Document
..J/L.; Ci_ <Z :_
NASA Technical Memorandum 4465
!
] " _ , / 7_ i -Z=-2W -
Summary of the Effects of Engine
Throttle Response on Airplane
Formation-Flying lities
l " C. -
i
: .;-:;- - Z. Z_ Kevin R. Walsh MARt H t ....
(',_ A-! ,'-44_5) SUY -_A_Y 'J_ T_" -F.-CTS _F F_INE TH_3TTLE q_zSPONSE
N
-].. AIRPL_'_E FqR,_ATION-FLYI_G Uncl as _--" L_UAL ITILS (_ASA) 2o p 0153700 N HI/O8
NASA Technical Memorandum 4465
Summary of the Effects of Engine
Throttle Response on Airplane
Formation-Flying Qualities
Kevin R. Walsh Dryden Flight Research Facility Edwards, California National Aeronautics and Space Administration Office of Management Scientific and Technical Information Program
SUMMARY OF THE EFFECTS OF ENGINE THROTTLE
RESPONSE ON AIRPLANE FORMATION-FLYING QUALITIES
Kevin R. Walsh* NASA Dryden Flight Research Facility P.O. Box 273 Edwards, California 93523-0273 Abstract validated. No criteria for defining good or bad engine throttle response characteristics currently exist.
A flight evaluation was conducted to determine the In the past 10 years, engine throttle response prob- effect of engine throttle response characteristics on pre- cision formation-flying qualities. A variable electronic lems have been encountered in several airplanes, for ex- throttle control system was developed and flight-tested ample, the F-15 (McDonnell Douglas Corporation, St.
on a TF-104G airplane with a J79-11B engine at the Louis, Missouri) with the developmental F100 engine model derivative (Pratt & Whitney, West Palm Beach, NASA Dryden Flight Research Facility. This airplane Florida), _ AV-8B (McDonnell Douglas Corporation, was chosen because of its known, very favorable thrust St. Louis, Missouri), 3 and F-18 (McDonnell Douglas response characteristics. Ten research flights were Corporation, St. Louis, Missouri, and Northrop Cor- flown to evaluate the effects of throttle gain, time delay, and fuel control rate limiting on engine handling quali- poration, Hawthorne, California). Problems in these ties during a demanding precision wing formation task. examples ranged from excessive initial time delay or Handling quality effects of lag filters and lead compen- lag in the F-15 and AV-SB airplanes to high throttle sensitivity in the initial F-18 installation. Such exam- sation time delays were also evaluated. The Cooper and Harper Pilot Rating Scale was used to assign lev- ples illustrate the need for handling quality guidelines els of handling quality. Data from pilot ratings and or design specifications for advanced engine control sys- tems. These engine design specifications are analogous comments indicate that throttle control system time to those developed over the last two decades for ad- delays and rate limits cause significant degradations in vanced flight control systems. As a result, data are handling qualities. Threshold values for satisfactory required to develop handling qualities criteria to quan- (level 1) and adequate (level 2) handling qualities of tify the effects of thrust response dynamics on a pilot's these key variables are presented. These results may provide engine manufacturers with guidelines to assure ability to complete precision flight control tasks.
satisfactory handling qualities in future engine designs.
A brief flight research program was conducted at the NASA Dryden Flight Research Facility (DFRF) Introduction to investigate the effects of varying engine throttle re- sponse on airplane handling qualities. An electronic The ability to conduct such precise flying tasks as variable-throttle response system was developed and close-formation flight or aerial refueling is strongly af- fected by the engine throttle response, that is, thrust installed on a two-seat TF-104G airplane (Lockheed response caused by throttle changes. With the ad- Corporation, Burbank, California). This airplane was vent of digital engine control systems, control soft- an ideal choice because its J79-11B engine (General Electric, Lynn, Massachusetts) responds to throttle ware has become commonly used to modify engine re- changes extremely quickly. The variable-response elec- sponse characteristics) The engine throttle response tronic throttle enabled the pilot to evaluate throttle may be degraded if such modifications are improperly system degradations and resulting effects on engine handling qualities.
"Aerospace Engineer.
Copyright (_)1992 by the American Institute of Aeronautics This program provides initial data for developing and Astronautics, Inc. No copyright is asserted in the United handling qualities criteria and design guidelines for at- States under Title 17, U.S. Code. The U.S. Government has taining satisfactory (level 1) throttle response of high- a royalty-free license to exercise all rights under the copy- performance airplane engines. Data were obtained at right claimed herein for Governmental purposes. All other rights are reserved by the copyright owner.
one flight condition: the airspeed was 300 kn, and the Figure 1 shows the TF-104G airplane. Notable fea-
altitudewas15,000 ft. Thisflightcondition represents
tures include the extremely thin ttight surfaces, the a formation-flying task.
short and straight wings with 10 ° anhedral, and a con-
This papersummarizes the resultsof the flight re-
trollable horizontal stabilizer mounted at the top of
search program anddescribes the variable-throttle re-
the vertical stabilizer. The wings have leading- and sponse system asinstalledin the TF-104G airplane.
trailing-edge flaps and a boundary-layer control sys-
Timehistorydataforthe representative pilot evalua-
tem used with the trailing-edge flaps to reduce landing
tionsanda compilation ofpilotcomments withrespect
speeds.
to the time historydata arepresented. These data
showthe differences between satisfactory, adequate,
andinadequate (level1,2, and3) handlingqualities
on theCooper andHarper Pilot RatingScale. 4 Data
showing time delayandrate limit thresholds for the
different levels ofhandling qualities arealsopresented.
Additionaldata showthe effects of a first-order lag
filter andofa lead-lag filter in combination with addi-
tionaltimedelay onhandling qualities.
TheworkofGeorge E.Cooper andRobert P.ltarper,
Jr., is gratefullyacknowledged. 4 This workwasin-
strumental in developing thisengine handling qualities
flighttestprogram.
Nomenclature AX aircraft longitudinal acceleration, 9 9 acceleration of gravity, ft/sec 2 EC 80-12366 HQR handling quality rating Fig. 1. The TF-104G airplane.
L/L lead-lag LVDT linear variable-differential transformer The J 79-11B is an axial-flow, high-pressure-ratio tur- bojet engine with variable-inlet guide vanes, variable- MTE mission task element stator vanes, and single-rotor compressor. This engine N1 compressor speed, rpm has a cannular combustor, a three-stage turbine, and a PCM pulse code modulation fully modulating afterburner with a variable-area con- PIO pilot-induced oscillation verging and diverging exhaust nozzle.
PLA power lever angle, deg Variable-Response Throttle System PLACMD power lever angle command, deg Description PLAFH power lever angle position feedback, deg The electronic variable-response throttle control sys- rpm revolutions per minute tem was developed by Calspan (Buffalo, New York) RVDT rotary variable-differential transtbrmer specifically for the DFRF TF-104G airplane. Neal and TCU throttle control unit Sengupta described the implementation and operation of the throttle control system. 5 The throttle in the AT' change in time delay, msee forward cockpit was modified to command the exper- 6x throttle position limit commanded by imental system. Figure 2 shows the main components throttle control unit, deg of the system. These components consist of an elec- re L/L denominator time constant tronic throttle control unit (TCU), integrated servomo- rn L/L numerator time constant tor and clutch assembly, and position sensors. A cable w,_ second-order lag natural frequency linkage connected the servomotor with the engine fuel controller. The throttle in the aft cockpit remained in ¢ second-order lag damping ratio the production configuration and served as the safety backup system.
Airplane and Engine Description Figure 3 shows a simplified block diagram of the The test airplane was a TF-104G: a high- modified propulsion control system. Additional throt- performance, two-place, trainer-fighter-interceptor air- tle system dynamics were generated by the TCU and plane with a maximum Maeh number of 2.0.
ORIGINAL PAGE BLACK AND WHITE PHOTe(3,RAPI-I Throttle control Servomotor Rotary variable- Linear variable-_diff'erential transforme_'_ '_: ..........
differential transformer 920645 Fig. 2. Electronic throttle system hardware.
Figure ,1 shows a functional block diagram of the Pilot_ electronic throttle control system. The descriptors out- side the polygon represent, the mechanical system, and
,)
the e.lcctronic system is shown inside the polygon.
PLACMD Mechanical System r i In the production TF-104C, airphme, a conventional cable and pulley system connects the forward arm aft throttles to the engine fuel control. The two throttle
E
handles are linked so that, whe.n one throttle handh'.
is moved, the other tracks its position. As a re, suit, TCU variables the position of the forward and aft throttles match at all time.s.
To incorporate the electronic throttle system, the forward or evaluation throttle was disconnected by re- Fig. 3. Modified propulsion control system.
moving the throttle linkage to the fuel control cables.
A rotary variable-differential trarisf(_rmer (I{VI)T) was the servomotor. For example,, the power lever angle conrw.cted to the throttle handle through gears and in- commanded (t-'LACMD) by the pilot is modified by stalled in the throttle housing to sense forward throb the TCU to command the servomotor. The resulting tie position, l)isconnecling the forward throttle from change in the power lew.'r angle (f)LA) caused by the the cable system eliminated the inherent friction on the servomotor is the input to the fuel controller. This an- throttle handle; therefore, an ad,justabh.' friction device gle provides PLA position feedback (PLAFI3) to the was installed in the housing to maintain throttle feel.
electronic control system. Measurements of compres- This device, a small phenolic block and bracket, created sor speed (N1) and aircraft longitudinal acceleration drag against the throttle-handle axle. The device was (AX) are also represented. Throttle gain or sensitiv- intended to be adjustable to any friction level; however, ity, 9/deg; transport time delay, sec; and rate limiting, deg/sec were the primary variables for the experiment. the throttle stick force was adjustable to a maximum of 2 lb or approximately one'-half of the normal stick The secondary variables included first-order lag filter force for the production throtth:.
time constants and lead-lag time constants.
EILACK AND WHITE PH_OT_9(_PH ThrottleControlUnit (located in aft cockpit) Commandsignal filtem RVDT (lead-lag,second-orderlag, -- position /--- sensor timedelay, or cascade _I Vadeble- combinations) -I commandgain +'-+,_: Forwardcockpit throttle command __ Step Input _ (electrical) Linear mode o --" commend (nofilters) o
_ ( ,CUtest
Verlable.lrequency sinewave generator Slnueoldal input Aft cockpit throttle command L Variable- Variable- (mechanical) command --_ command Servomotor! (Cablelinkage) _R Servomotor L__ .-b /- Double I position rote assembly I / drive pulley i limit limit Tachometer!
I ddve amplifiersI ate feedback F---I valve l
i tJ ''0on*ro'l
LVDT Positionfeedback position sensor Originalcable linkage 1 920191 Fig. 4. The throttle control system.
system were optimized by adjusting the rate feedback A servomotor, clutch, and cable assembly was in- gain.
stalled in the engine bay to position the filel controller.
A linear variable-differential transformer (LVDT) was Figure 5 shows the TCU panel which the aft pi- used to sense PLA at the fuel controller.
lot used to activate the system, select the operating The TCU was installed in the left-hand console of mode, set the system variables, and enter test signals.
The TCU inserted time delays, rate limits, lead-lag the aft cockpit and provided the desired throttle re- time constants, first- and second-order lags, and po- sponse variations for the experiment. The backseat sition limits into the command path. These variables pilot could disengage or override the electronic throt- acted directly on the pilot throttle commands. A ro- tle control system at any time. The mechanical con- tary switch on the TCU panel was used to select the trol system was always functional from the aft cock- desired test variables. Thumbwheel switches on the pit. Since the aft cockpit throttle was mechanically linked to the fuel control valve, the rear throttle po- control panel were used to set the value of the test vari- sition tracked the throttle commands. The front seat ables. The variables could be tested individually with- out altering other variables. They could also be cas- pilot had no way to control the engine if the TCU were caded, such as testing a time delay followed by a lead- disengaged.
lag filter. Positive and negative rate limits could be ad- Electronic System justed independently to simulate an engine that would increase revolutions per minute (rpm) at a different The block diagram inside the shaded polygon of rate than it would decrease rpm. The output signal of Fig. 4 shows the electronics of the throttle control sys- the circuit selected was then rate and position limited.
tem. The RVDT generated an electrical signal which was amplified by a gain factor in the TCU. The TCU Circuits in the TCU could selectively generate two checked the amplified signal against position (ampli- test inputs: a step signal and a sinusoidal signal. The tude) and rate limits. This signal was compared with step input had either positive or negative polarity, and the actual throttle position at the fuel control valve the sinusoidal input could vary from 0.0 to 2.9 Hz. The which was measured by the INDT. The difference be- test inputs were used during the flight program; how- tween the RVDT command and the LVDT feedback ever, the data are not presented in this paper. Re- signals was an error signal used to advance or retard fer to Ref. 5 for additional information concerning the the servomotor position. A tachometer on the servo- implementation and operation of the variable-response motor assembly provided a rate feedback to the TCU.
electronic throttle system in the TF-104G airplane.
The damping characteristics of the electronic control _n _d Test frequency Thumbwheel switches Lead-lag for setting control system variable parameters
NN
_ mn Test Second-order lag
ND
AT 5x max Command gain I I delay position limit ,I /- Red LED indicates Red LED Indicates servoclutch has [_ [_ V power to servomotor been powered --_ Ra_'e limit (Sx max) CON/] _ Yellow LED indicates 0 r----Tellt--_ AT -_ //.*L/L-ZIT OSTBY .if power to TCU Engage ISin Poa [ 2nd \ //2nd-,,IT ON J electronics _- Sequence select switch for choosing mode of operation 92O192 Fig. 5. The throttle control unit panel.
Instrumentation in a TF-104G airplane. According to U.S. Air Force military specification MIL-F-8785C, the TF-104G air- Information from the airplane, engine, and TCU plane, being a highly maneuverable fighter and inter- was obtained on a pulse code modulation (PCM) ceptor, is considered a class IV aircraft. 7 data acquisition system. The serial PCM data were Formation flying was the mission of interest for the telemetered to the ground, decoded and formatted for evaluations. The specification classifies formation fly- real-time display on cathode ray tubes and strip charts, ing as a category A flight phase. 7 This flight phase and recorded for postflight analysis. Instrumented pa- requires rapid maneuvering, precision tracking, or pre- rameters were measured at 200 samples/sec.
cise flightpath control.
Measured values of pressure and temperature were The primary variable for obtaining handling qualities used to generate computed values of airspeed, altitude, data was thrust control. Other typical handling quality Mach number, static temperature, and standard day variables, such as airframe stability and control char- temperature at altitude. These calculated parameters acteristics as well as cockpit interface elements, were were also formatted for real-time display. Tables 1 and not considered.
2 in Ref. 5 show the airplane, engine, and TCU param- eters that were measured for this experiment.
The five pilots who participated in the program were experienced in evaluating handling qualities of class IV Development of Handling Qualities airplanes. Their role was important in determining Experiment pilot and vehicle performance during the task. Fig- ure 6 shows a modified Cooper and Harper Handling This discussion describes the development of the Qualities Rating (HQR) Scale. This scale defines sat- electronic throttle handling qualities experiment and isfactory, adequate, and inadequate (level 1, 2, and 3) is based on the work of Cooper and Harper. 4 To perform ance.
achieve reliable data and comparable ratings among the pilots, care was taken in developing the experiment Mission Description objectives, mission description, mission task elements The required operation for the formation-flying task (MTE) 6, rating criteria, pilot assessments, and flight was to attain and maintain fore and aft position rel- test procedures.
ative to the wing of the lead airplane by using visual Objectives references. This high gain, precise, closed-loop track- ing task requires the pilot to devote full-attention to In general, the experiment obtained highly definitive airplane control and to use the throttle as the primary handling qualities data for a modified J79-11B engine control input.
Demandson the pilotin Pilot ] Handling'_ Adequacyforselected task or requiredoperation ) C Aircraftcharacteristics " selectedtask or requiredoperation' rating I quality ,,) • Excellent, • Pilot compensation not a factor highlydesirable for desiredperformance Good, Pilot compensation not a factor LevelI, 2 satisfactor negligibledeliciencies • for desiredperformance Fair, soma mildly • Minimalpilot compensation required _ unpleasantdeficiencies for desiredpedormance Yea • Minor but annoying Desiredpedormance requires deficiencies • moderatepilot compensation Deficiencies Level2, Moderatelyobjectionable• Adequateperformancerequires without warrant deficiencies considerablepilot compensation adequate improvement Very objectionablebut • Adequateperformancerequires _ tolerabledeficiencies extensivepilotcompensation Yes Adequateperformancenot attainablewith maximum Majordeficiencies • tolerablepilot compensation; controllabilitynot in question Deliclencies'_ Level3, attainablewith Majordeficiencies • Considerablepilot compensation 8 inadequate is requiredfor control Intensepilot compensation Majordeficiencies • 9 _.. is requiredto retain conlrol Yes Controlwill be lostduring some Majordeliciencies • it controllable_ portionof required operation Yea Pilot decisions Fig. 6. Cooper and Harper Pilot Rating Scale.
The lead airplane for the formation task was either keeping. This subphase was intended to be represen- a T-38 (Northrop Corporation, tiawthorne, Calilbrnia) tative of the precision required in similar operational or an F-18. Nominal flight conditions used for the test tasks, such as in air-to-air refueling or close-formation program were an altitude of 15,000 ft above mean sea flight under adverse instrument conditions. Without the introduction of suitable task perturbations to chal- level and an indicated airspeed of 350 kn. This flight condition was chosen so that throttle motion was al- lenge the pilot and the throttle response system, the ways in partial power. Only in the worst case TCU task would not provide the desired degree of handling configurations did the throttle hit the military or idle qualities discrimination. Close-formation flight with a power detents. smooth leader does not, in itself, provide the necessary discrimination; the evaluation pilot cannot separate the Mission Task Elements optimum from the marginally acceptable cases. Note, for example, that even large relatively ponderous ships Developing a well-defined MTE with handling qual- ity performance criteria reduces the uncertainties and can fly accurate close formations during refueling tasks at sea.
extrapolation to the real-world equivalent required by the pilot, ttaving the pilots consider the real-world Since the emergence of full-authority electronic en- equivalent during the evaluation lends itself to some gine control allows the designer some flexibility, the pilot extrapolation. Such extrapolation differs for each task must allow definition of the satisfactory (level 1) pilot because of variations in training, knowledge, ex- throttle response characteristics. The MTE selected, perience, and ability to assess beyond the specific task.
therefore, involved precision wing station keeping during small and unannounced step throttle changes The flight subphase chosen after preliminary evalua- tions of several formation tasks was close-wing station by the formation leader. F_rther, the evaluation pilot Pilot ratings of one-half (for example, Pl{ 4.5) indi- was to maintain relative position at all costs. In short, cate an indecision or reluctance to assign either of the the task could not be abandoned when the pilot's per- adjacent ratings to describe the throttle configuration.
formance degraded. The initial transient of the lead Ratings of 3.5, 6.5, and 9.5 are generally not used be- airplane did not need to be followed exactly, but the cause they represent important boundary conditions.
new position was to be recaptured quickly and accu- The boundary conditions are based on yes or no de- rately. The real-world equivalent would be an emer- cisions to obtain satisfactory, adequate, or inadequate gency air-to-air refueling or formation recovery where (level 1, 2, or 3) handling qualities.
it is imperative to stay in position despite turbulence or inadvertent thrust changes by the leader.
Pilot comments were recorded during the evaluations to avoid having the IIQR represent the entire quali- Rating Criteria tative assessment. Pilot comments are one means of Pilot rating is a measure of the handling performance identifying good or deficient qualities of a configura- of an airplane, that is, pilot and vehicle performance.
tion. During an evaluation, the pilots reported what Cooper and Harper define performance as "the preci- they saw and felt and described their difficulties in com- sion of control with respect to airplane movement that pleting the MTE. Such comments were in response to a pilot is able to achieve in performing a task. "4 For this a questionnaire developed for the MTE. All evaluation flight evaluation, the pilot ratings were applied to the pilots participated in preparing the questionnaire. This MTE as previously described and were based on how questionnaire helped ensure thal important or suspect quickly and accurately the evaluation pilot acquired aspects were considered, the reason for the assigned and maintained formation position. The precision of rating was given, the tradeoffs that the pilot must con- control required by the pilot to maintain formation was tend with were understood, and any supplementary an approximately 5-ft wingtip clearance in the horizon- comments that help describe the pilots' evaluation of tal plane.
the configuration were provided.
Figure 7 shows an example of the precision and Flight Test Procedures performance requirements for obtaining adequate pi- The backseat pilot acted as the test conductor and lot performance for the MTE. The following criteria safety pilot for the research flights. Throttle config- were applied for obtaining the HQI{: urations were defined for evaluation according to the mission plan before each test flight. These configura- • Satisfactory performance was +1 ft with one over- tions were written on flight cards for the backseat pilot shoot allowed.
to use in selecting the TCU settings. This pilot could disengage the variable-throttle response system at any • Adequate performance was -t-3 ft with two over- time and use the unmodified rear cockpit throttle. In shoots allowed.
an emergency, this pilot could also overpower the throt- tle system servomotor to command the desired throttle Another important consideration for determining inputs.
HQR was pilot workload. Cooper and Harper define The evaluation pilots were unaware of the TCU con- pilot workload as "the integrated physical and mental figuration during the MTE evaluations. This method- effort required to perform a specified piloting task. "4 ology was used to avoid pilot preconceptions of what In this case, physical effort is the motions and forces effect such configurations as throttle time delay would imposed on the throttle by the pilot during the MTE.
have on airplane handling qualities.
No auxiliary tasks were required during the MTE so Each throttle configuration was evaluated in two that only thrust control characteristics were evaluated.
stages. The set up of each test configuration was made Additional workload imposed by auxiliary tasks would with significant distance between the test and lead air- have interfered with the pilots' evaluation of tile thrust planes. After entering the test configuration, the evalu- control characteristics.
ation pilot maneuvered into a close-formation position Pilot Assessments with the lead airplane. The pilot evaluated the throttle control system during the join-up and formation phases Cooper and Harper note that the pilot rating is a of the maneuver until satisfied that a valid llQl{ could "shorthand representation of the handling qualities of be made. Then upon command from the evaluation an airplane in the performance of a defined mission and pilot, the lead airplane began a series of small, random task. ''4 The evaluation pilot continuously considered throttle excursions called throttle jinks. These throt- the rating decision process to obtain the pilot rating tle jinks increased or decreased the separation distance during the MTE. This process involves a series of di- and the rate of separation. The evaluation pilot's tasks chotomous decisions based on the adjective descriptors of the pilot rating scale (Fig. 6).
Test aircraft -_ two ov3rsft_loots_ _ 5"ft clealnce 920646 Fig. 7. Precision and performance requirements for adequate pilot performance.
were to aggressively reacquire and maintain the initial which the task was performed, and evaluate the pilots' separation and to evaluate the task as previously de- assessment of the task. Only data from four of the pi- scribed. This pilot provided comments pertinent to the lots are presented in this paper because the data from questionnaire during the evaluation.
the fifth pilot does not apply to the results discussed.
During the flight test program, 113 data points con- The evaluation pilot briefly returned to the so-called sisting of pilot ratings and comments were gathered.
benchmark configuration between each MTE. Note that the evaluation pilot knew that this was the bench- The primary goals of the flight test program were mark configuration. This configuration was essentially to investigate the thrust control sensitivity effects of the basic J79-11B engine (that is, the TCU in linear throttle gain, time delay, and rate limiting on airplane mode) modified by the throttle servodynamics, opti- handling qualities. In addition, the insertion of the mum throttle control gain, and change in throttle fric- electronic throttle control system changed the baseline tion. Evaluation pilots had the option of re-flying the throttle system dynamics. Such changes needed to be task in the benchmark configuration to ensure quality quantified. Secondary goals were to investigate the ef- control of each evaluation. Afterward, the pilots could fects of first-order lag filter time constants as well as change the HQR, but they rarely made such changes.
lead-lag filter time constants. Although the primary Since l0 to 14 evaluations were performed during each goals were met, insufficient data were available to draw flight, this benchmark procedure helped the pilots to conclusions about the secondary goals.
retain a good reference standard.
Time history data are presented for the representa- Upon completion of the task, control was returned to tive pilot evaluation tasks. In addition, a compilation the backseat pilot. Then, the evaluation pilot assigned of pilot comments with respect to the time history data is shown. These data show the differences between sat- an HQR and provided more comments regarding the performance of the configuration.
isfactory, adequate, and inadequate (level 1, 2, and 3) handling qualities (Fig. 6). Data showing time delay Flight Test Results and Discussion and rate limit thresholds for the different levels of han- dling qualities are also presented. The modified air- Five pilots flew 10 flights to evaluate the electronic plane response with the electronic throttle control sys- throttle control system. Two of the flights were tem as well as the effects of throttle gain, time delay, flown to check the variable-throttle response system rate limiting, lag filter, and lead-lag filter on engine functionally, define the MTE, critique the manner in handling qualities are discussed next.
Throttle Gain Effect
Airplane Response With Electronic Throttle
Control System The throttle gain was initially evaluated to select the optimum gain value. The optimum throttle command The dynamics of the electronic throttle control sys- gain for this electronic throttle control system was de- tem (Fig. 3) were estimated by applying frequency re- termined to be 1.5. This command gain closely resem- sponse analysis to flight test data. Determining what bled the baseline response of the standard TF-104G added effect the throttle servomotor had on system re- airplane. This value was then used as the benchmark sponse was important. The throttle servoloop repre- configuration and in the majority of the evaluation con- sents a feedback control system. Although its com- figurations. The following table shows the number of ponents were known, the equivalent input and output evaluations and the average of all the HQR's for the dynamics were best estimated from flight test data.
given command gain: An estimate of the airplane acceleration response to throttle inputs, including the servomotor, was reason- Gain setting No. of evaluations HQR ably modeled as a first-order lag with a pure time delay 1.0 5 4.60 within the frequency range of interest to the task. The 1.5" 15 2.65 baseline break frequency and time delay which applied 1.7 1 4.00 to the benchmark configuration were then estimated 2.0 4 4.25 as first-order break frequency = 5.7 rad/sec = 0.91 Hz, and time delay = 65 msec. The 65-msec time delay of *This gain equates to 0.008 g/deg of throttle motion the throttle system is included in the time delay stated for the nominal TF-104C, gross weight of 18,500 lb for the following data analysis discussions.
(=t=2,500 lb variation during a flight).
Throttle friction force was maintained at approxi- Throttle Time Delay Effect mately 2 lb for the experiment. This value was some- what light when compared with approximately 5 lb of The effects of additional time delay are summarized force for the standard TF-104G throttle system. Pilots in Fig. 8 showing HQR plotted as a function of time commented about the lack of friction in the evaluation delay. The data at 65-msee time delay represent the throttle throughout the flight test program, hut they did not think it was a factor.
Pilot O A DB AC
<>o
-_ Average for
° o
time delay
,eve, 2
HQR Level 1 I I I t I I I 300 400 500 600 700 800 900 0 100 2OO Time delay, msec _2o1_ Fig. 8. Effect of throttle control system time delay on pilot rating for command gain = 1.5 and rate limit -- =[=99 deg/sec.
benchmark configuration. This65-msec timedelay ac- Several cases were evaluated where the throttle gain
countsfor the throttle servomotor andthe airplane
was changed to determine if the handling quality ef- fects of a time delay could be minimized. No trends
whichwaspreviously estimated.Ratelimit wasset
could be established for the limited cases studied. Ap- for +99 deg/sec for all time delay evaluations.
parent]y, time delay problems could not be solved by The data trend shows a steep degradation of tlQR suitable command gain changes. Note, however, that a with added time delay. A +1 HQR band was super- throttle command gain of 1.5 was chosen as optimum imposed on the data. Approximately 90 percent of the with an inherent time delay of 65 msec and a hysteresis data falls within this band. Significant thrust pilot- loop. All pilots commented that except for the lack of induced oscillations (PIO's) occurred during the evalu- friction, this configuration responded similarly to the ation task for time delay values greater than 250 msec.
standard TF- 104G airplane.
Data from this experiment suggest the following time delay thresholds: Figure 9 shows time histories and piloting comments representing satisfactory (level 1) handling qualities for • Level 1 time delay < t00 msec (HQR _< 3.5).
a time delay of 65 msec. The parameters are throttle command, throttle feedback, compressor speed, and • Level 2 time delay < 300 msec (3.5 < HQR 6.5).
.4 AX, I g L__L ...... l 1 i i 1 l _o_ ,___l w.4
[
N1, ?000 rpm 80 F r PLAFB, ?
deg 40 i 8O PLACMD, r- deg 40 I J. 1__ _ • _ __l_ _____l _ 1 __J 0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45 50 Time, sec (a) Time histories of pilot A, HQR = 3.0. (b) Time histories of pilot B, HQR = 2.0.
Pilot Comments HQR A 3.0 I guess there was no time delay. Performance was as desired. Most of the time, I was within a foot and seldom had an overshoot.
B 2.0 Good response for the throttle inputs I put in. Easy to fine tune final position with small inputs. When I follow the lead airplane's thrust changes, it takes me a cycle or two to get in phase with it. That's probably more my technique than anything.
C 3.0 Very responsive. Able to track changes very well, however, there is a little bit of delay in the initial response compared to perfection. The initial re- sponse requires a little compensation, like a quick overdrive of the throttle to catch the rate and move into position.
D 2.0 Good relationship between throttle angle and rpm change. Not much ap- parent hysteresis. Would prefer more friction in the throttle -- I am used to working against the friction with the standard F-104 airplane.
920195 Fig. 9.
Level 1 handling qualities for time delay = 65 msec and rate limit = +99 deg/sec.
longitudinal acceleration. Note the small amplitude compensating for the delay. The second time, pilot A and low frequency of the throttle command. Pilots required extensive compensation to complete the task.
commented that the throttle was very responsive and Pilot B reported adequate performance but only with that it was easy to control fore and aft position. In extensive pilot compensation. This pilot could not be addition, they noted a good relationship between the precise about the task, especially when the lead air- throttle angle and engine rpm. plane changed thrust.
Figure 10 shows time history data representing ad- Figure 11 shows time history data for inadequate equate (level 2) handling qualities for a time delay (level 3) handling qualities for a time delay of 465 msec.
of 225 msec. The parameters are the same as in The amplitude of the throttle command increased, Fig. 9, Note the amplitude and frequency increase of while the frequency decreased. Both pilots moved the the throttle command. Pilot A recognized the time throttle around a great deal and were unsure how to delay, and pilot B interpreted the configuration as compensate for this configuration. Performance was a lag. Pilot A evaluated the 225-msec configuration inadequate. Pilot A's airplane oscillated more than :/:3 ft with more than two overshoots. Pilot B moved twice. The first time, pilot A did not have ditficulty .4 AX, 0 g __ i . _ L .__ .... • _i i • _ L_ a -- d [ .__ 1 t _ i _ L • _- 1 *__ * --.4 N1, i rpm ___t __ t t t t * J J PLAFB, deg 40 i 8O PLACMD, deg 40 L.. I 1 t 1 _ t. l t I 0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45 50 sec Time, (a) Time histories of pilot A, HQR = 6.0.
(b) Time histories of pilot B, HQR = 6.0.
Pilot Comments HQR A 4.0 (a) A small delay occurs, but I'm not having any difficulty compensating.
Gross acquisition is great; fine tracking is not all that bad. Desired performance achieved with moderate compensation.
A 6.0 (b) I think there's a time delay here, and this delay causes a great deal of pilot compensation to do an adequate job. There may be some rate deficiencies too.
I was working just about as hard as I could and still just barely hung in there.
6,0 Seems like there's some lag there. I don't detect it so much by the vibration and the sounds -- the audio cues that I think I use in flying formation and a feel for power changes. Seems like I put the throttle in a position to respond, and then a little bit later the airplane moves and seems to catch up. Then I'm backing the throttle in the other direction. I'm not in a classic PIO, but I seem to be just moving back and forth slowly. 1 can't be precise about the task, especially when the lead aircraft is changing thrust. I guess I can adequately perform but only with extensive pilot compensation.
920196 Fig. 10. Level 2 handling qualities for time delay = 225 msec and rate limit = ±99 deg/sec.
.4 AX, g --.4 i m_ N1, rpm I I I I I l I 1 J I i J i i i _L _ c ___ . L --I PLAFB, deg 1 I I I 1 I L I____ • 0 - t i i L__ _L _ _ J PLACMD, deg I I Z L I i L _ J___J 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45 50 Time, sec (a) Time histories of pilot A, HQR = 7.0.
(b) Time histories of pilot B, HQR = 10.0.
Pilot HQR Comments A 7.0 Working the throttle, experiencing some throtOe overshoots, but don't see much airplane motion fore and aft. Don't understand this -- isn't what I thought a time delay would look like. Performance was not adequate -- greater than +3 ft with more than two overshoots.
B 10.0 At first, I didn't think it was all that bad, but "after the lead aircraft started making throttle changes, I broke loose. I'm just oscillating, moving the throttle all around.
Once I get in position, I'm not sure what's going on. Seems like a big lag -- moving fore and aft on the lead airplane a lot more than I'd be comfortable with.
g_o1g7 Fig. 11. Level 3 handling qualities for time delay -- 465 msec and rate limit = +99 deg/sec.
fore and aft on the lead airplane to the point of be- produced minimal change in engine handling qual- coming uncomfortable. Inability to stay with a tanker ities. These data indicate a sharp degradation in caused pilot B to rate this configuration as a 10.0. handling qualities with rate limit values of less than +40 deg/sec. The HQR's at 4-40 deg/sec ranged from Throttle Rate Limiting Effect 2 to 6, thereby suggesting an initial handling quality To investigate the effect of throttle system rate limit degradation. Significant PIO occurred for rate limits on engine handling qualities, the rate at which the less than +25 deg/sec. Suggested threshold rate limit values are as follows: throttle could be commanded was progressively re- duced from that of the benchmark configuration. The benchmark rate limit was +99 deg/sec, and the lowest • Level 1 rate limit > -t-40 deg/sec (HQR < 3.5).
limit tested was +10 deg/sec. For example, throttle • Level 2 rate limit > +30 deg/sec (3.5 < HQR < position can change 10 ° in 1 sec for a step input with 6.5).
a +10-deg/sec rate limit. The commanded longitudi- nal acceleration in this example would be 0.008 g/see Figure 13 shows time history data representing ad- for the first, second at the nominal TF-104G weight of equate (level 2) handling qualities for a rate limit of 18,500 lb and throttle gain of 1.5.
-t-40 deg/sec. The parameters are the same as in Figs. 9 Figure 12 shows the results of the rate limit eval- to 11. Note that the amplitude and frequency of the uations. The HQR is plotted as a function of throt- throttle command is approximately the same as ade- tle control system rate limit. The data point at -t-99- quate (level 2) handling qualities with additional time deg/sec rate limit is for the benchmark configuration.
delay (Fig. 11). Pilot A evaluated the -I-40-deg/sec con- Note that decreasing throttle rate limits to +45 deg/sec figuration twice. The first time, pilot A did not have Level 3 O A B _ C D - -_- Average for time delay - _ Level 2 HQR
- o
i
Level 1 I I l I I I I I I I 0 10 20 30 40 50 60 70 80 90 100 Rate limit, deg/sec 92o_98 Fig. 12. Effect of throttle rate limiting on pilot ratings for throttle time delay -- 65 msec and command gain = 1.5.
effects. There was no question that a tanker would difficulty compensating for the rate limit and rated the have been damaged, so an HQR of 9.0 was given.
task as a 2.5. The second time, pilot A interpreted the rate limit as reduced gearing (Fig. 13(a)). Moderate Asymmetric rate limits were tested in addition to pilot compensation was required to obtain desired per- the symmetric rate limits. The lower rate dominated formance, so an HQR of 4.0 was given. Pilot C had the evaluation in each case. Pilot ratings and com- difficulty anticipating the amount of throttle motion ments were comparable to those given for symmetric required when the lead airplane made power changes rate limits. For example, a 99 and -20-deg/sec rate (Fig. 13(b)). Controllability was not in question, so an limit resulted in inadequate (level 3) handling qualities.
HQR of 5.5 was given based on pilot workload. Pilot By combining the pilot ratings for time delay and D said that at times the engine response was pretty rate limit (Figs. 8 and 12), boundaries were approxi- decent but otherwise there was a tendency to overcon- trol (Fig. 13(c)). The final response was unpredictable; mated to describe level 1, 2, and 3 handling qualities.
however, the airplane was controllable with adequate Figure 15 shows the level of handling quality bound- performance, so an HQR of 6.0 was given. aries for rate limit versus time delay. The boundary for level 1 handling qualities is well defined such that Figure 14 shows time history data of inadequate throttle time delay should be less than 100 msec, and (level 3) handling qualities for a rate limit of 20 deg/sec.
rate limit should be greater than 40 deg/sec. The outer The amplitude of the throttle command increased sig- boundaries between levels 2 and 3 and between level 3 nificantly, often from throttle stop to throttle stop and uncontrollable are not well defined and vary greatly while the frequency decreased. Pilot B experienced between pilots. As a result, the outer boundaries shown a PIO tendency and moved the throttle from stop to in Fig. 15 represent rough estimates.
stop in attempts to get the required engine response Lag Filter Effect (Fig. 14(a)). Because a pilot would not be able to stay on a tanker boom in such cases, an HQR of 10.0 was The effects of first-order lag filters on the throttle given. Pilot C could not achieve adequate performance response handling qualities were evaluated. Unfortu- and thought that someone could be hurt, so an tIQR of nately, an insufficient number of lag time constants 9.0 was given (Fig. 14(b)). Pilot D commented that the were tested to investigate a suitable range of handling initial response was slow (Fig. 14(c)), noted that there qualities.
was some time delay, and was unable to obtain original Figure 16 shows the effect of first-order lag filters on formation position. Pilot D was never able to position pilot ratings. Except for one test point, the data re- the throttle to obtain the desired response. The PIO sulted in inadequate (level a) handling qualities. The was in the form of moving the throttle with no apparent .4 AX, g _.4 80O0 N1, rpm 6000 I I 1 i L I L L L 1 !
PLAFB, deg PLACMD, deg .I t I I I i 1 I 1 1 0 5 10 15 20 25 30 35 40 45 50 Time, sec _oI_ (a) Time histories of pilot A, HQR = 4.0.
.4 AX, g --.4 N1, rpm I I I I l I I J J 8O PLAFB, deg PLACMD, 4O deg 0 5 10 15 20 25 30 35 40 45 50 Time, sec _o2oo (b) Time histories of pilot C, HQR = 5.5.
Fig. 13. Level 2 handling qualities for rate limit = +40 deg/sec and time delay = 65 msec.
same 4-1 HQR band obtained from the time delay eval- to the degradation from increasing time delay. Various uations (Fig. 8) was superimposed on the first-order degrees of PIO occurred in each of the inadequate (level lag evaluations. The degradation in handling qualities 3) cases. The character of the amplitude and frequency of the PIO tended to be similar to that of the +20- from increasing the lag time constant appears similar deg/sec rate limit evaluations.
.4 AX, g 1 I 1 I I I 1 I I J N1, rpm ..... I ....... 1 J 8O PLAFB, deg _1 __ I I 1 I I I L J 8O F PLACMD, 4O deg • .... I I I I I I I I I 5 10 15 20 25 30 35 40 45 50 Time, sec 920201 (c) Time histories of pilot D, HQR = 6.0.
Pilot HQR Comments A 4.0 The throttle feels like the gearing is reduced. More throttle motion is needed for a given thrust change. Took some compensation to get enough thrust to move fore and aft. An HQR of 4 was given based on moderate pilot compensation to get desired performance.
C 5.5 I'm falling into the typical trap of trying to figure out what's going on and have no idea, so I'll admit it. The confusing thing was that there were periods where I had pretty decent performance and then some periods of overcontrol that left me less than happy with the predictability of the basic airplane. So the performance overall was adequate with a reasonable amount of throttle motion. With the initial response, it felt like I was overcontrolling it. At times, I was getting it moving when I wanted it, but generally the final response was not as predictable as I would like. The airplane was controllable with adequate performance.
D 6.0 A little bit of trouble anticipating how much power to take off and when to put it back on as the lead aircraft made changes. The throttle had more motion than I would have liked. Initial and final thrust responses were O.K. I might have been overdriving the throttle a little bit to get the amount of thrust that I wanted out of it in both directions. Controllability was not in question, had adequate performance.
Steady state station keeping was no problem. An HQR of 6 was given based on pilot workload. [This pilot initially rated this point as a 5, then changed it to 6.]
920202 Fig. 13, Concluded.
AX, g -.4 _-----_l__L _ k___ 1 L__J_ j N1, rpm 6OO0 J----L__l 8O PLAFB, deg k__l PLACMD, deg --i--J-----l_____l__.J_ j 0 5 10 15 20 25 30 35 40 45 50 Time, sec 9_O203 (a) Time histories of pilot B, HQR = 10.0.
AX, g -.4 ' J _--/--L___L j I I I J N1, 70O0 rpm 6000 _ , , , __j PLAFB, 4O deg 0 I _J I I U [ I I J PLACMD, deg --J-__.l I I I --. I I ---J__l J 0 5 10 15 20 25 30 35 40 45 50 Time, sec _oao4 (b) Time histories of pilot C, HQR = 9.0.
Fig. 14.
Level 3 handling qualities for rate limit = 20 deg/sec and time delay = 65 msec.
.4 AX, g --.4 N1, rpm .... I I I I L I PLAFB, 4O deg I I I I I I t I I J 8O PLACMD, 4O deg I • I I I I L_ __ I I _.__ 0 5 10 15 20 25 30 35 40 45 50 Time, sec 920205 (c) Time histories of pilot D, HQR = 9.0.
Pilot HQR Comments B 10.0 Even for the steady state station keeping I'm making pretty big inputs and lots of motion and not getting the response I expected. There was a PIO tendency. With the lead aircraft changing thrust, I'm just stop-to-stop with the throttle. I have to give that an HQR of 10. If you were on a boom, you'd be off it pretty quick. You'd have a hard time staying with the guy in any formation flying.
C 9.0 The boomer's eyes are very big! Station keeping was really no problem without the throttle jinks. But it seemed like there was 1/2 in. to 3/4 in. of absolute deadband in the throttle. So, I could not achieve desired performance, not even close. And, I think I could hurt somebody. So we're talking 8, 9, or 10 again. It's not clear that I'd have to abandon the task, but it's very close to that.
D 9.0 I had a lot of throttle motion anytime I wanted to do something. The initial thrust response was pretty slow. It appeared to be some sort of a time delay or something that exhibited time delay type characteristics. In the final response, I was never able to nail down where I wanted to leave the throttle. There were no PIO tendencies because I never seemed to get in any one place at any one time. The only PIO was me moving the throttle with no apparent effects. The chase pilot and I had a little discussion about whether that one would have damaged the tanker, and there's no question that it would have. _o2o6 Fig. 14. Concluded.
1.0 .9 Uncontrollable .8 .7 -- ::i::ii_ii!i:!:iii:iiiii:iiiiii?_!_i?_i?_i_ii!;_i?_i!ii:_i!iii_i!iii_i!i_!iii!_iii;i:ii_!_iii_i??_i_$i$_i_:?
- ::S .6 .:::::: Time delay, .5 - .S _:': Level 3 sec .4 - .:,:j" ::#," .3 Sii? _l!j_iiiiiil Level 2 i/iiii/i/i/iiii/iiiiiiiiiiiiiiiii .2 ::::::::::::::::::::::::::::::::::::::::::::::: ::::::::::::::::::::::::::::::::::::::::::::::::::::::: : .1 I :' I .4::x_:::: J2::::.._::._::::::::::::::::2::::::::::::_:'_::_._:_:i_i_$_!:!:!:!:!:!:!_!:_:!:!::::::::::::::::::_:::_::::::::::_: 0 10 20 30 40 50 60 70 80 90 100 Rate limit, deg/sec 920647 Fig. 15. Level 1, 2, and 3 engine handling quality boundaries for time delay and rate limit.
10 A Pilot A B Level 3 [] C A O Trend from time 7 O delay results from Fig. 8 HQR Level 2 Level 1 I I I 1 I I I I 0 .4 .8 1.2 1.6 2.0 2.4 2.8 3.2 3.6 Lag time constant, sec _o_o_ Fig. 16. Effect of first-order lag filter on pilot rating for throttle time delay = 65 msec, command gain = 1.5, and rate limit = +99 deg/sec.
Figure17shows timehistorydataandpilotC com- 0.25 to 0.20 cycles/sec. The pilot could not control the ments for a lag time constant of 0.3sec. '['his time engine response and was from stop to stop with the constant represents the only adequate (level 2) han- throttle (Fig. 18(b)) for the 1.0-see case. As a result, dling qualities data point. The effects of the first-order an HQR of 10.0 was given.
lag filter are similar in throttle motion and magnitude Figure 19 shows two sets of time histories and pi- to the pure time delay effects (Fig. 10(a)). Pilot C com- lot comments for lag time constants of 1.0 and 3.0 see mented that station keeping was not too bad, but it got which describe inadequate (level 3) handling qualities worse when attempting to match throttle changes by for pilot A. Note the similarity between the throttle the lead airplane. This pilot said that the throttle felt motions of pilots A (Fig. 19(a)) and C (Fig. 18(b)) for as though he/she were PlO'ing the throttle. Extensive a 1.0-see lag time constant. Figure 19(a) also shows pilot compensation was required; therefore, an tIQR of that pilot A moved the throttle from stop to stop. Pilot 6.0 was given.
comments were also similar. Pilot, A noted large thrust delays and overshoots in both directions and was un- Figure 18 shows two sets of time histories of inad- able to obtain adequate performance. Note that pilot equate (level 3) handling qualities and pilot C com- A rated this configuration a 7.0 compared to a 10.0 for ments. Comparing Figs. 17 and 18 shows the effect pilot C. Figure 19(b) shows a 3.0-see lag time constant of increasing the lag time constant from 0.3 through 0.6 to 1.0 see. Increasing the lag time constant sig- time history for pilot A. The frequency of throttle mo- tion decreased significantly from 0.2 to between 0.1 and nificantly increased the throttle motion by the pilot.
The pilot noted Pie and much throttle input for the 0.05 cycles/see. Pilot A commented that the rate was so slow that it was impossible to stabilize the airplane 0.6- and 1.0-see lag time constant configurations. Ob- in formation and rated this configuration a 7.5.
serve that the frequency decreased from approximately .4 AX, g I ..... _ L L.__ • i I I --.4 N1, rpm t ___1 __ h 1_ I I L • [ I J 8O F PLAFB, 4O deg L l L._ _ k L L i i _ 1 F PLACMD, deg I 1 I I L. L ___ L _J J 5 10 15 20 25 30 35 40 45 50 Time, sec Station keeping isn't too terribly bad, but I'm putting in an awful lot of inputs in the throttle, and pretty big inputs too. I'm slowly going forward and aft with the throttle. I'm not seeing a classic Pie, but I am PIO'ing the throttle.
Things get worse when trying to match throttle changes of the lead aircraft.
Deficiencies certainly warrant improvement. Extensive pilot compensation is required. An HQR of 6 was given.
920208 Fig. 1T. Time histories and pilot C comments describing level 2 handling qualities for lag time constant = 0.3 sec, command gain = 1.5, time delay = 65 msee, and rate limit = :t:99 deg/sec.
.4 AX, g _ J. 2_ J J 2 J .L J _- J I 1 I I I 1 ] I I --,4 N1,
r
_m I I I I I I I 1 I I I I L I I I I I l ± • 600O 8O PLAFB, deg PLACMD, deg 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45 50 Time, sec (a) Lag time constant = 0.6 sec, HQR = 8.0. (b) Lag time constant = 1.0 sec, HQR = 10.0.
Lag time HQR Comments constant, sec 0.6 8.0 I don't know what the problem is. But every time I get in a stabilized station keeping position, I see some acceleration or deceleration. It just falls apart on me. It definitely needs improvement when trying to track throttle changes of the lead aircraft. I was really oscillating and PIO'ing and lots of throttle input.
I consider this to be a major deficiency; therefore, and HQR of 8 is given.
1.0 10.0 I really can't control it. I'm about 50 ft away and almost stop-to-stop with the throttle now. I'm really trying to slow my gain down in manipulating the throttle, but I'm still seeing a PIO. I have to give that a 10. I don't think that if I'd been out over the Atlantic, and I needed fuel, I'd have been able to do that safely.
rm just seeing the throttle go stop-to-stop and that's completely unsatisfactory.
920209 Fig. 18. Time histories and pilot C comments describing the effects of increasing lag time constants for level 3 handling qualities for command gain = 1.5, time delay = 65 msec, and rate limit = 4-99 deg/sec.
evaluated. Superimposed on this figure is the 4-1 HQR Lead-Lag Filter Effect band from Fig. 8. Unfortunately, the database for this Handling quality evaluations were conducted to experiment was insufficient for use in drawing conclu- study the effects of appropriate first-order lead-lag fil- sions. The data remained within the 4-1 HQR band.
ters in combination with additional time delay. This The HQR trends suggested some improvement with study was to determine if suitable lead compensation the lead compensation for the 165-msec cases, but the could offset the handling qualities degradation of throt- few evaluations flown showed inconsistencies between tle system time delay.
some HQR's and pilot comments. The overall ratings remained essentially unchanged with the added lead Figure 20 shows the data obtained using the lead-lag compensation filters for the larger 265-msec cascs.
filter to compensate for additional time delay. Con- figurations with 165- and 265-msec time delay were 2O .4 AX, g * l t l t e L I l J --.4 N1, rpm i i | t t i " i i i i L _ I t i i i • i i J r PLAFB,
deg \_j \/
L__ . • • L J.. i I t t i i PLACMD, deg i L 1 t i t l i i i [ i 0 5 10 15 20 25 30 35 40 45 50 0 5 10 15 20 25 30 35 40 45 50 Time, see (b) Lag time constant = 3.0 sex:, HQR = 7.5.
(a) Lag time constant = 1.0 sex:, HQR = 7.0.
Comments Lag time HQR constant, SCC 1.0 7.0 There was a big delay and big overshoots in both directions. 1 couldn't do an adequate job. Even with all that I could bring to bear on it, I couldn't get adequate performance.
3.0 7.5 I'm not even in formation but get the impression that the rate is very slow.
Particularly when you add throttle, the rpm doesn't change very fast. I wish you could see this on TV. This is wild. The rpm rate is so slow that it's just impossible to stabilize this in formation. 92o2_o Fig. 19. Time histories and pilot A comments describing the eft'cots of increasing lag time constants for level 3 handling qualities for command gain = 1.5, time delay = 65 msec, and rate limit = ±99 deg/see.
I Pilot [] B A C D [] None • 30/20 [] 40/20 [] 50/20 HQR 5 • 20/10 2 Level 1 I I I I I I I I 0 .1 .2 .3 .4 .5 .6 .7 .8 .9 Time delay, sec 920211 Fig. 20. Effect of lead-lag compensation for additional time delay on pilot ratings for command gain = 1.5, time delay =: 65 msec, and rate limit -- -t-99 deg/sec.
Concluding Remarks tlandling quality ratings at -1-40 deg/sec ranged from 2 to 6, thereby suggesting initial handling quality degra- A flight evaluation was conducted at the Dryden dation.
Flight Research Facility to investigate the effects of varying engine throttle response on airplane handling An insuflicient number of lag time constants were qualities. An electronic throttle control system in a evaluated for determining satisfactory (level 1) or ade- TF-104G airplane was used to evaluate the effects of quate (level 2) airplane handling qualities. Except for throttle time delay, rate limit, first-order lag time con- one test point, the data resulted in inadequate (level 3) stant, and lead-lag time constant. The Cooper and handling qualities. When the same 4-1 handling quality Harper Pilot Rating Scale was used to describe satis- rating band obtained from the time delay evaluations factory, adequate, and inadequate (level 1, 2, and 3) was applied to the lag time constant evaluations, the handling qualities. A precise longitudinal, close-wing degradation in handling qualities from increasing the station keeping task was flown to simulate such oper- lag time constant appeared similar to the degradation ations as air-to-air reflmling and close-formation flight from increasing time delay.
under adverse instrument conditions. The formation The attempt to identify the effect of adding lead- leader introduced small and unannounced step throt- lag time constants with additional time delay to deter- tle changes to challenge the evaluation pilot. These mine if lead compensation could offset handling quality changes provided the desired degree of handling qual- degradations was inconclusive. Data for the lead com- ities discrimination. Highlights of tile result are sum- pensation configuration remained within the 4-1 rating marized next.
band of the uncomperlsated configuration data.
For level 1 handling qualities, throttle time de- References lay should be less than 100 msee; and rate limit should be greater than 40 deg/sec. Thrust pilot- l Burcham, l_ank W., Jr., Lawrence P. Myers, induced oscillations occurred for time delay values and John R. Zeller, Flight Evaluation of Modifica- greater than 250 msee and rate limits less than 25 tions to a Digital Electronic Engine Control System in deg/see. A steep degradation of handling quality an F-15 Airplane, NASA TM-83088, 1983. (Also avail- rating occurred with added time delay. Decreas- able as AIAA-83-0537, Jan. 1983.)
ing throttle rate limits to +45 deg/sec had min- imal effect in changing the engine handling quali- 2Myers, Lawrence P. and Frank W. Burcham, Jr., ties. A sharp degradation in handling qualities oc- Propulsion Control Experience Used in the Highly curred with rate limit values less than +40 deg/sec.
Integrated Digital Electronic ContTvl (ItlDEC) Pro- SNeal, Bradford and Upal Sengupta, 7"he haple- gram, NASA TM-85914, 1984. mentation and Operation of a Variable-Response Elec- tronic Throttle Control System for a TF- 10d G Aircraft, aWalker, Laurence A., "ltarrier II---Digital Engine NASA TM-101696, 1989.
Control Flight Tests," 30lh Symposium Proceedings, 6Hoh, Roger It., "Unifying Concepts for ttandling Society of Experimental Test Pilots, Sept. 24 27, 1986, pp. 49-70. Qualities Criteria," AIAA-88-4328, Aug. 1988.
rU.S. Air Force, Militar?j Specification Flying Qual- 4Cooper, George E. and Robert P. Harper, .lr., 7'he Use of Pilot Rating in the Evaluation of Aircraft Han- ities of Piloted Airplanes, MIL-F-8785C, Nov. 1980.
dling Qualities, NASA TN I)-5153, 1969.
Form Approved REPORT DOCUMENTATION PAGE OM8 No ozo4-o788 PLIbhc teDorling burden ]or Ibis COlleCtiOn ol information is eslimaled 1o average 1 hour par response. _ncludnng the lime for reviewing mslruCtlons, searching ex=shng dais sources, galharing arid mair_lainin_ tha data needed, and compleIing and reviaw_ng the collection o1 informatiorL Sefld commerits regatdirig lhis burden eslimats or any Other aspecl of Ih_s COIkDction of informatnon, including suggestions for reducing this burden, to Washington Headcluarlers Services, Directorate for intormahon Operal=ons and Raporls, 1215 Jellerson Davit Highway, Suile 1204. Arlinglon, VA 22202-4302, and 1othe Office of Management and Budgel, Paperwork Reduction Pro act (0704-0188). Washmglon, DC 20503 1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED March 1993 Technical Memorandum 4. TITLE AND SUBTITLE 5. FUNDING NUMBERS Summary of the Effects of Engine Throttle Response on Airplane Formation-Flying Qualities WU 307-05-01 6. AUTHOR(S) Kevin R. Walsh 8. PERFORMING ORGANIZATION 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) REPORT NUMBER NASA Dryden Flight Research Facility P.O. Box 273 H-1880 Edwards, Califomia 93523-0273 10. SPONSORING/MONITORING 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) AGENCY REPORT NUMBER National Aeronautics and Space Administration NASA TM-4465 Washington, DC 20546-0001 11. SUPPLEMENTARY NOTES Presented as AIAA-92-3318 at the AIAA 28th Joint Propulsion Conference and Exhibit, July 6-8. 1992, Nashville, Tennessee.
12b. DISTRIBUTION CODE 12a. DISTRIBUTION/AVAILABILITY STATEMENT Unclassified -- Unlimited Subject Category 08 13. ABSTRACT (Maximum 200 words) A flight evaluation was conducted to determine the effect of engine throttle response characteristics on precision formation-flying qualities. A variable electronic throttle control system was developed and flight-tested on a TF-104G airplane with a J79-11B engine at the NASA Dryden Flight Research Facility. This airplane was chosen because of its known, very favorable thrust response characteristics.
Ten research flights were flown to evaluate the effects of throttle gain, time delay, and fuel control rate limiting on engine handling qualities during a demanding precision wing formation task. Handling quality effects of lag filters and lead compensation time delays were also evaluated. The Cooper and Harper Pilot Rating Scale was used to assign levels of handling quality. Data from pilot ratings and comments indicate that throttle control system time delays and rate limits cause significant degradations in handling qualities. Threshold values for satisfactory (level 1) and adequate (level 2) handling qualities of these key variables are presented. These results may provide engine manufacturers with guidelines to assure satisfactory handling qualities in future engine designs.
15. NUMBER OF PAGES 14. SUBJECT TERMS Electronic engine control system; Engine throttle response: Formation-l'iying 16. PRICE CODE qualities; Handling qualities criteria: Throttle variables A03 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT 17. SECURITY CLASSIFICATION 18 SECURITY CLASSIFICATION OF ABSTRACT OF REPORT OF THIS PAGE Unclassified Unclassified Unclassified Unlimited NSN 7540-01-280-5500 Standard Form 298 (Rev 2.89_, Prescribed by ANS_ Sto Z39-18 298102 NASA- Langley, 199:1