriner were designed to haveimilarcontours.Theprinciple of Lamilloy is representedschematically in figure 5. Cooling air enters the linerthroughholesand passes between themetal ayersthroughetchedpassageways.Thisair is eventuallyconducted hroughonemetal ayer o he next byholes. The process is repeated until the air leaves theliner on the combustion gas side. Cooling s achievedprimarilyby onvectionwithinheiner,houghtheremaybesomecoolingbecause he exiting airprovides a film barrier to heating. Thedesign chosen

T A B L E 1. COM BUSTO R LI NER DESI GNR E Q U I R E M E N T S

Maximum operat ing condit ionIInlet air temperature, K .............................................. 900

Liner exit average temperature ,K 2500Inlet air pressure, atrn 40System pressure loss different ial , percent ......................... 10Lamilloy l iner maximunl metal temperature, ............... 31 0Step-louver l iner maxirnum metal temperature.K ............1260Desired l i fet ime at maximunl operat ing condit ions, hr ...... 100Allowable cooling airf low rate, percentFlow factor range, W\ T / P , a 44-65(percent of total 18

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

....................................................................................a Where W is total airflow in kg/sec, T i s nominal inlet airtotal tempe rature in K , an d P s nominal inlet air totalpressure in atm.

used three layers of material each 0.0508 centimeterthick. The selection of these dimensionsas based onvendor experience in fabricating combustor liners fora variety of applications.A large enough passage izewas selected t o allow passage of foreign material inthe combustion air through the liner. Considerableimpurities are in the combustion air supply, but theyare usually n theformof a very ine ronoxidepowder hat hould easily pass hrough he linerwithout plugging the passages.Theheat ransferanalysis was performed at hemaximum operating condition, as shown in table I .Thenalysisssumedhatheverageastemperature close to the fuel modules as 2300 K butthat he emperaturefartherdownstream was nearstoichiometric, or 2650 K . Maximumot-spottemperatureserelsossumedoehestoichiometric temperature. The designed combustorpressure loss and liner differentialpressure werebased on calculations and on data supplied from testsconducted with conventionalilm-coolediners.Oncehenalysisadeterminedheequiredcoolinglux and pressure loss distribution,heLamilloy permeability ould e etermined.Thepermeability, expressed as discharge coefficients, forrequired lows is shown in figure6.Figure6alsoshowsheoefficientalues of thehreepermeabilitieselected andhe linerength verwhich thesevaluesapply.The highest permeabilitywas used in theexhaust ransition egionas hisregion requires the greatest cooling.The material selection was based on the followingfactors:(1 ) High-temperature strength(2) High-temperature oxidation resistance(3) Buckling resistance(4)High modulus of elasticity(5) Ease of fabrication and repair

Combustion gas side

Q

Cooling G-J:k' ElectrochemicallyetchedFigure 5. - Sketch of Lamil loy construction and airflow path.

4


7/21

.011

. 0 8V"n

.- .ooUa,0Ucu

.-c

.ooc"v).-n

.oo

D -

8 -

6 -

4 -

2 -

0 -

""" Values used or abricat ion.= CalculatedaluesExhaustinst rumentat ionplane-,

I

,-Fuel module- y' discharge lane- I

I

- 1

I

I I

U I I I0 4 8 12 16 20 24Axial distance rom uel module discharge plane, cm

Figure 6. - Lamil loy permeabil i ty select ion.

Ofheseactorsigh-temperaturetrengthndoxidation resistance were judged to be of the greatestimportance in this application. Hastelloy X , Haynes188, and thoria-dispersedTD)ickel-chromiummaterials were evaluated.Haynes 188 was selectedfor use as it hasbetterhigh-temperaturestrengththanHastelloy X without hedifficult abricationand welding problems of TD nickel-chromium.Detailed stress and buckling analyses were appliedtohe liner esigns.Where ppropriate a finiteelementtressnalysiswassed.heucklinganalysis led to heconclusion hat five tiffeningrings were required on the outer liner, and these ereadded to the liner at the positions indicated by theanalysis. The stiffening rings are shown in figure 1.The stimated life ofheLamilloy linerwascalculated by using the strain-life curves basedon theworkofManson ref. 5). Strain-lifecurves weredeveloped for a variety of material properties andsurface conditions. The results indicate that a linerlife of 109 hours is achievable for reduced materialpropertiesand a notchedsurface.Increases n ifewerecalculatedbyassuming mooth ather hannotched material.Reference 6 describes the first investigations withthe Lamilloy liner. Temperature-indicating paint wasused todeterminemetalemperatures.Theestsindicated that he axial andcircumferential welds

required to fabricatehe LamilloyanelswereconsiderablyotterhanhedjacentLamilloy.Because ofheotter weld temperatureshreethermocouples were nstalled on each liner in thewelds, and dditionalests were conducted.Tnethermocouple positions are shown in figure 7 .The film-cooled step-louver liner (shown inhe testcombustor in fig. 3) was designed to withstand theoperatingconditionsshown n able I . Theaddedprotection urnished by a thermalbarriercoatingtha t was applied to the iner combustion gas sidewasnotconsidered in thedetailed hermaland tressanalysis. This liner had undergone extensive testingto exhaust temperatures in excess of 2200 K with noapparent damage or deterioration. However, testinghad beenimited to pressuresof 8 atmospheres,which was a facility limitation.Compared withconventional aircraft engine liners the metal of thislinerasuitehick,eingabricatedf0.20-centimeter-thick Hastelloy X material. As withmost combustors of this type there were no dilutionair holes in the liner. All air except that required forliner cooling passed directly through the array f fuelmodules. The conditions shown in table I are quitestringent,and he low availability of cooling irbecause of near-stoichiometric operation makes theliner design very critical.Fabrication.-The Lamilloy liner is composed of aseries of Lamilloy sheets welded together. The seamsbetween adjacentsheetsarearranged to be at anangle to the flow so that there is always film airflowacross, rather than parallel to, the narrowweld joint.The liner panel portions, composed of Lamilloy of

Fuel module discharge plane

7 Outer iner

, . ' 1 plane---,"-0 4 8 126 20 24Axial distance rom uel module discharge plane, cm

Figure 7. -Ax ial pos it io n of hermocouples nstal led on Lamil loyl iner.

5


8/21

varyingermeability, were hydroformedoheproper shape and then welded together to form thecomplete liner. Hydroforming was necessary, ratherthanpinning,ecausefhearticularinergeometryhosenorhisombustor. A lessconvoluted liner could probably be spun, which is asimpler abricationprocedure hanhydroforming.The finished liners are shown in figure 8 (the innerliner n ig.8(a)and heouter iner n ig. 8 0 ) ) .Stiffening rings can be clearly seen in figure 8(b).Film cooling air calibration."Beforehecombustion tests were begun, both the Lamilloy andthe tep-louver inerwere ilm-airflowcalibrated.This was done by installingboth nnerandouterliners on a flow stand and measuring the airflow ratethrough the liners at varying pressure drops acrossthe liner. A sketch of the calibrating stand is shown

Upstream end

.. . ..

(a) Inner liner.

Upstream end

C-77-3756(b)Outer iner.

Figure 8. - Lamilloy iner with temperature-indicating paint.

in figure 9. The Lamilloy liner was calibrated as aunit. For the step-louver liner some of the individuallouvers were calibrated one at a time by unmaskingorntapingheariousoolingoleowssequentially. The flow curve for the Lamilloy linerspresented in figure lO(a), and the flow calibrationsfor the step-louver linern figure 100). Table 1 givesthe characteristics of the inner and outer step-louverliners. The table lists the number of holes, the holediameter, he otal flow area,and hecalibratedvalue of the area of the unblocked film cooling airholes imes he lowcoefficient ACd and c d forsome of the individual panels. TheCd value for theentire Lamilloy liner is compared with values for thestep-louver liner in table 111.Test Facility

The investigation was conducted in a closed-ductfacility. The flow path and the arrangement of themajor components of the combustion air system areshown in figure 11. The combustion air s heated toamaximum of 589 K in an outside preheater and isdelivered tohe cell through 1.5-centimeter-diameter ASME orifice run. Upon reaching the testcell the air can be elivered to the test combustor ortcan be first passed through heat exchangers having

Inner inerStaticressure, \inlet annulus -, \\

\ -K T - --\

ILStatic pressure,exhaust annuluse.

Airflow orifices

CD-12408-20Figure 9. - Liner cooling airflow calibration ixture. Single or multiplerows of holes can be calibrated.

6


9/21

InstrumentationA cross-sectional sketch of he est rig, showinginstrumentation planes and dimensions, is presentedinigure 12. Theombustor inlet air averagetemperature wasdetermined rom eight Chromel-Alumel thermocouples mounted at plane 2 (fig. 12).Figure 13(a) shows the position dimensions of thesethermocouples, whichwere nstalled at centersofequalareas.The ndicated hermocouple eadingswere taken as true values of total temperature. Infigure 13@) are shown the dimensions of the eightinlet total ressureakes four robesach,installed at centersfqualreas,ndhedimensions of 16wall taticpressure taps, eightequally pacedaroundboth he nnerandouterwalls. Inlet pressure instrumentation was mounted atplane3(fig.12).Exhaustgas nstrumentation wasmounted at plane 4 (fig. 2). Dimensions aregiven infigure3(c). There were eightixedakesach

containing five totalressurerobesndiveplatinum-plus-13-percent-rhodium/platinum asira tedhermocouplesoreasur ingotaltemperature. All were mounted at centers of equalareas. Static pressure was measured by four wedgestatic probes equally spaced around the annulus atarea centers. Four gas sample rakes, each with threearea-centered probes,were equally spaced around thecircumference. The rakes were plumbed so that gassamples could be obtained from any individual rakeor combinations of wo, hree, or four. The hreeprobesof each rake wereall ubed to acommonmanifold.Four additional hermocouples were nstalled onboth he nnerandouterLamilloy iners, givingatotalof seven oneach.Their axialpositionsareshown in figure4.herithmeticveragetemperature of the 14 thermocouples was consideredthe liner metal average temperature.Chromel-Alumel thermocouples were installed onthe tep-louver iners o ndicate hehot-gas-sidesurface temperature of the metal, onhich a thermalbarrieroatingadeenpplied.ightthermocouples were equally spaced circumferentiallyaround he number 4 panel(fig.3), straddling hecenterlines. One thermocouple each was installed at0" and 180" on panels 1 and 7 . The outer liner had

one additional thermocouple at 180" on panel. Thestep-louverinerveragemetalemperature wasdeterminedfrom hearithmeticaverageof he 25thermocouples.

p.

Bo th nne r andou te r i ne rs -

01 (a ) 1 I I d.amil loy iners.ACd orboth iners, 0.2352.

-0 1 2 3 4 5fi $m3 kPa(b ) Step-louver iners.

Figure 10. - Cal ibra ted i lmai r f low hroughcombustorl i ne rs as a func t i on of ai r densi ty and pressure d i f fe rent ia lacross l iners.

the capacity to heat the air to22 K . Fixed probes atthe test combustorntrancemeasurehe inlettemperatureandpressureprofiles.Thecombustorairflow rate and pressure are set with an inlet valveand an exhaust valve. Before the hot gas enters theexhaust valve, it is cooled to 355 K or less by a seriesof quench water sprays. eyond the exhaustvalve thegas flows into hecentralatmosphericoraltitudeexhaust system. A more complete description of thefacility is given in reference 7 .

ProcedureThe operating conditions for the evaluation of thecombustor liners are listed in table IV . Most of thetests were conducted at a nominal test pressure of 5


10/21

TABLE 1 1 .- FI LM -AI RFLOW CAL I BRA TI ON OF STEP- L OUV ER LI NER S(a) Inner l iner"

Total f i lmhole areaper panel ,c m 2E__"8.8659.OOO10.4519.0006.3586.5024.049-

.25 cm2 /liner~

Calibrationvalue ofAC:

0.0316,0375.03 19"""

Calibrationvalue offlow coefficientcd

0.00357.00359,00354"""

Nu m b e r o ffilm holesper panelDiametelof filmholes,cm

Panel

14213 213213 214 216 217 8

0.282.295,318,295,239.226,170

21 (upstream)34567 (downstream)

"""""" """"""0.00344.18664.2ll panel holes unblocked

(b) Ou ter line10.00360,00394

,00370,00363,00374,00373,00395

"

"21 (upstream)34567 (downstream)

22 025225 822 822 0208196- ~

10.70611.28312.9431 1.5538.2458.3484.458

0.0385.0444,0479,0419,0308.03 1 1,0176-~

0.249.239,253,254,218,226,170

Al l panel holes unblocked 167.536rn*/l iner I 0.2488.00368 1a W=ACd,map, where W is film cooling airflow in kg/sec; A is area of unblocked filmcooling alr holes in cm2 ; C, is flow coefficient; p is density of film cooling air entering linerin kg/m3, and A P is pressure differential across liner in kPa.

TA BLE 111.-COMPARISON OF CALI BRATI ONFI LM - AIRFLOW VALUES O F AC, FO RLAM I LLOY A N D STEP-LOUVER LI NERS operate at higher fuel-air ratios,

it was necessary toremove the thermocouple rakes and rely entirely ongas analysis measurements o determine combustorperformance.The test operational procedures were asfollows:The inlet air emperature was raised to he desiredlevel, andhe inlet pressurendirflow wereadjusted to the desired values for ignition. Once thecombustor was lit, heparameters ofairpressure,inlet airemperature,ndirflowate weremaintained as close as possible to the values shown intable IV . Data wererecordedateachfuel-airratiosetting, witharefulttentionaidohe linertemperatures. If the liner temperatures were otconsidered to be dangerously high, the fuel-air ratiowould be increased anddata ecordedagain.Thisprocedure was followed for the tests with each liner.Smokenumbers weredeterminedas ollows:Theabsolute reflectivity of thetainnWhatmannumber 4 filter paper, obtained from the exhaust gassample, was measuredwith heWelchDensichronusing a black background.TheDensichron wascalibrated with a Welch Gray scale. The smoke indexwas determined from the following equation:

inner linerOuter linerInner andouter liners Larnilloylineriner Film airflow values ofAC","""0.23524414~

a W = AC, \ I Pa Pwh e r e W is film coolin g airflow inkg/sec; A is area of unblocke d film coo ling airholes in crn2; C, is flow coefficient; p is densityof film co oling air entering liner i n kg/rn3 ; an dA P is pressure differential across liner in k P a .

atmospheres with an inlet air temperature of 894 K .The combustor was operated over a range of fuel-airratios. This range was limited to a maximum fuel-airratio value ofbout 0.021 to 0.023 when thethermocouple rakes were installed in the exhaust. To8


11/21

Figure 11. - Combustor test facility showing flowpaths and equipment arrangement.

Instrumentationlane 2 k 2 4 . 5 9 4 k 3 . 8 1(inlet air thermocouples)Figure 1 2 - Schematic cross section of combustion est rig showing co mbustor iner test section. (Dimensions are in centimeters. )

9


12/21

rThermocouples (8 places,:; equal ly spaced)/ I Top dead cen ter

, I

(a) Inl et air Chromel-Aiumel thermocouples (fig. 12, piane 2).

Diffuser nlettotal pressure probe(sample positions) > Top dead center

b.7 diam 45.31 diamI'.

Diffuser inlet wallstatic pressures (8outer diameter, 8 Diffuser inlet totalinn er diameter; pressure probes (8

Top dead cente r

(b) Dif fuser nle t air total and static pressure nstrumentation (fig. 12, piane 3)

Top dead centerExhaust wedge static I ,,9 posit ions)Gas sample probe

(c) Exhaust gas total and static pressure probes, exhaust gas sample probes, and exhaust gas thermocouples (fig. 12, plane 4).Figure 13. - Combustor test instrumentation. LAlviews looking downstream. Not to scale. Dimensions are in centimeters. I

10


13/21

Fuel module discharge plane Results and Discussion

0 4 8 12 16 204Axial distance rom uel module discharge plane, cm

Figure 14. - Axial position of thermocouples nstal led onLamil loy iner. Seven hermocouples on each iner. )

TABLE 1 V. -NOM I NALT E S T C O N D I T I O N S

[Inlet total temperature, 894 K;diffuser inlet Mach nu mber,0.419; l iner reference Machn u m b e r , 0.038; f i lm coolingairf low for model 2combus tor , 15.4 a n d 9.2percent of total airf low fors tep- louver and Lami l loyliner, respectively; filmcooling airf low for model 2combustor with venturis ,13.7 percent of total airf lowfor Lam il loy l iner .]

Inlet totalpressure,

0.505

.794

Inlet airf low,kg/sec

10.81

16.99

Yo absolute reflectivity of samplex L1 - Yo absolute reflectivity of clean paperSmokeumbers werehenalculatedy theprocedure given in reference 1.

Liner Metal TemperatureThe results of the first testsith the Lamilloy liner

are presented in reference 6 and are compared withstep-louver liner data in this section.The first Lamilloy test was conducted at an inletairressuref 7.9 atmospheres,n inlet airtemperatureof 894 K , and an averagecombustorexhaust gas temperature of1700K . The test time wasonly about 10 minutes, long enough to ensure thatthe aintwould ndergoheppropriateolorchanges.Whenexaminedafter his est,both heinner and outer linerswere uniformly the same color,with the exception of some slight color differential innarrow regions at each weld joint. The color changeindicated that the hottest portions of the liner wereonlybout 180 kelvins abovehe inletirtemperature. The remainder of the iner showed onlytheuniformcolor ndicativeof very ow thermalgradients. his testwas not severenough topinpoint areas, other han liner weld joints , wherethermocouples hould beocated.Theinerwascleaned and epainted with emperature-indicatingpaint preparatory for testing atigher combustor exittemperatures. Three thermocoupleswere installed onboth the inner and outer liners. These thermocoupleswereplaced n iner weld joints betweenLamilloypanels as shown in figure 7.The operating conditions for the second test werean inlet air pressure of 7.9 atmospheres and an inletair emperatureof 894 K; thecombustoraverageexhaust temperature was varied upward from 1700 Kto 2215 K in several steps. Substantial color changeswere obtained this time and are shown in figure 15.Figure 16 is a plot of the readings of the inner andouter liner thermocouples t he wo ombustorexhaust asemperatures.Two test points wereobtained:heirsttnveragexhaustastemperaturef 1719 K toonfirmi ththermocouples the temperatures indicated by paint inthe previous test, and second at an average exhaustgas temperature of2215 K to force pronounced colorchanges in the paint. The liner temperatures at thelower combustorxhaustasemperature weregenerallyngreementwith thoseemperaturesindicated by the paint. That is, theeld areas were atleast 180 kelvins hotter than the nlet air temperature .At the higher combustor exhaust gas temperature theinner liner hot spot (fig. 16(a)) was about 285 kelvinsabovehe inlet airemperature,rmetaltemperature of 1180K . The results obtainedwith theouter liner arehown inigure 16(b). Twothermocouples on this liner failed during the higher-

11


14/21

Typicalthermocoupleinstallation-,

weld!Axia\\\\ _ .

(a) Inner iner.C-78-512

(b ) Outer l iner.Figure 15. - Lamilloy iner with temperature-indicating paint after test to2215 K average exhaust gas temperature.

combustor-exhaust-gas-temperature est.Theouterliner tended to operate slightly hotter than the innerliner, ut s eforehe ottest regionwas thecombustor exit portion of the liner. The maximumindicated temperature was 380 kelvins above the nletair temperature, or a metal temperature of 1272 K.Linermetal temperaturesfrompreliminary estswith the Lamilloy liner of reference6 were comparedwith data for he step-louver iner, both using themodel 1 fuel modules (fig. 4(a)). The data,presentedin figure 17, were obtained over a range of fuel-airratios and test pressures. The data presented are theaverage of all he liner thermocouples (both nnerandouter liners)minus the inlet air emperature.Also shown is the maximum liner temperature that

2 100E% LLa,c.-d

Averageexhaust ga stemperature,K

0 17190 22153

(a) nner iner .-

0

I I00 I I I4 8 12 16 20Axial distance from fuel module dischar ge plane,cm4 8 12 16 20Axial distance from fuel module dischar ge plane,cm

(b ) Outer l iner.Figure 16. - Lamil loy iner emperatures,asindicated by hermocouples. obtained during

tes ts at tw o combustor exhaust gas temper-atures. Nominal nlet air otal emperature,895 K nominal nlet air otal pressure,7 .9 atm. (Fro m ef. 6. )

was measured. These data were obtained at pressuresfrom 5 to 8 atmospheresanddonot ndicateanyeffect of pressure over this range. The Lamilloy linermetal temperatures were somewhat higher than thetemperatures obtained with the step-louver liner.For n 0.043 fuel-airatiohe veragemetaltemperatures for the Lamilloy and step-louver linerswere 295 and 265 kelvins, respectively, above inlet airtemperature, or actual average values of 1189 and1159 K. Thetep-louver linerverageetaltemperature was only about 30 kelvins ower thanthat obtained with the Lamilloy liner even though thefilmcoolingairflow ate was abou t 1.8 times heLamilloy value (fig. 10 and table 111).Theombustionxhaustasemperature,determined from gasanalysis data, is shown in figure18 as a function of fuel-air rat io. For a fuel-air ratioof 0.043 the combustion exhaust gas temperatureas2215 K for the particular operating conditions. TheLamilloynd step-louverineretalveragetemperatures of 1 189 and 1 159 K, respectively, were

12


15/21

L ine romina ln l e t a i r totalpressure,MPa0 Lamil loy 0.194A Step louver .505

- v Step ouver .649e l Solidsvmbolsdenotemaximum ocal

L ineromina ln le t a i r totalpressure,MP a0 Lamilloy 0.794A Step l ouve r .5052400 v Step louver

t400

1200 1 I I I I 2.02 .03 .04 .05 .06Fuel -a i r a t ioFigure 17. - Lami l loy inerands tep - l ouve r i ne rdifferential metal emperatures as a fu nct ion of fuel-a i r a t i o - for uelmoduleassembly model 1 Nomina lin le ta i r o ta l emperature , 894 K.

obtained with a combustion exhaust gas temperatureof 2215 K .As a result of these tests w o serious problemswerediscovered with the Lamilloy liner. First, the patternof hot streaks shown by the temperature-indicatingpaint was alined with the angled welds that joined thepieces ofLamilloy ogether o orm a continuoushoop. These angled axial welds and the hot streakscan be seen in figure 15. Second, the Lamilloy linercooling airflow was below the design value becauseof a change in the overall combustor pressure losscharacteristic. This change occurred when the moreopencounterswirl uel-injectormoduleassemblies,models 1 and 2 (figs. 4(a) and (b)), were substitutedfor a less durable design thathadhigherairflowblockage.Hot-streak problem.-Thealinementof hehotstreaks with the angled axialwelds was considered tobe very serious. The only cooling of these welds wasby conduction and a washing over of the film airformed by coolingflowsexiting heLamilloy ustupstream of each weld. Hot streaks alined with theinsufficiently cooled welds could result in damage.The hot streakswere observed o be wirling, and this

al-L

m01 1800mxal

Fuel -a i r a t ioFigure 18. - Combustorexhaustgas emperatureas afun cti on of uel-air atio,determinedbygasanalysis,obtained during estswith both Lamil loy and step-louver

l i n e r s - for uelmoduleassemolymodel 1 . Nominalin le ta i r o ta l emperature , 894 K.

was believed to be caused by the arrangement of theouter air swirler of the type A fuel-injector modules.As shown in figure 4(a) these swirlers reinforce eachother to induce coswirl flow. The model2 assemblywas constructed o eliminate his problem. Type Bfuel-injector modules were installed on the inner rowas shown in figure 4(b) to prevent the formation ofany net swirl in the airflow pattern.The model 2 modules were ested over a limitedfuel-air ratio range. Figure 19compares average linermetal emperaturesobtained withmodel 2 for heLamilloy and step-louver liners. Though data wereobtained only to a fuel-air ratio of0.0235, the trendas shown in figure 19 is clearly indicated. Averagemetal temperatures of the Lamilloy liner were lowerthan those of the step-louver liner, and the maximummetal temperature as measured in one of the axialweldswas also lower thanhatbtainedwithmodel 1 assembly.Thus hearrangementof uel-injector modules comprising model 2 was successfulin minimizing he effect of any hot streaks on theaxial welds in the Lamilloy liner.

13


16/21

Combustorl i n e r

0 Lamilloy- D Stepouver

Sol idsymbolsdenotemaximumlocaldi fferential emperatureb

c Combustor5 l i n e rLamil loy /Stepouver /Lami l loyiner 1

designvalues

/.d/. 5 1 I I l l.1 .2 . 3 . 4 . 5 . 6Di f fuser n le tMachnumber

Figure 20. - Iso thermal nonburn ing) inerpressure l o ss as a function of di ffuseri n l e t Mach number - for uelmoduleassemblymodel 2. Var iab le n le tai rpressure, emperature,and flow.Fuel -a i r a t io

Figure 19. - Lamil loy ineran dstep- louverl inerd i f fe rent ia lmeta l emperaturesas afunct ion o f fue l -a i r a t io - for uelmoduleassembly model 2. Nomina l n l e ta i r o ta ltemperature, 894 K; nomina l n l e ta i r o ta lpressure, 0.723 MPa.

Lamilloy liner coolingflow.-A plot of combustorliner isothermalressure loss as a function ofdiffuser inlet Mach number is shown in figure 20 ofthe Lamilloy and step-louver liners with the model 2fuel-injector module assembly. Also included are thepressure loss values used for design of the Lamilloyliner. Liner pressure loss is calculated as follows:(Liner annulus av. st. press.) (Ex. v. tot. press.)

(Diffuser inlet av. tot. press.)The Lamilloy liner operated at a combustor pressureloss considerably less than design and hence was notflowing the required amount of cooling air. Also, ydesign, he film cooling airflow for he step-louverliner was greater than that for the Lamilloy liner atany particular combustor pressure loss.To determine if there is or is not an advantage tousing combustor liners fabricated from Lamilloy, thepressure loss values had tobe adjusted to be loser tothe design values.To increase the pressureoss acrossthe Lamilloy liner, it was necessary to increase theflow blockage of the fuel-injector module assembly.

4

Fuelmoduleassemblymodelo z (w i thou ten tu r i s )

6

4 -- -- - ami l loy iner5- 0 2 (w i then tu r i s )viVI designvalues

;:2 35s2-.- xk zS &2 1-E

. 8

z 3 -a n

L 9"W

.---cWL U

m n //0

.1 . 2 . 3 . 4 . 5 . 6Di f fuser n le t Mach numberFigure 21. - Pressu re loss acrossLamilloy

l i ne r as a func t i on of d i f fuse r n l e t Mach2 wi than dwi thou t en tu r i s .Var iab lenumber - for uelmoduleassemblymodelin le tai rpressure, emperature ,and flow.

14


17/21

This was done by adding mixing venturis o the outerswirlers of each module of the model 2 assembly, asshown in figure 4(c). These venturis reduced the flowarea fhe uter swirlerlightly andhereforeincreased thepressure ossacross he uel-injectorassembly. Theventuris lso served to directheairflow rom heouter swirler inwardandshouldresult n an improvedmixingof he uel and airstreams. The pressure osses across the Lamilloy linerwith the model 2 moduleswithandwithout hemixingventuris re ompared in igure 21. Theinstallation of he mixing venturis did ncrease hepressureossubstantially,ndressure lossslightlyhigher than heoriginal designvalue wasachieved. The increased pressure differential acrossthe liners resulted in an cooling airflow valueof 13.6percent of the total combustion airflow, as comparedwith 9.2 percentf totalirflowthe lowerdifferential pressure.The Lamilloyineremperaturesormodel 2moduleswithndwithout mixingenturisrecompared in figure 22. With hemixingventurisinstalled, tests were conducted to an overall fuel-airratio of about 0.064 to 0.065 without the liner metalaverage or he maximum ocalmetal temperaturelimiting themaximum fuel-air ratio. As shown infigure 22 the average iner temperatures with theventuriswere ower han when noventuriswereinstalled.This is dueprimarilyohe increasedcooling airflow rate through the Lamilloy liner. Theexhaustasemperaturethe 0.065 fuel-aircondition,determinedbygasanalysis,wasabout2430 K . At this condition the average liner metal andmaximum local temperature differentials above the

Y

Fuelominalnlet moduleirotalassemblymodel pressure,MPa

0 2 (w i thou ten tu r i s ) 0.7230 2 (wi then tu r i s ) .505

Solid symbols denote maximum ocald i f fe ient ia l temperature. m I

J.07Fuel -a i r a t ioFigure 22 - Lami l loy iner d i f fe rent ia l metal temperaturesas a unct ion o f fue l -a i r ra t io - fo r ue l moduleassembly model 2 wi th and wi thout mix ing ventur is .Nomina l n le tai r total temperature,894 K.

Lineromina lnlet a i r to ta lpressure,MPan Step louver 0.505v Step louver649

0 Lamil loy79 4

D I I"02 .0 304 .05 .06Fuel -a i r a t io

Figure 23. - Lami l loyan dstep- louver inercomhst ionefficiency as a unction of fuel-air ratio - fo r fue lmoduleassem5lymodel 1. Nominal nletair otaltemperature, 94 K

inlet ir temperature were 140 and 280 kelvins,respectively, or metalemperatures f 1034 and1174 K .Combustion Efficiency and Pattern Factor

Effect of liner type and fue l m odule mod el.-Comparisons of thecombustion efficiencyvaluesobtained with theLamilloyandstep-louver inerswith themodel 1 fuel module assembly are shownasunctionfuel-airatio in figure 23.Combustion efficiency was determinedrom asanalysis data.Ashown,here is nopparentdifference in efficiency values, which were near 100percent, over a fuel-air ratio range of0.016 to 0.043.For these tests inlet pressure was varied from 0.505MPa to 0.794 MPa; the inlet temperature nominalvalue was 894 K .Combustion efficiency and pattern factor data arepresented in figure 24 as a function of fuel-air ratiofor heLamilloy nd tep-louver linerswith themodel 2 fuel module assembly. For the range of fuel-air ratios tested combustion efficiency was nearly 00percent forbothiners (fig.24(a)).Figure 24@)presentspattern actordata.Fora uel-air atiorangeof 0.013 to 0.016 there was anappreciabledifference in patternactors betweeniners; thevalues from he step-louver inerwerehigher. Thepattern factor values varied from about 0.21 to 0.31.Pattern factor values were calculated as follows:

(Ex. m a . local temp.)- Av. ex. temp.)(Av. ex. temp.)- Av. combustion nlet temp.)

Pattern actordata were alculated rom da taobtained with thexhausthermocoupleakes(fig. 12, plane).Because of thexhaustas15


18/21

loor-L i n e r

0 Lamil loyp Step louve r

L

L i n e r0 Lamil loy

.,- p Step louve rVI! I0 96 (a )Combustionefficiency.-4r

. 1 .0 2 .03Fuel-a i r at io(b )Pattern actor.

Figure 24. - Lami l loyan dstep- louver inercombustion performance parameters as afunct ion of fuel -a i r at io - for uel moduleassemblymodel 2. No mi na l nl eta ir totaltemperature,89 4 K; nominal n leta i r ota lpressure, 0.723MPa.

Fuelominalnlet module air totalassembly pressure,model MPa

0 2wi thoutentur is) 0.7230 2 (wi thentur is) .505

100

96 -

92-5 1801 .0 2030405 .M .0 7Fuel -a i r at io

Figure 25. - Lamil loy inercombustioneff icienciesasa funct ion of fuel -a i r at io - for uelmodule assemblymodel withan dwi thout entur is.Nominal n letai r total emperature,894 K.

temperature imitationon he hermocouple akes,they were removed from the test rig during operationat fuel-air ratiosbovebout 0.021 to.023.Therefore there are no pattern factor data presentedat the higher fuel-air ratios.Results of nvestigationswith heLamilloy linerand the model 2 module assembly with and without

L inermoduleassemblymodel IA 1 Step louverLami l loyven tu r i s )

c 4P0.0 20304 .0 5 .M .0 7IFuel-a i r at io

Figu re 26. - Compar ison of moke number sobtainedwi thstep- louver ineran d fue l moduleassembly model1and hose obtained wi th Lami l loy iners and uel moduleassembly model withmixing enturis.Nominalnletair otalpressure, 0.505MPa.

themixingventuris are hown in figure 25. Themodel 2 modules with venturis, permitted operationto a fuel-air ratio of about 0.065 without exceedingliner maximummetal emperature limits (table I ) .The combustion efficiency for model 2 both with andwithout venturis was nearly 100 percent for a rangeof fuel-air ratios of 0.015 to 0.045. As the fuel-airratio wasncreasedo.065,heombustionefficiency decreased to about 91 percent.Smoke Density

Smokedensitydataaregiven in figure 26 as afunction of fuel-air ratio. Data nclude test resultsusing the step-louver liner and the model 1 moduleassembly and test results using the Lamilloy liner andthe fuel model 2 assembly with venturis. The nominalinlet conditions were 0.505-MPa pressure and 894 Ktemperature. The smoke numbers obtained from thetests with oth configurations were similar.Thenumbers were below 5 for fuel-air ratios of 0.024 to0.044 and increased to bout 25 (visible smokeregion) as the fuel-air ratio was increased to 0.058.S u m m a r y of Results

Comparison of performance was madebetweendata obtained with Lamilloy liners and data obtainedwith tep-louver liners with different fuel modulemodels.Data were obtained at pressures o 0.794MPa, an inlet air temperature of 894 K , and exhaustgas temperatures to 2430 K .16


19/21

The following results were obtained:I . For the Lamilloy liner operated with a nominalinlet air temperature of 894 K and an exhaust gasaverageemperature of 1700 K , temperature-indicating paint on the hot-gas side of the Lamilloyliners was of uniform color , except at the Lamilloypanelelds. Thisndicatesminimumhermalgradients throughout the liners. The paint color onthe welds indicated a temperature about 180 kelvinsabove the inlet air temperature (or a 1074 K metaltemperature).Operation of a step-louver inerwitha hermalbarrieroating,timilar peratingonditions,showedmetal emperatures from hermocouples)somewhat lower thanhosebtained with theLamilloy liner; however, the design cooling airflowfor he tep-louver linerwas about 1.8 times hecooling flow for the Lamilloy liner.2. For the Lamilloy liner operated with a nominalinlet air temperature of 894 K and an exhaust gasaverage emperatureof 2215 K , substantialcolorchanges and indications of streaking were obtainedwith theemperature-indicatingpaint. nnerinertemperatures were 285 kelvins above he inlet airtemperature,nduter liner temperatures wereabout 380 kelvins above he inlet air emperature.These represented liner metal temperatures of 1179and 1274 K , respectively. The Lamilloy liner coolingairflow was about 9.2 percentfheotalcombustion airflow.Thetep-louver linermetal temperatures,tsimilaroperatingconditions, were 300 kelvins and460 kelvins above the inlet air temperature , for theinner nd uteriners, respectively. The oolingairflowwas 15.4 percentof the otalcombustionairflow.3 . Achange in the fuelmodulesusedwith heLamilloy liner was made to decrease the hot streaksindicated by theaint.Anotherhange in themoduleswasmadeoncreaseheoolingirpressuredifferentialacross he inerand hus toincrease the cooling airflow rate. With these changesto the fuel modules the 894 K inlet air temperature,and the 2430 K exhaust gas average temperature, anaverage liner metal temperature of 140 kelvins above

the inlet air temperature was obtained, with a localmaximum s f 280 kelvins abovehe inlet airtemperature.In addition to increasing theamilloy liner coolingairflowrom 9.2 to 13.7 percent,he newuelmodules probably improved the mixing of the fueland irtreams, whichwould tendo ecreasestreaking.4. Comparable smoke density data were obtainedwith both the Lamilloy and step-louver liners. Thesmokenumbers werebelow 5 fora uel-air atiorange of 0.024 to 0.044 and then increased to about25 as the fuel-air ratio increased to 0.058.

Lewis Research CenterNational Aeronautics and Space AdministrationCleveland, Ohio, September 5 , 1980

References1. Envi ronmenta l Protec t ion Agency: C ont ro l of Ai r Pol lu t ionorAircraft Engines-Emission Standards and Test Proceduresfor Aircraft . Fed. Regist . , vol . 38, no. 136, July 17, 1973, pp.2 . Gleason, C. C. ; Rogers , D. W; and Bahr , D. W.: ExperimentalC l e a n Co m b u s t o r P r o g r a m , Ph a s e 11. (R76AEG422, Genera lElectr ico .;A SAont rac tAS3-18551. )A SA

19088-19103.

CR-134971, 1976.3 .Gl e a s o n ,C .C .;a n dBa h r ,D .W .:T h eExper imenta lC l e a nCo m b u s t o r r o g r a m ,hase 111. (R79AEG410,Genera lElectr ic Co.; N A S Aont rac t NAS3-19736. )A SACR-135384,1978.4. E rcegovic, D avid B.: Effect of Swirler-Mounted Mixing Venturion Emissions of Flame-Tube Com bus tor Us ing Je t A Fuel .NASA TP-1393, 1979.5 .M a n s o n , S. S.: Fatigue,AComplexSubject-SomeSimpleAp p r o x i m a t i o n s ,Ex p .M e c h . ,vol .5 ,n o . 7 ,Ju ly 1 9 6 5 ,p p .6 .Wear , e r ro ld D. ; Trou t ,Ar t h u rM .;Sm i t h ,J o h nM .;a n dJones ,Rober tE .:Des ign ndPre l iminaryResul t s fSemitranspirat ionCooledLamil loy) inerorHigh-Pressure High-Tempera ture Combus tor . NASA TM -78874,1978.7 . Adam , Paul W.; and Norr i s , James W.: A dvanced Je t Engine

193-226.

Com bustor Test Faci l i ty. NAS A TN D-6030, 1970.


20/21

3. Recipient's Catalog No.2. Government Accession No.." -

I .. " . .PERFORMANCE OF SEMI-TRANSPIRATION-COOLED LINERIN HIGH-TEMPERATURE-RISE COMBUSTOR

7. Author(s)...

M. Trout, and John M. Smith- _ . - . . . . .

9. Performing Organization Name and AddressNational Aeronautics and Space AdministrationLewis Research CenterCleveland, Ohio 44135

12 . Sponsoring Agency Name and Address~" . . . - . - -

National Aeronautics and Space AdministrationWashington, D . C. 20546

. . ~" - . l

.. "_ .5. ReportDateMarch 1 9 8 1.6. PerformingOrganization Code5 0 5 - 3 2 - 3 28. PerformingOrganizationReportNo.

. . - -

E - 4 9 410 . Work Unit No.I _ ~"11 . Contract or Grant No.

. .13 . Type of Report and Period CoveredTechnical Paper

~ ~"1 4. Sponsoring Agency Code . .

" - . . . - -". . ~~~ ~16 . Abstract ... -. . - - "" . .A combustor liner fabricated from Lamilloy was tested at inlet air pressures to 8 atmospheres,an inlet air tempe ratur e of 894 K, and exhaust gas temperatures to 2430 K. Results obtainedwith the Lamilloy liner ar e com paredwith results obtained with a conventionally designed,film-cooled step-louver liner. Operation of the Lamilloy liner with counterrotating swirlconlbustor fuel modules with mixing venturis was possible to a fuel-air ratio of 0 . 0 6 5 withoutobtaining excessiv e line r meta l temp erat ures . At the 0.065 fuel-air condition the average linermetal temperature was 1 4 0 kelvins and the maximum local temperature 280 kelvins abovethe inlet air temperature. Combustion efficiency, pattern factor, and smoke data are included.

- " " -17 . Key Words (Suggested by Aut hor ls) ) 18. Distribution StatementCombustors;Combustionefficiency;Liner

stoichiometric combustionSTAR Category 07emperatures; Fuel njection; Near-Unclassified - unlimited

19 . Security Classif. (of this report)Unclassified" . 1I.0. Security Classif. (of this page)Unclassified 21 . No. of Pages19

* Fo r sa l e by th e N a t i o n a l T e c h n i c a l I n fo r ma t i on S e rv l ce . S p r l n e f l e l d . V l rg rn l a 22161 NASA-Langley, 1981


21/21

National Aeronauticsnd SPECIAL FOURTH CLASSA1 LSpace Administration BOOKWashington, D.C.20546Official Businesspenalty for -:..--- c?nn

70 1 I U , A , 030281 5 0 ~ 9 0 3 0 5DEPT OF THE AIR P O B C E-AF WEAPONS L A B O R A T O R YATTN: T E C H N I C A L L I B R A E Y (SUL)K I R T L A t J D A P B NB 87117

m

Postage and Fees PaidNational Aeronautics andSpace AdministrationNASA451

[ -

POSTMASTER: If Undeliverable (Sectio n 158Postal Manual) Do Not Return

nasa transply report

Documents