A real time Pegasus propulsion system model for VSTOL piloted simulation evaluation

N O T I C E

THIS DOCUMENT HAS BEEN REPRODUCED FROM

MICROFICHE. ALTHOUGH IT IS RECOGNIZED THAT

CERTAIN PORTIONS ARE ILLEGIBLE, IT IS BEING RELEASED

IN THE INTEREST OF MAKING AVAILABLE AS MUCH

INFORMATION AS POSSIBLE

NASA Technical Memorandum 82770

AIAA-81-2663

L1.1

4-

A Real Time Pegasus Propulsion

System Model for VSTOL Piloted

Simulation Evaluation

Nb1-1314V

,ASUS

ki

A

(MA.:A-rM-62110) H k A L

BULEL kVki VSTUL NILUTEU

SYSTEM

i?BUPJLSICN

18 P (NASA) EVALUATICM SINULATICN 6uc:ias LSCL21L A01 Aj)lMk HC

08419

G3/U7

Ar

r ,

James R. Mihaloew

'

Lewis :Qesearch Center

Cleveland, Ohio

V,

and

sy, F

r ►

Stephen P. Roth and Robert Creekmore

S%aa 3'

Pratt and Whitney Aircraft Group

West Palm Beach, Florida

Prepared for the

VSTOL Conference

cosponsored by the American Institute of Aeronautics

and Astronautics and NASA Ames Research Center

Palo Alto, California, December 7-9, 1981

S A REAL TIME PEGASUS PROPULSION SYSTEM MODEL FOR VSTOL PILOTED SIMULATION EVALUATION R. Mihaloew' .iylmas National Aeronautics and Space Administration Lewis Research Center Cleveland, Ohio Stephen P. Roth r and Robert Greekmore Pratt and Whitney Aircraft Group Government Products Division West Palm Beach, Florida Introduction Abstract During low speed operations, VSTOL aircraft depend An emphasis on increased aircraft and propulsion not only on the propulsion system for lift, but control system integration and piloted flight also for the forces and moments needed for flight simulator evaluation has created a need for high path and attitude control. Thus highly coordinated fidelity real tine dynamic propulsion models. In integrated flight and propulsion control systems recognition of this need, a real. time propulsion are critical and necessary to the success of these system modeling technique suitable for use in advanced aircraft.

man-in-the - loop simulator studies has been developed by NASA-Lewis and demonstrated using Simulation, with its inherent flexibility, will flight simulator facilities at NASA - Ames. This play a key role in the development of these technique provides the system accuracy, stability propulsion control systems.

integrated aircraft - and transient response required for integrated These simulations will provide a comprehensive aircraft and propulsion control system studies.

source of qualitative and quantitative information concerning the characteristics of aircraft and A Pegasus - Harrier propulsion system was selected as propulsion systems in a dynamic state. They will a baseline for developing mathematical modeling and also serve as essential tools for the analysis and simulation techniques for VSTOL. Initially, static synthesis of control logic and as test vehicles for and dynamic propulsion system characteristics were control software and hardware development.

modeled in detail to form a non-linear aerothermodynamic digital computer simulation of a . t, under contract with the Pratt and Whitney Aircra:` Pegasus en gine. From this high fidelity NASA-Lewis Research Center, has made a conceptual simulation, a real time propulsion model was evaluation of propulsion control systems for VSTOL formulated by applying a piece-wise linear state - aircraft in order to define critical control variable methodology. A hydromechanical and water requirements and to identify critical technologies injection control system was also simulated.

pertaining to the integration of aircraft and propulsion controls. One of the major technolog-f The real time dynamic model includes the detail and areas of this program includes the development of flexibility required for the evaluation of critical mathematical modeling and simulation techniques control parameters and propulsion component limits applicable to the design of VSTOL integrated over a limited flight envelope. The model contains propulsion controls.- The long range - line computational aircraft - approximately 7.OK bytes of in objective of the program is to conduct a real time 7k of block data. It has an 8.9 ms code and 14 .

piloted simulation evaluation of an integrated cycle time on a Xerox Sigma 9 computer.

aircraft - propulsion control on the NASA-Ames flight simulator facilities.

The model has been programmed for interfacing with a Harrier aircraft simulation at NASA-Ames.

Since the advent of piloted simulators and the Typical propulsion system simulation results are presented. growing emphasis for systems integration, there has been an increasing need for higher fidelity real-time propulsion system models. Propulsion and integrated control system e^rsluation of VSTOL aircraft on flight simulators will require that propulsion system simulations be realistic and The high and low pressure comps"`^sor spools are independent, c"raxial and counter-rotating include significant dynamics as well as important Counter-rotation minimizes gyroscopic effects which internal parameters. In recognition of this need, a dynamic digital real-time model of an advanced is an important consideration in hovering propulsion system Its* developed which is operations. The three sta e fan is driven by a two been g suitable for use in man-in-the-loop simulatot stage turbine at a design speed of 6500 rpm at a studies. This model provides the engine-control pressure ratio of 2.31. The high pressure compressor uses variable inlet guide vanes and is system accuracy, stability and transient response required for the intended studies. These studies an eight stage compressor driven by a two stage might include the evaluation of critical control turbine at a design speed of 10500 rpm at a parameters, system response, system environmental pressure ratio of 5.6.

effects and critical propulsion component At a maximum design thrust of 19500 lbf, thrust $,a aerodynamic, mechanical and thermodynamic limits.

divided ivenly between the fan and core nozzles, The model may also be used to analyze propulsion control failure modes and effects. Pairs of nozzles rotate and deflect the nozzle flow from both the fan and turbine exits through a range In the VSTOL Propulsion Con;;rol Analysis Program of 0 to 98.5 degrees. The four nozzles are mechanically linked to each other to ensure that reported in (1), ' on engine model was used to the vertical-to-horizontal angular position is explore the merits of using a combined simulation of aircraft-propulsion systems for armlysis of identical for each thrust vector.

propulsion control requirements. Simudtions of this nature Integrated into the des:tgn scheme In order to provide aircraft attitude control, up to 22 percent of the high pressure compressor exit provide an important cost effective ool in airflow is available for ducting to a remote specifying, generating and conveying control requirements for the next generation VST01. designs. reaction control puffer jet system.

A Pegasus 11 propulsion system was chosen as the baseline VSTOL engine for developing mathematical M romechanical Control Configuration modeling and simulation techniques. The real time engine model is a piecewise linear state variable The engine fuel control modeled is the Dowty representation derived from a detailed hydromechancal control used on the Pegasus 11 aerothermodynamic simulation of a typical Pegasus engine. The fuel control is designed to meet 11 engine. Dynamics included in the simulation are engine performance throughout the flight envelope.

e.tgine fan and compressor rotor dynamics, engine At the same time, the fuel control unit ensures burner heat transfer dynamics, engine control that the engine limitations are not exceeded, dynamics, and sensor and actuator dynamics. This T,e engine fuel control system is shown model provides steady state and transient characteristics for various engine pressures, schematically in figure 2. Fuel flow to the engine °temperatures, flows, stall margins and thrust, The is regulated by two metering devices; a metering model calculates transient performance by numerical valve and a throttle shutoff valve. The metering valve is effectively a variable orifice under the integration of time-dependent differential state control of a low pressure compressor speed equations and contains the dynamics necessary to simulate aircraft forces resulting from engine governor. The metering valve normally controls the thrust. The real time model also represents tne. fuel flow to the engine in the high fan speed range above 87 percent. The pressure drop across the engine fuel control system the water injection orifice is controlled by a pressure drop regulator.

system and flight simulator interfaces, The throttle shutoff valve meters fuel flow The following sections present descriptions of the directly to the burner and is also used for propulsion 4ystem, control system, real time shutting the fuel off completely. The pressure drop across it is maintained constant by a flow methodology and model capabilities. Typical control pressure difference regulator, The results from the propulsion system oimulation are throttle valve normally controls the fuel Clow to also presented.

the engine in the low fan speed range below 87 percent.

opulsion System Description Pr A number of engine limiting _ functions are also Engine Configurat ion included. These include an acceleration control unit, a jet pipe temperature limiting control, an engine predaure ratio limiter, a combustion chamber The engine modeled in this program is a nonmixed, p ressure limiter and an airbleed reset unit for twin-spool Pegasus 11 as shown In figure 1. The inlet fuel compensation due to reaction control bleed. A engine weighs approximately 3540 lbm with an diameter of 46 in and a total uninstalled dry manual fuel flow control, is also provided in the thrust of 19300 lbf. Total design airflow is 435 event of fuel control unit failure.

lbm/sec divided between the fan duct and engine ratio of core stream with a bypass 1.35.

Simulation Techuiaue Water Infection System Aerothermodynamic Detailed Simulation

high load configuration, if an attempt was

In A

made to increase speed and thrust to Nat vertical

A detailed nonlinear aerothermodynaoic simulation takeoff requirements, the turbine inlet temperature of the baseline

propulsion system forms the base

could be exceeded. To avoid this, a water Injection system was provided to allow engine speed for the real time propulsion model development.

to increase for a given turbine inlet temperature.. This nonlinear simulation is a high fidelity model that represents each component in the engine and Water is introduced at the turbine inlet.

control. Heat transfer dynamics, rotor dynamics

Provision is made to carry 62.5 gallons of water

and aerothermodynamics are modeled. This detailed which is sufficient for 90 seconds of operation.

digital simulation includes complete component The water injection system is shown schematically performance maps and gas flow balance equations.

The components ate

In figure 3. It it controlled by a selector switch matched for aerodynamic stability from detailed stability audits that

in the cockpit, float level switches in the tank, a

throttle control microswitch and a water pressure consider surge line and operating line destabilizing switch. Setting the selector switch "on" arms the influences for steady state and transient operations.

system, raises the fan speed mechanical governor setting 4 percent when the throttle is in excess of A nonlinear propulsion systemsimulation such as

87 percent fan speed and energizes a fuel bypass

solenoid. The bypass this produces a model of high frequency fidelity

solenoid provides

supplementary fuel flow to increase fan which, however, does not run in real time.

speed. If Extension to an all digital format

there is sufficient water in the tank, moving the for 'piloted

simulators would require high sampling rates (small

throttle beyond a position which gives 92 percent time steps) to maintain calculational stability.

fan speed operates a microswitch in the throttle

Real . ,,.me would be virtually impossible. The

control linkage which opens a solenoid valve to

admit engine bleed air to the pump turbine. When general approach taken in the real time digital water pressure reaches 240 psi the pressure switch simulator model presented here was to represent dynamic response over a reduced frequency range but

operates to increase the jet pipe temperature

to maintain as much control system detail as

limiter and engine life recorder datum settings and

indicates a green light in the cockpit. possible. For the level of steady state and

dynamic complexity required to meet this objective, When 15 seconds of water is left, a float operated steady state accuracy does not have to be compromised over detailed models.

switch rinses to indicate a low level warning in the coct.pit. An empty level switch operates a relay to isolate the pump control circuit after the The real time model must cover a wide range of tank has been emptied. The system will continue to frequencies as shown in fig-Are 5. Expansion on the bandwidth is possible, but would involve trade-offs operate until all water is used, the throttle is between real time capability and control or retracted, or the selector switch is turned off.

There is also a jettison feature in the system. interface detail.

Real Time Methodology Reaction Control System The aircraft is equipped with both aerodynamic The real time model is based on a piecewise linear controls and a reaction control system state variable 'technique reported in (2).

as shown in

figure 4. Aerodynamic controls on the aircraft are standard control surfacers. But, these supply The state variable form is shown in figure ;6, where X is the vector of state variables, X is the time negligible control during vertical, hover and

of derivative of the state variables, U is the control

transition modes because the low velocities in

these modes. A reaction control system consisting input vector and Y is the vector of observed or output parameters. A is the plant

of six fully modulating puffer jets located at the matrix and its

elements are the partial derivatives of each state wingtips and at fore and aft fuselage locations is variable to the time derivative of each state required to provide thrust control for roll, pitch and yaw motions during these modes. These puffer variable. Elements of the output matrix C define jets are mechanically linked to their respective the effect of each state variable on each output aerodynamic control surfaces to accomodate control variable. The control matrix B and the direct

couple matrix D define the effect of

transfer during transition from vertical to each control variable on each state variable time derivative and horizontal flight. A master shutoff valve is linked to the nozzles so that bleed air from the each output parameter.

engine is turned off when the nozzles are in the

horizontal flight position. A modal analysis was used to determine which states in the nonlinear aerothermo model are required to adequately represent the system within the desired bandwidth. The nonlinear aeruthermo model was linearized to obtain the -system A, B, C, and D matrices. An eigenvector-eigenvalue analysis of the A matrix was performed. The eigenvectors were a high fidelity propulsion model which provides examined to associate allonvalues with states. steady state and transient characteristics for Nigh frequency states outside the control bandwidth drtsired engine pressures, temperatures, airflows, thrusts and rotor speeds: The were eliminated. A mode controllability matrix was surge margins, simulating the engine calculates also defined. States which were uncontrollable by computer program both steady state and dynamic engine the inputs were eliminated. The final model states which were characteristics that are representative of a considered only within a 0.1 to 10 Hs bandwidth Pegasus 11-402 engine.

Modeling large transient excursions efficiently and The state variable technique as shown In figure 7 generating a set of matrices or point accurately in the state variable form depends on involves the number of models selected. With real time various levels within engine

models for the

computation as a requirement, an optimum number of operating range from minimum (7 percent) to maximum state models is required. Initially, a piecwise (109 percent) power. These point models are then linear fl• t of the steady state operating line was linked together by scheduling the matrix elements performed to define the minimum resolution. These In each model as a function of both low and high then augmented by additional models to -wise linear models were compxessor rotor speeds to form a piece accurately define transient response through the representation. For large power excursions from 7 full power range. to 109 percent, the coefficients of the

as

differential equations vary continuously both The matrix partial derivatives are generated using rotor speeds increase. After ccmputing the small an offset derivative technique. This technique is changes or deltas in states and controls as they automated on the baseline nonlinear aerothermo deviate from the known steady state operating simulation. In Chin process, each X Is perturbed characteristics, the differential equations are Integrated using a 6imple Euler integration to one at a time whiles holding all other X's and all U's constant. This allows calculations of the A compute the transient engine response.

and C matrix partial derivatives. Each U is then perturbed one at a time while holding all other U's Figure 9 shows an overview block diagram of the important state s0eduled parameter state variable and all X's constant. This allows calculation of B and D matrix partial derivatives. Several model logic. The initial steady state point is different levels of perturbations on the states and calculated from the output and state operating Inputs are used. lines. The initial time point of any transient run is assumed: to be in a steady state condition.

To provide real time capability, the state variable models must be connected efficiently. The type of Transient operation occurs as follows. The last Interpolation 1s flexible and could vary in each time step values for the rotor speed states are application.. The interpolation is controlled by used to calculate the state scheduling parameter scheduling the matrix elements with an independent (SSP). This parameter contains information about variable from the input vector U Applying this the dynamic states at any time during the methodology results in reducing several linear transient. DX's and DU's from the model points models to one nonlinear model as shown in figure 7. above and below the SSP are computed. A relative distance weighting scheme is used to combine a

Engine Model delta

delta from the model point above with the from the model point below. The SSP is also used Using the piecewise linear state variable to schedule the A, B, C, D partial derivatives.

methodology previously discussed, a real time These matrix elements are stored in a linear propulsion model of the Pegasus 11 engine was equation form which allows for rapid interpolation formulated. The engine model consists of 14 state between the discrete model points. State variable models with 6 inputs, 3 states and 21 derivative computations are performed by the matrix outputs. Steady state and dynamic operations from multiplication of the A and B elements with the ground idle to maximum power (7 to 109 percent) up DX's and DU's computed earlier. The derivative to 5000 feet altitude and 0.3 Mach number are vector is then integrated by Euler integration.

searesented. The state variable vectors used are shown in figure 8. The steady state model output is calculated from the operating line curves. The operating line Control Model output levels change as the SSP varies. Transient deviations from the operating line moist be

The control model was developed directly from calculated. This is done through the output

detailed information on the hydromechanical fuel equation which makes use of the same DX and DU

control and water Injection system. All runcL

^ns vectors input to the state derivative equation.

In the real control were modeled. The DY vector represents deviation from the operating line. Summing the steady state operating line values and DY elements gives the output PROGRAM DESCRIPTION vector.

The flight simulator state variable engLie model is Transients are shown in figures 10 and II that exhibit the accuracy with which the state variable nozzle angle and reaction control thrust dynamics

model match*# the aerotherso Pegasus 11 are included. This engine model is referred to as

f representation for to 100 percent power the simple engine model. In the simulation study

a 7

at

F excursion at sea level with water injection and presented her&, this simple engine model is

replaced by the state variable model. A simplified

an altitude of 5000 feet.

model of the landing star including vertical force, braking force and pitching moment is also used.

Agglieation To Piloted Simulator

The ship model used was that of a 6pruance clams

To satisfy the requirements of real time piloted destroyer. Environmental conditions could be

modeling is calm sea state 6. A ship air-wake

simulation, innovative mathematical varied from to

required (3,4). One must attain the desired level turbulence model was included;, Ship dynamics are of fidelity yet have the computations accomplished modeled as six degree of freeljom sinusoidal motion.

In a limited amount of time. Also, since the The ship was assumed to hav'z a fixed mean position propulsion system is only a part of a larger about which it oscillates. Wind over the deck is simulation, only a fraction of the total. composed of a steady induced wind equal to the ship computation List is available for the propulsion speed plus a separate north and east component of system calculations. Real time, then, in the independently specified natural wind. No context of overall simulation requirements, implies

turbulence model designed specifically for VSTOL

that the propulsion simulation must be faster than A model developed for aircraft exists.

real time. conventional carriers is used. This :aodel

calculates free air turbulence as Well as ship wake

The program structure consists of the piece-wine turbulence which may be varied in amplitude. The linear engine representation, the engine control wake intensity is calculated as a function of P model and met of propulsion system force and range, altitude and lateral position relative to

a

balance equations. The control system model the flight deck.

provides the interface with the aircraft-flight control simulation. Figure 12 schematically Flight Control identifies the interfaces with the propulsion system. The aircraft and engine elements are A state rate feedback implicit model-following type combined by means of interfacing logic that controller (6) is used in the basic flight control.

provides fan and core nozzle thrust calculations, Thin flight control concept was applied to all axes reaction control ayatem thrust calculations,

of the aircraft model. Power management controls

ekternal environmental disturbance effects and and pilot displays were designed to match the pilot stick :movements in terms of roll, pi,tch,yaw various modes of control provided by this type of and height requests. flight controller.

Aircraft System Within the overall flight controller' are two Type 1 control system employs variants. The The aircraft model usrd was a typical Harrier control augmentation in all degrees of freedom.

AV-8A. The model Includes nonlinear aerodynamics, Included here, in the transition flight mode, is a engine and reaction control response,. stability vertical axis pilot control based on vertical augmentation, actuator dynamics and a simplified In this type the propulsion velocity command.

landing gear model. The model consists of a group system is within the closed flight control loop.

of basic subroutines applicable to any aircraft and a set of specific aircraft model subroutines which The Type 2 system employs control augmentation in have been configured to represent the AV-8A the attitude degrees of freedom only. Translation, Harrier.

or flight path control, is through thrust and thrust vector angle inputs from the pilot to the The basic subroutines handle trim initialization, propulsion system. Velocity command is not coordinate transformations, the integration of provided in this system..

differential equatiure and the interpolation of

tabulated nonlinear fuunt$ons. These were adapted a head-up display (HUD)

Each control variant used from simulation programs used at the NASA Ames that included flight director information. A Research Center to perform real time simulations.

complete description of the flight control concept is given in (6).

Control inputs, forces and moments and disturbance inputs are determined by user supplied routines Simulator which must be varied to represent specific aircraft types and environmental conditions. The aircraft A small fixed base simulator (C06) was the model includes an aerodynamics subroutine which principal tool used in evaluating the state determines three force and three moment variable propulsion system simulation. The coefficients by interpolating_ tabulated values of simulator was driven by a Xerox Sigma 9 digital ar,_odynamic functions. A separate subroutine is computer in conjunck;ion wi:.ft a PDP-11 for HUD used to represent the Pegasus 11 engine operating generation.

low altitude and Hach number. Engine speed, at ' and Overall engine model operation is stable. Core Test Plan power closely with fan thrusts follow lever inputs The primary objective in the simulation effort was second order effects of bleed superimposed by fuel to evaluate the propulsion system model performance control compensation.

for use in future piloted simulations involving t1,4 In the transition phase the most significant Harrier aircraft and advanced integrated effects are the smooth nozzle angle action and flight -propulsion control concepts. To achieve phase to demanded this objective within limited time available bleed flow. bleed flow in this the for the simulation exercise, a %tringent test from reaction control system which in turn is At hover, condition was chosen with a minimum of test commanded from the flight controller.

bleed activity increases significantly due to parameters. The basic piloting task was to fly an by the pilot. At this I!R curved approach transition at 120 knots to an reaction control demand station keeping point 12000 feet down range point the nozzle angle is fixed at slight forward Initial a and land on the destroyer at a fixed sea position to account for wind and ship velocity, state 6 with wind over deck from the east at 25 knots. The At touchdown the powerlever is brought to idle and only variables in the test were turbulence and the aircraft "drops" to the ship deck. Bleed flow control variant type.

continues to vary due to the action of the reaction this standard flight task, 4 test control system commanded by the flight controller In addition to for maximum control power at hover w,ja run to which is responding to the ship's roll, pitch and yaw motions. Normally the flight control would be determine reaction control system forces and at touchdown.

moments for comparison with published aircraft disengaged data. A small perturbation test was also made to compare the new state variable model thrust Figure 15 shows a series of engine parameter transients for the standard flight task using, response to the simple engine model.

however, the Type 1 flight controller. With this controller, as descrilied previously, the engine is Simulation Results within an altitude flight controller loop. The power lever angle is demanded as a function of Model Performance. The state variable propulsion altitude to provide a prescribed flight path system model performed within the aircraft model commanded from the flight director. The control environment over the full operating range without gains and implicit engine model time constant are run-time Fortran errors or missed intervals during the test program. The model exhibited a high level the same as those used with the simple engine of calculational stability. For a simulation frame model.

time of 50 ms the propulsion system model executed for a real-to-execution time ratio of As shown, the engine breaks into a limit cycle In 8.9 as oscillation. Expanded time scale traces of the 5.6. This compares with a 3.8 ms execution time for the simple engine. oscillation indicate that the engine fuel control is responding to power lever angle demand from the flight controller. This kind of interactive Ensine Performance. A comparison of the state response is typical of integrated flight-propulsion variable engine model forward and rear nozzle thrusts to that of the simple engine is shown in controls where the engine control response is figure 13 for a power lever step increase. from 60 within the flight control bandwidth and the flight controller is analyzed without the advantage of to 70 degrees. For the simple engine both the realistic engine response characteristics.

forward and .rear nozzle thrusts exhibit a lag response with a time constant of about 0425 the state variable model, on the Figure 16 shows the same typical flight task except seconds. For other hand, the thrust exhibits a first order lag that the implicit model in the altitude flight ;1 controller has been modelled to approximate the response with time constant of 0.5 seconds in the state variable engine model response. The implicit front nozzle and a moderately damped, second order model time constant was chosen to approximate the response with a time constant of about 0.25 engine fan thrust response and the flight the rear nozzle. These characteristics seconds in controller gain was reduced by a factor of 4. As are important in designing integrated flight shown, the transient indicates no evidence of controls that include the engine in a closed loop.

oscillation. However, it was evident from the flight data that the altitude controller, although Control Performance. Figure 14 shows a series of acceptable, was not as responsive. Other similar engine parameters as a function of time for typical flights indicated that designing the implicit landing task flights using the Type 2 flight controller to represent the state variable engine controller. Within this flight controller, model response also gave satisfactory results.

propulsion system vectored thrust is controlled directly by the pilot through the power lever and nozzle angle. Altitude information is communicated Concluding Remarks to the pilot through a flight director via the HUD.

Attitude is maintained through the flight A state scheduled state variable propulsion system controller.

model of a Pegasus 11 engine has been developed for 6, real time flight simulator application. The model a limited flight program using a was exercised in simulation with an implicit Harrier aircraft model- following flight controller.

The propulsion system model performed very Well within the flight environment exhibiting excellent stabilty and satisfactory cycle time calculational characteristics. No run-time errors or missed Intervals occurred. The engine and control transient characteristics were typical of a turbofan engine.

The propulsion system exhibited an oscillatory characteristic within the closed loop implicit model -following) flight controller. Further analysis of th1s flight control within the context model is required to provide of the state variable satisfactory flight performance.

References 1. Roth, S.P. at al, "VSTOL Propulsion Control Analysis Phase I Final Report", NASA CR^165208, 1981.

at al, "Real Time Dynamic Propulsion 2. Roth, S.P.

Model for Ma r i -in-the-loop Simulators", PWA FR-9433, 1977.

J.R. and Hart, C.E., "Real Time 3. Mihaloew, Digital Propulsion System Simulation for Manned Flight Simulators", NASA TM-78958, 1978, "A Nonlinear Propulsion System 4. Mihaloew, J.R., Simulation Technique for Piloted Simulators", NASA Technical Memorandum 82600, 1981.

5. Nave, R.L., "A Computerized VSTOL/Small Platform Landing Dynamics Investigation Model", Report No. NADC-77024-30, 1977.

6. Merrick, V.K., "Study of the Application of an Model-Following Flight Controller to Implicit Lift Fan 'VTOL Aircraft", NASA Technical Paper 1040, 1977• C N1 3841 Figure 1. - Rolls-Royce Pegasn 11 propulsion system.

W,1Y,

I

^ n.w .r..^ rrw uw^..

.ri.w Figure L - Fuel flew control system.

wy ' Mw1 1r ^^ M MI.Iw.A rM•m.

A M M rw _w www •«w w•• wl w.w rr•w^w .. ^^^ w.w^ w.r .ww_ w .ww.f figure I - Water injection system.

FRONT PITCH 1VE PITCH VALVE CORF NOZZLE figure 4. - Reaction control jet nozzle system.

MODELING REALTIME MODEL L -- — • ENGINE L------- ^ AEROTHERMOCOMROLS ------J AEROI400 DYNAMICS METAL TEMPERATURE DYNAMICS ROTOR INERTIA DYNAMICS

AIRCRAFT

AUTOMATIC FLIGHT CONTROLS VSTOL AIRCRAFT' 10-A 1 10 FREQUENCY- HERTZ Figure 5. - Modeling frequency range of Interest.

MATRIX OUTPUT EQUATION AY • C' AX+D • AU

UTPUTS

CONTROZ=TiSTATES

AUAXb

UC == Y =4

A. MATREr RIA"TS 1. MATR IX ELEMENTS A. B.C.D MATRICES GENERATED MAX - AT EACH MODEI POINT T= - All oil ,E)

iAl

SSP SSP C. MATRIX ELEMENTS TEADY STATE OPERATING D. MATRIX ELEW. t .

HARACTERISTIC

i

i

CII I DII MIN SSP SSP l _I _ SSP • RENGINE tPEEDSI M14 MAX ENGINE FUEL FLOW FNUre 1. - State %(.heduled parameter nonlinear real time model.

CONTROL VECTOR STATE VECTOR • FUEL FLOW • LOW ROTOR -,PEED *WATER IN )E C I ION FLOW AX • • HIGH ROTOR SPEED • REACTION CONTROL BLEED FLOW • BURNER METAL TEMPERATURE AU *ENGINE FACE PRESSURE • ENGINE FACE DELTA TEMPERATURE • AMBIENT PRESSURE OUTPUT VECTOR • CORE NET THRUST • FAN NET THRUST • FAN PHYSICAL • AIRFLOW • HIGH COMPRESSOR PHYSICAL AIRFLOW • FAN AVERAGE PRESSURE RATIO • HIGH COMPRESSOR PRESSURE RATIO • FAN NOZZLE DUCT TEMPERATURE • FAN NOZZLE DUCT PRE SSURE • HIGH COMPRESSOR INLET TEMPERATURE • HIGH COMPRESSOR INLET PRESSURE AY • BURNER INLET TEMPERATURE • BURNER INLET PRESSURE *HIGH TURBINE INLET TEMPERATURE *HIGH TURBINE INLET PRESSURE G LOW TURBINE INLET TEMPERATURE G LOW TURBINE INLET PRESSURE • CORE NOZZLE DUCT TFMPERATURF • CORE NOZZLE DUCT PRESSURE *THRUST SPECIFIC FUEL CONSUMPTION • FAN SURGE MARGIN" • CORE SURGE MARGIN' AUXILLARY MODEL OUTPUT COMPUTATIONS Figure B. - Real time model slate i,ar adle vectors.

- A ITERATE AN STATE D SET x STATE VECTOR OXIt^ INI TIAL EQUATION CONDITIONS A41j) X(0) A11,1,L) IDIJ,Li STATE IDENTIFY X CALCULATE CALCULATE FEEDBACKS COMPUTE AND MODEL IL) SSP "$ AX STATE BRACKET MATRIX U AND SCHEDULNG ROL MO D EL ODEL 6 U,S POINT d MENTS

M

PARAMETER INPUT M VECTOR LIT C (1, J, L) D li„ I, L) AXIL) AMP STATE SCHEDULING PARAMETER CALCULATE )U OUTPUT WEIGHT OW RATE MJ VECTOR AX AND AL' AXwIIt STEADY STATE, EQUATIONS - , AuM,la)

YitJ)

COMPUTE OPERATING

j

LINE OUTPUTS Figure 9. - State scheduled state variable engine model

P

^a

XW S 2 t n LU p ^ O

U

a OC 0AER HERMO MODEL

a`

7 VARIABLE MODEL ib)

3 z

a _ I I I I i l LL 90 100 ^ e

501 1 D

0 1 2 3 4 5 6 7 0 1 2 3 4 6 TIME, sec Figure 10. -Full range f uel transient with water injection At 02 percent Nan speed, sea level static, s P k k f ^ 3916 t9A6

t

1.116 d La yl .116

m

-.014 _a 100 100 4ENKOMO EL /CO37 d 50 IZ 50 STATE VARIABLE MODEL

tbl

o s^ u Z i 0 a K Z^ X G a z z U.

'_ 0 2 4 6 8 00 2 4 6 8 TIME, sec Figure 11. - Full range fuel transient at 5000 feet i'ltitude.

NOZZLE NOZZLE PILOT THRUST VECTORED THRUST NOZZLE ANGLE A A^tNEaHANISM SPLAY VECTOR ANGLE ANGLE COMMAND ACTUATOR EFFECTS THRUSTS PROPULSION ENGINE FUEL PROPULSION SYSTEM PICOT FUEL FLOWENGINE HEIGHTAND WATER SYSTEM FORCE AND INJECTION MODEL THRUSTS AND WATER FLOW BALANCE COMMAND CONTROL MOMENTS EQUATIONS ENGINE FEEDBACKS E XIT BLEED CON DITIONS REACTION CONTROL' PUFFER IT BLEED FLOW DEMAND SYSTEM NOZZLE AREA AND COMMANDS THRUST REACTION CONTROL JET THRUSTS CALCULATIONS Figure 12 - Propulsion system model for aircraft Integration.

z6^ 13 0 H

H

Oa

11 WN

J A 10

^ZCDlZ -

a 9 0

1 1 1 1 ° 100 B. J ^NLn Ej

!

j6 . 12NC^^

oz°

N

g 6-z loo to L s

I

I I 11 1 t—L I I I I I I II

1 1 1 1 1 STEP INPUT

c

° 12 50 lq m "'_' c 50 -^ 13 p°a

N "'

9 yk

to N 10 u' t U.

1x- 0 $ 0 1 2 3 4 5 TIME, sec LU^J t

f

step Figure 13. - Model response to thrott input 140 160 J80 200 220 20 } 40 60 80 100 120 .

1 TOUCHDOWN START HOVER TRANSITION TIME, sec Figure 14. - Approach and landing engine transients using type 2 flight controller.

1 000 1 000 W ti I 10 000 N ^ ^Ny S2 a 5 000 CC N 1 1__ 1 1 .1.--- I _- -I I I _! __J Q 10 000 t N N GC ti O[ to < 0 G 20 000 W 10 000

U

O 1 I 1 l 1 1 I I 1_ 1 l 1 ^_ ( 1 ! 1 1 ^_ t 10 000

r

5000 —

Z

111 fill 111111111 i__j

200 r o^ W N a O LL --I—J Q _._ ._.1._.

15 000 1-500

LL c `

Wa

0 1

1 1 1 1 f 1 1 ( _^ 1 1 I

W J O h ti O z 0 I I I 1 ( 1 1 I cc W z p 4 1 1 1 I J_ I I ____ I 1 I I 1 I I I I_ L I 0 20 40 60 80 100 120 140 160 180 10u 220 START TRANSITION HOVER TIME, sec Figure 15. - Approach and hover engine transients using type I flight controller with simple engine control constants.

i so -_ 5.0

p 1tI111IIIII

o ^, to 100 -

^ H ~ i ; ; SO °v Wd N z 111 11H Jam.

W

p 10 0 / JV pp C z de

I I I I l i I I 1J

200 220 20 40 60 80 100 120 140 160 180 START TRANSITION SHOVER TIME, sec Figure 16, - Approach and hover engine transients Ono type 1 flight controller With state variable control constants.

s'