RESEARCH MEMORANDUM

ALTITUDE OPERATIONAL CKARACTERISTICS OF A PKOTOTYPE

MODEL OF THE J47D (RX1-1 AND RX1-3) TURBOJET ENGINES

WITH INTEGRATED ELECTRONIC CONTROL

By E. William Conrad, Harry E. Bloomer, and Adam E. Sobolewski

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS WASHINGTON January 8, 1952

...... I ,...... CINCLASSIFIED

1N NACA RM E5U?,O8 L

U It

w0 ;j MODEL CIE' ICBE J47D (RX1-1 AND RXl-3) TURBOJEIT

By E. William Conrad, Ea- E. Bloomer, and Adam E. Sobolew8ki

SUMMARY The altitude operatianal characteristicsof a prototype modelof the J47D (RXl-1 and RXl-3) turbojet engines, which includesan after- burner, a variable-area exhaust nozzle, anand integrated electronic control, were investigatedin the NACA Lewis altitude wind tunnel at altitudes from10,000 to 55,000 feet ata flight Mach numberof 0.19 and at fli@t Mach numbers from 0.19 to 0.89 at as altitude of 25,000 feet. Data obtained with oscillograph recorders and conven- tioa instrumentation are presented to showfollowing the charac- teristics: (a) Compressor stan

cb) Combustor blow-out (c) Acceleration (d) Deceleration (e> Altitude starting

[f) Afterburner ignition

(g) Afterburner transients (h) Afterburner blow-out

For both of the engines investigated(FZL-1 and RIEL-31, it was found that the compressor-s€alldata plotted as single curveson coordinates of compressor pressure ratioand corrected engine speed. TKO distbct types of stall appear to exist with the transition occur- ring at corrected engine speeds between5250. and 6250 rpn. The com- pressor unstall characteristics areshown on the same coordinates. 2 NACA F44 E51Eo8

rpm, For corrected engine speeds above5800 the pressure ratlo neces- +L sary to unstall the capressor lower was than %he pressure ratio for either stall or steady-state operation.As the altitude wss increased, the compressor unstalled at slightly higher pressure ratios; however, flight Mach nuniber had no apparent effect within the range investigated. P Combustorblow-out data for .dlflight conditions investigated with 8 the RXL-1 engine were plotted aas single curveon the same coordinates and coincided almost exactly with the compressorstall curve. Acceleration time from the toidle the rated thrust condition without afterburning Fncressed from14 to 22 seconds as the altitude was increased from 15,000 to 45,000 feet ata flight Mach number of 0.19. At an altitude of 25,000 feet, &~1increase in flight Mach number from0.19 to 0.75 reduced the accelerationtime from idle to rated thrust without afterburningfrom 14.4 to 6.0 skconds. When a minlmum fuel limit of 450 jpunds per hour ms used, lean combustor blow-out could not be obtainedduring deceleration for the range of flight conditions covered by the investigation.

’ Ignition in all combustors was obtained at.windmilling speeds from ,1300 to 1500 rpu at an altitudeof 50,000 feet usingMIL-F-5624 (AN-F-58) fuel ata temperahre of about 70° F and inlet-air tempera- tures ofOo to -So F. At 35,000 feet, ignition occurredFn all com- c bustors up to,3500 rpm,. the highest windmilling speed obtainable. With the useof MIL-F-5624 fie1 (treatedto give a l-pound Reidvapm pressure) ata temperature of apprdximately90° F and inlet-air tem- peratures from-200 to 300 F, ignition was possible inall combustors at 49,000 feet .upto a windmilling speed .of about JE.0rpn. At 25,000 feet, however,.starts were possible upto 2200 rpn. At an altitude of 40,000 feet, theoptimum fuel flow for starting appeared to be about650 pounds. per hour for ME-F-5624 f’uel with a 7-pound vapor pressure. merburner starts by autoignitionwere obtained at altitudes up to 53,000 feet at a flight Mach numberof 0.19 using MIL-F-5624 fuel. The tail-pipe fuel-air.ratios required for autoignition increased with.. altitude d.at35,000 feet decreasedas the burner-inlet temperature was raised. The tail-pipe _fuel-air ratio at leanwhich blow-out of the afterburner occurredwas increased as altitude was raised. The width of. the blow-out- band- also .increase& with altitude. RACA RM E5lXO8 3

The improvement in performance characteristicsof turbojet engines effectedby the applicationof a$terburning and a continuously variable-area exhaust nozzle have been well established during the past few years. Manual control of a turbojet engine equipped withan afterburner and variable-area exhaust nozzlemad, however, placea heavy burdenon the pilot or flight engineer.To relieve the pilotof the extreme complexityof engine operation and 'the need for constant surveillance of mahy engine operation limits,a completely automatic control system is required.

Accordingly, a prototype modelof the J47D (RXI-1 and RXL-3) tur- bojet engineswas provided withan afterburner, a variable-area nozzle asd an integrated electronic control. The enginewas installed in the NACA Lwd-6 altitude wind tunnel to obtain the performance character- istics and insight into the engine's operational problems. The steady- state performance of the engine without- afterburning isin presented references 1 and 2. me operational characteristics are presented herein.

The integrated electronic control usedthe with prototype model of .. the J47D (RXl-1 and Rx1-3) turbojet engineswas designed to (I) provide - s-e-lever thrust control over the full range of operation from starting tofull afterburning condition, (2) schedule all services required, and (3) give the maximum acceleration and deceleration rates

I possible without exceeding engine speed and temperature limits or causing cdustor blow-out or compressorstall. To attain the latter objective, the ccmgressor stall and combustor blow-out regions. were investigatedwith the control inoperative. The limits maintained by I the control were then adjustedas necessary and the transient per- formance of the controlled engine was evaluated.

Oscillograph traces axe presented hereinto show the typical behavior of these variables during compressor stall, comp1et.e and partial conibustor blow-out during acceleration, controlled accelera- tion and deceleration, and afterburner ignition. Compressorstall and combustor blow-out limits have been correlatedto show the effects of altitude, flight.Mach numberand corrected,engine speed. The effects of altitudeand flight Mach number on engine acceleration snd deceleration arealso given, as wellas steady-state operational char-. acteristics suchas engine and afterburner ignition apd aFterburner 4 NACA HM E5lEO8

lean blow-out. Data were obtained at altitudes from 10,000 to 55,000 feet at a flight Mach.number of 0.19 and at flight Machnumbers from .. 0.19 to 0.89 at an altitude of 25,000 feet: .

DESCRIPTION OF ENGIXE

The J47D RXl-1 and RXl-3 engines used in this invesfiigation were aerr>"xLly the same as the J47D engine. - The manufacturer guaranteed static sw-level performaace of.tk J47D engine is 5670 pounds thrust at 7950 rpm'ana an exhaust-gas telqperature of 1275O F. The main chpnents include a 12-etage axial-flow compressor with a pressure ratio of about 5.1 at rated conditions, eight cylin- .. . drical direct-flow combustors, a single-stage turbine, a diffuser, an afterburner maustion cbber, a variable-&ea exhaust nozzle, and an integrited electronic control. The over-all length of the engine including the afterburner is about 217 inches', the maximum diameter is approximately 37 inches, and the total weight is about 3000 pounds. A view of the turbojet engine installed in the test section of the altitude .wind tunnel is given in fLgure 1 and a schematic drawing of the engine ie presented in .figure 2. Two combustor configurations were used with the RXL-1 engine. The original configuratfon included ,conventional-type spark plugs,. a 20,000-V0lt, coil and vibrator unit, and cross-flre tubes I&inches in diameter. The second configuration,referred to as the Iff modified combustor," had cross-fire tubes that were 2 inches in diameter. The modified combustor liners also had semlcylindrical shroud8 projecting 1/4 inch radially inward from the damstream half of each of the first six rows of holes. These ch&nges,- as well a8 other ininor hifferences between the original and modified combustor liners are shown in figure 3. The ignition systems used with the modified configuration were (1) qpposite polarity spark plugs mounted as shown in figwe 4 'and (2) an ignition system which produced a potential difference approximately twice that of the standaxd system. Only the modified configuration was used with the Rx1-3 eane. For both combustor configurations, two sets of spark plugs, located fa diametrically opposite combustors, were. used, . . ..

The afterburner shown schematically in figure 2 wa6 camprised of a diffuser 43 inches in length, a ccmibustionchamber 50 inches in length ' which tapered from a J%-inch1 diameter at the flame holder to a 29-inch diameter at the e&ust-nozzle section, and a variable-area exhaust nozzle whfch was 16 inches in Length in the open position. Flame-holding surfaces were provided- with 2-rbig, V-gutter flame PTllCA RM E51EO8 5

holders used in conjunction with a d-inch diameter pilot cone. The 2 location of the flame holders was varied from 2 to 7 inches downstream of the pilot wne. Fuel was injected by means of 24 radial-spray bars (two sets.of 12) located and 13 inches upstream of thepilot cone.

The function of the integrated electronic control system is to cause the engine to operate at an optimum point determined from performance, operational, and safety considerations at asy given operating condition. - A compreheneive study of the control is given in reference 5. A block diagram from this reference is given in figure 5 to show therelation of the control components. Detailed - functions, some of which are discussed more fully in appendix A, my be listed as follows:

1. Maintain a-set enginespeed irrespective of changes in alti- tude or flight Mach nmber and maintain rated engine speed within close limits during steady-state operation

2. Prevent serious overshoot above rated enginespeed as a result of tranaients

3. Prevent serious loss of engine speed when ignitionoccurs in afterburner

4. Maintain rated turbine-outlettemperatee (12750 F) at the rated thrust position, irrespective of flight conditions

5. Preventexcessive tuibine-outlet temperatures during after- burning operation when the exhaustnozzle is fully open . or if the exhaust nozzle locks in a partially open position

6. Provide a minimum idle speed at a value above the blow-out speed and at a value from which satisfactory accelerations can be made

7. Permit the most rapid accelerations lpssible uithout combustor blow-out, compressor stall, or exceedingtemperatures above the transient Ustit

8. Give near optimum spcific fuel consumptions at thesteady- state conditions . 6 - NACA RM E5m8 9. Permit the maximum rateof deceleration possible without com- bustor bldw-out

10. Provide ti schedule of me1 flows awing the starting cycle to avoid overtemperature

11. Schedule the necessary services, such coollng as air to the exhaust nozzle, before afterburning begins

12. Fail safelyso that failure of anypart of the controlsystem will permit engine controlto .continue uninterrupted under the remainlng componentsof the:main system or emergency 6ySteEl

mine speed, exhaust-nozzle area, and afterburner flow fuel during steady-state operation are scheduledas functio& of thrust- selector position,as shown in.figure 6. Exhaust-nozzle area (in the- nonafterburning regiononly) and afterburnerfuel flqw are scheduled by potentiometers linked directlyto the tlgu6t selector lever.In the af'terburning region, the exhaust nozzle is adjusted by the control to maintain limiting turbine-outlet temperature. Although the area is affected by f1.igh.t condition,a typical curve of exhaust-nozzle area is given .by the broken line.Engine speed is governed by changes in fuel flow. The relation between engine speed and exhaust-nozzle area during steady-state operation is given .in7., figureIt will be noted thata large reduction in exhaust-nozzlearea occurs at rated engine speedas the thrust selector changes-fYom 70° to 90°. "his large changein exhauet-nozzle area was usea to permit a relatively large thrust changeto be made very rapidly without the necessity of changing engine speed. During large-accelerations, the exhaust nozzle remainsin the initial position-until the appmximatefinal speed or until a speed of 7800 rpm is reached, whereuponit returns to the steady-statesched- ule. A maxim dellimit, scheduled asa function of the compressor- outlet pressure,is imposed during accelerationsto prevent both com- pressor stall andcmbimtor blow-out. A transient temperaturelimit of about 150Q0 F at the turbineoutlet is also imposed3 excessive temperatures will cause the fuel valveto close.

INSTALLATION

The enginewas munted on a wing spaning the test sectionof the altitude wind tunnel. Dry refrigerated airwas supplied to the engine from the tunnel make-up .air systema duct. through connecte-d to the .NACA RJH E51E08 7

v engine inlet (fig.1). Manually controlled butterflyvalves in thLs duct meused to adjust total sir pressures at the engine inlet: A slip joint witha frictionless sealwas used in the duct, thereby making possible-the measurementof thrust and installation drag with the tunnel scales and with strainon gagesthe engine supports.

Iqstrumentation for measuring pressures and tempratureswas installed at various stationsin the engine{fig. 2) to determine the steady-state performance andto calibrate the transientperformance data. Lnstmentation (dynamic responseand design featuresof the .transient instmentation are aven in reference 4) for measuring transient performancems also installed,as given in the following table: ......

Measured Transient instrumentation Steady-state quarltltg 1 (calibrated from steady-state instrumentation)instrumentation I Semior Recorder Dsnamic lag Inetrument (equivalent time constant) : (see)

@ne speea Tachometer genera- Multiple- 0.04 Tachometer generabr, $or, direct current channel direct- alternate current inking oscillo- gmph with associated amplifiers i Aneroid-type 0.02 Bourdon-type gage pressure sensor I til-pipe tem- unshieldea loop 0.25 at Six themcouples in :rake (skbion 6~ thernao couple sverage 6ea- parallel on self- level mass balanclng Bmvn flow potentiometer lgine mmt Strain gzqe on main ~rce(function engine support Jet thrust)

@.ne fuel-valve Wire-wound poten- mition (also tiometer !heat fuel-valve rsition) I bust-nozzle Wire-wound poten- xa Itiometer ua pressure ratio Anemid-type pressure senaor

...... -:N NACA KME5lEO8 . 9

The air flow through the make-up air duct was throttled from approximately sea-level pressure to a total pressure at the engine inlet corresponding to the desired flight Mach number at a given 0 altitude. A list of symbols is given in appendix B. Inasmuch as the J throttling valves-were manually controlled, it was impossible to maintain constant engine-inlet pressure during all transients, partic- ularly at high altitudes where the v8,l.v-e~were'almost closed; however, an attempt wasmade to maintain the desired preseure as nearly as possible. The static pressure in the tunnel test section was -main- tained to dorrespond ta the desired altitude. Because of the large tunnel volume, the tunnel test-section pressure did not vary appre- ciablyduring transients'. The engine-inlet-sirtemperatures were held at approximately NACA standard values corresponding to the simulated flight conditions, except for high altitudes and low fJ-ight Mach numbers. Bo inlet-air teqeratures below -20° F were obtained. In addition, three data points were obtained at an inlet-air temper- ature of 140° F.

Compressor stall and coabustur,blow-out during accelerations were investigated by imposing step-f&tio-n increases in fuel flow to the engine with thecontrol inoperative. Successively larger steps were used until eikher stall or blow-out occurred. In general, after a stall was obtained, the fuel flow was reduced, permitting the compres- sor to unstall. Stall data. were obtained in the range of corrected enginespeeds from approximately 4000 to 8000 rp. Combustor blow-out data during accelerations were obtained at corrected eane speeds from about 5000 to 7800 rpm. This phase of the investigation covered a range of altitudes _from.1O,,ooO to 55,000 feet -+d flight Mach num- bers from 0.19 to 0.89.

With the exception of a br-ief study to determine the fuel flows required for starting at an altitude of 4Uy~oO0feet, the remainderof . the investigation discussed herein was conducted-with the control in operation. The terms "throttle burst" and "throttle chop" are used to denote accelerations and decelerations, respectively, wherein the thrust-selector position was changed as rapidly as possible. In addition to these data, several runs were made wherein the throttle was first chopped and then a throttle burst was made while the enghe speed was decreasing rapidly.

. A large number of starts were attempted with the control operative over a wide range of windmilling speeds at altitudes from 25,000 to 45,000 feet using MILF-5624 fuels-having Reid vapor pressures 10 " NACA RM EsIEO8 of 1 and .7 pounds.These starting ,attempts-were yade by setting wind- milling speed at the desired value and then advancing the thrust selec- tor to the idle range (from loo to ZOO) . Whether the control w&s operative or inoperative during starts, if ignition was not obtained within 30 to 40 seconds, €he attempt W&S considered unsuccessful.

Afterburner ignition and blow-out limits were obtained during the normal course of the investigation of afterburner performance. AS the afterburner fuel flow was gradually Increased, the fuel flow at which autoignition occurred was noted and.simtlarly the lean blow-out data were obtained as the afterburner fuel flow was gradually decreased prior to shutdown...... - . ..

RESULTS AND DIWUBSIOB

Many of the results to be discussed are In the form of oscillograph traces. To familiarize the reader wlth the typical behavior of several important engine variables, an oscillograph .record is. given in figure 8 for a throttle-burst acceleration of the controlled engine from idle to f" dry thrust. Arrows are shorn with each trace to Zndicate the direction of increase of each variable. The fuel flow increased almost instantaneously to the maximum fuel limit imposed by the control at r point A and thereafter followed the nmxbum fiel-limit curye until, at point B, the flow was reduced by the control because of a progressive reduction in the transient tqerature limit with engine speed af'ter anengine speed of about 7200 rpm was reached. Thfe limit, generally set- at 13000 F, was set at 12500 F during the transient shown. Tur- bine-outlet temperature followed the trend of the fuel flow and can- pressor-outlet pressure was afYected by both f'uel flow and engine speed. Engine speed increased at almost uniform- rate. %til the fuel flow was reduced at point 3, after which the rate of acceleration W~S reduced. No appreciable speed overshootoccurred. The exhaust-nozzle area remained locked in the initial open position until a speed of about 7700 rpm was .reached. The nozzle -then closed to the final posi- tion in about 0.9 second. As mentioned previously, the ram pressure ratio could not be held constant during transients of this magnitude and rapidity. As will be shown, the effect of a -decrease in ram pres- sure ratio on the engine during an acceleratfaa is an increase in the acceleration time as compared to the time taken with a constant ram pressine ratio: In general, veiations in ram pressure ratio of the magnitude encountered in this inveatigatfon had little effect on the operatiow characteristics of the engine; however, this Variation was twn into account tn ~LLcorrelatians of the stall, upstall, and conibustor blow-out data. . NACA RM El308 11

Compressor Stall s Compressor stall. Fn turbojet engines is usually encountered when excessively fast accelerationsare attempted, asthough some engines encounter stall withinthe normal steady-state operating region.. Compressor stall in a turbojet engine is characterizedby a sudden reduction and severe fluctuationof pressure throughout the engine, a decrease of.air flow, and excessively high turbine-outlet tempera-. tures. Acceleration of a turbojet. engine requires that the enthalpy drop through the turbinebe increased to exceed the enthalpy rise through the com.pressor,'therebypm7iiding the excess ppwer required for acceleration. This is, of course, accomplishedby increasLng the fuel flow and consequently the turbine-inlet temperature. Inasmuch as the compressorpressure ratio is a function of the t-bine-inlet temperature, the campressor-outlet pressure-immediately increasesto some value abovethe steaajr-state value at the startof an acceleration . and, unlessstall occurs, remains higher throughoutmost of the tran- sient. The presa-we. ratio whichcan be tolerated by the compressor without flow breakdown (stall) is limitedthe effectsby of the increased adverse pressure gradienton boundary-layer flowand by the angle of attack of the b-s. As a result, the rateof acceleration of a turbojet engine islimited by thestall characteristics of the compresser. As explainedrunder PRO-, pmpessively larger .stepsfn fuel flow were .madewith the control inoperative untilstan wag encoun- tered. Two SUeh runs are shorn in figure 9. In the first run, (fig. 9 (a) ) the fuel flow was increased from 800 to 3900 pounds per hour, producing a smooth acceleration until fuelthe valve was retmc- ted to prevent overspeedof the uncontrolled engine. In the second ruzl {fig. 9(b) ) , a slightly larger fuel-valve step 800from to 4150 poutuls per how was used, resultingin compressor stall. The effects of stall may be clearly seen by comparing the traces of figure 9(a) with figure 9 (b) . The stall is evidenced by the sudden reduction (point A) foUowed,bya rapid fluctuation.in the compressor- - - " outlet pressure trace, a by large increasein the turbine-outlet temperature, and by a break in the slopeof the engine-speed trace. .. It wlll be noted that the engine speed continuedto increase during the period of stall; however, the accelerationcould not be completed . - because of thehigh turbine-outlet tempera" (2160° F]. The manner in which stall. affectsengine air flow, is indicated by the tracesof . .-..." ram pressure. -Inthe fi,rstrun, the increase in airflow associated with the engine acceleration decreased theram pressure unttlmanual adjustment by the tunnel operatortoad be made. Ip the secondrun, a sudden increasein ram pressure occurredat the stall..point, indi-. cating a marked.redqction in air flow. Analysis of other-traces not shown indicates that duringstalled operation the air flow gradually 12 - NACA RM E51M)B decreases despite the fact that engine speed slowly increases.A few runs were made witha chart speedten times as fastae the traces shown. These datashow that the frequencyof the compressor-outlet pressure pulsation8was from 37 to 55 cycles per second, however, the respnse of the instrumentswas not sufficiently fastto permit accurate determinationof.the amplitude. It should be noted that the compressor unstalled at pointB followihg a decrease in fuelflow and temperature. The compressor stall curvefor the J47D (W-3) engine is given in figurelO(a) for altitudes from10,000 to 35,000 feet at a flight Mach numberof' 0.19, for flight Mach numbers up0.89 toat 25,000 feet, and forhigh (140° F) inlet-air temperatures-atl5,OOO feet. Inasmuch as air-flow measurements were not available during transients, cor- rected engine speedis used as the abscissain figure 10 instead of corrected airflow. It should be noted that the compressorstall data reduce4x1 a single curvefor the entire range.offlight conditione investigated. A limited amountof data werealso obtained atan altitude of 45,000 feet but are not included in figure lofa) because data for uorrectionof t& variation in ram pressure ratio were unavailable. When consideredon ag uncorrected basis, however, these data for an altirtude of 45,000 feet correlatedto a single curve with the data at lower altitudes. The stall-limit curye appearsto be divided into two distinct segments,with a transition occurring at corrected engine speeds between5250 and 5500-rpm. ' Similar data for the RXl-1 engine (fig. 10(b) ) corroborate the trends shown in figure lO(a), however, the transition occurs at slightly higher cor- rected engine speeds. Although the compressor..designwas the same for the two engines, theremay have been slight differencesin the tip clearanceof the v&Tlous stages becauseof manufacturing toler- ances. Various types of compressor stall are discussed in reference5 and explainedin terms of the blade velocity triangles. .According to this reference, the stall at speeds 5500 above rpm is probably due to excessive angle of attack (positive stall.) of the latter compressor stage whereasst lower speeds itis due to positive stallof the early stages. I

Compressor Unstall

The measures necessary unstall to a.compressor areof interest and accordingly the unstalldata have been correlatedon the same coordinates as the stall data.It might be expectedi;ht a slight reduction in compressor pressure beUwratio the..stalllimit would cause the compressorto unstalli however, from figure9 it may be seen thatduring the stall the pressure ratio.dropped far below the 13

stall limitand the compressor did not unstall.A furthur reduction in pressure ratio obtaineda reduction by in fuel wasflow required. Unstall characteristics Of the RXl-3 engine are given in figure11 to show the effects of altitudea flight at Mach number of 0.19 and engine-inlet tempera;ture. 15,OOOat feet.. .As the altitude increased from 10,000 to 25,000 feet, the canpressor unstalled at slightly higher pressure ratios. Although the amountof bats at 15jOOO feet ._ with engine-inlet temperatureof 140° F is Umited, two of the three data points indicate110 effect of engine-inlet temperatureon the unstall characteristic of the engine. Similar datain figure 12 sbw that chnging the fli-ghtMach number from0.19 to 0.89 at an altitude of 25,000 feet hizd-no eflect on the unstall characteristics. At high corrected engine speeds, the compressor unstalledlower at pressure ratios than were encounteredat stall or during steady-state operation. This is shownin figure 13, where both the stalland unstalldata are represented by single curves and corngareil with the-steady-state region of operation. The-shaded area represents the maximum region of steady- state operation possible by variations in exhaust-nozzleor area flight condition. Inasmu&. as theunstall melie.6 below this region over most of the practical ofrange engine speeds, it 1.s obvious that the compressor.cdt in *generabe unstalledby -openin@; the exhaust nozzle andthat a reduction of! fuel flowis usually required. The distance between the operating line {which lies somewhere within theshaded region) and the stall-limit curveis indicative of the margin-of excesspower available for acceleration; Itwill be noted that this margin increases abruptly as the corrected engine speed is raised above5250 rpu. The pathof EL typical throttle-burst acceleration with the control operativeto restrict operationto the region below the stall limit is bydenoted a' broken line. The rapid .. decrease in compressor pressure.ratio 8000at rpm is due toa reduc- tion in fuel flow calledfor by the controlto prevent exceeding the transient temperature limit.

Combustor Blow-Out During Acceleration In turbojet-engine operation, combustor blow-out during tran- sients, like compressorstall, is usually encountered during rapid acceleration. It is likely that the blow-out during acceleration is caused by excessively richregions Fn the primary zone, resulting from the sudden.-increasein-fhel flow and a reduction in a-lr flow. associated with the-negative slopecompressor of characteristic curves at high pressure ratios. For the engines investigated andfor . a flight Mach nuuiber of 0.19, stall was prevalent at altitudes below 35,000 feet and blow-out was prevalent at higher altitudesi however, a change in compressor design, combustor design, or flight Mach nuniber * would probablyshift this transition altitude.A few stalls (not shown) were encountered at.45,000 feet, and occasionally stalls at 35,000 feet were immediately followedby blow-out. - Oscillograph tracesof tm runs are given Fn figure 14 to permit comparison af a successM acceleration and an attdpted acceleration using a slightly larger step increasein fuel flow which resulted in combustor blow-out.The blow-out point ie obviousOIL both. turbine- outlet temperature and compressor-outlet pressure traces (fig.). 14(b) During the firstng (fig. 14(a)), a turbine-outlet temperature of 1300° F was obtained, however, during the secondrun, (fig. 14(b)) blow-out occurred atB2Oo F. The fuel flow -8 reduced manually shortly after the blow-outwas obtained. Engine speed and turbine- outlet temperature decreased very rapidly and went below of the limit pen travel on the recorder. Because of the reductionin turbine- inlet temperature. andthe flow conditione in the t'urbine nozzledia- phragm, the compressor-outlet pressure decreased markedly atblow- the out point and thereafter decreasedslowly as the engine speed decreased. Because of.differences in instrument response time, the compressor-outlet pressure trace indicated the blow-out approximately 0.1 second befare the turbine-outlet te%perature trace. On several occasions, combustor blow-out appeared to be incom- plete. Oscillograph traces of one-suchcase of partial blow-out are k given in figure 15. FoU-owing the parthl blow-out, the average turbine-outlet temperaturedecreased rapidly Prom 1254O E' to about 830° F and then increased about30' F in about 6.5 seconds during which time thef"X flow was consta+t. &pp~ximately1/2 second after the fuel flow was reduced andabout 9 seconds after thepartial blow-out, full combustion was restored and the average turbine-outlet temperature increasedto U6O0 F for about 4 seconds after which it decreased towarda' value commensurate witht& fuel-air ratio. It wiU be noted that complete combustion resumed at approximately the same f'uel flow as that required for steady-state- operation. Motion- picture recordsof individual thermocouples locatedjut downstream of the turbine behind each combustor revealed that during the partial I ' high blow-out, temperatures behind three com~ustors became very ." whereas the temperatures behind the rema-iive were between . . .. 500' and 600' F. The phenomenon occurringin five of the combustors providing a low temperature rise is believedto be dueto cessation of flame propagation away from the recirculating.In regionthe.campus: tor primary zone, which is dueto an excessive fuel-air ratio, and the resumptionof flame propagation following the restomtionof a *. more favorable fuel-air ratio. . The inordinate temperature value following complete cornbustionis probably the result of f'uel accumulated in the combustors during the period of partial blow-out. It should be noted that complete combustion was restored at an engine speed of about 5800 rpm in contrast to a maximum windmilling speed for ignition (to be discussed later) of about 2000 qm. From these observations it may be concluded that some cases of apparent combustor. blow-out during rapid accelerations &e .not actually a complete failure of combustion but .Instead the occurrence of a conibwtion pro- cess in which only part of the fuel 58 burned. In this event, complete combustion may be restored by a gradual throttle retraction which will reduce the fuel-air ratio to more favorable values.

Another interesting point is shown by the data of figure 15. Immediately following the fuel step, 8 quenching action occurred .. which reduced the turbine-outlet temperature as much as noo F. This \ quenching action occurred over a period of mre than 1 secona and was of sufficient magnitude to cause a reduction of 0.9 inch of mercury in compressor-outletpressure and 50 rpn in engine speed. Inasmuch as this effect becomes mre pronounced as altitude is increased and .. . engine.speed decreased, the flame may well be quenched completely at some flight conditions following a rapid increase in fuel flow.

Maximum values of fuel flow which may be used without causing combustor blow-out are defined by the data of figure 16 as a function of corrected initial engine speed for both original and modified conibustor configurations and a flight Machnumber of 0.19 at an alti- tude of 45,000 feet. Steady-state fuel-flow requirements and the margin of fuel flow available for acceleration are indicated. It will be noted that the margin for acceleration was not greatly dif- ferent for the two configurations and varied .from about 100 percent of the steady-state requirement at a corrected engine speed of 5500 rpm to 145 percent at -66900 rM. At 8200 xp~, the GFgi~Yiras69 percent for the original combustors and 65 percent for the modified conibustors.

Stall and Blow-Out Protection

It has been shown here ind elsewbere (reference 5) that compressor stall is a function of compessor pressure ratio and either engine speed or air DW. ~SQ,to a close approxiiatton, air flow 1s directly proportional to the compre.ssor-outlet pressure and inversely proportional to the tqbine-Wet taperature. By relating turbine- inlet.tem=rature to fuel flow, air flow, and compressor-outlet temperature and in turn relating compressor-outlet temp=rature to ressure ratioc. and inlet-air temperature, relations may be derived P appendix C) which my be used to provide protection against both compressor stall and combustor blowout using only two measurements, compressor-outlet pressure.ad engine-Met-air temperature. The man- ufacturer designed the control accorcllngly-to -prevent.stall or blow- out by limiting the maximum fuelfor any'givenflow value of compressor-out1et.pressure corrected for.inlet-air tempe-rature.The data obtained 1400at F inlet-air temperature (fig.10 1 indicate, however, that the correction for inlet-air-temperature variations can be neglected. . - . - ......

Because of the uncertaintyof the assumptions made in the deri- vations, it was ne-cessaryto establish experimentally the sbsolute values of themaximum fuel-limit. curve.Ws was done by plottfng fuel flow against compressor-outlet pressure for all stall and blow-out points for wh5ch data were available.. data These are given in figure17 for both engfnes and Combustorboth configurations. The highest permissible setting of maximum the fuel-limit curve alsois shown as well a steady-stateas operating line andpath the of a typical throttle-burst acceleration.Although the stalland blow-out data correlate reasonably well, maximum the permissible fuel flow appears to increase with. flight Mach Thenumber.. data are notsuf- ficient to establish this trend definit~&yj'ho%ever, inbst-of the data indlcate that atan altitude of25,000 feet aqda flight Mach number of 0.75 the margin available for acceleration-could be increased perhaps 35 percent over that obtainedusing -the-settin@; of the maximum fuel limit requiredto prevent -stall aat" Maoh number of0.19. Much of the scattershown is probably dueto variatiane in component efficiencies which were assumed constantFn the derivations. Although meager, the data obtainedan atinlet-air temperatureof laoF at 15,000 feet indi-ate110 apparent trend with inlet.temperatures.Also, the data show no difference between combustor configuratfms (denoted by differentsymbols) or between the two engines.

By making adjustments to the control,was it possible to shift the positionof the maximum fuel-limit curve,and considerable effort was devoted to obtaining the optimum position. The positionof the curve for the accelerationshown is givenby the lineBCD, the accel- eration starting at pointA. The-fuelflow_increased immediately from pointA to the limit curve, followed limitthe curve BCD as the compressor-outlet pressure increased -newlth speed, and then was reduced suddenly Dat because the turbine-outlet temperature reached the transient limit value. Excessively rapid actfonof the fuel valve in response to the overtemperature signal causedfuel the flow to undershoot the steady-state value in the E regionbefore equilibrium runnin@; was reached, F.at

Combustor blow-out data comnlyare correlated in terms of the pressure, temperature,and velocity at thecodustor inlet and the fuel-air ratio. Because of the relationsascussed prevlously, 3N NACA RM El308 " -IC 17

" combustor blow-out data may be correlated in terms of compressor pres- sure ratio and.corrected engine speed. aSuch correlation is.shown in figure 18 for both original and modified combustor configurations. Compressor. stall data, denotedby the solid symbols, are superimposed, and it will be noted.that thestaU and blow-out datafall around a 0 single curve. An alternative method of providing protection against 1yI d stall an& blow-out is thus afforded by the use of this correlation. A comparison of the stall -agd.blo-w-out protection correlationsas established by the manufacturer's control {fig.171 and the alternative method of protection derived from the analysisof the data {fig.18) indicates thata better correlationis obtained with the alternative method. Consequently, with the alternative method insteadof the manufacturer' s method, mre rapid accelerationof the engine can be made overa wfder range of operating conditions. However, with the alternative methodof protection, four measurements (compressor-inlet . .. pressure, compressor-.outlet pressure, engine speed, and inlet-air temperature) are requiredas compared with one (compressor-outlet pressure} used by the control.

Acceleration Characteristics

After thestall and blow-out regionshad been determined, the maximum fuel-limit curvewas aajusted to sart the lower side of these regions,thereby permitting the safe margin for acceleration. The acceleration and deceleration characteristics ofRXL-3 the engine were then evaluated overa wide range of flight conditions and engine speeds. Results of tws phase of the investigation are summarizedin figures 19 to 23. An oscilLograph record aof typical throttle-burst acceleration from idle speedto full dqy tkmt is givenin figure 19. Fuel flow increased to the maxirmnn fuel limit in about 0.1 second and thereaf'ter followed themaximui fuel limitas detemined from compressor-outlet pressure {fig. 17) until the accelerationwas almost completed. After an initial reduction todue the quenching effect, turbine-outlet temperature reflected the changesfuel in flow and inereased rapidly for about1 secohd merwhfch it increasedgradually to the final value of12800 F. Compressor-outlet pressure decreased slightlyand then increased aat mre or lessuniform rate for about 9 seconds after which the rate of accelerationwas reduced. Engine speed decreased slightly dueto the quenching effect and thereafter increased smothly until rated engine speedwas reached 16.5 seconds after the thrust selectorwas moved. The exhaust-nozzle area remained open until an engine speedabout of 7800 -rpm was reacheit. It then moved in about 0.8 second to the area requlred for limiting turbine-outlet temperature. It should be noted that the turbine-outlettemp=- ture was not drastically affectedby the lazge area change which occurred at approximately pnstantfuel flow. Following a small initial reduction, the tbrust increased graduallyWtll the exhaust nozzle closed. The decrease-in nozzle area resultedin a rapid thrust increase. The final-engine speedand thrust were attained approxi- mately 16 and 21 seconds, respectively, after the thrust selector was moved.

Inasmuch as thrust is the variableprime of importance during accelerations, the succeeding figureswill be discussed in termsof thrust acceleration time, whichis defined as the time requiredto change the engine mount force (a functionjet' thrust) of from the initial ta the final vklue where the final value is defined as the pink where thrust becomes approximately constant or where it begins to vary about a mean line. The effect of altitudean the time required to change both thrust and engine speed isin shownfigure 20 for throttle burstsfrom idle speed tofull unaugmented thrust for altitudes from 15,000 to 45,OOO feet 8% a flight Mach luzniber of 0.19. Both engine speed(fig. ZO(8)) and enginemount force (fig. 2O(b]) are expressed as percentof the rated values at the flight condition under consideration. It should be noted that the control caused the idleenghe speed to increasewlth altitude. The time required for increasing both thrustand engine sFed from idle to rated condi- tions increased w5th altitude. The thrust acceleration time (fig. ZO(b)') required varied from 14 seconds at an altitude of Y 15,000 feet to22 semnds at 45,ozx) feet, a ratlo of 1.57. If these accelerations had been started at the same engine speed or percent of rated-thrust, the ratio of acceleration time would have been much larger. It will be noted thata very slight speed overshoot occurred at 15,000 and 25,000 feet, resulting inan immediate reductionin fuel flow and thrust. Both the speed -overshoot and thrust reduction were of very short duration. A trend, which becomes more apparent with increasing altitude, is exhibited by the engine-mount-force curve45,000 for feet about 9 seconds after the start of the acceleration.The decrease in mount force shown was due to a change in ram pressure at the offace the engine resultingfrom increased sensitivityof the tunnel make-up air throttle valve highat altitudes where the valveis almost closed. If it had been possibleto maintain constantram pressure ratio, the engine mount force would probably have increasedsmothly along the . - ". broken line. . . . ." ..

Because of thehigh energy of the air at the -nei4et due to ram, and the corresponding variationFn idle speed it would be expected that acceleration characteristics would be improved with NACA RM E51E08 19

an increase in flight Mach nuniber. Thedata of figure 21 show tothis be true. At an altitude of25,000 feet, an increase in flight Mach nuniber from 0.19 to 0.75 decreased the thrust acceleration time from idle to full unaugmented thrustfrom 14.4 to 6 seconds. The thrust reductions occurring after3.5 seconds ata Mach number of0.75 and after 13.4 seconds ata Mach number of 0.19 are dueto reduction of fuel .flow made by the controla result as of overtemperature in the first caseand overspeed in the second. Throttle-burst accelerations from various initial thrusts to rated unaugmented thrustare compared in figures 22(a) and 22(b) for altitudes of15,000 and 45,oOO feet ata flight Mach numberof 0.19. For the throttle burst from10°’to 90° on the thrust selector at 15,000 feet (fig. 22(a)), acceleration was extremely slow below 30 percent of rated thrustand required 14 seconds. An acceleration fmm 42 percent to ratedthrust required only 3 seconds, and only 1.8 seconds was required to -e thrust from 69 percent to rated thrust. At 4S,ooO feet (fig. 22(b)), the time requiredmried directly with the increaee in tmt. The very rapid thrust changes associatedwith closure of the exhaust nozzle are apparent, and also it will be noted thatthrust overshoots occurwhich are .as much as15 percent of rated thrust. ..

The effectof exhaust-nozzle are8on acceleration time is given in figure23 for altitudesof 15,000 and 425,000 feet ata flfght Mach number of 0.19. For the accelerations at constant area, the exhaust nozzlewas locked ata position giving limiting turbine-outlet temperature at rate& engine speed, thereby simulating the performance of an engine equipped witha fixed-area exhaust nozzle.In view of the increasein turbine back pressure with the smaller fixed-area nozzle, and the corresponding increasein compressor pressure ratio and turbine work, would it be expected that acceleration charaater- istics would be penalized by reductionin exhaust-nozzle area inasmuch as the fuel margin available for acceleration is limited. This was found to be the case at 15,000-both and 45,000-fmt altitudes. At l5,OOO feet, thrust acceleration timewas 13.5 seconds for the variable-area nozzle as compared18 withseconds for the fixed-area nozzle, a decrdse of 25 percent. At 45,W feet, rated thrust vas obtained in 22 seconds with the variable-area nozzle as compared wLth35 seconds forthe ffxed-area nozzle. From a tactical point of view, these accelemtion times45,000 at feet are misleading because duringalmst the ent€re time required for acceleration with the variable-area nozzle the thrustwas higher with the fixed-area . nozzle. It should be noted, however, that at 15,000 feet the variable- area nozzlew&s superior with respectto both thrust leveland - acceleration time. 20 NACA RM E51Eo8

The effectof an improper settingof the maximum fuel Umit which will permit compressor stall isshown in figure24. Throttle-burst accelerations from idle to rated thrustan altitude at of 10,000 feet and a flight Mach number of.0.19 are compared. The solidUne denotes . the accelerationduring which first stalland then unstall oucurred owing to the excessivelyh,igh &mum fuel limit. Thebroken curve denotes an acceleration with themaximum fuel limitas high a8 possible without stall. For the first5.4 seconds upto the stall point the higher maximum fuel limit resultedin higher values of both thrust and engine speed.After the stall occurred, the thrust fluctuated violently but the general thrust level remained almost constant until unstall occurred14.9 seconds after the acceleration started. The amplitude of thrust fluctuations isof COUTB~ attenuated. Upon unstall, the engine thrust increased rapidlyi however, stable opera- tion had not been obtainedthe endat of an 18-seeond period. In contrast, the proper setting of maximum the fuel limit permitted the successful accelexatian to be completed14.4 in seconds. It vlll be noted thatduring the stall, the rate.of engine acuelerationwas reduced; however, the speeddld increase gradually. Also, the com- pressor press- ratio (not shown) was inFtiaUy reduced whenstall was encountered and thereafter increased slowlyas engine speed increased. As a result, the operating point droppedf'rom the stall limit (fig. 131 to a-point above the unstall line arid to moved the right and up until unsU the limit was .reached, whereupon the com- pressor unstalled. In many of the throttle-burst accelerations, the compressor did not unstall and ratedcould speed not be reached with- out exceeding turbine temperature limitation..

Deceleration Charaoterietlae

A series of throttle chopsfrom rated thrustto idlethrust were made at .altitudesfrom l5,OOO to 45,000 feet and a flight Mach number of-0.19 aha flight Mach numbers upto 0.75 at 25,000 feet to determine the suitabilityof the minim fuel limitof 450 pounds per hour for preventing lean combustor blow-out. At no time during the entire investigation was lean combustor blow-out encountered.The effects of altitude and flight Mach numberon deceleration characteristics are shown in figures25 and 26, respectively. Neither engine speed nor thrust decreasedas rapidly atan altitude of 45,000 feet as at 15,000 feet because themi~Lmum fuel flowwas a larger percentage of the steady-statefuel flaw requirement 45,OOOat feet thanat l5,000 feet and because the densityof the working fluidwas reduced while the rotor inertia remained constant. Although the final thrust and speed levels are different, the of.effect flight Mach number on deceleration characteristics is slight, (fig.26). . ". . NACA RM E51Eo8 - 21 As a more severe test of both maxim and minim fuel limits, a series of runs was &de over a wide range of flight conditions in which throttle bursts to rated thrust-were made during rapid decelera- tions and throttle chops to idle were made during rapid acceleratiorl. With the final setting of the maxirmun and mFnimum fuel limits, no stall or blow-out was encountered aid no appreciable time delay occurred in changing from acceleration to deceleration or deceleration to acceleration.

AI€itude Starting Characteristics

The superior altitude starting characteristics of the modified . .. combustor configuration as compared with the original is shown in reference 6. Only the starting data obtained with the RXL-3 engine using the mod;Lfied combustor configuration are predented.

The maximum windmilling speed and the corresponding flight Mach number at which ignition and flame propitgation are possible using the fuel flow scheduledby the control are shown In figure 27(a) as a $unction of altitude. MIL-F-5624(AN-F-58) fuel with a Reid vapor pressure of 7 poynds per square inch was used. The fuel was at a temperature of about 700 F and the engipe-inlet-air temperature c varied from Oo to -6O F. Wne fuel flow varied from an average of $00 pounds per hour at art altitude of 50,000 feet to 650 pounds per hour at 25,OOCfeet. At an altitude of 50,OOO feet, ignitionoccmed in all conibustors at windmillin@; speeds from E500 to 1500 rpm. Propa- gation was poor at 1550 rpm> and no ignition was obtained at wind- milling speeds of 1800 rp or above. As the altitude was reduced ' the maxhum stast;ing speed increased to 2300 rpm st 38,000. feet, and . .. at 35,000 feet star%s could be made up to 3500 rpm, the nmaximum wind- milling speed obtainable in the tunnel.

The starting limits' obtained with MIL-F-5624(AH-F-58) fuel, which was treated to give a 1-pound Reid vapor pressure, are shown in figure 27(b 1. The inlet-air temperature varied from -20' to 30° F and the fuel temperature was about 90° F. At an altitude of 49,000 feet, starts were possible at Wind~CULngspeeds up to about 1500 rpm~however, at 25,000 feet starts were not possible above 2200 rpm. The reduction in vapor pressure from 7 pounds to 1pound appears to hiwe had little effect at 49, OOO or 50, OOO feet but at 25,000 and 35,000 feet the maximum- starting speed was reduced considerably. 22 RACA IIM E51M)8

It was observed, however, 'that at altitudes40,OOO above feet, once ignitionwas obtained itW&S not possibleto accelerate the engine above about3100 rpwithout increasing theram pressure ratio. This may be explained by the existencea combustor of dead band similar to. that shown in reference 7. This dead band resultsfrom the fact that the temperature rise requirede@ne by fsthe greater than the temperature rise obtainablein the combustors. As the dead band is approached, the rateof acceleration drops to zero and some of the combus'cors blow-out whereas others to begin emit flame through the turbine. An increase in ram pressure ratio decreasesthe tempera- ture rise required by the engine and permits the accelerationto continue.

Previous work (reference 6 1 has shown the importanceof fuel flow on starting characteristics. Accordingly, a large numberof starts were attempted withMIL-F-5624 (AN-F-58) fuel with a vapor pressure of 7 pounds. ata 40,000-foot altitude usi!x~.various @ue_s. of..fuel flow . . ._. which were setmanually. These data, shown in figure28, define a range of f'uel flows -in which starts were consistently obtained.Thfe range becomes narrower as the windmilling is speed increasedj the maximum speed atwhich ignition was possible ip all combustors was approximately 2410 rpm. The regionof certain ignitionis bounded by a region in which ignitionand flame propagationwas possible in some combustorsand another regionis shown in which no ignition was possible. These data indicatean optimum Fuel flow 40,000at feet of about 650 pounds per how as contrastedto a value of about r- 450 pounds perhour scheduled by the control.

Temperature historiesof six of the eight combustors,as obtained from motion picture recordsof individual thermocouples located just downstream of the turbine behind each combustor, are given 29in figure to ,show the time required for flame propagation-md also the rate of temperature risefor starts atan altitude of 35,000 feet and wind- milling speedsof X00 and 1200 rpm with MIL-F-5624 CAN-F-58) 7-pound vapor pressme fuel. . Thermocouples for.tug combustors were .. ... burned out when these data were obtained. Propagation in occurred less than5 seconds for theruns shown.

The time required at various altitudesto obtain ignition in one combustor (denotedby circles) and in all combustors (denoted by squares) is defined by theshaded areas shown in figure 30. The numbers adjacentto each data point referto windmilling speed. As the altitude is increased, the time requiredfor ignition in one combustor and the time requiredflame for propagation to the remaining combusturs also increased. There appearsto be no consistent effect of windmilling speedon the time required for either ignitionor

" flamepropagation, . .. .- > ...... " . -+ ..+ NACA RM E51M)8 - 23 afterburner Operational Characteristics

During most of the previous afterburner investigationsst the NACA Lewis laboratory, fixed coni&l exhaust nozzles were used. When exhaust-nozzle sizes were sufficientto permit larget-t augmen- tation ratios, the burner-inlet temperaturewas 0- about 800° F at the timeof ignition. With such low burner-inlet temperatures, after- burner ignitionwa8 accomplished by the ofuse the 'kt-shot" imtion system discussed in reference8. For the more recent investigations, variable-area nozzles were available, permitting burner-inlet temperatures 12Wo of to 1300° F at the time of ignition.On two occasions in the revious Investigations wlth burner-inlet temperaturesof 1200° to 1300g F, autoignition was obtained with MIL-F-5572 (AH-F-48) grade-80 fuel highat tail-pipe fuel-air ratius. Becauseof the high tail-pipe fuel-air ratio required for autoignition with this fuel, theshrts were violmtj one start caused the engine combustorsto blow-out and the other resulted in damage to the afterburner. Subsequent experience withMIL-F-5624 (AM-F-58) fuel, whichhas a slightly lower surfaceimtion tempera- ture than MD-F-5572 fuel, has shown that autoignitionmay be obtained at reasonablylow tail-pipe fuel-air ratioswithout excessive violence. s Starting the afterburner by autoignition has therefore become comm0l~- place, although slightly higher fuel-air ratios are required than are - needed using the hot-shot system. For this investigation, afterburnerstarts, with either auto- ignitiop or using the hot-shot ignition system, were obtainedat all flight conditions investigated, including analtitude53,000 of feet where the absolute pressurein the afierburnerwas 388 pounds per square foot. -

The bazd of tail-pipe fuel-air ratiosin which autoignition occurred while thefuel. flow was gradually increesedis shown fn figure 31 as a f'unction of altitizde. The tail-pipe f'uel-air ratiois defined as the ratio of the afterburner fuelto the flows unburned air entering the afterburner. Differentsyaibols are used to denote various fuel-injection-systemand flame-holder configurations. A study of the symbols shows that the fuel-air ratio requiredfor auto- ignition wi-t;ha given configuration and altitude does not reproduce exactly. This lack of reproducibFlity is attributedto variations of as much as 50° F in the burner-inlet temperature.A comparison of these data. with preliminary ml.,culationsof tail-pipe combustion efficiency indicate that,in general, autoignition occurs at leaner mixtures for configurations havingthe best steady-state performance. 24 NACA RM E51M)8

Data are presented in figure.3.2 for.one-_typical..configuration to ...... show the effectof burner-inlet temperatureon autoignition character- istics usingMIEF-5624 (AN-F-581, 7-pound vapqr-pressure fuel. Atan altitude of25,000 feet and a .flight Mach numberaf 0.19, the effect.. of temperature was negligible over the range investigated. At 35,000 feet, however, the tail-pipe fuel-air ratio required-forauto- ignition. -creasedfrom about 0.008 at 1760' R to 0.038 at 1660O R, or increase of almost5 times f0r.a100° F reduction in burner-inlet temperature.

Oscillograph traces are gl-venin figures 33 to 35-to show a throttle burst fromf'ull dry thrust tofull afterburning, full d2y thrust to partial afterburning,and a thgottle chop from f'ull after- burning to full drfthrust at an altitude25,000 of feet anda flight Mach numberof 0.19. In figure 33, which shows a throttle buretfrom full dry thrust to full afterburning, a period of about-9 seconds elapsed before autoignition occurred, as noteda comparison from of the tracesof afterburner fuel Plow and engine munt force. When ignition occurred in the afterburner, the increaseIn turbine back . pressure (not shown) caused- the engine speedto decrease about160 rpn even though the enginefuel flow was increasing to mintain constant speed. AB a result of the increase in fuelflow and the reductionin engine speed, the- tgrbine-outlet temperature became excessive, reach- ing a value of1690° F before the overtemprat&e, underspeedcondi- - tion was alleviated by the opening of the exhaust nozzle. The exhaust nozzle required1.6 seconds to open. A reduction in the opening timeof the exhaust nozzleand a; reduction in themcouple' response time would reduce the amountof temperature oversh~ot., After th-exhaustnozzle had openedfully, the turbine-outlet temperature was still excessive. Accordingly, the afterburnerfuel flow was reduced by the control in responseto the turbine-outlet temperature signal to prevent damage to the turbine. At the time werethese data obtakned the responseof the afterburner fuel valveto overtemperature conditions was excessively ragid, resulting in the afterburner fuel original The flow being reduced almoet to ane-third.af the value. .. reduction in afterburner fuel flax caused.the turbine-outlet temperature todrop below the limitingvalue, thereby permitting the afterburner fhel valveta open. The af'tei-burner fuelagain increased to an excessive value causlng wertemperature,and a8 8, m6Id-t the cycle of events repeated with little attenuation.

A throttle burstfrom full dry thrust to partial afterburning shown in figure 34 was aceompgied .by a few cyclea whichin turbine temperature became .excessive and the enginewas reduced speed a8 a result of the simultaneous action of the exhaust nozzle; whlch tries to maintain constantturbb-outlet temperature, and engine blflow, 4N WCA RM E51E08 25

which attempts to maintain constant engine speed. Because the turbine- outlet temperature was controlledby the exhaust nozzle (whichwas not open] rather than the afterburner fuel flow, the instability 8 exbibited by the dataof figure 33. was not encouptered. Equilibrium td running conditions were restored pintat A, 12.5 secopds after igpi- tion occurred. A throttle chop from fullthe afterburning conditionto the full dry thrust condition isshown in figure 35. The time required for .. . most of the variablesto return to equilibrium was about equalto the time required for the exhaust tonozzle close, approximately 8.5 seconds; men the afterburner fuelflow was reduced, the turbine- .. . . outlet temperature-droppedfrom 13200 to loO@ 3'. Simultaneously the engine speed increased about100 rpn. -ne fuel flowwas modi- fied by the controlto restore speed and at s- the time, the exhaust- nozzle areawas reduced to restore the turbine-outlet temperature.

The relatively slow closureof the exhaust nozzle is probably the result of continued burningof a small amowt of fuel in the afterburner, inasmuch as the flowfuel to the Edgterburner did not stop comgletely for about7 seconds. During most of the transient, the thrustwas less than the rated. value. Equilibrium turbine-outlet temperature, indicatedon the oscillograph trace after the transient, was 30° F lower than the original valueWl atafterburning, probably as a result of slight discrepancies between the indicatingthem- couples and those usedby the control.

For operation aat flight Mach number of 0.19, afterburner auto- ignition delay, measuredTram oscillograph traces, varied with alti- tude as follows :

Lean blow-out limitsfor several afterburner configurations, comprised ofchanges in fuel distributionand flame hslders, are shown as a function of altitude in figure 36 for a flight Mach nmiber of 0.19. The configuration changeswhich were made had no significant effect on the results) however, shouldit be pointed out that the .flame-holder block& ar" was not altered. Theminimum tail-pipe fuel-air ratio for lw...blow-out increased from 0.004 at an altitude of 15,000 feet to &aut Q,Q13.at -50,000 .feet.. -%e. adth of the blow- ., out region was not markedlychanged with altitude. Rich blo2r-out limits were not obtalnedbecause at low altitudes the exhaust-nozzle size limited the maximum fuel-air ratio and at high altitudes opera- tion was not attempted beyond the fbel-air ratio Tpriducin@;the paximum exhaust-gastemperature .- - j. .. .

From an investigation of J47D (FKL-1)and (RXl-3) turbojet engines (with in'tegrat.ed electronic controls) in the PLACA Lewis altitude wind tunnel over a range of altitudes up to-55,TXX, feet at a flight Machnumber of 0.19 and flight Ma& numbersup to 0.89 at an altitude of 25,000 feet, the fallowing results were obtained.:." .

1. For the complete range of sltitu&s and f'lLght Mach numbers investigated, compressor data. seduced. cuPves for both stall @>-we" . ." -_ engines on plots of' csmpressor pressure ratio againat corrected engine" speed. The turves for bothengines were eimflar and each shared two distinct segments indicating stall in different portions of the com- pressor. The trangition, however, accurred at. slightly different en gine speeds engine for "the two engines. " .- . . 2. On the same coordinates atgiven a corrected engine speed, - the compressor unstalled at slightly higher pressure ratios as the altitude was increased. Flight Mac& number had a0 apparent effect on campressor unstd characteristics within the range investigated.

3. The unstall compressor pressure ratio occurred at a lower value than either th stall or steady-state-operation compressor pressure ratios for given corrected engine-. speeds above 5800 rpm. .. 4. A maximum fuel llmit scheduled as a Rrnction of compressor- outlet pressure prodded adequate protection agaiet both compressor stall and combustor blow-out and required the neisuremen-b of only one variable; however, the limit appearsconseyyative for operation at high flight Mach numbers.

5. Both combustor blow-out andcompressor stall data reduced to a single curve on coordinates of compressor pressure ratio against corrected engine speed, thereby proriding a relation which could be used for protectinn_against these dlfficuties. 6. The time required to accelerate the controlled enginefrom idle to rated thrust increased from 14about seconds at15,000 feet to 22 seconds at45,000 feet for operation a atflight Mach number of 0.19. At an altitude of25,000 feet, an increase in flight Mach co0 number from0.19 to 0.75 reduced the acceleration time 14.4from to rj 6 -seconds. 7. For the complete rangeof flight conditions investigated, lean combustor blow-out could not be obtaineda constant using minimum fuel limit 450of pounds per hour.

8. Using MIL-F-5624 (Am-F-58)fuel with a 7-pound Reid vapor pressure ata temperature of about 70° F and inlet-air temperatures from 0: to -60 F, ignition duringautowtic starts was possible inall . combustors at windmflling speedsfmm 1300 to 1500 rpm at an altitude of 50,000 feet. At these conditions ignition was possible in some . combustors up to2100 rpm. At 35,000 feet, ignitionwas possible in all combustors upto 3500 rp, the highest uindmilling speed obtain- able. At altitudes above40,OOO feet, however, the presenceof a combustor dead band prevented acceleration3100 above rpm at the IQwer.

flight Mach numbers. Mach flight "

9.. Using MIL-F-5624 (AK-F-58) fuel (treated to givea 1-pound vapor pressure) ata temperature of approximately90° F and inlet-air temperatures from-ZOO to 380 F, ignition was possible inall cam- bustors at49,000 feet upto a windmilling speedof about 1500 rp. At 25,000 feet, however,starts were not possible above2200 rpm. 10. At an altitude of 40,OOO feet, the optimum fuelflow for starting appearedto be about650 pounds per hour for MIL-F-5624 . (Am-F-58) fuel.

11. Afterburner starts by autoignitionusing MIL-F-5624 (AH-F-58) fuel were obtained at altitudes 53,000 up tofeet ata flight Wch number of 0.19. The tail-pipe fuel-air ratio requiredautoimtion for increased with altitude and35,000 at feet decreased as the burner- Wet temperature was raised. .

12. The tail-pipe fuel-air ratioach at lean blow-out. of the afterburner occurredwas increased as altitude was raised. The width of the blow-out band remained constant over ofthe altitudes. range

Lewts Flight Bopulsion Laboratory, Hational Advisory Committee- for Aeronautics, Cleveland, Ohio. 28 - NACA RM E5lEO8

DESCRIPTION AFTD OmTION OF INTEcraATED ELECTROlPIC CONTROL Principles of Operation

At all operating coaditions, the engine is controlled by mauls-' tion of fuel flow. As notedpreviously, steady-state exhaust-nozzle area is scheduled as a f'uncticm of thrust-selector.posit1o.u (fige. 5 and 6) for operation in-the nonaf'terbming region. Under after- burning conditions, afterburner fuel flow (corrected for altitude and ram) is scheduled against thrust selector position, and limitkg turbine-outlet temperature. is maintained by modulation of the axkaust-nozzle area. The detailedflmctions of thecontrol were listed previously; the merin which the cantrol was designed to accomplish the more important objectives will be discussed in the f ollowlng parwapha: Ehgine speed control. - mine speed, scheduled as .a function of thrust-selector position, is controlled by suitable modulation of engine fuel flow in response tu a speed-error signal. The speed- error signal is the difference between two voltageej one voltage from the speed selector (fig. 5) is proportional tg the desired speed, whereas the other voltage is the output of B, tachometer unit (fig. 5) driven by tlie engine and consequently represents the actual speed. This speed-error signal is amplified by the sem amplifier unit (fig. 5) and used to- drive a motor-actuated fuel valve located in the main fuel control. If the .desired engine epead called for by the thrust selector is higher than the actual engine speed sensed by the tachometer, a positivespeed-error signal results. In re.sponse to a positive error si-, the valve opens, .increasingthe .fuel flow and producing an acceleration until the speed-error signal is reduced to zero when thedesired speed is reached. Similarly, a negative speed-error signal causes the valve to close,reducing the engine speed.

An important requirement'of a control is that it maintain constant enginespeed under changing flight conditione. This is accomplished by the same system of error signals just discussed. An increase in flight Mach nmiber, for example, will cause the actual speed to increase above the speed set by the thrust selector, producing a negative speed-error ..signal which reduces .the fue1.f.U.w.and .restores or maintains the initial engine speed. .

Overspeed protection is afforded by a tuned circuit which pro- duces a negative speed-error Signal; the error-signal voltage increases very rapidly in the overspeed region, causing a rapid mACA RM 29

- reduction in fuel flow. This protective circuit attempts to hold rated engine speed withip20 rpm. Further protection is provided by an overspeed governorworking .on.the fly7ball principle. This gover- nor is set bypassto fuel backto the pump inlet 8050at rpm~however, full by-pass is not obtained until the engine areaches speed of. 3 8300 rpm. .. S Ebgine temperature control.- With the exhaust-nozzle area sched- - ule used (fig.6) steady-state turbine-outlet temperatures are well . .. below the limiting valueof 12750 F at all engine speeds except rated engine speed. At rateddry thrust (900 thrurrt selector position) even thougha given valueof exhaust-nozzle area is scheduledby the nozzle-area selector, theexhaust nozzle, actuatedby the nozzle actuator, is permitted todosed be only until limiting temperature is reaehed. Becauseof the effectsof Reynolds rider on component efficiencies, the,exhaust-nozzle area producing limiting turbine- ' outlet temperature 5ncreases with altitude.From figure 6 it may be seen,, therefore, that rated atthrust high altitudes will be obtained at thrust-selector positionsslightly below SOo and thata &ea& band on the thrust selectorwill result athigh altitudes. To keep t.hisdead Bad narrow, the scheduled area is changed rapidly in this region.

During eagine starts and accelerations-at speeds below7200 rpm, a turbine-outlet temperature limit(a8 measured by the thermocouple d unit (fig. 5) of about 15W0 F is imposed. If the temperature tends to exceed this limit, the control reduces engine flow.fuel A smooth transition of the temperature limit is proatledfrom the valueof " - 150O0 F at speeds upto 7200 rp to the valueof 1275O F in effect .- at rated speed. Under afterburning conditions, the exhaust-nozzle area is mdulatea to maintain-limiting- temperature. In the eventof nozzle failure,or if the nozzle is open,wide an overtemperatureof 20° F will cause the afterburner Rzelflow to be .reduced ato value consistent with the temperature limit. Acceleration stall and blow-out protection.- As noted previously under RESULTS AMD DISCUSSION, it is possihle to provide protection against compressor stalland- combust& blow-out by scheduling the maximum and m-Lnium fuel flows {corrected for temperature) a6 functions of compressor-outlet pressure. Experimentaldata given in figure 17 showed that the maximum fiel-limit curve shduid be comprisedtwo of straight-line se-ents. The relation between the fuel limitand steady-state operating lines for high both and low altitude8 is - given in the following sketch: - 30

Fuel-flow margin available for

operating line, low altitude

,Fuel-flow deficiency during deceleration

-. \ Compressor-outletpressure .. -

The marginof fuel flow in excess of the steady-state requirements . " .- available for accelerationis also indicated.by the vertical distance " between the steady-state operating axid line the'miximuii rUel limit. The optimum acceleration.characteristics.areof course obtainedwith

f'uel bord.er the. of Sta-Uand rich the maximum .limit set ta re&o:ri- ..~ .- blow-out. Both the.maximumFuel limit and the maxirmun temperature . limit discussed previously arein effect..sim.vlbeouslyand either " may impose a restriction on the rateof acceleration. At low engine speeds, the maximum fuei limitusuaily.encouiitered.first is whereas near rated engine speeds, maximum the temperature LLdt (which " decreases near rated isspeed) generally ""the .controlUng . . .. lmt. . " Fuel-flow deficiency during deceleration represents the differ-.. ence between thef'uel flow required. to maintain steady-stateConditions .. and thefuel flow YequiPed to prevent .lean-c6mbustor~biow~out. This .- deficiency is indicative of the forces'tmding.to reduce the.engine - NACA RM E5W8 31

speed. As noted in the text, 1- combustor blow-out was notobtained xith the minimum fuel limit set at the value of 450 pounds per hour. When accelerations of large magnitude are made and speed-error aignals equiva1ent”tq 400 rpm are present, the exhaust-nozzle area is locked at the initial area until the approximate,final speed is reached or until a speedof 7800 rp is reached. After release, the exhaust nozzle .closes to the steady-state exhaust-nozzle schedule.

Provision of services. - All services required in the various regions of operation are scheduled against tlq-ust-selector position or engine speed. The following are some examples-l When thethrust selector is advanced from Oo to the idle position of loo, fuel flow is provided, the engine starter is sctivated, .and ignition is supplied. At an enginespeed of about, 2ooo rpm the starter “cuts out”. Just above the 90° thrustcselector position, the afterburner fuel shut-off. valve,the valve supplying air to the afterburner-fuel pwp, and thevalve supplying mling air to the exhaust nozzle open. A relay controlling the afterburner fuel shut-off valve is closed at 7200 rpm, preventing afterburner operation at lower speeds during a burst from low speed into the afterburning region.

Stabilization. - A two-phase motor is used with a gear reduction to mve the engine fuel-flow control valve. This mtor is driven by the amplified error signal. A tachometerconnected to the motor suppliesa voltage proportional to motor speed. This voltage modulates the error signal, which is the input to the amplifier. When the error signal is small, the kchometer output tends to prevent small oscil- lationsj however, when the error signal is large, the tachometer output is overpowered and the-mtor is permitted to operate at full speed. ..

A ptentiometer i-s also connected to thegear reduction. The outputaf this potentiometer is used to oppose the error signal. When a speed. change -is called for by the thrust selector, the motor begins to mve the me1 Valve to a new position. At the beginning of the transient the err& si-1 is -large and the signal from the potenti- ometer has little effect; however, when the actual engine speed approaches the set speed the error signal becomes sufficiently small so that the signal fmm thepotentiometer is effective. TRis signgl tends to stop the mtor before the set speed is reached. The poten- tiometer feed-back signal is slowly reduced to zero in a short time to permit the set speed to be ohtafnedj however, the anticipatory action provides stability as the engine appmaches the set speed. 32

AppENDrx B SYMBOLS

The following synibols are usedin this report: effective area at turbine-nozzlediaphrw f’uel-air ratio

acceleration due to gravity,32.174 ft/secz

constants

M Mach number N engine- speed P total pressure, lb/sq ft absolute P static pressure,lb/sq ft absolute R gas constant, ft-lb/(lb)(OR] .

T total temperature,OR t static temperature,OR air flow, a/sec

gas flow, lb/sec

Wf fuel flow, lb/hr ‘1, adiabatic compressor efficiency

%I burner efficiency

qt adiabatic turbine efficiency r‘ ratioofspecific heats e ratio of absolute static temperature at engine toinlet absolute static temperatureat RAGA standard atnmOs- pheric sea-level conditions 5N EACA RM EU08 L 33

- Subscripts :

e engine

0 free air stream $2 d 1 engine inlet

3 compressor outlet

4 turbine inlet

6 turbine outlet 34 II NACA RM Es1EoB

A relation whichmay be used asa basis forstall and blow-out protection may be developedfrom a eonsideration of the flar conditions that exist at the turbine nozzle diaphragm. For operationlow at power levels, the turbine nozzle diaphragmis not choked, and gas flat is given by thefollowing equation:

For steady-state operating conditions,if the various specific heats are assumed constantand the Rrel flow is neglected, the turbine total pressure ratio Pq/P6 may be shown to be relatedto compressor total pressure ratio P3/P1 by the following expression:

As a first approximation, assume that totalthe temperatures and pressures in equation(2) are equalto the static temperatures and pressures, respectively,and that the turbine-Inlet static pressure p4 i8 equal to the compressor-outlet static pressurep3. A study of equations (1) and (2) reveals thatfor a given engineoperating condftion (Wg = K) and flxed flight conditions(pl and tl constant) the compressor-outlet pressure ais function of the turbine-inlet temperature.

For most of the normal engine operatingcondAtions, the turbine nozzle diaphragmis choked and thefollowing ie true:

During an acceleration, however, equation(3) is valid for lower engine speed than that encounteredin steady-state operation, because the turbine pressure ratfo.must be increasedto producean acceleration. . NACA RM lElXO8 II) 35

Turbine-inlet temperature is given by

t4 = A% + 5 where the combustor temperature riseAS is

I = %(fmeK1 and r

or if qc is assumed equalto 1 throughout the transient,

therefore,

*

From equation(4) it may be seen that theflow fuel arid the . compressor-outlet pressure (correctedfor inlet temperature)may be scheduled in such a manner that the compressor pressure willratio remain belowthe value producingstall. Stall data obtained at an engine-inlet temgemture corresponding to NACA standard condition and inlet tempera.ture of140° F (fig. 121 indicate that the effect - of engine-inlet temperaturemay be ignored. The impositionof a limit on fuel flow for a fixed compressor-outlet pressurealso limits the combustor fuel-air ratio@nd thereby tendsto prevent combustor blow- out. Because of the assumptionsmade and the approximations used, experimental dete-nation of the proper scheduleof fuel flow against - compressor-outletpressure was necessary. 36 " NACA RM SIEO8

1. Baari, M. J., and Wintler, J. T.: Altitude-Wind-Tumel Investi- gation of Performance Characteristicsof J47D (RX1-1) Turbojet mine with Fixed-Areamust Nqzzle. NACA RM E5lBO6,

3. Wells, B. A., FTillFams, J. F. , and Yates, C. G. : Study of Inte- grated Electric Control for Turbo-Jet Engines. Rep.TR-55410. . Aero. and Ord. Sptems Mvs., Gen. Eiec.' Co. (Schenectady, N.Y. ) 4. Delio, Gene J., and -went, Glennon V. : Instrumentation for Recording Transient Perfarmance for Gas-Turbine Engineand Control Systems. nACARM ElD27, 1951.

5. Constant, Hayne: Gas Turbines and Their Pmblems. Todd publishing Group, Ltd., (Iondon), 1949.

6. Golladay, Richard L. , and Bloomer, Harry E. : Investigation of Altitude Starting and Acceleration Characteristicsof 547 Turbojet -mine.. NACA RM E5oGo7, 1951. 7. Fleming, William A,: Altitude-Wind-Tunnel Investigationof West- inghouse 19B-2, 19B-8, and 1SxB-l Jet-Propuls-bn Engines. I - Operatinnal Charaoterfstics. NACA RM E8J28, 1948.

8. Conrad, E. William, and Prince, WilliamR.: Altitude Performance and Operational Characteristicsof 29-rnch-Diameter Tail-Plpe Burner wlth Several Fuel Systems 6nd Fleme Holder8 on 535 Turbo jetE!ngine. NACA RM E9G08, 1949. . - ......

2uo "

...... I ......

...... I 2180 , an I

w W

..

JN. .

0 CO

Figure 3. - Original and modified ccrmbustar liners. .. L,,

. ." .

. 43

...... , ...... - ...... 2180

Servo amplif ier unit presaure

I I I I I Thmocmple unit I

I I””“”

Bigwre 5. - Block diagram of integrated electronic control.

...... 46

3 Thrust-selector position, deg " " Figure 6. - Schedule of enginespeed, exhaust-nozzle area, and after- burner fuel flow with- thrust-select.or position......

I 8 am I

Figme 7. - Schedule of percentage of rated erhawt-nozzle area to engins speed.

......

...... ,. ' .I ,. ' .a , , .. r:'.;, ;:

B

-...... " ..

I...... -. .. I t

...

.. . - ......

I 4

I

CorreoCedengj (a) (RX1-3) engine. Figure 10. - Correlation of canprssaor atall with ccmpreaso~total-preesure ratio and cmrected engine weed far J47C engines. 54 I, NACA RM E5m

. .. .

..

..

Corrected e-e epea, IT/@, rpm Figure '11. - Effect .of altitude and engine-inlet texperature on cam- pressor unstall characteristics of J47D (RX1-3) engine. Flfght

1. .. . - 4 5- E 7 a 9x103 " Corrected eng1r.s spead, n/@, r;3m Figure 12. - Effect of flight Mac& nuxber on compresscu' unstall charac- teristics of 347D (RX1-3) engine. Altitude, 25,000 feet. .. . 1 ON NACA RM E51E08 55

7.0

I

6.0

5.0

4.0

3.0

2.0

1.0 9x10’3 56

. "

.. . 11N NACA RM FS1Eo8 - 57

Blgure 15. - Behavior of several engine varhblea aUrlng aoaeleratlon whloh meULt0d fiw Etcl, inorease in fwl flow sholtlng phenwenon of Incomplete ombustor blow-out. Altitude; 45,000 feet, ilight Naoh number, 0.18.

...... -..

2N NACA RM E51Eo8 " 59

"-

.. .

Figure 16. - Lirnlta on engine fuel flow imposed by combustor blow-out. Alti- . "" . tude, 45,000-feet; flight Mach. number,. 0.19...... 3N NACA RM EX308 61

"

.. .

.~ ."

.. .

Figure 18. - CorP37-ation of compressor stall and canbustor blow-out with compressor.preesure ratio and corrected e&ne speed for J47D (m-1)engine.

...... : .: , . I ,. I. : .. . ." .. . .. , . I: I ' I: , , ,4 I' I' ' 'I .. ., ,

%!

-...... - .. 5N NACA RM E51M)8 - 65 120

a al 2 100

80

k

60

40

x)

0 (a) Engine epeed.

X)

0

Figure 20. - Effect Of altitude on variation of engine speed and engine mount force during thst selector burste frcm 10 to 90 degrees. Blight Mach number. 0.19. - 66

(a) Engine speed.

0 2 4. 6 8 la 12 14 16 Time, eec (b) Engine mount force. Figure 21. - Efreot of flight Maoh number on variation or engineepeed and engine mount force during thruat saleator bursts from 10 to 90 degrees. Altitude, 25,OOO feet. NACA RM E53308 I 67

I.

. 68

100

0 c1 e :: 40 u a P :. u 220 " 8 :e 0 69

..

..

" .

..

..

:.

(a) =pine apeed. Altitude, 45,000 feet.

c 70 NACA RM E51E08

- ......

". .- I- ."

c

...... NACA RM E51E08 - 71

100

80 .. .

60

40

20

0 (a) Engine agead.

..

Figure 25; - Effect of altitucie on variation of wine speed and enghe mount force during that selector chope from 90 to 10 degrees. Flight Mach num- ber, 0.19. 72

. "...... -......

b I 2180

all ombuatms

n I =no00 " .. "

I

20,000 800 loo0 3500 xxx) 3000 SSOO 4OQO Wlndnilling speed, N, rpm I I -20 .30 .M .50 .BO .70 .ED -90 1.00 1.10 1.20 Blight Waoh number, (a) Reid vapor pressure, 7 wundu.

Bim27. - Altitude starting charaoteristioswing llIL-I?-5624 (AN-F-58) m01.

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

I

t I

;I ...... 2180 I , I

.70

...... 76 .

1200

800

400

a

1200

800

400

(b) Windmilling speed, 1600 rpm.

Figure 29. - Variation OF turbine-outlet temperatures behind indi- vidual combuetors with time for two typical start6 at altitude of 35,000 feet. NACA RM =DO8 - 77

L

50,000

45,000

40,000

cl - 2 35,000 9 -P *+I 2

30,000

25, OOO

20,000

Figure 30. - Variation of time required far ignition and flame propagation with altitnae for altitude tart^ a*.&Ou8 VindmillFng speeds. Valuesby data points Indicate initial engine speed at wMch i-iticm eyetern turned on. 78

Tail-pipe fuel-air ratio, (f/a),, Figure 31. - Tall-pipe fuel-alr ratios reguired for autot@tion with eeveral afterburner canifgumtione. Fuel, MIL-F-5624 (AN-F-58); burner-Inlet tempera- ture, 1250° to l3OO0 F; flight -Maoh number, 0.19. - ...... -.

1 I 2180

Tail-pipe fuel-alr ratio, (f/a)a,.,

, 'I

......

I I' !

Figure 33. - Behavlor of eeveral engine variables Uurlng automatically controlled acaeler ...... , r'

I ......

. . ..

.

Figure 34. - Behavior of several enRine variables during automatically controlled acceleration Prom ...... -. . -. - . - -

dry thrust to partial afterburning. Altitude, 25,000 feet; PutMch mmber, 0.19.

I...... - ...... I

......

19N

I ...... , ...... ZON XACA RM %I308 - 87 .

- I"

50,000

40 000

30, OOC

;-r 4 20 J om

10,000 v 3 .005 .010 .ox -020 Tail-pipe fuel-air ratio, (f/a)ab

Figure 36. .- Lean blov-out limits obtainedwith several con- f~gurationsusing MIL-F-5624 (AN-F-58) fuel at burner- inlet temperatures fran 1250° to 13000 F.. Flight Mach number, 0.19.

NACA-Lpnplep - 1-8-82 - 428 - SECURITY INFORMATION

I 4