Document
VINIDEIIA ' N Old l , ",iV H VS_ i N ' ) ,FJVNEt 11 E!3J..Ng,') _;..,W3SL4_I A3"IDNV'I s_ H "q .&) sB af pu_ aHpo H "f pa_q:)!8
aau_mso t aad a_mS - II
au .t _uq o u.tq a nL s _D
aI .tqo m o mv p a p _s_ d y l
*
_aauq t o lu a maa_ da G t o
o u !q s nj . O A t. Z G - z o ssa z d mo D
t o aau_mzo ts ad s ._V-ploD
2 b/, _ 'aaa> ra 'b/
818gg-IAIJ_-VSVN I-D-g 8 aso d oEI I_a.mqaoJ . 8 1 8 9 8 mnpu_a ° molAII_a.mqaoJ .
NODOVSAV VSVN
' ": . 2 :
NASA AVRADCOM
Technical Memorandum 82818 Technical Report 82-C-1
Cold-Air Performance of
Compressor-Drive Turbine
of Department of Energy
Upgraded Automobile
Gas Turbine Engine
II- Stage Performance
Richard J . Roelke Lewis Research Center Cleveland , Ohio J effrey E. H aas Propulsion La b oratory A VRADCOM R esearch and T echnology Laboratori es Lewis R es earch Cent er Cl eveland , Ohio National Aer o nautics and Space Administration Scienti f ic and Technical Information Branch 19 8 2
A/ _ 3 - //d 63_€-
NOTICE This rep o rt was prepared to document w ork spons o red by t he Uni t ed States Government. Neither the United States nor its agent, the United States Department of Energy,nor any Federal employees,nor any of their contractors,subcontractors or their employees, makes any warranty, expr e ss or implied, or assumes any legal liab i lity or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights.
Summary engine. A general description an d some of the design
features of the UGT engine are given in reference 1.
The aerodynamic performance of the compressor-drive Lewis was assigned the tasks of the aerodynamic design turbine of the DOE Upgraded Gas Turbine (UGT) engine and testing of the compressor , the compressor-drive was determined in low - temperature air. The nominal turbine , and the power turbine. The aerodynamic designs turbine-inlet temperature was 320 K. Inlet pressures were of these components are described in references 2 to 4.
varied between 0.4 and 2.4 bars absolute. The turbine The aerodynamic design of the compressor-drive- blading used in these tests consisted of duplicates of the turbine inlet manifold and the mechanical designs of all stator and rotor castings used in development engine the turbomachinery components were performed by the tests. The as - received cast rotor blades had a significantly Chrysler Corporation. The experimental evaluation of thicker profile than design and a fairly rough surface the compressor is still under way , and the performance finish. Because of these blade profile imperfections three determination of the engine configuration of the power- turbine rotor configurations were evaluated. These were turbine is presented in reference 5. The experimental the as-cast rotor , a reduced - roughness rotor , and a rotor evaluation of the compressor - drive turbine is covered in with the blade profiles thinned to near the design profile, this report plus references 6 and 7. Reference 8 is the final Tests to determine the effect of Reynolds number on the contractor ' s report for the UGT engine program.
turbine performance were also made. The experimental evaluation of the compressor-drive The turbine efficiency with the as-cast rotor at design turbine was performed in two phases. The first phase was equivalent speed and work factor was 0.783. Test results an evaluation of the inlet manifold and stator assembly to showed an increase of 1.1 points in efficiency from determine (1) the losses within the manifold , (2) the smoothing the surface finish and another 3.1 points from conditions of the flow entering and leaving the stator , thinning the rotor profiles. The turbine performance of and (3) the stator blading performance. The results are both the as - cast and reduced - roughness rotor presented in reference 6.
configurations changed with Reynolds number. There The second phase of the program , which is the subject was very little effect of Reynolds number on turbine of this report , was an evaluation of the overall stage performance with the smooth , thin rotor blades. An performance. The turbine blading used in these tests equation commonly used to predict the change in consisted of as-cast hardware representative of the stator efficiency with Reynolds number did not satisfactorily and rotor castings used in the test and vehicle engines.
predict the measured results. A comparison between the Initially , aerodynamic performance of the as-cast blading UGT compressor - drive turbine and the compressor - drive was obtained over a range of turbine equivalent speed turbine of the DOE baseline gas turbine engine showed and pressure ratios. Because inspection of the rotor that the UGT turbine achieved a 2-point improvement in blading before the start of the turbine component tests efficiency, showed significant deviations from design in the profile shape and a fairly rough surface , two additional turbine builds were tested in which the as-cast rotor blading was
In t rodu ct ion modified. One configuration had reduced rotor blade
surface roughness , and the other configuration had the The Department of Energy (DOE) is sponsoring an rotor blade profiles reworked to more nearly approach engine research program to demonstrate an automobile the design profile. The stator was not modified in these powered by a gas turbine engine of contemporary design tests. The measured effect and an analysis of these with drivability characteristics and fuel economy that can blading modifications on the turbine performance are compete with those of a conventionally powered reported in reference 7 and summarized in this report.
automobile. In 1972 the Chrysler Corporation ' s sixth - Finally , all three rotor configurations were tested over a g e neration prototype gas turbine engine was selected as range of inlet total pressur e s at design equivalent speed to the baseline engine. Under an interagency agreement the e valuate Reynolds number effects.
NASA Lewis Research Center obtained the The stage performance of the as - cast turbine blading turbomachinery comp o nents from the baseline engine was d et e rmine d with air at a n o minal inlet temperature of and w e re given the responsibility for evaluating the 320 K and an inlet pressur e of 0.8 bar absolute.
aerodynamic performance of the components. The DOE Performance data were taken at total - to-total pressure contract with Chrysler was later amended to include the ratios from 1.3 to 2.7 and rotative speeds from 5 0 to 110 design , building , rig testing , and road demonstration of percent of equivalent design speed. During the an Upgraded Gas Turbine (UGT) engine. This was a new performance tests with the two modified rotors , data engine designed for a smaller vehicle and int e nded to wer e taken only at equivalent design speed. Rotor-exit meet the F e deral emission standards with a significant radial surveys of total pressure , total temperature , and improvem e nt in fuel economy ov e r that of the baseline flow angle were made at equivalent design speed and d esign work fact o r for all three rotor configurations. The h hub Reynolds number tests were conducted at different values rn mean of turbine-inlet pressure. The inlet pressure was varied meas measured from 0.4 to 2.4 bars absolute , resulting in Reynolds numbers from 1.2 × !0 5 to 8.0 × 105. sur survey The aerodynamic performance of the compressor-drive t tip turbine is presented in terms of equivalent mass flow , 1 subscript used in eq. (1) torque , specific work , and efficiency. A comparison is 2 subscript used in eq. (1) made between the aerodynamic performance of the 4. 5 station at manifold inlet (fig. 1) subject turbine (with the reworked rotor blades) and the 5 station at stator inlet (fig. 1) compressor-drive turbine from the baseline gas turbine engine. 5 . 5 station at stator exit (fig. 1) 6 station located about half an axial chord downstream of rotor (fig. I)
Symbols 6.3 station located about three axial chord lengths
downstream of rotor (fig. 1) A coefficient used in eq. (1) Superscripts: AR bla d e aspect rati o based on actual ch o r d length ' abs o lute t o tal state B c o effic i ent used i n eq. ( I) * U.S. standard sea-level c o nd i t io ns ( temperature , c actual ch o rd , cm 288.15 K; pressure, 1.013 bars) A h spec i fic w o rk, J / g
m mass fl o w rate, kg / sec T ur b i ne D es i gn
N ro tative spee d , rpm p abs o lute pressure, bars The UGT c o mpress o r- d rive turb i ne i s a sing l e-stage, Rx rotor reaction, ( ,o 5. 5 - P 6.3) / ( p _. 5 - P 6.3) axial-flow design having a rotor tip diameter of 11.15 cm and stator and rotor blade heights of nominally 1.12 cm.
r radius, m A cross section of the turbine as it appeared in the test rig Re turbine Reynolds number, m/l z rm is shown in figure 1. The duplicate engine parts used in s blade spacing, cm the test rig are the inlet manifold, the stator ring, and the T absolute temperature, K U blade vel o city, m / sec Instrumentation V abs o lute gas vel o city , m / sec stati o ns A Vu change in absolute tangential velocity, m / sec 5 5.5 6 6.3 W relative gas velocity, m / sec c _ absolute gas flow angle measured from axial _ direction , deg 4.5 3 relative gas flow angle measured from axial di rect io n, d eg 11.2 0 cm 3 ' ratio of specific heats 6 ratio of inlet total pressure to U.S. standard sea-level pressure, p, i. 5/p * function of 3 ' used in relating parameters to those using air inlet conditions at U.S. standard sea- level conditions, ( 0 .74 0 / 3" ) [(3'+ 1) / 2]v / (v- 1) rt' efficiency based on total pressure ratio p ,_. 5/p _.3 0 cr squared ratio of critical velocity at turbine-inlet , Stator temperature to critical velocity at U.S. standard j sea-level temperature, ( V cr /V _cr)2 //
!
# viscosity, kg / m sec f' Inletmanifold '_ r torque, N-m
I /
Subscripts: _---_-_ - , _........
cr c o nditi o n corresp o nding to Mach 1 Figure 1. - Cross section ofupo r ade O c0mpress0 r -drive turbine.
rotor. The instrumentation stations shown in figure 1 are resonance in the engine operating range of the 56 - blade discussed in the section Research Equipment and rotor. Details of the 62-blade rotor design are presented Procedure. in table Ill. The aspect ratio is 1.219 and the average The hot - engine and equivalent design conditions are trailing - edge blockage is about 11.8 percent. The rotor listed in table I. The turbine hardware was fabricated blades were designed with incidence angles ranging from undersized so that the flow passage would expand to the + 3.1 ° to - 0.4 ° and rotor reaction R x ranging from design area when the engine was operating at the design 0.134 to 0.347 from hub to tip, respectively. The design inlet temperature. Thus, it is necessary to show the rotor blade surface velocities are shown in figure 5. These equivalent flow conditions for both hot and cold were obtained by using the computer program described hardware. The design mass flow rate is 0. 5 98 kg / sec at in reference 9. Diffusion is indicated on the pressure the inlet temperature and pressure of 1325K and 4.0 bars surface of all three sections and on the suction surface at absolute. The design efficiency is 0.85 at a work factor of the tip section. The rotor is unshrouded and operated in 2.1. The turbine design velocity diagrams are shown in the component performance tests with a tip clearance of figure 2. The stator - exit flow angle averages about 66.5 °, nominally 1.7 percent of the rotor blade height.
and the rotor-exit relative flow angle averages about The turbine exhaust is a constant - area duct as shown in 54.5 °. figure 1. In the engine the compressor - drive turbine The stator and rotor profiles are shown in figure 3. The exhaust is a diffusing interstage duct, which is the inlet to inlet manifold, stator , and rotor are shown in figure 4. the power turbine. However, for the component tests, a The manifold is volute shaped with a single entry port, constant - area exhaust duct was used so that the and this imparts swirl to the flow. The swirling flow instrumentation could be located a distance downstream leaves the volute and is accelerated in an axisymmetric where there were no rotor wake effects and where the duct to the stator inlet. The stator has 15 blades, an static pressure would not be affected by the diffusion aspect ratio of 0.484, and a small amount of blade process.
camber. Table II lists design parameters for the stator.
Initially, two rotors that met the aerodynamic
performance requirdments were designed - one with 56 Research E quipment and Procedure
blades and the other with 62 blades. Reference 3 details the stage aerodynamic design with the 56 - blade rotor. The apparatus used in this investigation consisted of However , as discussed in reference 8, this design was not the research turbine , an airbrake dynamometer used to selected but was replaced with a 62 - blade rotor because of control the speed and absorb and measure the power high untwist blade stresses and the probability of blade output of the turbine, an inlet and exhaust piping system TABLE I. - TURBINEDESIGNPARAMETERS Parameter Hot engine Equivalent a Hot hardware Cold hardware Turbine-inlet temperature, K 1325 288.2 288,2 Turbine-inlet pressure, bars 4.0 1.01 1.01 Massflowrate,kg / sec 0.598 0.335 0.325 Rotative speed, rpm 58 500 27 673 27 673 Specific work,J / g 198.1 44.4 44.4 Torque, N-m 19.3 5.1 5.0 Power,kW I18.2 14.9 14.5 Turbine totalpressure ratio, 1.94 2.01 2.01 ' / ' P4.5 P6.3 Totalefficiency, n' 0.85 0.85 0.85 Workfactor, AVu / U m 2.1 2.1 2.1 Reynolds number, m / fir m 2.44xi05 3.74xi05 3.69xi05 aThe turbinehardwarewas fabricatedundersized so that the f l ow passagewould expand to the design area when the turbinewas operatingat the hot-enginetemperature. Thus the distinction betweenhot and cold hardware.
o 5 - 48 .7 - 0.40 3 o 5 - 45. 6 {VIVcr_ 5 - O. 38 1 I _ . 5" 3 7 .6 V lV c r)5.5" ( 1 851 _ 5. 5 = 45. 8 '/ Vc r ) 5 5" O.929 05 . 5 . _ . 5 c_ . 5 " _ '3 .
(W / W) 5 . 5" O . 446 ( W / W c r )5. 5 " O . 559 (W / W cr)6 . 3 • O.85 1 % . 3 = - 1 9 . 9 ° 6 .3 = - 21. 1 _ o.3 = -56 . 4 (W I Wcr) & 3 " O . 81 7 _ . 3 = - 54 .7 ( V / Vc r)6. 3 = 0.5 18 ( V l V c r)6. 3 - 0.5 2 1 ( a )Tip . (b) Me an .
_ . 5 " 5 2 . 3 'V c r)5. 5 " 1 . 026 o 5 . 5- _ 6 (W / Wcr) 5 . 5 % . 3 " -2 1 . 8 (W l Wcr)6. 3" O. 781 _ . 3" - 5 2.7 F (V l Vcr)6. 3" O. 522 (c)Hub.
Fi g ure 2. - D e si g n velocit y dia g ra m s.
Flow (a )Hub. (b)Mean. ( c ) Tip.
F i gu r e 3 . - B l a d ing prof i les andf lo wpassa g es.
(al Inlet manifold.
C-8H6% (e) Rotor.
(bl Stator.
Figure 4. - Compressor-drive turbine parts.
TABLE II. - TURBINESTATOR DESIGNPARAMETERS TABLE III. - TURBINEROTOR DESIGNPARAMETERS Parameter Hub Mean Tip Parameter Hub Mean Tip Profile radius, cm 4.445 5.004 5.563 Profile radius, cm 4.445 5.010 5.575 Actual chord, cm 2.091 2.311 2.525 Actual chord, cm 0.981 0.927 0.879 Axial chord, cm 1.067 1.168 1.270 Axial chord, cm 0.978 0.914 0.851 Leading-edge radius, cm 0.0635 0.0635 0.0635 Leading-edge radius, cm 0.0356 0.0330 0.0305 Trailing-edge radius, cm 0.0191 0.0191 0.0191 Trailing-edge radius, cm 0.0191 0.0191 0.0191 Trailing-edge blockage, 5.0 4.3 4.0 Trailing-edge blockage, 12.8 11.8 II.I percent percent Inlet blade angle, deg 50.1 46.7 43.6 Inlet blade angle, deg 49.2 45.7 38.0 Incidence, deg -2.0 -2.0 -2.0 Incidence, deg 3.1 0.1 -0.4 Exit blade angle, deg 65.5 65.2 65.5 Exit blade angle, deg -48.3 -50.4 -52.4 Solidity, c / s 1.12 I.I0 1.08 Solidity, c / s 2.18 1.83 1.56 Blade number 15 ....... Blade number 62 Blade height 1.118 Blade height 1.13 Aspect ratio, AR 0.484 Aspect ratio, AR 1.219 Radius ratio, rh /r t ...... 0.799 Radius ratio, rh / r t 0.797 including flow controls, and appropriate corresponded to the station locations in the UGT test instrumentation. Figure 6 shows a schematic of the engine. Stations 5 . 5 and 6.3 were added for component facility and a photograph of the test installation. The testing. Instrumentation at the manifold inlet (station rotational speed of the turbine was measured with an 4. 5 ) measured wall static pressure , total pressure , and electronic counter in conjunction with a magnetic pickup total temperature. At both the stator inlet (station 5) and and a shaft-mounted gear. Mass flow was measured with the stator exit (station 5 . 5 ) static pressures were measured a calibrated venturi. Turbine torque was determined by with six taps , with three each on the inner and outer measuring the reaction torque of the airbrake, which was walls. The inner and outer wall taps were located mounted on air trunnion bearings , and adding opposite each other at different intervals around the corrections for the turbine bearings and seal losses and circumference.
the coupling windage loss. These tare losses corresponded There were two measuring stations at the rotor exit , to about 7. 5 percent of the measured torque obtained at stations 6 and 6.3. At station 6, located about half an design equivalent speed and work factor. The torque load axial chord length downstream of the rotor , static was measured with a commercial strain-gage load cell. pressures were measured with six taps , with three each on The turbine instrumentation stations are shown in the inner and outer walls. At station 6.3 , located about figure 1. Figure 7 shows the instrumentation at each three axial chord lengths downstream of the rotor , static station. Stations 4. 5, 5, and 6 were chosen because they pressure , total presgure , total temperature , and flow 1 . 2 m m __ b Suction surfa c e 1.0 -- ._ .8 -- -_ .6 -- Pressure b surface -: .4 --
.2 I I I I I I I I I
. 2 . 4 .6 .8 1. 0 0 .2 .4 .6 .8 1 .0 0 . 2 .4 .6 .8 1 .0 Posi ti on,fractionof axialchord (a) H ub. (b)Mean . (c)Tip.
Figure5. - Rotorbladesurfacevelocitydistributions.
6 ,
r;::.====== - Pressurized air
Air heater " Flow venturi , Gas filter Pressu re control va Ive Test turbine To low-pressure Pressurized air -::{)I<::J:==:::::: exhaust system (a) Facility schemat;c.
(b) Turbine test apparatus.
Figure 6. - Test installation.
angle were measured. The static pressure was measured instrumentation could be located where there would be with six taps, with three each on the inner and outer no rotor wake effects.
walls. Three self-aligning probes located around the The stage test program consisted of three parts: Part circumference were used for measurement of total one determined the turbine performance with the as-cast pressure, total temperature, and flow angle. The location blading over a range of equivalent total pressure ratios of station 6.3 was determined by using a hot-wire and rotative speeds. The manifold-inlet-total-to-rotor- anemometer survey probe at several axial locations exit-total pressure ratio was varied from 1.3 to 2.7 and downstream of the rotor so that rotor-exit the speed from 50 to 110 percent of equivalent design locations at stati o n 6.3. These mass-averaged values were Flow_ then arithmetically averaged to obtain overall values. The survey probes were then positioned with one each near the tip, near midspan, and near the hub so that the
o bt ai nedf
average flow angle from these three positions would Manifold inlet Stat0r inlet correspond closely to the o verall mass-averaged value ( s t ation4.51 ( sta t ion5) obtained from the survey• Performance data were then The stage evaluation was conducted in air at nominal inlet conditions of 320 K and a range of turbine-inlet pressures from 0.4 to 2.4 bars absolute. The turbine was _ obtained at other o perating con d itions.
rated on the basis of total efficiency. The actual work was Stator exit Rotor exit calculated from torque, speed, and mass fl o w (sta t ion 5.5) (st ati on 6) measurements. The i dea] work was based on the • Static pressure manifold-inlet (stat i on 4.5) and rotor-exit (station 6.3) o T otalpre ss ur e * Totaltem p erature total pr e ssures w er e calculated from mass flow , static o Totalpressureandflowangle p re ssur e, total t e mperature , and flow angle. For the Instrumentation manifold-inlet-t o -r o t o r-exit total pressure ratio. The calculation of manifold-inlet total pressurethe flow angle Rotor exit was assum e dto be zer o .
( s tation 6 . 3 ) Figure 1 . - Flow pathi ns trumentation , viewed looking downstream.
Resul ts and Di s cu ss io n
spee d. The t es t s were conduct e d at t he h ot- e ngin e Re yno l d s num ber t h at i s li s t e d in ta b l e I. Pa r t t w o The tu rb ine per fo r manc e res ult s f r om t h i s d e t er min ed t he t u rb in e per fo r manc e of t he re duc e d- expe rim e ntal inv es tigation a re prese nt e d in fou r se ction s : b lad e-s u r fac e - r oug h n ess and t he rew o rke d- r oto r - pr o fi l e The fi rs t se c t ion p r ese nt s t he m e a s u re d per fo r manc e of con fi gu r ation s . T hese t w o modifi e d r oto r con fi gu r ation s t he a s -ca s t b lading. M a ss flo w an d to r qu e data a s we ll a s we r e e valuat e d ov er a r ang e of tu rb in e press u re ratio s at a tu rb in e m a p and t he resu lt s of a r oto r - ex i t s urv e y a re eq uival e nt d es ign spee d• Pa r t t hr e e w a s a Re ynold s pr e se nt e d. The se con d se ction prese n ts t he tu rb in e num b e r evaluation of al l t hr ee r oto r configu r ation s • per formance c h an ges , at 1 0 0 p e rc e nt of d es i g n s p ee d, a s Re ynol ds num ber w a s c h ang e d f r om 1. 2 x 10 5 to 8.0 x 10 5 t he r oto r b la de pr o fi l es were , firs t, p oli she d to re duc e t he ov er a r ang e of t u rb in e press u re r atio at d es ign e quival e nt s urfac e r oug h n ess and, se cond, r ew o rke d t o mo re n e a r ly spee d. Re ynold s num ber w a s c h ang e d b y va r ying t he matc h t he d es ig n p rofil e . The t h i r d se ction prese n ts t he tu rb in e -inl e t tota l press u re , e ff e ct of Re ynold s num ber on s tag e per fo r manc e fo r e ac h In e ac h p a r t of the t es t pr og r am a roto r - e xit r adia l of t he r oto r con fi gu r ation s . In t he la s t se ction a s u r v e y w a s f i rs t conduct e d at e quival e nt d es ign valu es o f com p a r i s on i s mad e be t wee n the rew o rke d U GT spee d an d spe cific w o rk . M a ss -av er ag e d valu es of f lo w com pr e ss o r -d r iv e tu rb in e and t he com press o r-dr iv e ang le , t otal t e m per atu re , and to t al press ur e were tu rb in e of the b a se lin e ga s t urb in e e ngin e .
• 3 4 -- Te s t s p e ed, pe rc entof Design-_..
equivalent u ° design s peed -- _ F60 r50 _ .30 -- g _o Cll 0 ClO0 -90
•26 I I I I I I I
1.3 1 . 5 1 . 7 1 . 9 2 . 1 2 . 3 2 . 5 2 . 7 Equivalent total pressureratio , p_.5 1 P_ . 3 Figure8. - Variationof equivalentmassflowwith total pressureratioandspeed; as-castblading.
Perfor m ance of As-Cast Binding Equivalent total pressureratio , M ass flo w . -The v a ri a tion i n eq u iv a len t m a ss flow Pa. 5 / P6.3 Equivalent de s ign with equiv a lent tot a l pressure r a tio a nd rotor speed is 50x103 speedand w0rkfact0r-a 110 shown i n figure 8 . The equ i v a lent des i gn m a ss flow g i ven \ _ z . 5 i n the f i gure i s the des i gn m a ss flow v a lue for cold 9° "_ i L'"_ z .3 h a rdw a re th a t i s l i sted i n t a ble I. The d a ta i n figure 8 p e r cento f equivalent show th a t a t the two lowest rotor speeds tested the st a tor 45 des i gn speed Test s peed , choked before the rotor but as the rotor speed was incre a sed the rotor choked first. At 1 0 0 percent of __ z .1 / ,__ . , 771 : ,' ° I equivalent design speed and at the equivalent design total pressure r a t i o of 2 . 01 , the me a sured m a ss flow w a s 0 . 306 4o " 1 ; " / , _ " i J"x I the stator a nd rotor thro a t a reas were measured and , b, 6c kg / sec , which is 6 percent less than the design flow. Both _ _ L9 when compared with the cold design areas , were found to be 4 . 1 a n d 6.3 percent s ma ller , respectively . Th i s < 35 j _-- - -_. /f _1.8 mismatch in stator and rotor flow areas caused a
' 11 / I
t_ redistribution of pressure in the turbine, which in turn ._ 1._- ] - _--_ 1.7 changed the flow characteristic of the machine. It _ 30 appears, however, that most of the mass flow deficit was caused by the undersized hardware. ._ 1.6 Torque . - The v a r ia tion o f eq ui v a lent torque w it h _ / / equivalent total pressure ratio for the equivalent speeds 25 tested is shown in figure 9. The design value of torque shown in the figure is the design value for a cold turbine.
At the equivalent design conditions of 100 percent speed 20 1.4 a nd a total pressure ratio of 2.01 the measured torque Total efficiency, q' was 4. 3 6 N - m . T h i s i s 1 2.6p e r cen t l ower tha n the de s ign t or q u e . S i x p er c e n to f th i s d e fici t was d u eto the shortfa ll i n m ass flow o f the t u rb in e, a nd the rema ining 6.6 15 I I I I percent was caused by increased aerodynamic losses. 4 5 6 7 8 9 10 103 P er fo rm an ce m ap . -The perform a nce m a p shown in Mass flow - spee d parameter , _mN / 6 , kgrpm l sec figu re 10 was g e n erate d fro m the m assflow a nd tor qu e Figure 10 . -C0mpressor-dr i veturbine p erformance map ; as-cast blading.
Test speed, 8 -- perce n t ofequivalent d a ta of f i gures 8 and 9 . At a given pressure ra t i o , a nd for designspeed each speed, smooth-curve values of torque and mass flow z - _ 5 0 were used to calculate the equivalent specific work 6_. .- --_ 60 A h/O cr, the mass flow - speed parameter _m N/ _, and the 7 0 The turbine attained efficiencies of 0.62 to 0.79 over 6 __...._ 8o total eff i ciency _7 ' .
E .n_o_ _ ^_ . . ._o90 the r a nge of test condit i ons . At the equivalent design which is 0.067 less than the design efficiency of 0.85. A _ 5 -- _ _ r ""'__r _ __ .. . . . .. .. . . . . _ _ y _ . 0"'_100110 speed a nd work factor the turb i ne efficiency was 0.783 , _ significant part of this deficit was caused by manufacturing imperfections in the rotor blade profiles , ._ a s is discussed la ter i n this report .
Roto r - ex it s urv ey . - The results of the r a d i al surveys a t station 6.3 of flow angle, total pressure, and total temperature are shown in fig u res ll(a), (b), and (c). The measurements were taken with the turbine operating at equivalent design speed and specific work. The data shown are the averages of the measurements of the three I [ I I I I I combination probes. With these measurements the radial 1.3 1 . 5 1. 7 1.9 2. 1 2 . 3 2 . 5 2 .7 Equivalent totalpressureratio , p_.5 / P_.3 variation in stage efficiency was calculated and is Figure 9. - Variationof torquewith totalpressureratioandspeed; as - cast blading , presented in f i gure 1 1 (d ). The dashed curves i n the figures -3 0 [- t _ _ efficiency ( fig. 1l(d)) reached a maximum value of 0.82 o _ _ toward both endwalls. The probable reasons for the :r-- __ at about 30 percent of passage height and then fell off . _ __ -2 0 decrease in efficiency are the same as those that caused # the increase in stage pressure ratio. The rotor-exit survey " 6 = -1 0 [ I I ] ] results were also used to calculate a mass-averaged turbine efficiency. The value calculated was 0.781, which (a)Flowangle.
2 . 4-- is in excellent agreement with the value of 0.783 obtained = o- _ from torque, speed , and mass flow.
_ =_ 2-2_ Effect of Rotor Blade Profil e I m perfection s __ As mentioned earlier the turbine blade rows used for _- Z . 0 I component testing were dupl i c a tes of the st a tor and rotor (b)l0tal pressure rati0, c a stings used in the engine. Inspection of the rotor _, .861--- blading before the start of the turbine testing showed / _•__q _.= F-Desig n significant deviations from design in the prof i le shape _.-_ o_ -- - ,' ... and a fa i rly rough surface. Figure 12 compares inspect i on I:: _ -,_: ;
_- _- .84 I I tr a c i ngs of the mean and tip sections with the design
( c l T emperature rat i o• profile of two r a ndomly selected rotor bl a des . Hub •9- section tracings were not obtained because the tracing _= - . - stylus was too large to fit in the small hub area.
. _ I._'" \\ . "_ _.
__ ,_ __ Bec a use of the condit i on of the a s-rece i ved rotor ___= .8 _ casting, two modifications were made to the rotor -_ blading after completing the performance tests of the as-
I I r I I
o . 7 0 20 40 60 80 100 In s pection trace Hub T ip D e s i gnp r of i l e P a ssage height, p ercent Figure11. - Turbine-ex i t s urveyat equ i valent d e si gn s peed k and s pecific work; a s -ca s t blading .
represent the calculated design radial variations at the / The flow angle measured was within ±2 ° of the design (d)Totalefficiency. _k 1 turbine exit. / variation over most of the blade span except near the , / //., hub. At the hub the flow was overturned. This , _ ,, / overturning may have been caused by secondary flows at the hub.
The st a ge pressure ratio (fig . l l ( b)) w a s signific a ntly (a) Tip se c tions.
higher than the design variation over most of the passage height. It is belie v ed that the thickened bl a de profiles ___ (x\ in a higher than design pressure ratio across the entire passage height. In addition, thick turbine-inlet boundary layers and stator incidence near the endwalls most likely X\ increased the rotor trailing-edge mixing losses, resulting __) \ caused higher pressure ratios in those regions. The 0.91-cm J manifold - stator tests ( ref . 6) i ndicated a turbine-inlet hub , ¢ a x ia l ch or d /I
t and positive stator incidence up to 20 °. At the tip the inlet
boundary layer and stator incidence, although reduced, // were still significant. / " # boundary layer equal to 20 percent of the passage height _ / / L // _ // The temperature measurements (fig. ll(c)) indicated nearl y constant work extraction radiall y and agreed well (hi M ean s ecti o n s.
with the design v a riation. The rad ia l v a ri a tion in st a ge Figure 12. - Com p arison of design and as-cast rotor blade profiles.
IO cast blading. The first modification consisted of reducing .34- Design-, the blade surface roughness by polishing the suction _ _ O , - R ew 0rked prof ile R e d uc e d ,' surface of each blade and applying a thin coat of lacquer _ _ r0u / to the pressure surfaces. Tests were then made with this _ _ .30-- g_ configuration. The second modification consisted of ±_'_
.,4 5
electric discharge machining the r o tor profiles to the "_E -- // 1_ As cast] design profile. Inspection traces at the mean and tip of _ _ . z 6 ] [ the machined rotor agreed closely with the design profile, (a)Mass fl0w.
but rotor throat measurements indicated that the hub 84- section was still thick. However, any further hub D e si gn ,- Rew0 r ked pr0file machining may have resulted in undersized profiles away - sp ecln cW Or K _._.. _ , from the hub, so no additional machining was attempted. _ 80 -- _ \ The respective suction- and pressure-surface roughness "_. / / _'_x"_ measuremen t s of the reworked rotor were es s entially t he _ m- , ,- Reduced r o u gh n e ss same as th o se measured after polishing and coating the __ 7 6 1/ _\ As ca st as-cast rot o r. Tests were then conducted on the rew o rked / \ -" rotor blading. Table IV lists selected geometric measurements of the three rotors. Reference 7 gives 7 2 ] ] ] ] 1.2 1.6 2.0 2.4 2 .8 additional details of the blading modifications and test E qui valent total pressurer a t i o, P}I. 5 1 P_. 3 results. The chief findings of the effect of these rotor blading changes on the turbine performance are (b ) Eff i c ie ncy.
summarized in the following paragraphs. F i gure 13. - V ar i a t i o no f mass fl0wan d eff i c i ency w i th pre s sur e rat i oat d e s ign s pe e d for three rotorcon - Mass flow and overall efficienc y . - The variations in f i gurations.
equivalent mass flow and efficiency with the stage total pressure ratio at equivalent design speed are shown in 0.783, 0.794, and 0.825 for the as-cast, reduced - figure 13. The lowest mass flow was measured with the roughness, and reworked - profile configurations, as - cast rotor and the highest mass flow with the reworked respectively. According to influence coefficients for the rotor, but the difference was small, only about 0.7 UGT engine cycle, an increase of 4 points in compressor percent at the pressure ratio of 2.01. The flow area turbine efficiency would have increased the engine power increase of the reworked rotor was 3 percent. These 7.5 percent.
results indicate that at this rotor speed the as - cast rotor Radial variation in efficiency . - The radial variations choked just before the stator but that with the reworked in turbine efficiency calculated from rotor - exit rotor installed the stator choked first and therefore measurements of total temperature, total pressure, and limited the stage mass flow. The difference in efficiency flow angle are shown in figure 14. As can be seen the between the as-cast and reduced-roughness rotors was largest increase in efficiency obtained by improving the nominally 1 point and that between the as-cast and surface finish and reducing the blade thickness occurred reworked-profile rotors was nominally 4 points. At the from midspan out to the tip. This may have occurred design value of specific work the turbine efficiencies were because, as mentioned earlier, it was difficult to improve TABLE IV, - ROTOR GEOMETRIC CONPARISON Parameter Design As cast Reduced Reworked roughness profile Surface finish, pm: Suction surface (a) 1.35 0.33 0.33 Pressure surface (a) 1.35 0.95 0.95 Average trailing-edge 0.038 0.053 0.053 0.042 thickness, cm Average trailing-edge ll.8 16.5 ]6.5 13.0 blockage, percent Profile tolerance, mm ±0.I ±0.15 max ±0.15 max ±0.025 aSpecification unknown.
II .9 - Design-,x increase in efficiency was measure d for Reynol d s numbers above the turbine design value, and only a slight _= . - '''_ - '-0-,_-_ _ , _,......43.._... , _ , _ Reworked • _ _ . _=:_:_--,_._ - _ - _ X__pr0file decrease was noted at lower values.
_ .8 ._"_ ---Reduced Several equations have been used in attempts to = [_ _ roughness correlate turbine efficiency with Reyn o lds number. A -_ _As cast frequently used form from reference 10 is
o, . 7 I I f I I
20 40 60 80 100 H ub Tip l--r /; (Re2'_ 0"2 Passage height, percent I -- r / '2 -- A + B \ _II , ] (1) Figure14. - Radial variationin efficiencyat designspeed and specific work forthree rotor configurations, where A + B = I. The coefficients A and B are used to the bla d e profile near the hub. Also to be noted is that the pr o porti o n the turbine losses between viscous an d nonviscous effects. The subscripts 1 and 2 correspond to radial variation in efficiency for the reworked rotor approached the same shape as the design variation separate points on a performance curve of efficiency although the level was lower. Mass-averaged efficiencies versus Reynolds number. The reference suggests that A were calculated from these data and compared with the ranges from 0.3 to 0.4 and the corresponding range for B corresponding efficiencies shown in figure 13. The is 0.7 to 0.6. This equation, with A =0.4 and B =0.6, maximum difference between the two methods of overestimates the effect of Reynolds number for all the calculating the turbine efficiency was 0.8 point, rotor configurations for the range of Reynolds number Causes of performance changes .-The 1-point investigated. It may, however, provide a reasonable improvement in stage efficiency resulting from correlation at lower Reynolds numbers. The preceding smoothing the rotor blade surfaces was attributed to a equation appears to oversimplify the correlation of loss with Reynolds number. A dditional but at this time lower profile friction loss. A number of possible causes were considered for the additional 3-point gain with the unknown factors, perhaps both aerodynamic and reworked rotor. Possible causes considered in reference 7 geometric, are needed to provide a better correlation of were stator and rotor reaction changes, rotor incidence turbine loss with Reynolds number.
changes, and differences in the rotor trailing-edge losses.
Test data were used to calculate stage velocity diagrams Comparison of Upgraded and Baseline for all three rotor configurations. On the basis of those Com p ressor-Drive Turbines calculations it was concluded that the reaction and The performance of the reworked upgraded incidence changes were not large enough to be a major compressor-drive turbine was compared with that of the factor contributing to the change in performance. An baseline compressor-drive turbine to determine if a analysis of rotor trailing-edge losses, however, did performance gain was realized. The performance of the indicate that most of the performance gain between the baseline turbine is reported in reference 11. Most as-cast thick and thinned rotor blades, both with the experimental procedures used to evaluate the two same surface finish, was due to reduced trailing-edge turbines were identical except for the method used to losses of the reworked blades, calculate the turbine pressure ratio, which in turn was used to calculate the efficiency.
E ff ect o f Reynolds Number The baseline turbine pressure ratio was determined Reynolds number tests were made for all three rotor from measurements of total pressure at the manifold inlet configurations. As mentioned earlier the Reynolds (station 4.5) and at the mean radius 1 / 2 blade chord number was varied by varying the turbine-inlet pressure, downstream of the rotor (a location corresponding to For each pressure the Reynolds number and the turbine efficiency were calculated , from smooth-curve data, at / r-Rew0rked profile the design work factor of 2.1. The results are shown in 84 -- / figure 15. The turbine design Reynolds number at hot- _A _ Reduced roughness engine conditions was 2.44 × 105. _ 80 _ \ The turbine efficiency with the as-cast and reduced- "5 _As cast roughness blading showed stead y improvement with -o 7 6- . / -DesignReyn01dsnumber i / " Reynolds number over most of the range of Reynolds _ [ I I I [ I I number tested. The as-cast blading efficiency leveled out z 2 2 3 4 5 6 7 8 105 at a value of 0.796 when the Reynolds number was Reynolds number, m l pr m 6.0 X 105or higher. The effect of Reynolds number on the Figure 15. - Variationof turbine performance with Reynolds performance of the reworked rotor was very slight. No number forthreerotorconfigurations.
station 6 in fig. 1). This measured pressure ratio was then a slightly lower aerodynamic loading as indicated by the adjusted to agree with the mass-averaged pressure ratio work factor. The smaller size of the upgraded turbine obtained from radial surveys of the flow at station 6. resulted in more rotor trailing-edge blockage and a lower Station 6 was used for the baseline turbine because there Reynolds number. These aerodynamic features of the was a diffusing interstage duct downstream of that upgraded turbine, with the exception of the lower work station, factor, increased the difficulty of obtaining high turbine As mentioned earlier the upgraded turbine was tested efficiency. However, the upgraded turbine achieved a with a constant-area exhaust duct, and hot-wire 2-point improvement over the baseline turbine. This measurements were taken to find an axial location improvement is probably due to the use of improved (station 6.3, fig. 1) where no rotor wakes were design computer codes between the time the baseline distinguishable. Radial surveys of the upgraded turbine turbine was designed and the time the upgraded turbine were taken at stations 6 and 6.3, and the turbine was designed.
efficiency was calculated from mass-averaged as well as
calculated total pressures at the two stations. The Concluding Remarks
calculated total pressure was obtained from continuity and energ y considerations as described in the section The results of this program are encouraging and Research E quipmen t and Procedure. The upgraded indicate that attaining the effciency goal of 0.85 for an turbine efficiency based on the mass - averaged pressure at axial turbine the size of the upgraded turbine is not station 6.3 and the efficiencies based on the calculated unrealistic. A number of design and hardware total pressures at stations 6 and 6 . 3 were nearly identical deficiencies have been identified, and correction of these and were nominally 2 points lower than the efficiency would improve the turbine performance. The corrections calculated from the mass - averaged pressure at station 6. include reducing the inlet boundar y layers and stator This difference in efficiency was attributed to the loss incidence, fabricating more accurately the stator and caused by rotor wake mixing, which was not complete at rotor profiles, and improving the profile surface finish .
station 6 . Therefore, it was concluded that the Fewer, longer chord blades would reduce the rotor efficiencies reported for the baseline turbine (ref . 11) did trailing - edge loss, which was found to have a significant not include the total rotor wake mixing loss. Because of effect on performance. Because this change would also that the baseline turbine efficiency was recalculated by reduce the blade aspect ratio, a balance between the two using a calculated total pressure at station 6. It is that effects is necessary. Finall y , the application of stator baseline turbine efficiency that is compared with the endwall contouring and nonuniform radial work upgraded turbine efficiency, distribution may also be beneficial.
Selected aerodynamic parameters of the baseline turbine and the upgraded turbine with the reworked
profiles are given in table V. The turbine efficiencies Summary of Results
listed are the efficiencies with the turbines operating at their respective equivalent design speed and design work The aerodynamic performance of the compressor - drive factor. As can be noted, the upgraded turbine is smaller turbine of the Department of Energy Upgraded Gas than the baseline turbine, has less rotor reaction , and has Turbine engine was determined in air at nominal inlet conditions of 320 K and 0.8 bar absolute. Three TABLE V. - COMPARISON OF UPGRADED AND BASELINE modifications of the same rotor design were tested: an as- COMPRESSOR-DRIVE TURBINES cast rotor, the same rotor with reduced surface roughness, and the rotor with thinned blade profiles.
Reynolds number tests were made with all three rotors by Parameter BaselineUpgraded varying the inlet pressure between 0.4 and 2.4 bars absolute. The results of the investigation were as follows: Massflowrate a, kg / sec 0.558 0.308 1. The turbine efficiency at design speed and work Tip diameter, cm 14.0 11.1 factor with the as-cast blading was 0.783. The efficiency Tip clearance, percent 1.7 1.7 increased to 0.794 after the blade surface roughness was Rotor trailing-edge blockage, 10.8 13 re d uce d and reached 0.82 5 after the r o tor profiles were percent thinned.
Design work factor 2.4 2. l 2. The performance of the turbine with the as - cast Design Reynolds number, m / iJr m 3.3x1 0 5 2.44xi 0 5 blades varied with Reynolds number. The efficiency of Rotor mean reaction a, Rx,m 0.31 0.20 the as - cast turbine leveled out at a maximum value of Measured stage efficiency a, n ' 0.803 0.825 0.796 at a Reynolds number of 6.0 × 105. There was very little effect of Reynolds number with the rotor blades aFrom measurements taken at design work factor, thinned and smoothed.
Reyno l ds number, andspeed.
3. The change i n efficiency with Reynolds number for 3. Roelke, R. J.; and McLallin, K. L.: The Aerodynamic Design of a C o mpress or -Drive Turbine f o r Use i n a 75 kW Aut o m o t i ve all turb i ne c o nf i gurati o ns tested was not satisfacto r ily Engine. NASA TMX-71717 , 1975.
pre d icte d by the equati o n 4. K o fskey , M. G.; Katsanis, T.; and Schumann, L. F.: Aer o dynamic Des i gn o f a Free P o wer Turbine f o r a 7 5 kW Gas Turbine ' / { Re2 "_0.2 Automotive Engine. NASA TM X-71714, 1975. 1-- r/l - Z + B / 5. McLallin, K. L.; K o fskey, M. G.; and Wong, R. Y.: Cold-A i r I - _7_ \Rel , , ] Performance of a 15.41-cm-Tip-Diameter Axial-Flow Power Turbine with Variable-Area Stator Designed for a 75-kW where r t' is the efficiency based on total pressure ratio, Automotive Gas T urbine Engine. NASA TM-82644, DOE / NASA / 51040-30, 1982.
the subscripts 1 and 2 correspond to separate points on a performance curve of efficiency versus Reynolds 6. Roelke, R. J.; and Haas, J. E.: Cold-Air Performance of Compressor-Drive Turbine of Department of Energy Upgraded number, Re is Reynolds number, and the coefficients Z Automobile GasTurbineEngine. I - Volute-Manifold and Stator and B are used to proportion the turbine losses between Performance. NASA TM-82682, 1981.
viscous and nonviscous effects. 7. Roelke, R. J.; and Haas, J. E." The Effect of Rotor Blade Thickness and Surface Finish on the Performance of a Small Axial 4. At their respective design conditions the upgraded Flow Turbine. ASME Paper82-GT-222, 1982.
turbine achieved a 2-point improvement in efficiency 8. Wagner, c. E.; and Pampreen, R. C." Upgraded Automotive Gas over the baseline turbine. Turbine Engine Design and Development Program Final Report.
(COO-2749-43-VOL-2, Chrysler Corp.; EY-76-C-02-2749.)
NASA CR-159671, DOE / NASA / 2749-79 / 2-VOL-2, 1979.
Re f erences 9 Katsanis, T.: FORTRAN Program for Calculating Transonic
Velocities on a Blade-to-Blade Stream Surface of a Turbomachine.
1. Ball, G. A.; Gumaer, J. I.; and Sebestyen, T. M." The NASA TN-5427, 1969.
ERDA / Chrysler Upgraded Gas Turbine Engine Objectives and 10. Glassman, A. J.: Turbine Design and A pplication. NASA SP-290, Design. S A E Paper 760279, 1976. Vol. I, 1972, p. 60.
2. Galvas, M. R.: A Compressor Designed for the Energy Research 11. Roelke, R. J.; and McLallin, K. L.: Cold-Air Performance of the and Development Agency Automotive Gas Turbine Program. Compressor-Drive Turbine of the Department of Energy Baseline NASA TM X-71719, 1975. Automobile Gas-Turbine Engine. NAS A TM-78894, 1978.
1. R e port No. NASA TM-82818 2. G ove r nment Ac c e s sion No. 3. Recipient ' s C a talog No.
AVRADCOM TR 82-C-I 4 . T i t le and Sub t itle COLD-AIR PERFORMANCE OF COMPRESSOR- 5. Repor t Date DRIVE TURBINE OF DEPARTMENT OF ENERGY UPGRADED OCTOBE R ]982 AUTOMOBILE GAS TURBINE ENGINE 6. Performing Organization Code II - STAGE PERFORMANCE 505-32-2B 7 . Au t hor (s) 8 . Perf o rm in g Org a nization Report No.
Richard J. Roelke and Jeffrey E. Haas E-1165 10. Work Unit No.
9. Performing Org a nization Name a nd Addr e ss NASA Lewis Research Center and AVRADCOM Research and Technology Laboratories 11. Contract or Grant No.
Cleveland, Ohio 44135 13 . Ty pe o f R e por t a nd P e rio d Cov e r e d 12 . Spo ns ori ng A g e n cy N a me and A ddr e ss Technical Memorandum U.S. Department of Energy Office of Vehicle and Engine R&D 14. Spon s ori n g Agency _ Report No.
Washington , D.C. 20545 DOE / NASA / 1011-36 1 5. Supplementary N otes Richard J. Roelke , Lewis Research Center; Jeffrey E. Haas, AVRADCOM Research and Technology Laboratories. Prepared under DOE / NASA Interagency Agreement EC -77-A-31-1011.
16. Ab s tract The aerodynamic performance of the compressor-drive turbine of the DOE Upgraded Gas Tur- bine engine was determined in low-temperature air. The as-received cast rotor blading had a significantly thicker profile than design and a fairly rough surface finish. Because of these blading imperfections a series of stage tests with modified rotors were made. These included the as-cast rotor_ a reduced-roughness rotor , and a rotor with blades thinned to near design.
Significant performance changes were measured. Tests were also made to determine the effect of Reynolds number on the turbine performance. Comparisons a re made between this turbine and the compressor-drive turbine of the DOE baseline gas turbine engine.
1 7. K ey Wo r ds ( S u gge ste d b y A uthor(s ) ) 18 . D istr ib ution S tate m ent Automotive gas turbine engine; Small axial Unclassified - unlimited flow turbine performance; Effect of blade STAR Category 02 profile inaccuracies; Reynolds number DOE Category UC-96 1 9 . S ecu ri ty C lass i f . (o f thisre po rt ) 20. S ec u rity C la ss i f. (of th i s pa g e) 21. No . o f P a g es 22. P rice * Unclassified Unclassified 16 A02 * ForsalebytheNa t ional T echnical In f ormation Service, Springfield, Virginia 22 16 1 N A S A - Lang l e y , 198 2 Nat i onal Aeronautics and T H IRD- C L ASS BUL K R AT E Postage and F e e s P ai d #_=m=="_ Space Administrat i on N ational Aeronautics and Sp ace Adm i n i st r ation Wash i ngton, D.C . NASA-451 20546 Off ici al Bu si n ess Pen a lty for Privat e Us e , $300 __A P OSTMASTE R : I f Undel iv e r ab le (Sec ti on 15 8 Po s t a l Manua l ) Do N ot R e t u r n