NASA Technical Memorandum 4140 Measurement Effects on the Calculation of In-Flight Thrust for an F404 Turbofan Engine Timothy R. Conners SEPTEMBER 1989 NASANASA Technical Memorandum 4140 Measurement Effects on the Calculation of In-Flight Thrust for an F404 Turbofan Engine Timothy R. Conners Ames Research Center Dryden Flight Research Facility Edwards, California NASA ‘National Aeronautics and Space Administration Office of Management Scienttic and Technical Information Division 1989MEASUREMENT EFFECTS ON THE CALCULATION OF IN-FLIGHT THRUST FOR AN F404 TURBOFAN ENGINE ‘Timothy R. Conners* NASA Ames Research Center Dryden Flight Research Facil Edwards, California Abstract FGI AA study has been performed that investigates parameter PV measurement effects on calculated in-flight thrust for the = FN General Electric F404-GE-400 afterbuming turbofanengine ey ap ‘which powered the X-29A forward-swept wing research air- craft. Net-thrust uncertainty and influence coefficients were calculated and are presented, Six fight conditions were an- NWT alyzed at five engine power settings each. Results were ob- tained using the mass flow-temperature and area-pressure FR thrust calculation methods, both based on the commonly pyg used gas generator technique, Thrust uncertainty was deter- ‘mined using a common procedure based on the use of mea- 9 surement uncertainty and influence coefficients. The effects ‘of data nonlinearity on the uncertainty calculation procedure ypc were studied and results are presented. The advantages and disadvantages of using this particular uncertainty procedure. FT are discussed. A brief description of the thrust-calculation LPT technique along with the uncertainty calculation procedure Lypyp is included. M Nomenclature NI A cross-sectional area, in? N2 AE effective cross-sectional area, in? PLA ALT. altitude, ft Pr AP area pressure P CFG gross thrust coefticient R ci influence coefficient te) E ‘error in thrust due to parameter-measurement = T uncertainty IT ECU electronic control unit u FG ‘gross thrust, Ibf UF Thevospace engines: Member TAA. UFNAP Copyright © 1989 by the American Istiue of Acrnasic and Aso rauics, ne. No copyright i artered in the United Sates under Tide 17, US. Code. The US. Govemmenthasaroyaly.fieelcensetoexercve ali = UPNWT rights under the copyright claimed herein for Governmental porpores, All, cother rights are reserved by the copyright owner ideal gross thrust, Ibf fuel heating value, Buy/lbm net thrust, bf net thrust, area-pressure calculation method, Ibe ‘et thrust, mass flow-temperature calculation method, Ib ram drag, Ibf fan variable guide vane angle, deg gravity constant, 32.17 ft-lbmjlbf:s? total enthalpy, Buu high pressure compressor in-flight thrust low pressure turbine Jinear-variable differential wansformer ‘Mach number fan rotor speed, rpm high pressure compressor rotor speed, rpm power lever angle, deg. total pressure, Ibffin? static pressure, Ibf/in? as constant, Ibf-fylbm-°R root-sum-square static temperature, °R total temperature, °R ‘measurement uncertainty root-sum-square thrust uncertainty ‘uncertainty in net thrust, area-pressure ‘method ‘uncertainty in net thrust, mass flow- temperature method.v velocity, V5 vr ideal velocity, tus w ‘mass flo, Ibe/s WFAB afterburner fuel low, Ibm/s WFE engine core fuel flow, nm/s wr mass flow temperature 1 specific heat ratio, air F404 engine station identification numbers: ° froesueam 1 engine inlet 2 fan inlet 2s high pressure compressor inlet high pressure compressor discharge 4 combustor discharge Jow pressure turbine discharge 558 low pressure turbine discharge measuring plane 6 afterburner inlet 1 exhaust nozzle inlet 8 exhaust nozzle throat 9 exhaust nozzle discharge Introduction Knowledge of the uncerainy of calculated in-ight thrusts important in understanding the accuracy of irra performance values, including vehicle drag" Errors in en- Bine nd aircraft parameter measurements required forthe Calculation of thrust direcuy affect thrust uncertainty. The degree of influence that these errs have depends on the characteristics ofthe specific engine model and thrustcaleu- Jation method utilized. Other sources of uncertainty, nclud- ‘ng engine model error, affect thrust uncertainty, but these effects were not considered inthis investigation. Several studies have been undertaken in the past two decades at NASA Ames-Dryden Flight Research Facility in ‘vestigatng thrust calculation methods and their corespond- ing sensitivity to measurement errors. These studies were performed for various engines installed in several types of aircraft, including the XB-70? and the F-111.3 ‘The primary purpose of this investigation is to docu- ment the net-thrust uncertainty and influence of measure- ment error forthe F404-GE-400 afterburning turbofan en- gine. Problems in the analysis procedure, including inlu- ‘ence data nonlinearity, were also investigated to better un- ‘derstand the uncertainty methodology used. ‘The engine analyzed was installed inthe X-29A forward- swept wing research aircraft ship 1) which was flight ested at NASA Ames-Dryden, The results of this study arc ap- Dlicable to other aircraft powered by the F404 engine which utilize an instrumentation system similar to that used for this analysis, Previous work involving the F40S engine installed inthe X-29A includes a limited study of thrust sensitivity to ‘measurement changes. A limited thrust uncertainty invest gation was also performed, buta less accurate instrumenta- tion system was used An insight thrust (IFT) computer program was used 10 calculate thrust forthe F404 engine using measured parame- tervalucs a input This program was supplied by the engine ‘manufacturer? The program uses two variations ofthe cla- sical gas generator thnst-alculation technique: the mass low-temperature (WT) method and the ara-pressure (AP) method. The eror effects of ten important measured input parameters on thrust were determined fr both methods. In- Tet ram recover, bled air, and horsepower extraction were cstimated and usd in the thrust calculation. Tei effects on thrustuncerainty were not included inthis analysis since the rors in estimating these aircraft installation factors were unknown, Fora similar reason, engine model ero inherent tothe IFT program was not considered Actual measured test data was not available fr input ino te IFT program at the time the analysis was performed. ‘Therefore, this data was estimated using an F408 engine speciication computer simulation program. Actual X-29A aircraft and engine instrumentation accuracy dat, required in the thust- uncertainty analysis, was used in te study. ‘This report presents parameter measurement effects on calculated net thrust for the F404 engine which was installed inthe X-29A. Six simulated fight conditions, ranging from 10,000 ft to 40,000 ft in altitude and from 0.4 to 1.6 Mach number (M), were studied at throttle power lever angle (PLA) seuings from part-power (70" PLA) tomaximum af- terburmer power (130"P LA). Thrust uncertainty was deter- ‘mined using a common procedure based on the use of mea- surement uncertainty and influence coefficients, These in fluence coefficient values were also calculated and are pre sented. The effects of data nonlinearity on the uncertainty calculation were also studied. Results are presentod and the advantages and disadvantages of using this particular un certainty procedure are discussed. In addition, a brief de- scription ofthe thrust calculation technique and uncertainty calculation procedure is included. Engine Description ‘The F404-GE-400 engine is a 16,000-Ibf thrust class, low bypass, twin spool turbofan with afterburner. The engine incorporates a three-stage fan and a seven-stage high pres- sure Compressor, each driven by a single-stage turbine. The fan and high pressure compressor guide vanes utilize vati- able geometry. Bleed air extraction is provided atthe sev- enth stage of the high pressure compressor. ‘The combus- tor isa through-flow annular type utilizing atomizing fuel nozzles. The aterbumer can be fully modulated from mini- ‘mum to maximum augmentation and uses fan discharge air ‘and an afterbumer liner to maintain a low engine skin tem- perature. The hinged-flap,cam-linked exhaust nozzle is hy-raulically actuated. An engine accessory gearbox is driven by the compressor rotor. This gearbox drives the lubrica- tion oil pump, the variable-exhaust nozzle power unit, the ‘generator, and both the main and afterburner fuel pumps. A schematic view of the F404-GE-400 engine with station designations is shown in Fig. 1 ‘The engine-control system consists of the uhrotle, main fuel contol, electronic control unit (ECU) and afterburner fuel control. Throttle (power lever) movement is mechani- cally transmitted to a power amplifier which positions the ‘main fuel control. During fight, PZ.A ranges from 31° (ight idle) to 130° (full power with afterburner). Interme- diate power (fll nonafterburning) occurs at 87° PIA At power settings below intermediate, engine inlet total temperature (T°P1) and throttle movement control the high ‘pressure compressor rotor speed (V2) through the main fuel ‘control. At intermediate power and above, fan rotor speed (1) is controlled by the ECU as a function of TI, while 'N2 remain essentially constant, The ECU senses engine and aircraft parameters, computes schedules, and maintains engine limits, The afterbumer fuel control schedules fuel flow to the pilot and main spraybars. ‘A single F404 engine is mounted in the fuselage of the X.29A and utilizes two side-mounted, fixed geometry inlets ‘optimized for transonic performance. Instrumentation ‘The engine location of measurements used in the thrust calculation are shown in Fig, 1 A resistance temperature device is usedto determine TTL while 171 is measured using an eddy-current instrument uti- Tizing magnetic pickup from the fan. Fan variable guide vane angle (F°VG) is measured using a linear-variable dif- ferential transformer (LYDT). ‘The low pressure turbine discharge total pressure mea- surement, P7558, a critical parameter in uhrust calculation, is obtained using four five-element total pressure rakes. The 20 PT'SS8 pressures are measured by a multi-port differ- ‘ential transducer referenced to a highly accurate absolute twansducer. The differential transducer is thermally con- tolled by a heater-insulation blanket to maintain a con- stant temperature at which the unit was calibrated, ‘The f- nal PT'SS8 value is an average of the 20 PT'SS8 measure- ‘ments, Values outside a specified tolerance are omitted from the average, The uncertainty in the PT'5S8 measurement is 1 root-sum-square (RSS) of the differential PT'SS8 uncer- tainty and the reference-pressure uncertainty. ‘The nozzle throat area, A8, is also measured using an LVDT. Volumetric flow moters are used to measure engine core and afterburner fuel flows (W FE and W F AB respec: tively). Fuel temperatures are measured in both the gas gen- erator and afterbumer fuel lines o permit conversion of vol metric values to mass flow. Fuel heating value (FHV) is 2 laboratory determined quantity ‘The thrust calculation also requires the measurement of freestream altiude (ALT) and Mach number (0M), ‘These measurements are oblained through the aircraft air- data system, ‘Table 1 presents the range and absolute uncertainty for ‘each measured parameter used in the thrust calculation. The ‘measurement-uncertainy values were ether supplied by the engine and instrumentation manufacturers or determined through in-house calibration. This instrumentation was in- stalled in the X-29 during the aircraft’ performance flight- testing phase. Calculation Procedures In-Flight Thrust Calculation ‘Thrust Calculation Methods. Thrust was calculated according to two variations ofthe commonly used gas gen- erator technique: the mass low-temperature (WT) method ‘and the area-pressure (AP) method. The gas generator methodology is based on classical momentum, energy, and continuity laws, The primary difference between the WT ‘and AP methods is due to the manner in which nozzle mass flow is calculated, ‘The simplest form of the equation for ideal gross thrust (FG) is based on the time rate-of-change of mass momen- tumat the nozzle of the engine, This equation assumes com- plete exhaust expansion to ambient pressure and therefore neglects the thrust pressure term at the nozzle exit. Using F404 station identification, the equation takes the form rar-Bv19 ‘where 178 is the nozzle throat mass low rate and V9 is the ideal nozzle exit velocity 19 js expanded further in the above equation to ob- tain the mass flow-temperature equation in ideal gross thrust form. A gross thrust coefficient (CFG) is used to ob- tain actual gross thrust; and ram drag (engine inlet airflow, W1, multiplied by freestream velocity, V0) is subtracted to give net thrust. The result is the net thrust mass flow. temperature equation x CFG- vo aThe nozzle mass flow term (W8) is further expanded in the above equation to give the net thrust area-pressure equation PwAP = [os pre >| “| ay* xcre- “yo @ 9 ‘The CFG modifies the ideal gross thrustterm ofthe equa- tions to account for incomplete expansion, two-dimensional expansion effects, nozzle friction, and other sources of thrust loss, The values for this coefficient are deter- ‘mined empirically through engine altitude facility and ‘ground-testing. Equations (1) and (2) illustrate the effects that nozzle and freestream parameters have on the thrust calculation, ‘The quantities in the equations are obtained through sev- eral measurements which are discussed in the following sec- tion, Complete derivation of both equations is presented in Refs. 3 and’. ‘The In-Flight Thrust Program. The thrust values used to oblain the results of this report were acquired us- ing an in-flight thrust (IFT) calculation program. This com- puter program was developed by General Electric for the Navy's F404-F-18 Propulsion System Integration Program. ‘The IFT program determines mass flow, pressure, and temperature atthe exhaust nozzle inet by modelling the en- gine as a gas generator. The gas generator procedure uses a ‘combination of engine performance models, engine compo- ‘ent ground:-test data, and actual measured engine and air- craft fight data. The IFT program uses these models and data to generate the values necessary for thrust calculation, ‘The use of actual Night data allows the program to adjust for cengine-to-engine performance variations. A schematic representation of the actual IFT calculation procedure and data flow is shown in Fig. 2. The 10 parame- ter measurements listed in Table 1 are used as input into the TFT program and are labeled in italies in the figure. Flight-condition measurements ALT and M, along with the measured engine parameters NI, TTI, and FVG, are used in the inlet model and airflow calculation, Utilizing these parameters, 771 is calculated, Altitude (41.7) is used to calculate freestream pressure (p0) based on standard day correlations. This value, along with W1 and M, allows the program to estimate inlet ram recovery using empirically de- rived data. Engine inlet total pressure (PT'1) is then calcu- lated. Freestream pressure (p0) is also used directly in both ‘equations (1) and (2). Next, using an energy balance between the turbine and compressor, along with N'1, 771, and PT‘, the total temper- ature, enthalpy, and pressure of the flow Teaving the com- pressor (T'T3, 13, and PT3 respectively) are calculated ‘Compressor discharge airflow (W3) is determined by sub- ‘acting estimated nominal bleed airflow from W1, Horse- power extraction, also estimated by the program, is used in the energy balance also. The measurements WE, PIV, and the compressor ‘model output are then used to calculate the energy rise across the combustor and turbine, Total temperature and enthalpy atthe afterburner inlet, 776 and #6, are then determined ‘Total pressure at this station, PTS, is equal to the PT'SS8 measurement, Afterburner inlet mass flow (W76) is caleu- lated by adding W PE to W/1 and subtracting bleed airllow. Ifthe afterburner is not operating, 777, 177, and W7 re- ‘main unchanged from station 6. Nozzle inlet total pressure (PT7) isobtained by subtracting afterburner frictional loss, based on ground-test data, from PT‘. If the afterburner is in operation TT7, 17, and 7 are then determined using FHV and WFAB measured values and station 6 condi tions. Pressure losses because of heating and friction are included to obtain PTT. Since the flow is isentropically compressed from the no7- Ale inlet to the throat, PTS and TTR equal PTT and TT7 respectively. Nozzle throat toil pressure (PT) is required in both equations (1) and (2) while 7°78 is used in equa- tion (1). Nozzle inlet mass flow (17) is adjusted for no2: le leakage giving mass flow at the throat, 178, which is required in equation (1), the mass flow-temperature thrust equation. ‘The measured nozzle throat area, A8, is adjusted using a flow coefficient to give effective throat area, AES. ‘This value is required as input for equation (2), the arca- pressure equation, Freestream pressure (p0), from ALT, is used to determine nozzle leakage and flow coefficients Ideal gross thrust (G1) is then calculated according to both the WT and AP thrust-calculation methods. The re- sults are adjusted using the CFG, which is determined from ‘empirical data and is based on nozzle operating conditions. ‘Actual geoss thrust (FG) is then calculated. Freestream ve- locity (V0), calculated using M and ALT, is multiplied with W1 giving FR. By subtracting this erm from FG, net thrust (FV) is obtained for both thrust-calculation methods, Certain aircraft installation effects, including inlet spillage and nozzle drag, were not included in the calcula- tion because they are independent of the net-thrust calcula tion procedure, ‘Thrust-Uncertainty Calculation ‘The common procedure by which thrust uncertainty (UP) js calculated, is to perform a root-sum-square on the indi Vidual thrust errors () each due to the measurement uncertainty of one parameter.* The thrust-uncertainty calculation ‘equation therefore takes the form 12 UF = (ER + ER +...+ Bly 8) Equation (3) contains ten squared terms, one foreach of the input parameters analyzed. ‘The individual thrusterror values are determined ana- Iytically by adding the associated measurement uncertainty (from Table 1) to the measurement-parameter value and in- Putting the result into the IFT program along with the other required, but unmodified, parameter measurements. The calculated-thrust value is then compared to the value that results using all unmodified input, and F; due to the single- ‘measurement uncertainty, is determined. This procedure is repeated for each parameter. The F values must be recal- culated for a change in flight or engine condition or for a change in thrust-calculation method. Although precise, this thrust-uncertainty calculation pro- cedure can be time-consuming and cumbersome to use if ‘multiple instrumentation systems, and therefore different ‘measurement uncertainties, are analyzed. This is because of ‘the need to rerun the IFT program for each change in a mea- surement uncertainty. Also, this procedure does not clearly [ Fav Fig, 1 F404 engine station and measurement locations. Tm Fve ee ae To, i i Fenand 7 «| — ori] rims amine as [acme] [ne alr calculation] W1 | energy rise | PT3 Jenergy rise| PTE | model [pr7 | ansiysis 7 awe wr we [ ace wt mts | pie cre. M2 Remareg rae MA ALT mf FR ME vdestgrossinvs [FOL A Yea Tagen xvo eaigross tmust HFCL gross thrust ESS FGsFolxcre Sere T aur J Fa Fig. 2. F404 in-flight thrust calculation flow chartMach to number (@ WT method. ‘Aitude, Mach umber on 08 12 08 18 rT) PLA, dog (b) AP method. Fig. 3 Netthrust uncertainty as a function of PLLA. 3OD rv = pve Weas Eo Ni we SS Ao MMM ALT CO Prsse OG mach Intermediate (a) WT method; 10,000 f1, M = 0.4. OD rH 3 Fe Wweas 4 Nt woe BG 1 i cy, percent 2 Smee aT Ta f 7 8 99 100 «110 «120130 PLA, deg (b) AP method; 10,000 f, MF 4 Fig. 4 Relative parameter Ci values.OD rv = Wve WFAB [J Nt 0 wre ™ Ea as Mmm ALT CO Prsse Ol MACH 4 nara ci, percent 70200901008 110) 20190) PLA, deg (© WT method; 10,000 fi, M = 0.8. 8. ALT Ptssa Ol] MACH Intermediate ci, 2 percent 7 8 9 100 110 120 130 PLA, deg (@) AP method; 10,000 ft, =08, Fig. 4 Continued,OOD Fev FVG WFAB Nt eas mar ° ES Pisse HI flacn Intermediate 7 8 9 100 110 120 190 (©) WT method; 30,000 ft, M = 0.9. (AP method; 30,000 A, M =0.9, Fig. 4 Continued, 16DOD Hv 5 rva Weas [a Nt 0 wre ™ fl As mm ALT. CO Ptsse Gill MACH ci, percent (@) WT method; 30,000 1, M = 1.2. (70; 00 fee 100) 410) 120) tra PLA, deg (h) AP method; 30,000 ft, Fig. 4 Continued. ”WOE Fav 3 eva Weas Fj wt 0 wre ™ AS. ALT, Prsss MN MACH Intermediate 70 80 9000110120130, PLA, dog ( WT method; 40,000 f, M = OD rv SS rvs Weas Ent Owe SS Ag ALT Co Prsse Om MACH ze [= (AP method 40,000 ft, M = 08. Fig. 4 Continued.FvG NY ™ mm ALT CO Prsse UH) MACH Intermediate d a ci, percent 1D rw 3 ne WFAB Nt COO wee ™ GS as MB at CO prsss = ENE mach PLA, deg (1) AP method; 40,000 ft, Mf = 1.6. Fig. 4 Concluded.© WT method 1 AP method Thrust change from baseline, percent 4 2 08 2 4 6 Nt change from baseline, percent (@) 1; 10,000 1,4 8, 130°PLA, Thrust o change from basetin. percont 2 (© WT method 10 AP method 3 $6 4 2 oOo 2 4 6 FHV change from baseline, percent (©) PIV; 30,000 ft, M = 1.2, 130°P LA. Fig. 5 Nonlinear data analysis examples 207 10 8 Theust 6 change gL trom 4 baseline, 2 percent ° 2 (© WT method 4 1 AP method 6 o 2 4 012 3 6 «5 6 PTS58 change from baseline, percent (©) PTS58; 30,000 ft, M =09, 87PLA. Fig. 5 Concluded, (© WT method 10 AP method 1294 567 6 9 01 UF(eq. 4), percent Fig. 6 Nonlinearity effects on equation (4) thrust uncertainty results, 21Report Documentation Page T Paper Ro T Governt Reson We T Recipients Caaog Wo NASA 14-4160 TT nd Sue 5 Rapon aie - ‘September 1989 ‘Measurement Effects on the Calculation of Fea BR In-Flight Thrust for an F404 Turbofan Engine ® Taha 1 Paoming Oraranion Ripa Wo ‘Timothy R. Conners H.1556 78 Work Unit Wa, 3 Peorming Orguizion Name ad Arse RTOP 53-02-55 TH ontect arr NASA Ames Research Center Dryden Flight Research Facility P.O. Box 273, Edwards, CA 93523-5000 TE Toe of Report and Pood Cored 7% Sponeoing Agcy Rare nd Adtrane Technical Memorandum TH Sponsoing Agency Cede ‘National Aeronautics and Space Administration ‘Washington, DC 20546 ‘Supienentary Nowe Prepared as ALAA Paper 89-2364 for presentation at the AIAA Joint Propulsion Conference, Monterey, Califomia, July 10-14, 1989, seat ‘A study has been performed that investigates parameter measurement effects on calculated in-flight thrust for the General Electric F404-GE-400 afterburning turbofan engine which powered the X-29A {orward-swept wing research aircraft. Net-thrust uncertainty and influence coefficients were calculated and are presented. Six light conditions were analyzed at five engine power settings each, Results were obtained using the mass low-temperature and area-pressure thrust calculation methods, both based on the commonly used gas generator technique. Thrust uncertainty was determined using a common procedure based on the use of measurement uncertainty and influence coefficients. The effects of data nonlinearity on the uncer- lainty calculation procedure were studied and results are presented. The advantages and disadvantages of| using this particular uncertainty procedure are discussed. A brief description of the thrust-calculation tech- nique along with the uncertainty calculation procedure is included. oy Words Sopp by Aura TE Oaiibaian Siren F404 turbofan engine; Influence coefficient; Unclassified — Unlimited ‘Measurement uncertainty; Performance; Thnsst calculation; Thrust uncertainty Subject category 07| Secary Casal ot spe? 5 Secary an Tat 0 peas? Ti Ne. of poe Ya. Price Unclassified Unclassified 24 A02 NAGA FORM Tea oct o For sale by the National Technical Information Service, Springfield, VA 22161-2171.