Aircraft Turbine Engine Control Research at NASA Glenn Research Center

at Lewis Field

Dr. Sanjay Garg

Ph: (216) 433-2685

FAX: (216) 433-8990

http://www.grc.nasa.gov/WWW/cdtb

email: sanjay.garg@nasa.gov

Chief, Intelligent Control and Autonomy Branch

NASA Glenn Research Center

Aircraft Turbine Engine Control Research at

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

Outline

Safety and Operational Limits State-of-the-Art Engine Control Logic Architecture Early Stages of Turbine Engine Control (1945-1960s) Maturation of Turbine Engine Control (1970-1990) Recent Significant Accomplishments (1990- 2004) Current Research (2004 onwards)

• • • • • •

The Engine Control Problem Historical Glenn Research Center Contributions Advanced Engine Control Research Conclusion

– – – –

Glenn Research Center Intelligent Control and Autonomy Branch

need to have

–

20,000 hours

at Lewis Field

–

high temperatures and

–

avoid stall, combustor blow out etc.

–

Provide smooth, stable, and stall free operation of

Basic Engine Control Concept

Reliable and predictable throttle movement to thrust Changes in ambient condition and aircraft maneuvers Harsh operating environment Safe operation Need to provide long operating life Engine components degrade with usage

Thrust cannot be measured

• response • • cause distortion into the fan/compressor • large vibrations • • • reliable performance throughout the operating life Objective: Issues:

Glenn Research Center Intelligent Control and Autonomy Branch

• the engine via single input (PLA) with no throttle restrictions •

Yes Power desired?

Sensor Measure produced power Control Logic No

at Lewis Field

Inject fuel flow into combustor Fuel nozzle Accessories Valve / Actuator Meter the computed fuel flow Pump fuel flow from fuel tank Control Determine operating condition Compute desired fuel flow

Basic Engine Control Concept

Pilot’s power request Throttle

T = F(N)

•

Glenn Research Center Intelligent Control and Autonomy Branch

Since Thrust (T) cannot be measured, use Fuel Flow WF to

• Control shaft speed N (or other measured variable that correlates with Thrust

8mm+ 40+ Bar Gas pressures 20000+ hours Shaft movement Between service 1100+ºC Metal temperatures

at Lewis Field

at 650+ºC 10 000rpm Cooling air 0.75m diameter 2000+ºC - 40ºC ambient Flame temperature acceleration to beyond 20kHz 50 000g centrifugal >100g casing vibration Buffeting 2.8m Aerodynamic Diameter 120 dB/Hz to 10kHz

Environment within a gas turbine

~2 tonne/sec Air mass flow Foreign objects Birds, Ice, stones

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

long life

–

Referred to as Accel/Decel Schedule

Operational Limits Mapped into Limits on allowable fuel flow (ratio of fuel flow to combustor pressure) as a function of Corrected Fan Speed -

N1, N2

–

R R 2 2 N N over-speed over-speed over-pressure over-pressure

Compressor and Fan

minimum fuel

–

–

Operational Limits over-temp Deceleration over-temp Deceleration

line line blowout blowout Operating Operating Acceleration Acceleration surge surge Maximum Fan and Core Speeds Maximum Turbine Blade Temperature Lean Burner Blowout Maximum Turbine Inlet Temperature

Adequate Stall Margin

• • • • •

Glenn Research Center Intelligent Control and Autonomy Branch 30 30

Wf Wf Ps Ps Structural Limits: Safety Limits: Operational Limit:

• • •

GE I-A (1942)

at Lewis Field

Proportional control gain or droop slope Max. power Safe operating region Min. flow limit Engine shaft speed Required fuel flow @ steady state

Historical Engine Control

Max. flow limit - Hydro-mechanical governor. - Minimum-flow stop to prevent flame-out. - Maximum-flow schedule to prevent over-temperature

Stall protection implemented by pilot following cue cards for

Idle power

• throttle movement limitations

Droop slope Fuel flow is the only controlled variable.

Glenn Research Center Intelligent Control and Autonomy Branch

Fuel flow rate (Wf) or fuel ratio unit (Wf/P3) •

at Lewis Field

we will focus on fuel flow

–

Typical Current Engine Control

- FAA regulations provide an allowable maximum rise time and

- There are many controlled variables

Engine control logic is developed using an engine model to provide

• guaranteed performance (minimum thrust for a throttle setting) throughout the life of the engine maximum settling time for thrust from idle to max throttle command

Allows pilot to have full throttle movement throughout the flight envelope

•

Glenn Research Center Intelligent Control and Autonomy Branch

20 20 20 n n p Phi max Ps30 hi limit p Ps30 low limit 10 10 10 Time (s) Phi min 0 0 0 10 15 20 20 30 40 50 100 200 300 400 500 VSV pos. (deg) Wf/Ps30 (pph/psi) Ps30 (psia)

at Lewis Field

Inputs/Outputs

demand 20 20 20

–

n n Nf limit T48 limit 10 10 10 Time (s) x 10 0 0 0 1 2 0.5 1.5 2.5 1000 1500 2000 2500 1000 1500 2000 Wf (pph) q Nf (rpm) T48 ( R) 20 20 20 n Nc limit 10 10 10 Time (s) 0 0 0 0 0 50 20 40 60 VBV pos. (% open) 7500 8000 8500 9000 9500 TRA (deg) Nc (rpm)

Burst-Chop Example

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

(1945-1960s)

Time-Domain Closed-loop Engine Analysis Corrected Parameters for Model Simplification Simplified Dynamic Model of Engine Response Real-time Engine Dynamic Modeling Using Computers

Early Stages of Turbine Engine Control

• • • •

Glenn Research Center Intelligent Control and Autonomy Branch

J47:

–

noise from

–

at Lewis Field

performed time-domain studies to identify the

–

Time-Domain Closed-Loop Engine Analysis

Hydro-mechanical fuel control for main fuel flow and electronic (vacuum tube) fuel control for afterburner Limit cycle observed in altitude testing at GRC speed sensor feeding into the “high gain” of the speed governor GE and NASA engineers worked together to solve the problem problem and reduce the control gain at altitude This established the strong industry/government partnership in engine controls development which continues to this day

Glenn Research Center Intelligent Control and Autonomy Branch

• • •

1948: GE tests the first afterburning turbojet in the world

from months to

–

at Lewis Field

performance parameters corrected to non-

Performance calculations at each operating point took several hours Considerably reduced the amount of time to develop engine steady-state performance model a few weeks, and also reduced the amount of experimental testing to be done

– – Corrected Parameters for Model Simplification

Engine steady-state performance obtained through cycle calculations using slide rule / desk calculators NASA engineers developed the concept of “Corrected Parameters” – dimensionless temperature and pressure w.r.t. sea level static conditions • • Corrected Parameters concept is being used to this day for engine simulation and analysis

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

The study showed that the transfer function from fuel flow to engine speed can be represented by a first order lag linear system with a time constant which is a function of the corrected fan speed: N(s)/WF(s) = K/(as+1) with a=f(N)

Simplified Dynamic Model of Engine Response

Dynamic behavior of single-shaft turbojet first studied at

Glenn Research Center Intelligent Control and Autonomy Branch • NACA Lewis Laboratory in 1948 This discovery significantly reduced the time required to design and validate the control logic Linearized -TN-2826.

turbine Engines,” NACA

at Lewis Field

Dynamic response of a

• turbojet to a step fuel input

Dynamics of Gas- GRC pioneered methods to perform closed-loop (integrating control logic with the engine model) dynamic simulation using analog computers

Simulations using Analog

From J.R. Ketchum, and R.T. Craig, “Simulation of

• Computers at NACA Lewis in the 1950s

•

With increasing complexity of engines, it became too complicated to design and evaluate control logic primarily through testing.

Real-time Dynamic Engine Modeling using Computers

•

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

(1970-1990)

Maturation of Turbine Engine Control

Digital Electronic Engine Control (DEEC) Multivariable Engine Control Development Sensor Fault Detection, Isolation and Accommodation

• • •

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

simplified hardware

Digital Electronic Engine Control

increased thrust, faster response times, improved afterburner operation, improved airstart operation, elimination of ground trimming, fail-operate capability, and Extensive PSL testing done for operation throughout the flight envelope to get clearance for flight tests on the F-15 Early 1980s NASA in collaboration with USAF and P&W conducted flight tests of DEEC on an F-100 engine Benefits of DEEC: Testing at GRC altitude test facility led to resolution of a nozzle stability and verification of a faster augmentor transient capability Flight testing at DFRC (now AFRC) demonstrated capability of digital control and led to introduction of Electronic Engine Control Units on military and commercial jets.

– –

• • • • NASA testing of DEEC was a critical step leading to current FADECs

Glenn Research Center Intelligent Control and Autonomy Branch

., and R.D. Hackney, “Multivariable

GRC developed a series Successful testing on an

• of LQRs connected with simple transition logic and several operating limits to achieve effective large- transient controls. • F100 engine in NASA’s altitude test facilities Joint Propulsion Conference, 1979, AIAA-79-

at Lewis Field

th F-100 LQR Control Block Diagram From: B. Lehtinen, R.L. DeHoff Control Altitude Demonstration on the F100 Turbofan Engine,” paper presented at the 15 1204.

Multivariable Engine Control Development

AFWAL and NASA-Lewis jointly sponsored the Multivariable Control

• Synthesis (MVCS) program from 1975 to 1978 to investigate applicability of emerging optimal control and Linear Quadratic Regulator control theory to jet engines.

Glenn Research Center Intelligent Control and Autonomy Branch

Studies concluded that performance benefits of MVCS were not commensurate with complexity of control design and implementation

at Lewis Field

In mid to late 1980s GRC investigated engine sensor failure detection approaches under the Advanced Detection, Isolation and Accommodation (ADIA) program ADIA algorithm consisted of 3 elements: i) hard sensor failure detection and isolation logic; ii) soft sensor failure detection & isolation logic; and iii) an accommodation filter ADIA algorithm was integrated with a microcomputer implementation of F-110 engine and a real-time evaluation was performed using a hybrid computer simulation Transient engine operation over the full power range with a single sensor failure detection and accommodation was demonstrated on the F-100 in the PSL. The ADIA development team received the prestigious R&D 100 award

• • • • • Many elements of the NASA ADIA approach are implemented in current FADECs

Sensor Fault Detection, Isolation and Accommodation

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

(1990- 2004)

Recent Significant Accomplishments

Integrated Flight Propulsion Control Research (IFPC) Practical Application of Multivariable Control Design Intelligent Life Extending Control (ILEC) High Stability Engine Control (HISTEC) Active Stall Control

• • • • •

Glenn Research Center Intelligent Control and Autonomy Branch

at Lewis Field

Integrated Methodology for Propulsion

–

Emphasis on Advanced Short Take-Off Vertical Landing Aircraft (ASTOVL) led to investigation of IFPC concepts GRC developed IMPAC and Airframe Control and applied it to a concept ASTOVL aircraft IMPAC based ASTOVL IFPC design was successfully demonstrated in piloted simulations in the transition flight phase

Integrated Flight Propulsion Control Research (IFPC) • • •

Glenn Research Center Intelligent Control and Autonomy Branch

IMPAC and GRC led industry IFPC design studies helped developed various concepts that were later applicable to F-35

at Lewis Field

GRC IFPC research led to development of methods for practical application of Multivariable Control Design, including problem formulation using robust control design techniques, addressing controller scheduling and Integrator Wind Up Protection (IWP)

•

Practical Application of Multivariable Control Design

Glenn Research Center Intelligent Control and Autonomy Branch

IWP design and application to an engine simulation

Technologies were transferred to industry through collaborative tasks and contributed to successful application of MVC for F-135 engine 0.628 0.718 0.931 0.481 Control (-37.2%) (-36.6%) (-44.2%) (-21.0%) Schedule Optimized ILEC 0.725 0.826 1.181 0.500 Control (-27.5%) (-27.0%) (-29.2%) (-17.9%) Schedule 1.0 1.132 1.668 0.609 Control Original Schedule

at Lewis Field

(With Varying Ambient Condition and Control Mode ) Design Condition With Typical Operating Conditions Hot Day Bias Cold Day Bias Comparison of Average TMF Damage Accumulation 16.6 16.4 16.2 Original Schedule ILEC schedule Optimized schedule 15.8 Time 15.6 15.4 15.2

Intelligent Life Extending Control (ILEC)

x 10

wing life extension through “Intelligent Control” continues to be an

How the engine is controlled during transients has a significant impact on life of the hot gas path components GRC research in collaboration with industry demonstrated that optimizing the engine acceleration schedule can result in significant increase in engine on-wing life 3.1 2.94 2.96 2.98 3.02 3.04 3.06 3.08 N2 Speed

Glenn Research Center Intelligent Control and Autonomy Branch

On- area of interest in the industry

• •

Engine Core Speed Acceleration Response (Comparison of Original, ILEC, and Optimized Control)

continues to be an area

–

at Lewis Field

Highly successful flight test program demonstrated the capability to estimate inlet distortion and accommodate it through setting the stall margin requirement online in real time

High Stability Engine Control (HISTEC)

In mid to late 1990s, GRC in collaboration with P&W, USAF, and DFRC, developed the HISTEC concept and tested it with the PW- 229 engine on a modified F-15 Aircraft

Glenn Research Center Intelligent Control and Autonomy Branch

•

System studies indicate that up to 2% SFC reduction can be obtained by intelligently managing the operating stall margin of research

Active Stall Control demonstrated with simulated recirculated flow

at Lewis Field

Active Stall Control

In mid 1990s to early 2000s, there was a tremendous interest in exploring the feasibility of extending compressor operation into a higher efficiency region by using active flow control GRC developed various technologies including stall precursor detection, optimum flow injector design, high frequency actuation. Technologies were demonstrated on a modern multi-stage compressor in collaboration with USAF and industry

• •

Glenn Research Center Intelligent Control and Autonomy Branch

System level benefits did not appear to be sufficient enough to warrant transition of technology to product

at Lewis Field

(2004 Onwards)

Current Research

Model-Based Engine Control and Diagnostics Engine Dynamic Models for Control and Diagnostics Technology Development Distributed Engine Control (DEC) Active Combustion Control High Speed Propulsion System Dynamic Modeling and Control

• • • • •

Glenn Research Center Intelligent Control and Autonomy Branch

Engine Fuel flow Pressures • • Rotor Speeds Temperatures • • Instrumentation Positions Actuator Thrust • Efficiencies

at Lewis Field Flow capacities

• Stability margin On-Board Model • & Tracking Filter • Engine Estimates Fault Codes Component Sensor Estimates Performance Control • Sensor Measurements Ground-Based Diagnostics “Personalized” Maintenance/Inspection Advisories •

Enhanced gas path diagnostics

Sensor

Real-time performance estimation

•

Validation & Fault Detection

•

Performance enhancing engine control

•

Selected Sensors Bleeds • Fuel Flow Actuator Commands • Level

Model-Based Engine Controls and Diagnostics

Ground On Board Variable Geometry •

Glenn Research Center Intelligent Control and Autonomy Branch

Fnet HPC SM Estimation OTKF Sensors Engine Disturbance/ Health Fnet Actuator HPC-SM Limiting Parameters Feedback Parameters Protection Logic

at Lewis Field

Thrust Controller Limit Controller e e An Integrated Architecture for Aircraft Engine Performance Monitoring and Fault Diagnostics has been developed and validated against engine test data Command Fnet Limiters SM HPC

typically number of sensors is less than the

–

Model-Based Engine Control and Diagnostics

Challenge is to have an accurate enough on-board model which reflects the true condition of the engine “health parameters” in the model GRC has developed a patented approach to determining a set of “tuners” which provide a good estimate of unmeasured engine performance and operability parameters in the presence of degradation with usage

Glenn Research Center Intelligent Control and Autonomy Branch

An architecture for Model-Based Engine Control has been developed and demonstrated on a nonlinear engine simulation to provide a tight control of thrust and stall margin

• •

at Lewis Field

(2010)

Commercial Modular Aero-Propulsion Sys Sim (2008)

–

Modular Aero-Propulsion System Simulation (2003)

Diagnostics Technology Development

–

Engine Dynamic Models for Control and

Simulation of a modern fighter aircraft prototype engine with a basic Simulation of a modern commercial 90,000 lb thrust class turbofan High fidelity simulation of a modern 40,000 lb thrust class turbofan This simulation is being used extensively in NASA sponsored • research control law: http://sr.grc.nasa.gov/public/project/49/ • engine with representative baseline control logic: http://sr.grc.nasa.gov/public/project/54/ • engine with realistic baseline control logic: http://sr.grc.nasa.gov/public/project/77/ • propulsion control and diagnostics research, and has been downloaded by over 30 external organizations

MAPSS C-MAPSS C-MAPSS40k

The following engine simulation software packages, developed in Matlab/Simulink and useful for propulsion controls and diagnostics research, are available to U.S. citizens from GRC software repository • • •

Glenn Research Center Intelligent Control and Autonomy Branch

Fnet out_Nf nf_sens nc_sens out_P2 out_T2 out_Nc out_P25 P2_sens T2_sens egt_sens out_P50 out_T50 out_T24 out_T30 P25_sens P50_sens T30_sens T25_sens P30_sens out_P30 out_Fdrag out_Fnet out_Fgross Fnet Fdrag Fgross T2 sens P2 sens Nlp sens T24 sens T30 sens T50 sens P25 sens P30 sens P50 sens Nhp sens Engine Model altitude dTamb Mach Nf _zro Nc_zro f uel f low VSV VBV out_Wf out_VBV out_VSV Nf_zro Alt Nc_zro Mach dTamb GUI driven operation Wf _act VSV_act VBV_act Controller alt mach NcR Nf R corrected speeds Engine Outputs Displays Alt Mach NfR NcR Simulation programmed in graphical language Simulation Inputs Health Parameters C-MAPSS40k PAX200 Commercial Turbofan Engine and Controller Models

at Lewis Field

Plotting and graphical analysis capability

Glenn Research Center Intelligent Control and Autonomy Branch

Commercial Modular Aero-Propulsion System Simulation 40k

C-MAPSS40k thrust and stall margin response to throttle movements Engine flight data used to tune physics-based model

Industry Partnership

–

at Lewis Field

High temperature electronics Communications based on open system standards Control function distribution

Challenges: • • •

Government

Distributed Engine Control Working Group

Distributed Engine Control

Reduce control system weight Enable new engine performance enhancing technologies Improve reliability Reduce overall cost

Glenn Research Center Intelligent Control and Autonomy Branch

Objectives: • • • •

Real-Time Target

Fid

-

High-Fidelity High

Real-Time Modeling

Collaborative Environment for

E DH

Applications on the Turbine Engine k k

Development of Distributed Control TE AR

ol ol ork ork LA DW

MU ontro ontro p

ARDWARE AR etwo etwo Control SIMULATEDH Network

Hardware-in-the-Loop

Concept

–

HIGH HARDWARE TEMPERATURE

Simulation

Validation

–

Stimulus

Distributed Engine Control Hardware-in-the-Loop

Modeling

Controls

Evolution, &

Technology Sustainment

Development,

Revolutionary

Flame Acoustics Fuel Injector NL Combustor Pressure Instability Pressure + + + Filter White Noise Pressure from Fuel Modulation Combustor Instrumentation (pressures, temp’s) Research combustor rig at UTRC Controller Phase Shift & Combustion Fuel lines, Injector

at Lewis Field

Fuel Valve Emissions Probe Advanced Control Methods Fuel Nozzle Assy

Active Combustion Control

Modulated Fuel Flow Accumulator Modulating Valve

Glenn Research Center Intelligent Control and Autonomy Branch

Fuel delivery system model and hardware In 2005, GRC demonstrated the feasibility of active suppression of combustion instability in a realistic aero-combustion environment High-frequency fuel valve In Fuel off on Controller Controller Fuel Valve Dynamic Characterization Rig Time, sec Time RMS=0.54, Mean=0 Time RMS=0.11, Mean=0 Increasing Fuel/Air Ratio Combustor Pressure Response to 0 1 2 0 1 2 -2 -1 -2 -1 Combustor

at Lewis Field

used to prevent growth of combustion instability Adaptive Sliding Phasor Averaged Control algorithm (other side) Fuel Actuator (2012) Sensor Pressure Low Emissions Advanced Combustor Prototype

Active Combustion Control

Instability suppression was Utilize Active Combustion NASA GRC developed Adaptive Sliding Active combustion control can suppress

Glenn Research Center Intelligent Control and Autonomy Branch

Objective: Control Techniques to suppress combustion instabilities. Approach: demonstrated using a high-frequency fuel valve to modulate the main fuel. Fuel valve dynamic characterization rig utilized to predict actuator authority prior to combustion testing. Impact: combustion dynamics and allow operation of a low emissions combustion concept at conditions that would otherwise be precluded due to unacceptably high pressure oscillations.

Results: Phasor Averaged Control able to prevent growth of combustion instability in a low emissions advanced combustor prototype with a separate pilot and main fuel stage.

Combustion Instability Suppression Demonstrated for Low NOx Combustor Configuration

at Lewis Field

Modeling and Control

High Speed Propulsion System Dynamic

GRC is performing research in Aero-Propulso-Servo- Elasticity (APSE) - developing higher fidelity propulsion system dynamic models to investigate coupling effects with flexible modes GRC is conducting research in Combined Cycle Engine (CCE) technologies with an emphasis on inlet design for turbine based CCE. Controls focus is on developing dynamic models and control for safe mode transition

– –

Future Supersonic aircraft are expected to be long slender body fuselage. There is potential for interaction between flexible body modes and propulsion dynamics which can negatively impact performance and ride quality NASA has interest in developing hypersonic air breathing propulsion systems for low cost access to space. Air Force is interested in systems to provide rapid access to space

• •

Glenn Research Center Intelligent Control and Autonomy Branch

response distribution Inlet axial pressure Nozzle frequency domain Supersonic Concept Vehicle Aero-Servo-Elastic wing displacements w/ and without propulsion system coupling 2D flow field – Inlet CFD Variable cycle engine thrust response Goals Approach Accomplishments Closed loop APSE Model

Aero-Propulso-Servo-Elasticity (APSE)

At NASA GRC the propulsion system models are developed for a Variable Cycle Engine (VCE) based on 1D gas dynamics for engine and quasi-1D CFD for the inlet and nozzle. Alternatively, parallel flow path modeling is developed that includes rotational flow to study dynamic performance of flow distortion. Developed atmospheric turbulence models, propulsion system 1D gas dynamics models, and quasi 1D inlet and nozzle models. Developed first VCE dynamic model with feedback controls and schedules to operate continuously with varying power level. Developed first closed loop APSE system.

• • • • • Develop dynamic propulsion system models, aero-servo-elastic & aerodynamic models, and integrate them in closed loop together with atmospheric turbulence to study the dynamic performance of supersonic vehicles for ride quality, vehicle stability, and aerodynamic efficiency.

Shock Position Estimator P Low-speed plug High-speed plug AIP Low-speed flow path AIP Process Geometry High-speed flow path

Overview

u Isolator Overboard Bypass

–

Controller Bleed ducting e

at Lewis Field

AOA Σ High speed cowl • t0 t0 P M T splitter • • • Free-stream Set Point Pivot for AoA Variable ramp High-speed flow path Pre-compression forebody plate Low-speed flow path Testbed GRC 10-foot x 10-foot Low to high speed flowpath transition control Shock positioning Fuel flow

Hypersonic Propulsion System Control

Glenn Research Center Intelligent Control and Autonomy Branch

• • • Combined Cycle Engine (CCE)

at Lewis Field

Distributed Engine Control

–

Model Based Engine Control, and

–

Concluding Remarks

These technologies are critical for achieving the challenging goals of increased aviation efficiency and safety, and reduced environmental impact

–

GRC contributions to engine dynamic modeling and control technology development were critical for advancement of turbine engines, and are reflected in the modern Full Authority Digital Engine Control (FADEC) on today’s aircraft GRC developed engine dynamic model software packages provide an important set of tools for the technical community to develop and demonstrate advanced propulsion control and diagnostics technologies Current GRC research will enable revolutionary advances in control logic architecture control hardware architecture GRC research in high speed propulsion system dynamic modeling and control is critical to enable future supersonic aircraft and hypersonic air-breathing propulsion vehicles

• • • •

Glenn Research Center Intelligent Control and Autonomy Branch