Document
NASA Technical Memorandum 108867
Moving BaseSimulation of an
IntegratedFlightand Propulsion
Control System for an Ejector-
Augmenter STOVLAircraft in Hover
Walter E. McNeill, William W. Chung, and Michael W. Stortz Ames Research Center, Moffett Field, California June 1995 National Aeronautics and Space Administration Ames Research Center Moffett Field, California 94035-1000 2DCD (aft) nozzle thrust, Ibf T2D Nomenclature ejector thrust, lbf TEj b wing span, ft ventral nozzle thrust, Ibf TVN _ wing mean aerodynamic chord, ft V true airspeed, ft/sec ACI,_a incremental rolling moment coefficient due to aileron deflection ship speed, knots Vs ACre,Be incremental pitching moment wind velocity with respect to Earth Vwind coefficient due to elevator reference, knots deflection wind velocity with respect to ship VWOD ACn,fir incremental yawing moment coefficient reference, knots due to rudder deflection generalized aircraft state variable _'ac Fcm d commanded nozzle thrust, lbf generalized state variable output of Yac Ixx moment of inertia about the X-body sensor compensation axis, slug-ft 2 generalized state variable command YCMD output of the Maneuver Command Iyy moment of inertia about the Y-body axis, slug-ft 2 Generator Izz moment of inertia about the Z-body generalized state variable error (input to YE axis, slug-ft 2 Regulator) 6, Lps total rolling moment due to propulsion generalized actuator command (output system and RCS thrusts, Ibf-ft of Regulator) Mps total pitching moment due to propulsion aerodynamic control actuator deflection, _ac system and RCS thrusts, Ibf-ft deg commanded aerodynamic control Nps total yawing moment due to propulsion _ac,cmd system and RCS thrusts, lbf-ft actuator deflection, deg t5B.c roll angular acceleration about the _a aileron deflection, (Be, R - Be,L)/2, deg X-body axis due to ailerons and elevator deflection, (Be, R + Be,L)/2, deg 8 e RCS nozzle thrusts, rad/sec 2 left elevon deflection, deg _e,L dynamic pressure, pV2/2, lbf/ft 2 right elevon deflection, deg 8e.R _1B.C pitch angular acceleration about the Y-body axis due to elevators and rudder deflection, deg 8r RCS nozzle thrusts, rad/sec 2 commanded thrust nozzle position, deg 0cmd fB.C yaw angular acceleration about the air density, slugs/ft 3 P Z-body axis due to rudder and RCS nozzle thrusts, rad/sec 2 wind direction with respect to Earth _l/wind reference, deg S wing reference area, ft 2 wind direction with respect to ship _WOD reference, deg iii LQR linear quadratic regulator Abbreviations and Acronyms Maneuver Command Generator MCG CG center of gravity moment reference center MRC CLM component-level model PIO pilot-induced oscillation CMG Configuration Management Generator RCS reaction control system DMICS Design Methods for Integrated Control Systems rms root mean square FCS flight control system RTM rapid thrust modulation GEAE General Electric Aircraft Engines SKP station-keeping point HPT High-Pressure Turbine STOVL short takeoff, vertical landing HQR handling qualities rating TDP touchdown point HUD head-up display VMS Vertical Motion Simulator (Ames) IFPCS integrated flight and propulsion control V/STOL vertical and/or short takeoff and landing system Variable Stator Vanes VSV Inlet Guide Vanes IGV 2DCD Two-Dimensional Convergent- LFWC Lockheed Fort Worth Company Divergent LPT Low-Pressure Turbine
Moving Base Simulation of an Integrated Flight and Propulsion Control System
for an Ejector-Augmentor STOVL Aircraft in Hover
WALTER E. MCNEILL, WILLIAM W. CHUNG, AND MICHAEL W. STORTZ Ames Research Center Engines (GEAE) to study integration of flight and propul- Summary sion controls.
A piloted motion simulator evaluation, using the NASA The main purpose of this effort was to validate Design Ames Vertical Motion Simulator, was conducted in sup- Methods for Integrated Control Systems (DMICS) (ref. 1) port of a NASA Lewis contractual study of the integration applied to a specific aircraft configuration--in this case of flight and propulsion systems of a short takeoff, verti- the E-7D, an ejector-augmentor powered-lift aircraft cal landing (STOVL) aircraft. Objectives of the study designed by Lockheed Fort Worth Company (LFWC), the were to validate the Design Methods for Integrated Con- subcontractor to GEAE for this study. In support of this trol Systems (DMICS) concept, to evaluate the handling combined activity, piloted simulations, both fixed and qualities, and to assess control power usage. A highly motion based, were conducted at NASA Ames Research nonlinear mathematical model of the E-7D ejector- Center. The motion simulation, using the Ames Vertical augmentor STOVL fighter design and a component-level Motion Simulator (VMS), is the subject of this report.
propulsion system model served as the basis for the simu- Other work by NASA on the subject of integrated flight lation. On the basis of preliminary fixed-base evaluation, and propulsion control was reported in reference 2.
wherein the design was seen not to be completely devel- opod over the full-flight envelope, motion simulation was A simulation of an earlier version of this configuration, restricted to hover and landing; all results and conclusions the E-7A (differing from the E-7D in details of the are limited to that flight phase.
propulsion system and utilizing a simplified engine model and a flight-control system not based on DMICS The closed-loop response of the E-7D with integrated methodology), was reported in reference 3. Although the flight- and propulsion-control system (IFPCS), which was results of that evaluation indicated that generally adequate designed following the DMICS procedure, exhibited to satisfactory handling qualities could be achieved in deficiencies which warrant further improvement of the transition and hover, the present simulation was con- design process to fully validate the DMICS concept.
ducted to explore the benefits of an integrated flight and With or without disturbances, handling qualities ratings propulsion system based on the hierarchical, decentralized (HQR) for the precision hover and shipboard landing methodology of DMICS. In addition to application of the tasks varied from satisfactory to adequate. Roll control DMICS approach, the E-7D simulation model included power usage data indicate that increasing MIL-F-83300 nonlinear aerodynamic data based largely on small-scale roll control power specifications for hover should be con- wind-tunnel tests, a sophisticated component-level engine sidered. Acceptable pilot workload appeared to be attain- model, and a propulsive-lift system utilizing mixed flow able with integration of flight and propulsion controls in at the ejector, ventral nozzle, and cruise nozzle.
the specified simulation environment; i.e., wind, turbu- A secondary purpose of this study was to assess the effec- lence, and task.
tiveness and pilot acceptance of different control- command combinations and sensitivities. Depending upon Introduction the particular phase of flight and the control mode being evaluated, the pilot used a particular cockpit control, or In connection with existing cooperative programs involv- inceptor, to command a given state variable (e.g., longitu- ing NASA and the aeronautical research establishments of dinal stick force to command rate of change of flightpath the United Kingdom and Canada to develop technology angle or vertical velocity and throttle displacement to applicable to advanced short-takeoff, vertical landing command longitudinal acceleration or longitudinal veloc- (STOVL) aircraft, NASA's Lewis Research Center had ity). Since these inceptors more nearly resembled those entered into a contract with General Electric Aircraft for a conventional fighter airplane as opposed to a powered-lift aircraft, and because allcontrol effector Aerodynamic data for the E-7D simulation mathematical
positions were determined bytheintegrated flight-and
model were derived from several sources: Clean configu- ration static data from wind-tunnel tests of a l/9-scale
propulsion-control system (IFPCS), reduced pilotwork-
load was expected tobeabenefit.
E-7A model (ref. 4), power-induced aerodynamics (with ejector doors open) from tests of a 30 percent-scale
The purpose ofthisdocument istoreport the results ofthe
powered model (ref. 5), RCS-induced and power-induced
VMSmotion simulation interms of pilotevaluations of
aerodynamics from a 15 percent-scale free-flight model
thehandling qualities, cockpit controls, and control-power
(refs. 6 and 7), and power-induced ground effects, such as
availability inprecision-hover and shipboard landing
suckdown, from the 15 percent-scale model configured
tasks. The state ofIFPCS development fortheE-7D atthe
for hover (unpublished data). Dynamic stability deriva-
time ofthepresent piloted simulation evaluation allowed
tives were estimated for low angles of attack by applica-
simulator motion tobeused onlyinhover. Therefore, all
tion of the USAF Stability and Control DATCOM (ref. 8),
discussion in thisreport of system performance and air-
extended to high angles of attack via trends shown by craft handling qualities is limited to that flight phase.
F-16XL and/or B-58 data. Landing gear aerodynamic effects were derived from F-16 data.
Aircraft Description Propulsion System The E-7D design used as a basis for the present simulation study was developed by LFWC as a The propulsion system was designed around a derivative potential supersonic fighter/attack aircraft with an of the General Electric Aircraft Engines (GEAE) ejector-augmentor powered-lift system to provide F110 turbofan engine, sized to meet the requirements of STOVL capability. This design differs from the the E-7D. Maximum gross thrust was approximately earlier E-7A (ref. 3) mainly in the propulsion- and 19,000 Ibf. In addition to a two-dimensional convergent- flight-control systems. The E-7A employed split divergent (2-DCD) vectorable aft nozzle for up-and-away flow, where fan air was ducted to the ejectors and flight, the powered-lift system comprised two sets of to the aft nozzle and core flow was routed to a ejector nozzles mounted in the wing close to the fuselage vectorable ventral nozzle; the E-7D propulsion and a vectorable ventral nozzle for pitch balance and system utilized mixed fan and core flow to all three fore/aft control in hover. Air flow to all ports (except the thrust nozzles (counting the ejectors as one). Air RCS) consisted of mixed fan and core flow. A maximum flow for the reaction control system (RCS) was of 7 percent of the compressor discharge flow was avail- supplied by bleed from the high-pressure compressor.
able to the RCS. Figures 2 and 3 show the arrangement of The E-7D flight control system (FCS) was developed the components of the propulsion system.
as part of an integrated flight- and propulsion-control The ejectors were fitted with upper and lower doors, system (IFPCS) based on a hierarchical, decentralized which were closed and faired during cruise flight. For methodology known as DMICS. DMICS uses modern powered-lift flight (transition and hover) both sets of control system design techniques, such as state space doors were open, providing an ejector thrust augmentation and optimal control theories, to optimize control power ratio of 1.7. Ejector thrust modulation was accomplished usage throughout the flight envelope. The earlier E-7A by means of the butterfly valve, and ventral nozzle thrust FCS was designed using nonlinear inverse methods, was varied by a set of shutters, which were closed for which solve the aerodynamic and propulsive control cruise flight. The 2-DCD nozzle was closed completely commands directly from the pilot control commands during hover operation.
(flightpath or velocity).
For simulation purposes, a component-level model (CLM) of the propulsion system was developed by GEAE Basic Aircraft (ref. 9). The model contained each major engine compo- Figure 1 shows the general arrangement of the E-7D. The nent, including inlet, fan and low-pressure compressor, fuselage, cockpit, and vertical tail were those of the high-pressure compressor, bypass duct, main burner, single-seat F- 16. The overall configuration was that of a high-pressure turbine, low-pressure turbine, bypass and tailless delta wing with a leading-edge sweep of 60 deg.
core flow mixing, and calculation of individual nozzle Table I presents dimensional data critical to the present thrusts, as well as a propulsion-control system with a simulation. Table 2 shows pertinent weight and inertia multivariable regulator and rapid thrust modulation data.
(RTM) by means of variable stator vanes. The propulsion stability with respect to MCG control commands. The
system was designed tomeet small andlarge magnitude
regulator was designed based on linearized aircraft stabil- thrust response specifications generated byLFWC.
ity derivatives and linear quadratic formulations with out-
Maximum coremass flowrate was 250 Ibm/sec; maxi-
put error weightings. Proportional plus integral control mum RCS thrusts (per nozzle) were: roll,400lbf;pitch, was used to ensure zero steady-state error.
300 Ibf;yaw, 500Ibf.Nominal thrust splitwas 40percent
The Control Selector converted generalized actuator
ejectors, 60percent ventral nozzle. Ingeneral, the propul-
commands to physical actuator commands, which sionsystem response bandwidth was10rad/sec.
included aerodynamic control surfaces, propulsive thrusts, and nozzle deflections. Pseudo-inverse transformation
FlightControlSystem
was used to make this conversion, since the number of physical control actuators was greater than that of the
The E-7DSTOVL FCS consisted ofsixmajor compo-
generalized actuator commands. Each physical control
nents (fig.4) which were the Maneuver Command Gener-
was weighted based on its effectiveness during the trans-
ator (MCG), Configuration Management Generator
formation. The pitch axis generalized actuator command
(CMG), Regulator, Control Selector, Actuators, and Sen-
distribution was given higher priority than were the longi-
sorCompensation. The control system design employed
tudinal and vertical command degrees of freedom. The
explicit model following withproportional and integral
Hanus anti-windup algorithm (ref. 10) was used to protect
gains, developed byLinear Quadratic Regulator (LQR)
system stability in the presence of the nonlinear character- synthesis about sixselected design points: twoincruise, istics of the physical actuators. The unmet generalized
twointransition, and twoinhover. Adetailed description
actuator commands were fed back to the MCG to condi-
oftheFCS anditsintegration withthe propulsion system
tion the control commands so that the generalized actuator isgiven inreference 9.
commands would always stay within the physical actua-
TheMCG handled the control mode selection made by
tors' linear bounds.
the pilot,shaped thepilotcontrol inputs, and generated
The Actuators contained the dynamics of all aerodynamic
mission-level control commands (i.e.,flightpath orveloc-
surfaces, reaction control system area control valves, ejec-
itycommands inresponse topilotcontrol commands). For
tor doors, and landing gear extension/retraction. All
design purposes, the flightenvelope was divided into
actuator dynamics were simulated with first-order transfer
three flightmodes: cruise, transition, and hover. Incruise
functions with appropriate time constants, rate limits, and
the pilotcommanded aftnozzle thrust, rollrate, pitchrate
position limits.
and sideslip. Intransition thepilotcommanded longitudi-
nalacceleration, rollrate, vertical flightpath (orflightpath
The Sensor Compensation provided aircraft states in iner-
rate) and sideslip. Inhover (the onlyflightmode for
tial and body coordinates for the closed-loop feedback
which results arepresented here), thepilotcommanded
control. First-order filters with corresponding cut-off fre-
longitudinal velocity, rollrate, vertical velocity, and lat-
quencies were used to remove high-frequency noise nor-
eralvelocity. The relationship ofinceptors tocommanded
mally seen on the aircraft.
variables isdescribed intheSimulation Experiment sec-
First-order complementary filters were employed to gen-
tionunder Experiment Configurations. Washout circuits
erate forward velocity and sideslip.
were included tosmooth thetransient response when
switching between flightmodes. Second-order command
generators were customized tomeet mission-level han-
Simulation Experiment
dlingqualities specifications forSTOVL aircraft in this
class.
A total of seven pilots took part in the simulator evalua- tion. All were highly trained and experienced test pilots.
The CMG defined thepropulsion system trimstates for
Five pilots represented NASA or its support contractors, theregulator based onthespecific flightconfiguration, while two were employed by LFWC. The only ones with
such aswing-borne withlanding gear up,wing-borne with
extensive powered-lift flight experience were four of the
landing gear down, approach, hover, and on-ground oper-
NASA-affiliated pilots.
ation. The trimthrusts and nozzle positions were sched-
uled asfunctions ofairspeed and flightpath angle. The
trim schedule was developed to minimize abrupt and large Simulation Facility amplitude thrust changes and attitude transients due to Ames Research Center's VMS (fig. 5) was used for the these thrust changes.
piloted evaluation. The VMS is a large-scale motion simu- The Regulator produced generalized actuator commands lator coupled to a VAX 9000 digital computer and an to achieve model following characteristics and control
Evans and Sutherland CT5A visual scene generator. Com-
the purpose of interpreting the Cooper-Harper rating scale plete descriptions of all components of the VMS system and assigning a handling qualities rating, allowable errors are given in reference I I. For the purpose of this experi- were established corresponding to desired or adequate ment, the cab and angular-motion support were rotated performance. These errors for the precision hover tasks 90 deg, so that the large motion along the beam became are shown in figure 12.
the longitudinal motion axis.
Figure 13 depicts the shipboard landing task. From an ini- Figure 6 shows an interior view of the cockpit, including tial height 50 ft above the landing deck and a position locations of the control inceptors. Figure 7 shows 400 fl directly aft of the station-keeping point (SKP), schematically the instrument panel layout. The three- which itself was 100 fl aft and 100 ft to port of the landing window visual display provided a choice of reference tar- deck touchdown point (TDP), the task was to translate gets for a precision hover task (fig. 8) or a small-ship forward at a constant height to the SKP (which was (DD963) scene for shipboard approach and landing always fixed with respect to the ship axes), stabilize there, (fig. 9).
then translate to a point directly over the TDP and descend to the deck. The deck landing area was 40 ft by All pertinent flight information was presented in a head- 70 ft. The target time from the SKP to touchdown was up display (HUD), the details of which are shown in 35 sec. The normal descent rate with respect to the deck figure 10. The symbology and drive laws for this HUD was approximately 4 fps. Allowable touchdown errors for were as described in reference 12, updated according to desired or adequate performance are also shown in reference 13.
figure 13.
Of the seven pilots who took part in the simulator evalua- Evaluation Tasks and Procedures tion, Pilots A, B, C, and G had extensive background and In order to evaluate the E-7D handling qualities and experience in jet-lift and other vertical and/or short take- IFPCS performance in hover, two different types of pilot off and landing (V/STOL) aircraft. Those with V/STOL tasks were used: (1) precision hover with reference to shipboard experience were pilots A and C. Pilots D, E, fixed targets (fig. 8) and (2) shipboard approaches and and F had considerable military experience in conven- landings aboard a Spruance class destroyer (DD963, tional fighters. Pilots B and G did not formally evaluate fig. 9). The Cooper-Harper handling qualities rating the precision hover tasks; all pilots evaluated the (HQR) scale (fig. II) (ref. 14) was used to evaluate these shipboard landing task.
tasks.
Figure 12 depicts the lateral and vertical precision hover Experiment Configurations tasks. For the lateral task, the airplane's initial position The experiment variables of concern in this simulation was 500 ft in front of the target array at a height of 50 ft.
were divided into two control system configurations and That position was established by setting the simulation either two or four disturbance conditions, depending on initial conditions. From that position, the task was to whether precision hover or shipboard landing tasks were translate to and stabilize at a position 90 ft in front of the being performed.
right hand target, translate laterally and stabilize in front of the left hand target, then translate and stabilize again at Figure 14 shows the control-inceptor/command-variable the right hand target, all within a time limit of 55 sec. The relationships for the "front-side" and "back-side" control 90-ft distance and the lateral/vertical position at each tar- modes. All three flight phases (cruise, transition, and get were established visually according to the sight picture hover) are included for clarity and completeness, though illustrated in figure 8(c), where the "rabbit ears" appear to only hover was evaluated formally. The front-side mode was so called because the manner of control was similar form a continuous horizontal bar, the corners of which match the inside corners of the end squares on the back- to that of an airplane on the front side of the power curve; board. Allowable deviations for desired and adequate per- that is, forward acceleration or velocity was controlled by formance were demonstrated visually to the pilot by throttle movement and flightpath angle or vertical velocity statically positioning the eye point prior to performing the was controlled with longitudinal stick. The back-side tasks.
mode owed its name to the technique used on the back side of the power curve, where flightpath or vertical For vertical precision hover, beginning at the same initial velocity was commanded with the left-hand controller position, the task was to translate to and stabilize at a (the throttle) and pitch attitude (and consequently forward position 90 ft in front of the lower target, ascend to and velocity) was controlled using longitudinal stick. In the stabilize in front of the upper target, then descend and present experiment using the back-side mode, forward stabilize again at the lower target, all within 75 sec. For
acceleration withvelocity holdinthetransition phase was
Except for pilots C and D, HQRs for the back-side mode commanded withathumbwheel (theantenna elevation in the lateral task for both calm and disturbed conditions knob) onthethrottle gripand, inhover, longitudinal were less favorable than for the front-side mode, in three velocity was commanded withlongitudinal stickforce.
cases moving from Level I to Level 2 (notably with
Switches andbuttons available fortrimming thecom-
pilot E). This difference was generally due to difficulty
manded variables arealso shown in figure 14. Figure 15
establishing desired longitudinal position aggressively,
shows thelongitudinal and lateral force vs.deflection
using the right-hand force controller. This controller
characteristics oftheright-hand controller. Thiscon-
received many complaints regarding high force gradients
troller, fromanF-16, had verylimited motion and is
and almost no inceptor position feedback to the pilot,
referred tohereinafter astheright-hand force controller or
causing a lack of smoothness of response (described as force stick.
"ratchetiness"), mostly in longitudinal translation.
Table 3shows the disturbance conditions used inthe
For the vertical task, as for the lateral task, using the experiment. Ambient windvelocity and direction areindi- front-side mode (fig. 16), the HQRs from the non- cated forthe precision hover task, withtheforward-aft V/STOL pilots were more favorable than those from the axis ofthehover targets ona heading of 163 deg. Forthe V/STOL pilots, again demonstrating the difference in shipboard landing task, bothambient wind andwind over acceptance of the front-side technique as it is influenced thedeck areshown, aswellasship speed and sea state. by differences in training and experience. The most unfa- Basic standards forspecifying these disturbance condi- vorable ratings fell deep in the Level 2 region. The major tions were obtained fromreference 15.
factors driving these Level 2 ratings were imprecise con- trol of heading and difficulty coordinating lateral and directional control inputs during the initial horizontal Results and Discussion iranslation to the vertical targets, cross-axis coupling (lateral into vertical) introduced using the right-hand force Handling Qualities Evaluation controller, and the need to coordinate left-hand throttle Precision hover- HQRs for the precision hover tasks are inputs (for forward translation) and right-hand lateral presented individually for pilots A, C, D, E, and F in inputs to arrive at the desired position in front of the verti- table 4 and are also plotted in figure I6. The data are bro- cal targets. Some difficulty capturing the upper vertical ken down into lateral task and vertical task for the front- target was also reported if the pilot used moderately side and back-side control modes, for two atmospheric aggressive vertical inputs; some pilot-induced oscillation conditions: calm air and a 15 knot wind at 30 deg from the (PIO) tendencies were seen, attributed to a somewhat long left forward quarter with a turbulence level of 6 fps root apparent time constant in the vertical response combined mean square (rms).
with the characteristics of the right-hand force controller.
In contrast, pilot F, who consistently rated this task/ For the lateral task in calm air, the E-7D simulation using control-mode combination Level 1, reported no adverse the front-side control mode was consistently rated as satis- characteristics other than a little cross-axis control factory, or Level I. All other combinations for the lateral coupling with the force stick.
task were distributed between Level I and Level 2 (adequate). (A complete discussion of Levels of flying In corresponding atmospheric conditions, the vertical task qualities and their relation to HQRs are presented in using the front-side mode was generally rated less favor- ref. 16.) For the front-side mode, the HQRs given by ably than the lateral task, partly because of the more pilots D, E, and F, whose backgrounds were in conven- extensive maneuver required to reach the necessary tional non-powered-lift fighters and who stated that their X-Y position in front of the vertical targets. The apparent training and experience made the front-side mode more long time constant of the vertical response, mentioned in natural to them (i.e., similar to the technique used in for- the preceding paragraph, was also a factor.
mation flying or aerial refueling, wherein longitudinal The vertical task using the back-side control mode was position is controlled with throttle and vertical position by rated consistently at 4.0. The principal reason for not rat- longitudinal stick) were consistently more favorable than ing these cases better than Level 2 was the persistent dif- those given by pilots A and C. Pilots A and C had exten- ficulty capturing precisely a desired longitudinal position sive V/STOL experience, where during low-speed and using the force stick. Good height control using throttle hover operation the use of throttle to control vertical displacement and relative ease of control of X-Y position velocity and right-hand center or side stick to control with a single controller, however, were reported.
translational velocity in the horizontal (X-Y) plane was the normal technique. For this reason the back-side mode On average, for these tasks using the front-side mode, the was more natural to them.
presence of wind and turbulence required enough
additional pilotcompensation tocause a deterioration of
using the throttle controller was considered good to about one HQR. In the back-side mode, there was no excellent.
influence of wind and turbulence on the ratings.
The addition of ship speed, sea state, and wind and tur- Shipboard landing- HQRs for the shipboard landing task bulence had an effect on mean level and spread of HQRs are presented for pilots A through G in table 5. Figure 17 similar to that observed for the front-side mode (fig. 17).
presents these data plotted for zero ship speed and as a In the more severe conditions (sea state 3 and 4 plus wind function of sea state and wind condition for a ship speed and turbulence) the majority of the pilots rated the back- of 10 knots.
side mode equal to or better than the front-side mode; the one exception was one of the non-V/STOL pilots, pilot F, In calm conditions and with zero ship speed, the front-side who consistently tended to favor the front-side mode for control mode showed a division between V/STOL and all tasks. The good vertical-velocity control using the non-V/STOL pilots similar to that observed during the throttle controller carried over into these disturbed cases.
precision hover tasks, though to a lesser degree, the aver- age HQR for the non-V/STOL pilots being just one rating For the back-side mode at sea state 4, a wide discrepancy point more favorable. Without disturbances of any kind, in HQR was noted between pilots A and B. Landing per- performance of the task with the front-side mode was formance (touchdown position error and vertical velocity) generally considered not difficult. Longitudinal velocity was within desired limits for both pilots. Pilot A com- control using the throttle was considered good to excel- plained of excessive workload in compensating for that lent. Where the HQRs were not better than Level 2, the amount of deck motion, coupled with difficulty making pilots complained of the workload associated with coordi- small, precise inputs using the force stick, whereas pilot B nation between the left-hand and right-hand controllers considered the workload minimal and described the when maneuvering in the X-Y plane, and of inadvertent maneuver from the SKP to touchdown as "not too bad."
cross-axis inputs using the right-hand force controller Both pilots did report difficulty using the longitudinal trim (usually lateral control coupling into vertical).
button on the force stick to relieve the steady forward force required to approach the SKP and match the ship's For the front-side mode, the effect of added disturbances velocity. When actuating the trim button with the thumb, on HQRs appeared to be largely due to the ship speed of it was necessary momentarily to relax hand pressure on 10 knots, with the increased sea state (and to some extent the stick, resulting in an unsteady approach to the SKP.
wind and turbulence level) increasing the HQR spread This initial phase, however, was not part of the formal into the Level 3 region. Being a displacement controller, evaluation task and was not used in arriving at the HQRs.
the throttle was found to be an effective means of com- manding forward velocity to match that of the ship (i.e., The vast majority of comments on the back-side mode the station-keeping point), almost lending itself to a "set were regarding the right-hand force controller and the and forget" operation. For these cases, some pilots con- manner in which it was integrated into the control system.
tinued to comment adversely on the need for left-hand/ At least three of the other pilots (not just A and B) found right-hand coordination as well as cross-axis contami- the need to hold force to match ship speed troublesome nation using the force stick. Additional complaints were unless trimmed out. As mentioned earlier, the method of also noted regarding the force stick. These were lack of trimming provided was not entirely satisfactory.
precision and smoothness of response (traceable to force Other drawbacks of the force controller installation that gradient or command sensitivity) and the need to hold were mentioned were force gradient or sensitivity prob- lateral stick force to match the ship velocity with the lems similar to those encountered in the front-side mode, aircraft heading into the wind, from 30 deg left of the leading to imprecise and "ratchety" control response, and bow. There was a provision for trimming out the torce, unwanted cross-axis inputs (this time in the X-Y plane). A but the requirement to release force on the stick grip in more optimal orientation of the force controller unit in the order to do so made that difficult. A similar problem was cockpit could have alleviated some of the latter problem.
encountered when attempting to trim out longitudinal force in the back-side control mode, discussed in the fol- Both the front-side and back-side modes were evaluated lowing paragraphs.
by pilots A and C, and the back side by pilot D, in 20 ft visibility and zero ceiling, with ship speed of !0 knots and For the back-side control mode in calm conditions, the sea state zero (fig. 17). These cases were given an HQR of overall average HQR was a small fraction of a rating point 5, the same as for unlimited visibility, the pilots citing the more favorable than tbr the front-side mode. The division same control deficiencies as before, with minimal effect in average HQR according to pilot background was also of reduced visibility. Even with unlimited visibility, these present for the back-side mode, but to a much smaller pilots relied mainly on the HUD for rate and position degree than for the front-side mode. Control of height guidance. Intherestricted visibility cases, theHUDwas been specified by MIL-F-83300 (ref. 16) and thesole source forrate and position guidance. AGARD R-577-70 (ref. 17) for hover and low speed flight operations. In reference 18 these specifications
Because it lentitselfreadily tothetask, allpilots relied to
have been related to control power and compared with
some degree ontheHUDhover mode forguidance tothe
later experimental data. These comparisons suggest
touchdown pointand descent tolanding. The degree to
that the roll control response specified in references 16
which each was able touse theHUDdepended onhis
and 17 may not be adequate for shipboard landings in
familiarity withthesymbology. Pilots A,B,C,and Ghad
severe weather conditions. One of the objectives of the
themost experience withthisdisplay, in this and previous
present experiment, therefore, was to examine the cont-
STOVL simulations. Pilots D,Eand F had priorexperi-
rol power usage requirements for the E-7D STOVL
ence withthedisplay onlyduring thefixed-base exercise
aircraft in adverse weather conditions.
leading tothepresent simulation intheVMSand, conse-
quently, less HUDfamiliarity coupled withrelative lack The control power usage is contributed by both the aero- ofV/STOL experience. Since there was insufficient time dynamic control effectors and the propulsive thrusts. The tobring these latter three pilots toalevel comparable with roll, pitch, and yaw control power usage for this experi- ment were calculated as follows:
that ofpilots A,B,C,and Gsome adverse effect ontheir
ratings would beexpected.
IbB.C = (qSbAC,,_5 a + LpS)?xx
Tosum upthepiloted evaluation, thefollowing state-
ments can bemade: Fortheprecision hover tasks, HQRs
(considering bothfront-side and back-side control) were
spread overarange covering LevelI and Level 2,with
more ratings fallingintheLevel 2 range inthe presence
ofmoderate windand turbulence. Fortheshipboard
The control power usage in the shipboard landing task for
landing task withzero ship speed, nodeck motion and no
various weather and sea state conditions is shown in
windand turbulence, HQRs covered thesame range
figure 18. It is shown that the roll control power usage in
(Level ! and Level 2)asforprecision hover. The addition
both calm air and adverse weather conditions has
of l0 knots of ship speed resulted inLevel 2 ratings in
exceeded MIL-F-83300 control power requirement for
practically every case; thefurther addition ofdeck
2 rad/sec attitude system with a damping ratio of 0.8944
motion, and moderate windand turbulence increased the
(ref. 18). The pitch control power usage in adverse HQR spread withinthe Level 2range and even beyond.
weather condition has also exceeded the specified
Single-axis response, withexception oftheyawaxis, did
requirement for a 2 rad/sec rate system.
notcause great problems; however, cross-coupling effects
were often present which were annoying oreven unsatis-
Conclusions
factory. Yawcontrol was characterized byanapparent
long time constant which, incombination withabsence of
A piloted motion simulator evaluation was completed in
aheading-hold loop, made capture ofadesired heading
support of a contractual study of the integration of flight
difficult. Several pilots (notably pilotsAand C)found the
and propulsion controls of a short takeoff, vertical landing
right-hand force controller difficulttouse and unsuitable
(STOVL) aircraft. The objectives of the study were to
forjetborne flyingtasks. Objectionable characteristics
validate the Design Methods for Integrated Control Sys- were found in the rudder pedals, attributed to their also tems (DMICS) concept, to evaluate the handling qualities, being force controllers. For jetborne maneuvering, those and to assess control power usage during hover and land- pilots highly experienced in V/STOL operations (A, B, C, ing. From this evaluation the following conclusions are and G) expressed a clear and sometimes forceful prefer- drawn: ence for the back-side control mode, wherein vertical velocity was commanded with a left-hand controller and I. The closed-loop response of the E-7D with the inte- longitudinal and lateral velocities were commanded by grated flight- and propulsion-control system (IFPCS), which was designed following the DMICS procedure, means of a single right-hand controller.
exhibited deficiencies which warrant further improvement of the design process to fully validate the DMICS Control Power Usage concept.
The control power has direct consequence to the trim- 2. With mild to moderate wind and turbulence, using ming, maneuvering, and stabilization of the aircraft in any either the front-side or back-side control mode, handling flight condition. The control response requirements for qualities were judged to be adequate (Level 2) to V/STOL, in terms of attitude change in one second, have marginally satisfactory forthe precision hover tasks. 6. Riley, Donald R.; Shah, Gautam H.; and Kuhn,
Under similar atmospheric conditions and with10knots
Richard E.: Low-Speed Wind-Tunnel Study of ofshipspeed, both thefront-side and back-side modes on Reaction Control-Jet Effectiveness for Hover and average were rated Level 2 fortheshipboard landing task.
Transition of a STOVL Fighter Concept. NASA TM-4147, Dec. ! 989.
Increasing sea state to3 and 4 resulted inacorrespond-
ingly greater spread ofratings.
7. Riley, Donald R.; Shah, Gautam H.; and Kuhn, Richard E.: Some Power-Induced Effects for
3.Rollcontrol power usage data indicate that increasing
MIL-F-83300 rollcontrol power specifications forhover Transition Flight Measured on a 15-Percent Scale operations should beconsidered. Further experiment is E-7A STOVL Fighter Model. NASA TM-4188, June 1990.
required todetermine if MIL-F-83300 pitch control power
specification isadequate.
8. Williams, J. E.; and Vukelich, S. R.: The USAF Sta-
4.Withnodisturbances, using thefront-side control
bility and Control Digital DATCOM, Volumes I,
mode, average handling qualities ratings (HQR) were in
I1 and III. AFFDL-TR-79-3032, Apr. 1979.
the satisfactory region (Level I) bothforthe precision
9. Adibhatla, S.; Cooker, P.; Pajakowski, A.; hover and shipboard landing tasks. Inthesame conditions, Romine, B.; Virnig, J.; and Bodden, D.:
using theback-side control mode, average handling
STOVL Controls Technology, Volumes I,
qualities ratings were borderline LevelI/Level 2 forthe
II, and III. NASA CR-195361, July 1994.
lateral precision hover task, Level 2 forthevertical
precision hover task, and Level1forthe shipboard 10. Hanus, R.; Kinnaert, M.; and Henrotte, J. L.: Condi- landing task.
tioning Technique--A General Anti-Windup and Bumpless Transfer Method. Automatica, vol. 23,
5.Inthefront-side orback-side mode, use ofthe limited-
no. 6, Nov. 1987, pp. 729-739.
displacement force controller (the right-hand side stick)
formaneuvering inhover resulted inobjectionably high
I I. Danek, George: Vertical Motion Simulator Familiar- pilotworkload. ization Guide. NASA TM-103923, Oct. 1991.
12. Merrick, Vernon K.; Farris, Glenn G.; and Vanags, Andrejs A.: A Head Up Display for Application to References V/STOL Aircraft Approach and Landing. NASA I. Shaw, P. D. et ah Design Methods for Integrated TM- i02216, Jan. 1990.
Control Systems. AFWAL-TR-88-2061, June 13. Merrick, Vernon K.: Some VTOL Head-Up Display 1988.
Drive-Law Problems and Solutions. NASA 2. Garg, Sanjay; Ouzts, Peter J.; Lorenzo, Carl F.; and TM- ! 04027, Nov. 1993.
Mattern, Duane L.: IMPAC--An Integrated 14. Cooper, George E.; and Harper, Robert P., Jr.: The Methodology for Propulsion and Airframe Con- Use of Pilot Rating in the Evaluation of Aircraft trol. NASA TM- 103805, June 1991.
Handling Qualities. NASA TND-5153, Apr. 1969.
.
Franklin, James A.; Stortz, Michael W.; Gerdes, 15. Fortenbaugh, Robert L.: Mathematical Models Ronald M.; Hardy, Gordon H.; Martin, James L.; for the Aircraft Operational Environment of and Engelland, Shawn A.: Simulation Evaluation DD-963 Class Ships. Vought Corp.
of Transition and Hover Flying Qualities of the Rep. 2-55800/8R-3500, Sept. 1978.
E-7A STOVL Aircraft. NASA TM-101015, Aug.
1988.
16. Chalk, Charles R.; Key, David L.; Kroll, John, Jr.; Wasserman, Richard; and Radford, Robert C.: 4.
Foley, W. H.; Albright, A. E.; Powers, D. J.; and Background Inlormation and User Guide for Smith, C. W.: Study of Aerodynamic Technology MIL-F-83300 Military Specification--Flying for Single-Cruise-Engine V/STOL Fighter/Attack Qualities for Piloted V/STOL Aircraft.
Aircraft, Phase II Final Report, Volumes I and II.
AFFDL-TR-70-88, Nov. 1971.
NASA CR-177367, Aug. 1985.
17. Anon.: V/STOL Landing, I -Criteria and Discussion.
5.
Banks, Daniel W.; and Gatlin, Gregory M.: Longitu- AGARD R-577-70, Dec. 1970.
dinal and Lateral Aerodynamic Data from Tests of an Advanced STOVL Fighter Employing a 18. Franklin, James A.: Criteria for Design of Integrated Powered- Lift Ejector. NASA TM-87672, May Flight/Propulsion Control Systems for STOVL 1986.
Fighter Aircraft. NASA TP-3356, Apr. 1993.
Table I. E-7D principal dimensions
Wingspan, ft 32.4
Wingarea, ft2 630.6
Aspect ratio 1.66
Taper ratio 0. I 15
_, ft 23.56 MRC, % Cw 30.00 CG, % Ew 35.97 _ieL,R (max) deg +30 _ir (max) deg +_20 Table 2. Weight and moments of inertia (hover, landing gear down) Weight, Ib 17,000 lxx, slug_ft 2 4,415 lyy, slug-ft 2 29,413 Izz, slug-ft 2 29,963 lxz ' slug_ft 2 -70.51 Table 3. Task environmental conditions At hover targets with heading At ship with heading 000 deg Wind 163 deg condition Vwind, _wind, rms turb., Vwind, _t/wind, Ship VWOD, _I/WOD, rms turb., Sea kt deg fps kt deg speed, kt kt deg fps state 0 0 0 0 0 0 0 Wl 0 0 0 8.11 292 10 15.04 330 3 0 W2 15 133 3 8.11 292 10 15.04 330 6 3 W2A 15 133 6 8.11 292 10 15.04 330 6 4 W2B 15 133 6 Visibility Visualrange, Ceiling, ft condition fl V I Unlimited Unlimited V6 20 0
Table 4.HQRs forprecision hover tasks a
Lateral
Task Vertical
Frontside Backside
Control mode Frontside Backside
ind
WI W2A WI W2A WI W2A W1 W2A
A 3 4 4 5 4 4
4 6
C 3 5 3 4 4 4
5 6
D 3 3 3 3 4 4
3 4
E 2 3 4.5 4.5 4 4
3 3 i
F 2 2 3 3 4 4
2 3 avisibility: V1.
Table 5. HQRs for shipboard landing task Frontside Backside Control mode Wl W2 W2A W2B Wl W2B Wind W2 [ W2A Vl,V6 VI VI Vi VI,V6 VI Vl Vl A 4 B 3 C 5 5 D 3 4 4.5 E 2 6 F G Win9 . Area 630.6 fl 2 • Aspect ratio 1.665 • Taper ratio 0.115 • Airfoil NACA 64A004 • t/c 4% Vertical tail • Ares 68.4 ft2 • Aspect ratio 1.294 . Taper ratio 0.437 _ _ Airfoi.____L__ i Biconvex 32 it 4,8 in;._.__ erall span.
..____ 49_15 in.
Figure 1. Three views of E-7D aircraft.
II Ejector //_ o (each side) . / //_1 °r 2-DCD /nozzle ....
__ v_.o,r., .o,_'_,to.
Pitch RCS nozzle RCS nozzle nozzle _20o TEj 110 o TVN Figure 2. Arrangement of propulsive nozzles and RCS nozzles.
Fan with inlet guide vanes (IGVs) Bypass duct Butterfly _. (2-DCD) nozzle Combustor Ventral nozzle Compressor with variable stator vanes (VSVs) Figure 3. Propulsion system schematic diagram.
FUGHT CONTROL SYSTEM _ Sensor I
/ compensation
-- I Aerodynamic 5a c v I actuators I I ' y commands I Maneuver Y_ I I _" I co, :trol I_- command Pilot I ctor r _, Regulat°r I_'I sel, I generator cmd ecmd' Fcmd , I " Yac Operational limits J PrcOPntlrSoi/n _.e, j Propulsionp
/ -.,- II "t"m I
Trim JConfiguration I Flight path | management | commands and airspeed _1 generator | commands ............................................... J Figure 4. Integrated flight- and propulsion-control system structure.
VMS NOMINAL OPERATIONAL MOTION Axis Displ Velocity Accel 16 24 Vertical ±30 Lateral ,20 8 16 ±4 4 10 Longitudinal Roll ±18 40 115 ±18 40 115 j j_J Pitch Yaw ±24 46 115 _ All numbers, units in ft, deg, sac Figure 5. Vertical Motion Simulator.
Figure 6. Simulator cockpit interior.
MODE SELECT SWITCHES MODE SELECT
B
ALSO AVAILABLE ON SIDESTICK TOUCHDOWN LIGHT LANDING GEAR HEAD-DOWN LIGHTS DISPLAY DOOR EJECTOR [-_ LANDING SWITCH GEAR HANDLE Figure 7. Instrument pane/layout.
]5 Lateral task Vertical task (a) Hover target arrangements Black White ed (c) Target appearance when in position (b) Target close up Figure 8. Precision hover target display.
]6 Figure 9. Small ship landing display.
m5 o 5o I I I ml0
©
3O 1. Horizon bar and pitch ladder 8. Height above ground/sea 2. Engine (fan) percent RPM 9. Vertical velocity 3. Resultant thrust vector angle 10. Horizontal velocity vector 4. Horizontal distance to touchdown point 11. Horizontal velocity predictor ball 5. Landing pad symbol 12. Vertical velocity predictor diamond 6. Airspeed 13. Allowable vertical velocity ribbon 14. Ground/deck bar 7. Ground speed Figure 10. STOVL head-up display; hover mode.
]8 Excellent Pilot compensation not a factor Highly desirable for desired performance Pilot compensation not a factor for desired performance Minimal pilot compensation _i_ Fair - Some mildly required for desired performance 77! _ unpleasant deficiencies Desired performance requires i_ Minor but annoying _,__:_ deficiencies moderate pilot compensation warrant Moderately objectionable Adequate performance requires • _:,_ deficiencies considerable pilot compensation improvement Very ob)ectJonabJe but Adequate performance requires tolerable deficiencies extensive pilot compensation Adequate performance not Major deficiencies attainable with maximum tolerable pilot compensation.
Controllability not in question Deficiencies Considerable pilot compensation improvement is required for control ii!iiiiiiii_ii Major deficiencies Intense pilot compensation Jsrequired to retain control Major deficiencies Control will be lost during some Improvement mandatory portion of required operation ::::: ::::: * Definition of required operation involves designation of flight I Pilot decisions I phase and/or subphases with accompanying conditions.
Figure 11. Handling-qualities rating scale.
]9 .!
_l---_----t_
-Ft .t-i
I I
°°1_
"o e- 1:: _ ' .aO -_" _:
i
o, ii
A_-..
o c ..i ,Ig Io m m e- I.- i ,,..,J 2O u) C_ e-, O vl i -"_l v |_'_l O O 2] • Pitch rate • Thrust • Accel/decel with velocity hold • Flight path rate/ flight path hold • Longitudinal velocity • Vertical velocity UHF/VHF switch • Roll rate • Roll rate/attitude hold • No function • Lateral velocity /_-_f • Accalldecal trim f • L_ngitudinal velocity trim llmw_ _mz_ Trim switch • Pitch/roll trim Targat management _ _'_ _ • Pitch/roll attitude trim • Nofunction \ ' ' / • Pitch attitude trim • No function _ J • Vertllat velocity trim Rudder pedels - • Sideslip • Sideslip • Yaw rate (a) • Pitch rate • Pitch rate/attitude hold • Thrust • Longitudinal velocity • Flight path /_* Vertical velocity • Transition • Roll rate J • Cruise • Hover • Roll rate/attitude hold • Lateral velocity _ mm_ t Trim switch • Pitch/roll • Pitch/roll attitude • Pitch attitude • No function • AcceVdecel with Target management velocity hold • No function • No function • No function • Velocity trim Rudder pedels - • Sideslip • Sideslip • Yaw rate/heading hold (b) Figure 14. Control modes and inceptor configurations. (a) Front-side mode, (b) back-side mode.
2.2 - Calibration I 0 _'_ \ \ \ \ \ _ t°leran_/ 0 5 -- 10 15 20 25 ---------.-L _a) Force, POUnds 30 35 .20 - .15 _ Calibration .os (b) Force, POunds 20 Figure 15. Right-hand controller.force gradients. (a) Longitudinal, (b) lateral.
10-- LATERAL TASK VERTICAL TASK Pilot OA Open symbols - frontside mode Oc Filled symbols - backside mode _D 8 m zI E Inadequate rl F (Level 3) 6-
n-
O -I- Adequate (Level 2) o @ 4 - A ee_l Satisfactory 2 - _1_] (Level 1) I I I I 0/0 15/6 0/6 15/6 Disturbance, wind/turb., klsfiVsec rms Figure 16. Handling-qualities ratings for precision hover tasks.
10 m Pilot O A Open symbols - frontside mode Filled symbols - backside mode []B Flagged symbols - ceiling zero, visibility 20 ft.
_c
D Inadequate E VS=10 kt VS= 0 (Level 3) F ® • 6 u
o n a _ _ _a
o
3: Adequate (Level 2) 4_ QA •A [] •• []
[]QA Eliza
Satisfactory (Level 1) 2 n (_ dll
I I I
0 2 3 Sea state Figure 17. Handling-qualities ratings for the shipboard landing task.
I I I DOra •
oee_
II II ,0 I I I _, ¢-- r_
._g
o _m C)
e cpe_ e
r_ E m r..)
e ee e ee_e
Q I I I o
REPORT DOCUMENTATION PAGE Form Approve_
OMB No. 0704-0188 Publicreporting burden for this collection of informationis estimated to average 1 hour per response, includingthe time for reviewinginstructions,searching existing data sources, gatheringand maintaining the data needed, and completingand reviewingthe collectionof information. Send comments regardingthis burden estimate or any other aspect of this collectionof information, includingsuggestionsfor reducing this burden, to WashingtonHeadquarters Services. Directorate for informationOperations and Reports, 1215 Jefferson Davis H ghway, Sute 1204, Ari ngton, VA 22202-4302. and to the Off ce of Managementand Budget. Paperwork Reduction Project (0704-0188), Washington, DC 20503.
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE 3. REPORT TYPE AND DATES COVERED June 1995 Technical Memorandum !4. TITLE AND SUBTITLE 5. FUNDING NUMBERS Moving Base Simulation of an Integrated Flight and Propulsion Control System for an Ejector-Augmentor STOVL Aircraft in Hover 505-68-32 6. AUTHOR(S) Walter E. McNeill, William W. Chung, and Michael W. Stortz 8. PERFORMING ORGANIZATION 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) REPORT NUMBER Ames Research Center A-950046 Moffett Field, CA 94035-1000 10. SPONSORING/MONITORING 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) AGENCY REPORT NUMBER National Aeronautics and Space Administration NASA TM- 108867 Washington, DC 20546-0001 11. SUPPLEMENTARY NOTES Point of Contact: William W. Chung, Ames Research Center, MS 243-5 Moffett Field, CA 94035-1000; (415) 604-1496 12a. DISTRIBUTION/AVAILABILITY STATEMENT 12b. DISTRIBUTION CODE Unclassified-Unlimited Subject Category - 05 13. ABSTRACT (Maximum 200 words) A piloted motion simulator evaluation, using the NASA Ames Vertical Motion Simulator, was conducted in support of a NASA Lewis contractual study of the integration of flight and propulsion systems of a STOVL aircraft. Objectives of the study were to validate the Design Methods for Integrated Control Systems (DMICS) concept, to evaluate the handling qualities, and to assess control power usage. The E-7D ejector-augmentor STOVL fighter design served as the basis for the simulation.
Handling-qualifies ratings were obtained during precision hover and shipboard landing tasks. Handling- qualities ratings for these tasks ranged from satisfactory to adequate. Further improvement of the design process to fully validate the DMICS concept appears to be warranted.
15. NUMBER OF PAGES 14. SUBJECT TERMS 3l STOVL, Integrated flight/propulsion control system, DMICS 16. PRICE CODE A03 18. SECURITY CLASSIFICATION 19. SECURITY CLASSIFICATION 20. LIMITATION OF ABSTRACT 17. SECURITY CLASSIFICATION OF THIS PAGE OF ABSTRACT OF REPORT Unclassified Unclassified NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89) Prescribed by ANSI Std. Z39-18 298-102